Text

Forwards CRGR Review of Proposed Resolution of Unresolved Safety Issue A-43, Containment Emergency Sump Performance. Draft NUREG-0869 & NUREG-0897 Re Unresolved Safety Issue A-43 Also Encl
	ML20154M176
Person / Time
Issue date:	08/07/1985
From:	Harold Denton; Office of Nuclear Reactor Regulation
To:	Stello V; Committee To Review Generic Requirements
Shared Package
ML20151L125	List: NUREG-0869, Forwards CRGR Review of Proposed Resolution of Unresolved Safety Issue A-43, Containment Emergency Sump Performance. Draft NUREG-0869 & NUREG-0897 Re Unresolved Safety Issue A-43 Also Encl;
References
	REF-GTECI-A-43, REF-GTECI-ES, RTR-NUREG-0869, RTR-NUREG-0897, RTR-NUREG-869, RTR-NUREG-897, RTR-REGGD-01.082, RTR-REGGD-1.082, TASK-A-43, TASK-OR NUDOCS 8603130350
	Download: ML20154M176 (420)
	v • d • e

{{#Wiki_filter:a nty

     +         bg                           UNITED STATES i

[ g NUCLEAR REGULATORY COMMISSION . 5 "j WASHINGTON, D. C. 20555

       ...../                              August 7, 1985 O

MEMORANDUM FOR: Victor Stello, Jr., Chairman Committee to Review Generic Requirements FROM: Harold R. Denton, Director Office of Nuclear Reactor Regulation

SUBJECT:

CRGR REVIEW 0F A PROPOSED RESOLUTION OF USI A-43

                              " CONTAINMENT EMERGENCY SUMP PERFORMANCE" We have revised the USI A-43 Regulatory Analysis to address: (a) questions and comments raised at CRGR Meeting No. 66 where the proposed A-43 resolution was discussed, (b) the staff's followup assessment of CRGR's views regarding inherent plant design features and operator capability to detect and mitigate a sump blockage occurrence, (c) utilization of results being obtained through

_ recent Severe Accident Sequence Analysis (SASA) studies regarding core melt consequences, and (d) recent studies regarding inherent containment overpressure capacity relative to the design basis pressure. The net effect has been a relaxation of the previous position which was to have all plants assess potential debris blockage effects using the guidelines provided in RG 1.82, Revision 1. In our revised position, we are proposing issuance of a Generic letter for information only to all licensees and applicants describing safety concerns dnd providing guidance with respect to assessing the effect of debris blockage on recirculation and containment spray pump NPSH margins during the ~ post-LOCA period. There are no requirements for operating plants or for plants under construction. Also, we are proposing to issue RG 1.82 Revision 1 and Standard Review Plan Section 6.2.2, Revision 4 providing specific guidance acceptable to the staff for assessment of sump and RHR suction intake performance which are to be implemented on standard plants and new construction permits only. Therefore, based on these studies and prior studies, and public comments received, we propose: (1) That the staff's technical findings (NUREG-0897, Revision 1) be issued for use as an information source by applicants, licensees, and the staff in addressing design and operation of PWR containment emerge _ncy sumps and BWR RHR suction intakes. 8603130350 850829 PDR REVQP NROCROR NEETINGOBO PDR

l August 7, 1985 (2) That Revision 1 of RG 1.82, be issued to reflect the techni'ca1 findings reported in NUREG-0897, Revision 1. In particular, the 50% screen blockage criterion would be removed. RG 1.82, Revision 1 provides specific guidance acceptable to the staff for assessment of sump performance and RHR suction intakes including debris blockage effects. (3) That the NRC Standard Review Plan (SRP, NUREG-0800) Section 6.2.2, Revision 4, " Containment Heat Removal Systems," be issued to reflect the guidance provided in RG 1.82, Revision 1 and the technical findings in NUREG-0897, Revision 1. (4) That a Generic Letter be issued to all licensees and applicants out-lining the potential safety concerns related to potential post-LOCA debris blockage and the non-conservatism of the current 50% blockage criterion contained in Revision 0 of RG 1.82. This action is significant because replacement of plant insulation has become common practice in operating plants and licensees and designers should be made aware of the potential for recirculation blockage from insulation debris following a LOCA. RG 1.82, Revision 1 and SRP Section 6.2.2, Revision 4 would become effective . 6 months from date of issuance and apply to new cps and standard plant designs. A draft copy of the proposed Generic Letter is provided in Appendix H of NUREG-0869, Revision 18 which is enclosed with this transmittal package. Enclosure 1 addresses issues raised by the CRGR in Meeting Number 66 (July 11, 1984). NUREG-0869, Revision 18 (Enclosure 2) reflects tne considerations and evaluations noted above and is labeled as Revision 18 to differentiate from the May 1984 version transmitted to you on June 14, 1984. NUREG-0897, Revision 1 (Enclosure 3) is the same document you have seen previously with the addition of two appendices which were provided by Diamond Power Specialty Company (manufacturer of Mirror (TM) insulation) and Owens Corning Fiberglass Corporation (manufacturer of NUKON (TM) fiberglass insulation). These appendices summarize the results of HDR blowdown tests conducted in the Summer of 1984 wherein these insulation materials were tested to obtain information on the damage which might be sustained under simulated LOCA conditions; and these findings have been referenced in the text. RG 1.82, Revision 1 (Enclosure 4) has been modified to reflect CRGR and BWR Owners Group comments received. SRP Section 6.2.2, Revision 4 (Enclosure 5) is the same as previously transmitted. The actions proposed above are Category 2, and do not warrant accelerated review by the CRGR. Therefore, it is requested that the CRGR complete review of these proposed actions within two weeks. ~

August 7, 1985 For further information on these reports, contact Karl Kniel, Chief, Generic Issues Branch (ext. 27359) or A. W. Serkiz, USI Task Manager (ext. 24217), rodI.d rector Office of Nuclear Reactor Regulation

Enclosures:

1) NRR Response to Issues Raised by CRGR in Meeting No. 66

2) NUREG-0869. Rev.18, USI A-43 Regulatory Analysis, June 1985

3) NUREG-0897, Rev. IB, Containment Emergency Sump Performance, June 1985 4 RG 1.82, Rev. 1 5 SRP Section 6.2.2, Rev. 4 6 Draft Generic Letter (see Appendix H of NUREG-0869,

ev. 18) I i L

6/26/85 ENCLOSURE 1 NRR RESPONSE TO ISSUES l RAISED BY CRGR IN MEETING NO. 66 (JULY 11,1984) REF: USI A-43 \ i

NRR RESPONSE TO 15 SUES RAISED BY CRGR IN MEETING NO. 66, JULY 11, 1984 At its meeting on July 11, 1984, CRGR reviewed the proposed resolution of Unresolved Safety Issue A-43. The CRGR noted their opinion that there was considerable conservatism included in the assumptions and analyses used in developing the proposed backfit requirement requesting an analysis of all plants and it declined to support such a position in view of the NRR result of enly moderate value/ impact. CRGR further indicated that it would be willing to reconsider the proposed A-43 resolution (or a modification thereof) after it receives respcnses to the issues raised at the meeting. The paragraphs below identify the issues raised by CRGR and describe the steps taken to resolve then. (1) Initiating Event Probabilities: CRGR: (a) stated that application of the initiating event probability data in Table 1, Appendix B of NUREG-0869, Revision 1 was unclear; (b) questioned the selected break locations from the viewpoint of probability of pipe displacement to result in a double, or single-ended flow model to predict insulation removal; (c) expressed concern that the tabulated data (in Table

1) included all kinds of disruptive failures, not just sudden explosive rupture.

Response to Item (1): The initiating event probabilities shown in Table 1 were derived by M. Taylor (DEDROGR staff) from a paper by Dr. S. H. Bush, " Reliability of Piping in Light Water Reactors," IAEA-SM-218/11, October 1977. USI A-43

                                            ~                                      ' . 5. Q 1

NUREG-0869, Revision 1 has been expanded to further clarify how such probabilities were derived, and to show the applicability of these values and more recent estimates of pipe rupture probabilities (see Appendix C of NUREG-0869, Revision 1B, formerly NUREG-0869, Revision 1). (2) Conservatisms in the Break Jet Model: CRGR believed that considerable undescribed conservatism existed in the model for determining how much insulation would be torn off, the effects of reducing the initiating event probabilities due to break selection, the use of a hemispherical model to predict stripping of insulation on the backside of targeted components, etc. Response to Item (2); The staff has again reviewed the' analysis methodology used in NUREG/CR-3394, "Probabilistic Assessment of Recirculation Sump Blockage Due to Loss of Coolant Accidents," and believes it is appropriate for this application. The pipe break probabilities were derived by M. Taylor (DEDROGR Staff) on the basis of a careful evaluation of the information available. The CRGR's views that stripping off of insulation on the backside of targeted components, the use of a hemispherical jet expansion model and other views that an overall conservative jet destruction model was employed are not supported by evidence of insulation destruction resulting from blowdown experiments conducted at the HOR facility in West Germany. Appendix C of NUREG-0897, Revision 1 was prepared to illustrate the explosive, and destructive, nature of a high pressure break and it's capability to destroy insulation on both the forward , and backside of targeted components, and to transport such debris con-siderable distances from the break location. USI A-43 A Nonetheless, further analyses were performed which evaluated debris from only that insulation installed on the primary coolant system piping out to the 7 L/D parametric envelop previously analyzed; insulation on targeted components was excluded. The results were similar to those obtained in NUREG/CR-3394 and did not significantly alter the conclusions regarding the extent of sump blockage that resulted. These findings have been discussed previously with DEDROGR staff and are incorporated into NUREG-0869, Revision 18, Appendix D, " Estimation of PWR Sump Failure Probability" to assist in understanding the analyses which were performed. (3) Conservatisms Regarding Debris Transport: CRGR stated that there was considerable conservatism in the assumption that all insulation debris torn off would transport to the sump, particularly when the analysis reported in NUREG/CR-2791 reported that for 4 of 5 plants analyzed the insulation debris would not significantly degrade sump operation. Response to Item (3): NUREG/CR-2791 was prepared to develop calculational models for estimating debris generation, transport and blockage effects and was based on the state-of-knowledge in 1982. These calculational meti.ods were applied to five plants and conclusions drawn accordingly. The plant-specific aspects were clearly demonstrated since: (a) Debris transport was calculated to occur in 2 of 5 plants analyzed, although only the Maine Yankee plant was judged to have possible blockage problems due to the small debris screen area available. Salem, the other plant in which debris transport was calculated, has an abnormally large debris screen, and therefore the analysis did not reveal severe blockage problems. US! A-43 (b) Shadowing effects and pipe whip were analyzed and the conclusion was that prompt transport (due to initial blowdown loads) was less significant than long term transport, if post-LOCA recirculation flow velocities were high enough. (c) The types and quantities of insulation employed were important, particularly fibrous type insulations. (d) For some plants (i.e., Sequoyah) direct, or prompt transport would occur since there were no intervening structures (or compartments) between the postulated break locations (i.e. , weld locations in the primary coolant systems piping) and sump location. Break jet loads on the sump screen were not evaluated. Subsequent testing at the Alden Research Laboratory revealed that: (a) Fibrous insulation debrir can transport at very ;ow velocities (i.e. , 0.2 ft/sec), deposit uniformly over debris screens and depending on the insulation material structural characteristics result in high pressure drop blockage (see NUREG/CR-2982, Revision 1, July 1982). (b) Interior metal sheets used in reflective metallic insulation, if niected by the break jet, can transport at low velocities (0.2 - 0.3 ft/sec, and result in blockage (NUREG/CR-3616, January 1984). The calculational models in NUREG/CR-2791 would not have predicted such transport. Thus, reciculation flow velocities (which are plant-specific) are a key parameter for assessing transport of debris. It is again noted that selective conclusions should not be drawn froo NUREG/CR-2791 (which was developed prior to September 1982) in view of the more recent experimental findings. USI A-43 M (4) Conservatisms Regarding Effect of Debris on the Sump and Potential Operator Intervention: CRGR noted two additional conservatisms, namely: (a) that the insulation debris distributed uniformly over the sump screens, (b) that operator intervention could reduce long term flow requirements to 300 to 1000 gpm rather than the 6,000 to 10,000 gpm utilized in the A-43 analyses. Response to Item (4): The assumption of uniform distribution of fibrous insulation debris over the surface of a sump screen has been shown in mineral wool debris transport tests conducted in Finland in 1979 and 1980, and also by fiberglass tests conducted at Alden Research Laboratory in 1983. The free fibers, which transport at near neutral buoyancy, are caught by the screen in a random fashion with a blanket covering being built up over a number of hours (i.e. , 4-6 hours) as the recirculating fibers are entrapped by the screen structure. Since considerable discussion at the July 11, 1984 CRGR meeting was directed at the potential for operator actions te offset a sump blockage situation, we have carefully reviewed such possibilities.* A summary of our findings is as follows:

Memorandum from R. W. Houston to F. Schroeder, " Operator Mitigation of Obstructed Flow From the Sump During ECCS Recirculation," September 17, 1984. Memorandum from W. Russell to F. Schroeder, " Operator Mitigation of Obstructed Flow From the Sump During ECCS Recirculation," September 2,1984.

USI A-43 _ G (a) _ Plant-specific operational procedures addressing the loss of the containment emergency sump due to debris blockage do not currently exist. Although generic emergency guidelines provide some guidance for coping with ECCS degradation, they do not currently address sump blockage per se. The concern about loss of recirculation flow due to a blocked sump has been identified as an open item in the staff SERs for all PWR owner group technical guidelines. Although this subject is repeatedly discussed at various times, neither licensees nor applicants have dealt directly with blockage effects since the current RG 1.82 (with the 50% blockage criteria) does not require addressing this concern based on plant-specific analyses. Rather, the a priori assumption of 50% blockage is made accompanied by unimpeded flow being allowed through the remaining unblocked half of the screen with the calculated pressure drops being very small. (b) Although for some plants instrumentation exists for detecting ECCS degradation (e.g., LPI pump motor current, flow rate meters, inlet and/or exit pressures), plant-specific procedures to identify sump blockage and possible corrective actions have not been developed at this time. ECCS alignments have been integrated with safety logic circuits which assume availability of the sump following transfer from the RWST and realignment options will require reassessing safety logic circuits. This is however, an option a licensee or applicant can elect to utilize to confirm availability of long term recirculation flow capability. (c) Any extended operation at low flows attempted by throttling with valves currently utilized in the RHR pump discharge lines may caused appreciable daniage to the throttled valve, the system piping and piping supports since the valves currently used are not designed for USI A-43 Q throttling beyond their Jesign range. Major potential sources of damage include downstream cavitation, vibration of valve internals and flow instabilities. Cavitation is produced as the valve is throttled and static pressure within the valve decreases to below the liquid vapor pressure while the static pressure downstream of the valve remains above the liquid vapor pressure. This results in the formation and collapse of voids within the valve body. For larger valves such as used in some RHR and low pressure core spray systems, there could also be valve positioner control feedback problems. Currently installed valves are oversized for the low flows considered necessary to keep the core covered (i.e. , <500 gpm) and would probably be operating in nearly a fully closed position (e.g., <5% open) for achieving these low flow rates. Reducing flow to these low values by throttling the valves in the ECCS discharge lines (which are not designed for use in this range) is not desirable in our opinion and is not recommended due to the hydro-dynamic loads which would be imposed. (d) Although various plant designs appear to have alignment and equipment utilization options that could be exercised to offset the loss of sump (or RHR suction intakes), it remains to be demonstrated that proper connections can be made, valves correctly aligned and safety interlocks overridden within a reasonable time under accident conditions. BWR Mark I plants are an illustrative example where realignment of containment sprays to take suction from an alternate source (such as the condensate storage tank) could be used to offset blockage of suction intakes used to feed containment sprays. (5) Conservatisms in Calculated Releases: CRGR cited conservatisms in the WASH-1400 release categories used in the A-43 analyses. USI A-43

                                                                                       - -- ~4@

Response to Item (5): With respect to the CRGR's concerns regarding conservatisms in the WASH-1400 release categories and probability of above ground containment failure, we have reexamined such postulates in terms of more recent containment survivability studies (see NUREG/CR-3653) and other SASA studies and have arrived at the following conclusions: (a) For large dry PWR containments that have safety grade fan coolers, it can be concluded that the containment response assumed in NUREG-0869 is conservative because the probability of containment failure due to overpressure has been overestimated. Large dry containments without safety grade fan coolers and subatmospheric containments have signi-ficant overpressure protection but no alternative way of removing decay heat if the sump is blocked. A similar survival conclusion for BWR Mark I containments could be made if it could be demonstrated that there is a high likelihood of the operator successfully exercising the option to align the containment sprays to take suction from the condensate storage tank (or other external water sources) should loss of containment spray suction occur. (b) For BWR Mark II and Mark III plants, we are not aware of design options which allow the operator to readily switch spray suction to an alternative water source. Although alternative water sources exist, and in some cases the piping appears to be in place, it remains to be demonstrated that the proper connection can be made, that valves can be aligned, and safety interlocks overridden within a reasonable time under accident conditions. Without the sprays, we would conclude that the probability of overpressure (or overtemperature) failure for these containments could be higher than previously calculated in NUREG-0869, for PWRs. USI A-43

                                                                    ,                                   -             _ Ed The 6WR Mark III containment is somewhat of an exception as its principal failure mode leads to significant scrubbing of fission products in the suppression pool. The reduction in conditional consequences that would result is offset somewhat by our conclusion that the probability of early failure for Mark III plants is higher than the NUREG-0869 estimates.        A similar mechanism for source term reductions for Mark I and Mark II containments would result if the operator can successfully relieve pressure by opening wetwell vents.

Ice condenser containments are the most susceptible to overpressurization failure due to their small volume and steam production at the time of estimated reactor vessel failure, or shortly thereafter. The consequences for such a containment failure are estimated to be higher (by perhaps a decade) over PWR dry or subatmospheric containments, should such failure occur due to loss of the containment sump (accompanied by loss of containment sprays). On the other hand, ice condenser plants: (a) employ primarily reflective metallic insulation on the primary system piping and majc'r components, (b) have high NPSH margins and moderate cize debris screens and (c) the majority of plants have low recirculation flow velocities (i.e., <0.2 ft/sec). These plant design features-would result in a much lower . probability of sump blockage and offset the increased consequences so that estimated releases (associated with sump blockage) are on the same order as for subatmospheric containments. The extensive use of reflective metallic insulation also alleviates concerns associated with blockage due to transport of fibrous insulation debris. The considerations stated above, along with allowing credit for over-pressurization margins inherent in current containment designs, have been used to revise the consequence analyses for USI A-43 (see Appendices E and F of NUREG-0869, Revision 1B). i r

USI A-43 - - _ _ - - __ _ , _ - _ . , _ _ . _ , _ _ _ _ _ - . _ - , _ . _ . - _ . _ . - - _ _ - . -

                                                                                      ._d l

(6) Value-Impact Assessment: The CRGR overall consensus was that: (a) questions raised at Meeting No. 66 regarding conservatisms in assumptions and in the A-43 analyses, (b) the moderate level of benefit versus cost impact and (c) the estimated applicability of this issue to only a few plants, all combined to argue against approval of a backfit requiring all licensees to expend resources to demonstrate that their designs are acceptable. Response to Item (6): The staff has reviewed the issues raised in Meeting No. 66 and has undertaken these additional studies and evaluations to address the CRGR's comments. The Regulatory Analysis, NUREG-0869, Revision 1A has been revised to reflect these more recent considerations such as containment survivability due to inherent overpressurization design caoabilities, use of controlled venting for BWRs, alternate means to dissipate decay heat (i.e., SGFCs in PWR large dry contain-ments), recognition of conservatisms imbedded in estimates of debris blockage probabilities leading to core melt, the estimated applicability of this safety issue to a limited number of plants, etc. The result has been to recommend issuance of a Generic Letter to all holders of an operating license or construction permit advising them of potential safety concerns regarding debris blockage and to rely on the plant owners to assess the plant specific significance of this safety issue and to take whatever actions they deem necessary to assure long term ECCS recirculation capability. USI A-43

                                                                         ~ 2_

m... 7/31/85 ENCLOSURE 2 DRAFT" COPY -

                                                         .NUREG-0869 REVISION 18 USI A-43 REGUIATORY AXALYSIS Regulatory Analysis for USI A-43,
          " Containment Emergency Sump Performance"                  .

Proposed Resolution Summary of Public Comments Received and Action Taken CRGR Minutes (Ref. USl A-43) Generic L'etter Draft Completed: June 1985 Date Published: Office of Nuclear Reactor Regulation ' V. S. Nuclear Regulatory Commission Washington, D. C. 20555 l e l

                          **'**~...._.._.

                                                                                 ,M i MM4 CONTENTS P_ age 1  Statement of Problem . . . . . . . . . . . . . . . . . . . . 1 1.1 Summary of Safety Issue . . . . . . . . . . . . . . . .      1

1. 2 Technical Findings. . . . . . . . . . . . . . . . . . . 2 2 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . 4 3 Alternatives . . . . . . . . . . . . . . . . . . . . . . . . 4 4 Consequences . . . . . . . . . . . . . . . . . . . . . . . . 7 4.1 PWR Dry Containments ................. 9 4.2 PWR Subatmospheric Containments . . . . . . . . . . . . 14 4.3 PWR Ice Condenser Plants. . . . . . . . . . . . . . . . 15 4.4 BWRs with Mark I and Mark II Containments . . . . . . . 16 4.5 BWRs with Mark III Containments . . . . . . . . . . . . 18 4.6 Estimated Occupational Exposure . . . . . . . . . . . . 18 4.7 NRC Operations Impact . . ............... 20 4.8 Impact on Other Government Agencies . . . . . . . . . . 21 4.9 Public Impact . . ................... 21 4.10 Other Constraints . . . . . . . . . . . . . . . . . . . 21 5 Decision Rationale . . . . . . . . . . . . . . . . . . . . . 22 5.1 Comparison of Regulatory Alternatives . . . . . . . . . 22 5.2 Rationale for Selecting the Recommended Resolution. . . 24 5.3 Recommended Regulatory Action . . . . . . . . . . . . . 31 6 Plan for Implementation. . . . . . . . . . . . . . . . . . . 31 7 Statutory Considerations . . . . . . . . . . . . . . . . . . 32 7.1 NRC Authority . . . . . . . . . . . . . . . . . . . . . 32 7.2 Need for NEPA Statement . ............... 33 8 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . 33 111

                                                             . , l l

I CONTENTS (continued) . Appendix A Summary of Public Comments Received and Action Taken Appendix B CRGR Meeting Minutes (Reference USI A-43) Appendix C Estimation of Pipe Failure Probability Appendix D Estimation of PWR Sump Failure Probability Appendix E Consequences of Loss of Recirculation Capability Appendix F Containment Survivability Appendix G Estimation of Costs to Replace Insulation Appendix H Draft Generic Letter 4 iv

                                                                   -. m a O LIST OF TABLES 4.1 Calculated Value-Impact Ratios for USI A-43 4.2 Assessment of Sump Blockage Probability                12 4.3 Radiation exposure to workers during replacement (in person rems)                                      19 5.1 Summary of Calculated Values and Impacts Associated with Various Plant-Containment Designs for Resolution of USI A-43                                           25 V

                                                                                     . _ wm REGULATORY ANALYSIS FOR USI A-43, CONTAINMENT EMERGENCY SUMP PERFORMANCE 1     STATEMENT OF PROBLEM 1.1. Summmary of Safety Issue Unresolved Safety Issue (USI) A-43 deals with a concern for the availability of adequate recirculation cooling water following a loss-of-coolant accident (LOCA) when long-term recirculation cooling from the PWR containment sump or the BWR RHR suction intake, must be initiated and maintained to prevent core melt. These safety concerns can be summarized as follows:

In the recirculation mode, will the sump design (for pressurized water reactors, PWRs) or the residual heat removal (RHR) suction intakes (for boiling Water reactors, BWRs) provide sufficient water to the RHR and containment spray system (CSS) pumps, and will this water be sufficiently free of LOCA generated debris and potential air ingestion so that pump performance is not impaired to the point of seriously degrading long-term recirculation flow capability. The USI A-43 safety concerns can be separated into three parts: l (1) sump (or suction intake) hydraulic pe.formance under post-LOCA adverse l conditions due to such effects as potential air ingestion and elevated j temperatures and break flow (2) LOCA generated debris resulting from a pipe break jet that destroys l large quantities of insulation, with this insulation debris then being l transported to the sump debris screen (s), and the resulting sump screen I i, 1 i i NUREG-0869, Revision IB t

(or suction strainer) blockage reducing NPSH margin below that required for the recirculation pumps to maintain long-term cooling (3) the performance capability of RHR and containment spray system (CSS) pumps to continue pumping when subjected to possible air ingestion, debria ingestion, and other effects such as particulate ingestion on pump seal and bearing systems Although USI A-43 was derived principally from PWR containment emergency sump performance considerations, the debris blockage concern applies to both PWRs and BWRs. The RHR suction strainers in a BWR are analogous to the PWR sump debris screen, and adequate recirculation cooling capacity is necessary for both BWRs and PWRs to prevent core melt.

1. 2 Technical Findinas Safety concerns have been investigated on a generic basis, and the staff's technical findings are reported in NUREG-0897, Revision 1. These findings can be summarized as follows:

! (1) Measurements derived from extensive, full-scale sump hydraulic tests have generally shown that low levels of air ingestion (less than 1% to i 2%) will occur and have also demonstrated that vortex observations alone cannot be used to quantify levels of air ingestion (as has been done in the past). These test results have been used to develop PWR sump and BWR suction intake hydraulic design guidelines for minimizing, or eliminating, air ingestion and have eliminated the need for plant-specific sump tests or model tests. (2) Plant insulation surveys, development of methods for estimating debris generation and transport, debris transport experiments, and information 4 i NUREG-0869, Revision IB 1 I provided as public comments have shown that debris blockage effects are dependent on the types and quantities of insulation employed, the

!                             primary system layout within containment, post-LOCA recirculation i

patterns and velocities, and the post-LOCA recirculation flow rates. It

is concluded that a single generic solution is not possible, but rather

;                              that debris blockage effects are governed by plant-specific design i                               features and post-LOCA recirculation flow requirements, i

i The results also show that the current 50% screen blockage assumption identified.in Regulatory Guide (RG) 1.82, " Sumps for Emergency Core ,

Cooling and Containment Spray Systems," should be replaced with a more l comprehensive requirement to assess debris effects on a plant-specific basis. The 50% screen blockage assumption does not require a plant-

{ specific evaluation of the debris-blockage potential and usually will l result in a non-conservative analysis for screen blockage effects. - i j -(3) Reviews of available data on pump air ingestion effects and discussions j with the U.S. manufacturers of RHR and CSS pumps show that low levels of ,

l air ingestion (2% or less) will not significantly degrade pump ! performance, and that the types of pumps employed in nuclear plants will J tolerate ingestion of insulation debris and other types of post-LOCA particulates that can pass through PWR sump screens or BWR suction i strainers. I i f i In summary, these findings show a concern for safety that is significantly l 1ess with regard to the effect of vortex formation and air ingestion than l that previously hypothesized, and that the potential loss of recirculation ! cooling capability, as a result of LOCA debris generation transport and screen blockage is potentially more significant. i i i i i > i 4 ! NUREG-0869, Revision 1B ! 1

                                                                                  , o The above findings were issued in NUREG-0897 For Comment in May 1983.

Information received during the public comment period on USI A-43 included technical data, plant-specific data, cost information, and viewpoints on the cost impact on industry. Appendix A summarizes the comments received and how the staff addressed them. Those comments that were principally technical in nature were incorporated into NUREG-0897, Revision 1. Plant specific design information and retrofit cost data received during this period were utilized in preparing this report. 2 OBJECTIVES The general objective of the proposed regulatory actions discussed below is to provide assurance that the safety concerns associated with USI A-43 will be adequately addressed in the licensing process. The goal is to meet the requirements of GDC 35, " Emergency Core Cooling and GDC 38, " Containment Heat Removal." The technical findings obtained as the result of the work performed for resolving USI A-43 have established a need to revise current licensing guidance on these mat +ars. Therefore, the staff's technical findings (NUREG-0897, Revision 1A) have been used to revise Regulatory Guide 1.82, Revision 0, " Sumps for Emergency Core Cooling and Containment Spray Systems" and the standard Review Plan Section 6.2.2, " Containment Heat Removal Systems." The issuance of these revised regulatory documents would clarify staff review practices in light of current technical findings. The issuance and need for implementation of the revised regulatory guide and standard teview plan section are discussed in Section 3, Alternatives, below. 3 ALTERNATIVES The following approaches have been considered as alternatives to resolve USI A-43. NUREG-0869, Revision IB

                                                                                   .1 (1) Issue RG 1.82, Rev. 1 and SRP Section 6.2.2, Rev. 4 and require all licensees and applicants to evaluate debris blockage potential effects (per RG 1.82, Revision 1) for confirmation of adequate NPSH margin.

(2) Issue RG 1.82, Revision 1 and SRP Section 6.2.2, Revision 4 and require an evaluation of debris blockage potential effects (per RG 1.82, Revision 1) for confirmation of adequate NPSH margin from only those licensees and applicants where loss of recirculation could lead to core melt and loss of containment integrity. The risk from offsite radio-active exposure depends both on the probability of core melt from loss of recirculation and the effectiveness of the containment in mitigating the offsite exposure. Plants for which it can be determined that containment integrity is preserved although core melt from loss of recirculation might occur, could be exempted from a requirement to evaluate debris blockage effects because of the low public risk. This alternative then involves making such a determination based on value-impact analyses for different types of plant containment (e.g., PWR dry containments, ice condenser plants, Mark Is and IIs and Mark IIIs) to determine if the benefits outweigh the cost impacts. This latter approach is used to assess both options (1) and (2) and is discussed in Section 4. (3) Issue a generic letter for information only to all licensees and applicants describing the potential safety concerns associated with insulation debris blockage of the sump screen which could lead to the loss of NPSH margin during recirculation. Make clear that the 50% blockage guidance provided in RG 1.82, Revision 0 is not a necesearily conservative way to perform an assessment of debris blockage, particularly if fibrous insulation is employed on the primary system piping and components. Provide (with this generic letter) NUREG-0897 Revision 1A, which contains the staff's technical findings, for NUREG-0869, Revision 18 _

information. State clearly that there is no requirement for analysis or modification for any oparating plant or plant now under construction. Issue RG 1.82, Revision 1 and SRP 6.2.2 Revision 4 for implementation on Standard Plants and new Construction Permit applications only, to be effective six months from date of issuance. In summary, under this alternative the significant safety information derived under the A-43 program would be provided to all licensees and applicants but there would be no requirement for any action from operating plants or plants now under construction. Standard Plants or new Construction Permit applications would be required to address RG 1.82, Revision 1 and SRP 6.2.2 effective six months from date of issuance. '(4) Make no revision to RG 1.82 or SRP 6.2.2; publish NUREG-0897, Revision 1A (the staff's technical findings for USI A-43) as an information only document. The need for the proposed actions can be summarized as follows: (1) Issuance of NUREG-0897, Revision 1 (the staff's technical findings) will provide a comprehensive description of the technical issues, along with an extensive data base for designing and assessing PWR sump designs and BWR RHR suction inlets. These findings (which show that vortex observations do not quantify air ingestion) will replace prior assumptions leading to previously required inplant tests (or model tests) and provide a common technical data base for licensees, applicants, and the staff to use, thereby reducing the regulatory burden in future assessments. (2) Revising RG 1.82 will bring that RG into conformance with more than 3 years of experiments and generic studies related to sump design and performance, and will remove the 50% blockage criterion which is an NUREG-0869, Revision IB _ ._m arbitrary assumption and not necessarily conservative from the viewpoint of sump debris blockage effects. (3) Revising SRP Section 6.2.2 will make the review considerations consistent with the USI A-43 technical findings and with RG 1.82, Revision 1.

4. CONSEQUENCES The purpose of this section is to assess the consequences (i.e., values versus impacts) of selecting each of the alternatives set forth in section 3.

Alternatives 1 and 2 which would require all (Alternative 1), or selective plants (Alternative 2) to perform analyses for debris blockage to determine if loss of recirculation capability might occur, and to undertake necessary plant modifications to reduce such potential risks are the subject of this section. Alternative 3, issuance of a generic information letter and implementation of revised licensing criteria in future reviews would not impact current licensees or applicants. The impact (cost) of including the revised licensing criteria in new designs is considered to be very small and the resulting value/ impact would be favorable and therefore no detailed quantitative analysis is deemed to be necessary. Alternative 4, do nothing, does not involve any impact. Because of the difficulty of treating all reactors as a homogencus group, both because of their design differences with respect to the sump and type of insulation used as well as the differences in containment capability subsequent to a blocked sump, the averted risk and value/ impact analyses have been developed for each of the five major types of plant containments. They are PWRs with large dry containments, PWRs with subatmospheric containments, PWRs with ice condensers, BWRs with Mark I and II containments and BWRs with Mark III containments. For many plants the staff expects that evaluations (per RG 1.82, Rev. 1) would reveal that adequate NPSH would exist despite the potential for debris blockage. For others a plant-specific value/ impact analysis would not support imposing backfit modifications. Without NUREG-0869, Revision 18 k performing the individual plant assessments, the staff cannot determine the number of plants in these two categories. Therefore, rather than attempting a value/ impact assessment for the total plant population, we have developed in the following sections estimates of releases and V/I's for a plant and associated containment type assuming that a significant probability of sump blockage exists. Thus for each such class we consider: (a) the potential reduction in core melt frequency, (b) the potential reduction in oublic risk if backfit modifications are required (i.e. , estimated averted releases in person-rem in the remaining plant life time), (c) the costs of such backfit modifications, and (d) the resulting value/ impact ratio of such modifications. Blockage of the containment emergency sumps (for PWRs), or of the RHR suction intakes (for BWRs) during the recirt.ulation phase following a LOCA, can lead to core melt and containment overpressurization unless alternate water sources can be made available to substitute for loss of recirculation flow capability. Core melt accompanied by containment failure can lead to radioactivity release and public radiation exposure. These consequences are dependent on: (a) pipe failure probability (i.e., LOCA probability) (b) sump blockage probability leading to loss of NPSH margin which then leads to loss of recirculation capability (c) containment overpressure structural survival LOCA probability is the controlling factor since this is the initiating event that requires recirculation and can lead to loss of recirculation. If pipe failure probabilities are judged to be extremely low due to such considerations as leak-before-break, etc., then these calculations would result in very low releases and backfits would not be supportable on the basis of value/ impact criteria. NUREG-0869, Revision 18

                                                                                  .. A Sump blockage probability is dependent on the amount and type of debris which could be generated by a LOCA, which is a function of break size and location, containment layout, sump location and design (i.e., size of the debris screens), recirculation flow requirements and available NPSH margins. The estication of sump blockage probabilities becomes highly plant-specific due to large differences in these plant design features and type of insulation employed. Arriving at a singular, or generic value is therefore not possible due to these individual plant differences.

Containment structural survival and the maintenance of containment integrity is also a key factor in determining consequences. Although core melt can be postulated as a result of loss of recirculation flow, containment integrity (or the ability to withstand an overpressure transient) would significantly limit the levels of radioactive release. Recent studies dealing with containment structural capabilities have shown that overpressure design margins do exist, and therefore containment integrity can be maintained through either inherent structural overpressure design margins, alternate containment cooling means, or controlled venting (as is being adopted by BWRs). Again, the containment design variability (e.g., PWR dry containment versus PWR ice condensers design, BWR Mark I versus Mark III design) preclude a singular conclusion to be drawn. Table 4.1 summarizes the number of plants with different types of containment. 4.1 PWR Dry Containments The PWR dry containment design concept is used extensively in U. S. nuclear power plants. There are currently 71 PWR dry containment plants licensed or in final license review (out of approximately 125 plants docketed). 57 of these 71 plants also utilize safety grade fan coolers (SGFC) to assist in the control of post-LOCA containment pressures and temperatures. NUREG-0869, Revision 18

                                                                           ,                                                             . a The pipe break probability is the first key factor to consider in evaluating sump blockage probability since the LOCA is the debris generator. A range of estimated pipe failu're probabilities was developed,for USI A-43 and these are discussed in Appendix C. Theestimatedvalueswerd3x10-6 /Rx yr for large pipes (> 28 inches) to 3 x 10~4/Rx yr for small pipes (2 to 6 inches).
                                          .                                                                                        These estimated pipe break frequencies do not include more recent leak-before-break i       considerations. Table 4.2 outlines the other factors needed to develop an estimated sump blockage probability many of which are plant-specific.

Appendix 0 provides a more detailed discussion and examples of the calculations which were used to arrive at sump blockage probabilities. Salem-1 was utilized'as a reference PWR to model piping layouts, weld j locations, break locations, insulation distribution and major insulated plant I components. Due to the variability of plant and sump designs, and l operational requirements, the following design specifications were analyzed ! parametrically: ) s Recirculation Flows = 6,000 to 10,0t0 gpm Available Debris Screen Area = 50 to 200 ft2 l Available NPSH Margin (w/o blockage) = 1 to 5 ft Hg0 These design ranges are representative of 19 PWR designs for which we have detailed information. The calculational methods and results are reported in NUREG/CR-3394. l The principal purpose of calculating sump failure probability is to estimate core melt frequency. For the above range of parameters the estimated PWR sump failure probabilities (see Appendix D) ranged from 3 x 10 -6 to 5 x

         -5 10 /Rx yr. Assuming that sump failure leads to core melt (which is the l

i result if other actions cannot be or are i not taken), core melt frequencies NUREG-0869, Revision 18 L

m. Table 4.1, Types of Nuclear Plant Containments and License Review Status Type of Containment Number of Plants OL Issued (1) PWR Dry w/SGFC(2) 57 51 PWR Dry w/o SGFC(2) 14 PWR Subatmospheric 7 5 PWR Ice Condenser 10 7 PWR TOTAL 88 63 BWR Mark I 23 21 BWR Mark II 10 7 i BWR Mark III 8 1 ! BWR TOTAL 41 21 i INDUSTRY TOTAL 129 92 l i (1) Estimate is based on OL issuance status in December 1984. (2)SGFC is an abbreviation for safety grade fan coolers. i NUREG-0869, Revision 18 Table 4.2, Assessment of Sump Blockage Probability Event Technical Consideration (s) Safety Implication Pipe Break - Break Probability If break probability is

                                                                - Break Size                       extremely low, debris blockage
                                                                - Breck Type                       potential is negligble.

Debris - Break Size & Location Small Pipes (< 10" diam) Generated - Target (s) Location generate small amounts of

                                                                - Type (s) of Insulation          debris and therefore debris
                                                                - Extent of Jet Damage (L/0)      blockage effects produced by small pipes are not significant.

Transport - Break Location If U < 0.2 ft/sec, transport of Debris - Plant Layout & Sump Location not likely to occur; therefore

                                                               - Type of Debris                   blockage would not occur.
                                                               - Recirculation Velocities (U)

Debris - Amount Transported Fibrous insulation debris Screen - Available Screen Area transports and coats total Blockage - Type of Blockage screen area. If debris screen areas are large; pressure drop is minimized. NPSH Impact - Type of Blockage If AHB > NPSHA, loss of

                                                               - Type of Debris                  recirculation can occur.
                                                               - Recir Flow Required
                                                               - Blockage Head Loss (AH )

B NUREG-0869, Revision 18 t

a from a blocked sump would be 3 x 10 -6 to 5 x 10-5/Rx yr. Two significant points to be noted are that: (a) these sump failure probabilities were based on the assumption that all fibrous debris was transported to the sump and lead to blockage (this would not be the case for PWR's which have recirculation velocities less than 0.2 ft/sec, which many large dry containments are known to have) and (b) no credit was given for detection of blockage buildup and operator corrective actions in these sump blockage estimates. For purposes of estimating core melt frequency resulting from loss of NPSH due to sump blockage, the assumption was made that in 50% of the cases this would lead to core melt. For the remaining 50% of the cases it was assumed that the operator would detect the onset of blockage and could take action to maintain recirculation flow. Thus the estimated core melt frequency is 1.5 x 10 -6 to 2.5 x 10-5/Rx yr for this type of plant. The second approach to assessment of the safety significance of this USI is based on estimating the public radioactive exposure consequences (or potential releases) associated with sump failure. Appendix E discusses consequences associated with loss of recirculation capability. The estimated conditional consequences from core melt for PWRs with large dry containments without safety grade fan ecolers is estimated to be 5 x 105 person-rem. Using the estimated core melt frequency of 1.5 to 25 x 10-6 , and an outstanding reactor life span of 25 years, results in an estimated averted risk range of 20 to -300 person-rem /Rx. These are low levels of averted risk and indicate the safety issue to be of moderate to low signi-ficance. 1 NUREG-0869, Revision 18 The value/ impact ratio (V/I) range which can be calculated for PWRs with large dry containments (based on estimated cost for corrective action of

  $1.5M/Rx for replacement of insulation) is 10 to 200 person-rem /$million. If less severe retrofit actions are required (based on an estimated cost of
  $0.4M/Rx), the value/ impact range is 50 to 800 person-rem /$ million. Plant retrofit cost estimates are discussed in Appendix G, and are based on a composite average of industry estimates received during the public comment period. The V/I is in all cases less than the criterion of 1000 person rem /$

million. PWR dry containments with SGFCs have an additional safety system capable of rejecting post-LOCA containment heat loads. SGFCs are designed to operate independently in the post-LOCA environment and would therefore not be directly affected by loss of the sump or containment sprays. This independ-ent heat rejection capability will assure that containment overpressure failure will not occur. Thus, although core melt could still be postulated, prevention of loss of containment integrity (by the SGFCs) will assure that the radionuclides are contained and the public risk is estimated to be very low (see Appendix E). 4.2 PWR Subatmospheric Containment The number of PWR subatmospheric containment plants is only 7, of which 5 have received an OL at this time. Most of the considerations and discussions presented in the preceeding section for PWR large dry containments apply to i this plant class, except that subattospheric containments do not have safety grade fan coolers. i l l NUREG-0869, Revision IB _ a.Il The estimated range of core melt frequency resulting from sump blockage discussed in Appendix D applies and is in the range from 1.5 to 10 -6 to

               -5 2.5 x 10 /Rx yr. As noted before, these numbers were derived assuming fibrous debris would transport to the sump and that operator corrective action to maintain recirculation would be obtained in 50% of the cases.

The estimated averted risk (for those plants where sump design problems may exist) is 20-310 person-rem /Rx. The value-impact ratio is the same as for PWR large dry containments, without safety grade fan coolers, namely, 10 to 20 person-rem /$M (at $1.5M/Rx) or 50 to 800 person-rem /$M (at $0.4M/Rx). 4.3 PWR Ice Condenser Plants Ice condenser plants are judged to be most prone to eventual overpressure failure if loss of recirculation occurs (see Appendix E). Although hydrogen igniters would protect against deflagration effects, containment failure due to steam production could occur at time of vessel failure or within a few hours thereafter. Therefore the consequences of sump blockage in an ice condenser plant are determined to be higher than for PWR dry and subatmospheric containment designs. There are ten ice condenser plants constructed as five twin units by three different utilities. As a consequence, there is considerable design similarity among these units including the details of insulation, sump design, and interior layout. Information on these design features was ! specifically obtained by the staff and is discussed further in Appendix E; however, the major points of similarity are summarized as follows: l l (a) All ice condenser plants utilize reflective metallic insulation (RMI) on primary system piping and major components. Sump blockage effects i associated with RMI are less severe than from fibrous insulation debris. l l l NUREG-0869, Revision IB

                                                                                .. m s    .

(b) The majority of these plants have approach velocities in the vicinity of the sump of less than 0.2 ft/sec. Therefore debris transport and blockage are not likely. (c) NPSH margins for the majority of these plants are in excess of 5 ft H 0, 2 whereas 1-5 ft H2O blockage loss criteria was employed for deriving PWR sump failure probabilities. The net effect is that the sump failure probabilities employed for the PWR dry and subatmospheric containments should be reduced when applied to the ice condenser plants. It is estimated that sump blockage probability leading to loss of NPSH could be reduced to 1 to 9 x 10 -6 /Rx yr, or lower. The

                                                   -6 estimated core melt frequency is 0.5 to 4.5 x 10 /Rx yr.

The averted release would then be based on an estimated consequence value of 6 5 x 10 person-rem, and for a 25 year plant life would be 60 to 560 person-rem /Rx. This estimate is in the same range as other PWRs dicussed previously. The estimated V/I ratio for ice condenser plants is 160 to 1400 person-rem /$M (at an estimated cost = $0.4M/Rx) and is 25 to 380 person-rem /$M (at estimated cost = $1.5M/Rx). 4.4 BWRs with Mark I and Mark II Containments The potential blockage of BWR RHR suction intakes is similar to that estimated for PWRs, particularly since BWRs are reinsulating with fiberglass l and newer BWRs are going on line with fiberglass insulation being used on the primary pressure boundary piping. BWR intakes have suction strainer areas of typically 50-150 f t2 (on the lower side of sump debris screen areas in PWRs) and somewhat higher suction flows (i.e., 8,000 to 12,000 gpm/ train). On the NUREG-0869, Revision 1B

                                                                                       -.2.a 4 '

other hand, suppression pool velocities are generally low (<0.2 ft/sec for bulk pool velocity) and the drywell versus wetwell design and separation will tend to inhibit insulation debris transport. Therefore, although a detailed BWR RHR intake blockage probability analysis (similar to that reported for PWRs in Appendix D) has not been undertaken, it is qualitatively estimated that BWR intake blockages will be somewhat lower and therefore an estimated core melt frequency (equivalent to intake blockage prooability) of 2 to 10 x 10-6/Rx yr was used in the calculations which follow. Such a reduction is supported by recent analyses received for Limerick 1 (using the proposed RG 1.82, Revision 1 analysis guidelines) which showed adequate NPSH margins (due < principally to high NPSH availability due to plant layout features). The estimated conditional consequence associated with core melt for Mark I and Mark II type of containments (see Appendix E) is 5 x 10 6person rem. The use of an outstanding reactor life of 25 years and the estimated RHR intake blockage probabilities noted above, results in averted releases of :*0 to 1250 person-rem /Rx; this value applies only to those plants where blockage leading to loss of NPSH has been determined. The estimated value/ impact ratios for Mark I and Mark II containments (with assumed containment failure) are therefore approximately 630 to 3,100 person-rem /$M (at an estimated cost of $0.4M/Rx) and 170 to 830 person-rem /$M (at an estimated cost of $1.5 M/Rx). If the estimated conditional consequence is analyzed to give credit for containment spray recovery and containment venting (which would assure containment integrity despite loss of recirculation due to blockage) the consequences are reduced by a decade to 5 x 105 person-rem (see Appendix E). The estimated averted releases would then be reduced to 25 to 125 person rem /Rx. NUREG-0869, Revision IB b

The estimated V-I ratios would reduce accordingly to 63 to 300 person rem /$M (at an estimated cost = $0.4 M/Rx) and 17 to 80 person-rem /$M (at an estimated cost = $1.5M/Rx). 4.5 BWRs with Mark III Containments Blockage considerations for Mark III containment are similar to those discussed in Section 4.4, and could be perhaps argued to be somewhat lower due to the wetwell versus drywell structural design. For these calculational purposes an estimated blockage probability of 4 to 20 x 10-6/Rx yr was utilized. Estimated core melt frequency is 2 to 10 x 10 -6 /Rx yr. The estimated conditional consequence for Mark IIIs is 5 x 105 person-rem (see Appendix E). This reduction (relative to Mark Is and IIs) results from the fact that fission products would bubble through a subcooled suppression pool, thereby yielding a significantly reduced source term. Therefore, the estimated averted releases for Mark IIIs are 25 to 125 person-rem /Rx (assuming again an outstanding reactor life of 25 years). The calculated V/I ratio are 60 to 310 person-rem /$M (at estimated cost =

       $0.4 M/Rx) and 17 to 80 person-rem /$m (at estimated cost = $1.5 M/Rx).

l ! 4.6 Estimated Occupational Exposure

Estimates of in plant radiological exposures associated with insulation replacement can be derived from actual experience during the steam generator repair and replacement at the Surry and Turkey Point plants. Table 4.3 shows the work categorics applicable to insulation replacement (as reported in NUREG/CR-3540) and the attendant exposure.

l l t NUREG-0869, Revision 1B . - .

r - ... A Table 4.3 Radiation Exposure to Workers During Insulation Replacement In (person-rems) Work Surry 2 Surry 1 Turkey Pt 3 Turkey Pt 4 Installing 46.5 40.9 9.95 34.19 scaffolding Removing 15.16 19.35 70.80 63.64 insulation Reinstalling insulation 57.80 6.30 85.72 4.17 Total 119.46 65.55 166.47 102.00 The benefit of preplanning (or learning from the first plant) is evident, as is the variation between plants. It should also be noted that the scaffolding installed at these plants was designed to remove steam generators, and that entire steam generators were stripped of insulation and reinsulated. Thus, the average of these exposures, 115 person-rem, is considered to be high with respect to insulation replacement needed to resolve A-43. Discussions with Surry site personnel during the For Comment period indicated that a 50 person-rem exposure level for insulation replacement is realistic if the job is thoroughly preplanned. On this basis the occupational exposure for major insulation replacement is estimated to be 50 person-rem per plant. Exposures associated with alternate actions (such as increasing debris screen size) would be less. t i NUREG-0869, Revision 18 --- -. -. --

Utilizing the estimated averted risks developed in Appendix E, and _ replacement exposures noted in Table 4.3, the following net retrofit radiological effects can be developed: Plant Backfit Estimated Averted Net Averted Type Exposure Risk Exposure (person-rem /Rx) (person-rem)/Rx (person-rem /Rx) PWR Ice 50 40 to 560 -10 to 510 Condenser PWR Dry w/o 50 20 to 300 -30 to 250 SGFCs and Subatmospheric Mark I and II 50 250 to 1250 200 to 1200 Mark III 50 25 to 125 -25 to 75 It is estimated that only a small number of plants would be represented by the higher values of net averted exposure (i.e., less than 10 plants). 4.7 NRC Operations Impact With respect to NRC staff review time, the impact of the proposed actions will be minimal. The guidelines in Appendix A of the revised RG 1.82 and in NUREG-0897, Revision 1, (and its supporting references) provide the technical l information and specific guidance that is needed by the staff reviewer to perform an evaluation in a reasonable time. It is estimated that about 2 weeks of staff and related licensing review time per plant will be needed (estimated cost = $5000 per plant) for review of analyses submitted. Assuming 5 to 20 such detailed responses, the estimated staff cost would be

 $25K to $100K.

NUREG-0869, Revision 18 _ _ _ . _ - - - . ,

The experimental data and generic plant information and calculations reported in NUREG-0897, Revision 1 (and supporting references) represent an investment of nearly $3.0 million by NRC and the Department of Energy. This information is of value to both the NRC and industry. In addition, this extensive hydraulic performance data base provides the basis for eliminating unnet.essary in plant testing or sump model tests designed to examine vortex formation which has been previously required for most plants at the OL stage. 4.8 Impact on Other Government Agencies Because nuclear plant design review and acceptance are done solely by the NRC staff, no impact on other government agencies is projected. 4.9 Public Impact If the recommendations herein are adopted, therew'ould not be an impact to the public. Rather, there would be a value to the public which would be added reassurance that adequate sump designs exist for assuring operability in the recirculation mode following a postulated LOCA. As discussed previously, only a small number of plants is expect 9d to be susceptable to LOCA generated debris. Issuance of a generic letter and the staff's technical findings to licensees and OL applicants outlining potential safety I concerns associated with insulation debris would add to the

       " defense-in-depth" concept, which ensures that the health and safety of the public is being maintained. Since insulation is periodically being replaced in operating plants, these findings could be used in the selection process for insulation selection and replacement.

l - t l 4,10 Other Constraints The Commission has proposed to amend its requirements governing the i backfitting of commercial power reactors and certain licensed nuclear l NUREG-0869, Revision IB l

facilities (see NRC Notice 84-137, dated November 30, 1984). This regulatory analysis is consistent with the factors to be considered that are noted in that release and the intent of the proposed A-43 resolution is to maintain consistency with these guidelines with respect to any backfit considerations.

5. DECISION RATIONALE The regulatory alternatives pertinent to the resolution of USl A-43 were identified in Section 3. The consequences in terms of values and impacts were discussed in Section 4, in terms of the variabilities associated with different plant designs. This section presents a comparison of regulatory alternatives incluaing consideration of the value impact analysis results for concluding this safety issue.

5.1 Comparison of Regulatory Alternatives The regulatory alternatives can be compared as follows: (1) An evaluation of adequate NPSH margin from all licensees and applicants would give the NRC a determination of which plants are most susceptible to debris blockage problems, and would identify plants that would benefit from corrective action. This option would however result in the

~~

greatest overall industry cost impact since it is expected that only a small number of plants (i.e. , 5-10 p1 ants) would identify a debris blockage potential requiring retrofit action. On the other hand, all licensees and applicants will incur analysis costs. (2) Requesting evaluations from only the licensees for those plants that are judged to have a high probability of containment failura focuses this safety issue on minimizing (or averting) the public risk of radiation , exposure as the result of loss-of-recirculation capability. This option NUREG-0869, Revision 1B ,

reduces the industry impact and would concentrate the analysis effort on those plants subject to containment failure that would benefit from corrective action. Identification of only those plants which have a high probability of debris blockage requires a plant specific analysis due to the plant-to plant variabilities discussed in Section 4. A generic conclusion (or identification) is not possible. Therefore, this option is similar to Option (1) but involves a fewer number of plants; both options are discussed as a singular option in the decisionale rationale presented below. (3) The forward fit of RG 1.82, Revision 1 to standard plants and new cps will address sump design and debris blockage effects in new plants at a very low incremental cost. Continued ues of RG 1.82, Revision 0 (which has the 50% blockage criterion) does not adequately address this issue, and is technically inconsistent with technical findings developed for the resolution of USI A-43. This option includes the issuance of a generic letter for information providing the industry with technical findings which would be useful for industry safety assessments regarding routine changeout of insulation materials. (4) Issuance of NUREG-0897 only, without using current technical findings to revise RG 1.82 and SRP Section 6.2.2, runs contrary to addressing safety concerns using current and the most reliable technical findings and a need to remove the current 50% blockage criterion in RG 1.82, Revision 0. 5.2 Rationale for Selecting the Recommended Resolution NUREG-0869, Revision IB Options 1 and 2 The rationale for selecting a resolution position is based in part on the values and impacts discussed in Section 4. Options 1 and 2 are discussed together since Option 2 is a modification of Option 1. Section 4 presents the consequences for these options considering plants grouped in separate containment design categories. Conclusions are drawn based on core melt frequency contribution, potentially averted releases (AR) and value/ impact (V/I) ratic utilizing the 1,000 person-rem /5M criteria for backfit consideration. Table 5.1 provides a summary of values and impacts developed in Section 4. PWR Dry Containments: These type of plants fall into two categories, namely with safety grade fan coolers (SGFCs) and without safety grade fan coolers. PWR dry containments with SGFCs have an additional capability to reject post-LOCA decay heat and prevent containment overpressurization, thereby assuring containment integrity. Furthermore, large dry containments are least susceptable to abrupt failures due to hydrogen burns, steam spikes, etc. Therefore, it is the staff's opinion that even if loss of sump should occur, postulated core melt effects would be contained and public releases maintained with current design values. Thus, backfit action (including analyses) for PWR dry containments with SGFCs is not justifiable and Option (1) which includes such a requirement for PWR dry containments is not attractive. NUREG-0869, Revision 1B _ __ _. -. ._ -. -

_ Table 5.1, Summary of Calculated Values and Impacts Associated with Various Plant-Containment Designs for Resolution of USI A-43 Estimated Core Calculated CalculatedVge/ Type Melt Probability (1) Risk Averted, AR Impact Ratio Containment (1/Rx yr) (person-rem /Rx) (person-rem /$MJ PWR Dry w/o 1.5 to 25 x 10 -6 19 to 310 48 to 780 SGFCs and 12 to 210 Subatmospheric PWR Dry 1.5 to 25 x 10 -6 _________ __________ w/SGFCs PWR Ice Condenser (2) -6 60 to 560 0.5 to 4.5 x 10 160 to 1400 25 to 380 PWR Dry w/o 1.5 to 25 x 10 -6 2 to 31 5 to 33 SGFCs and 1 to 21 Subatmospheric w/ Spray Recovery Mark I and I'I 2 to 10 x 10 -6 250 to 1250 630 to 3100 170 to 830 Mark III 2 x 10 -6 25 to 125 60 to 310 17 to 80 Mark I and II 2 to 10 -6 25 to 125 60 to 310 w/ Venting and 17 to 83 Spray Recovery (1)The estimated core melt frequency is based on the conditional consequences discussed in Appendix E, the sump blockage frequency estimates discussed , in Appendix 0, and the assumption that 50% of the time that blockage l occurs leads to loss of NPSH and core melt follows. This assignment of a conditional core melt probability of 0.5 is felt to be realistic from the viewpoint of potential detection of flow degradation, potential operator followup action to correct this situation and sump design variability. (2)A separate estimate of sump blockage probability was made for the ice condenser plants (see Appendix E) which takes into account their spe"ific (3) design features. The value/ impact ratios have been calculated for an estimated cost of

                    $0.4M/Rx (this assumes retrofit costs would be minimal) and for an i                    estimated cost of $1.5M/Rx (this cost assumes replacement of troublesome I                     insulation (s)); see Appendix G for a discussion of estimated costs.

NUREG-0869, Revision IB - - _ . . _ . _. . - ,_ _ __ __ -.

PWR dry containments (w/o SGFCs) have been evaluated for consequences (See Section 4.1) and the results are as follows: (a) PRA studies reported in section 4 have concluded that the core melt probability from this class of plants for the sequence involving a blocked sump is in the range 1.5 x 10-6 to 2.5 x 10 -5 For a number of reasons previously discussed we believe the best estimate of core melt frequency for this sequence is at the lower end of this range, about 3 x 10 -6 . This level of core melt frequency does not support a backfit requirement. (b) The calculated range of averted risk (assuming containment failure as per WASH-1400) is approximately 20 to 300 person rem /Rx. This is of a low-to-moderate level value. If corrective operator action to restore containment sprays (should debris blockage be encountered) is affected before containment failure, then the estimated averted risk is 2 to 30 person-rem /Rx. (c) The calculated value/ impact ratios were: 50 to 800 person-rem /$M (low retrofit cost) 10 to 200 person-rem /$M (high retrofit cost) Utilization of a 1,000 person rem /$M criteria does not support a backfit requirement. (d) If operator detection of onset of blockage and taking corrective actions is introduced into the above consecuences, the estimated releases noted above are reduced by a decade and the V/: ratios would also be reduced by a decade. t l I l NUREG-0869, Revision 1B i i k

                                                                                     . . . . =4 Therefore, based on the above assessment, it is the staff's opinion that all PWR dry containments can be excluded from any backfit requirements.

PWR Subatmospheric Containments: The release consequences due to an assumed failed containment associated with PWR subatmospheric containments are estimated to be the same as for PWR dry w/o SGFCs (see Appendix E). Sump blockage probabilities are judged to be in the same range also. Therefore, the averted releases and V/I estima'tes are the same as noted in the preceding section. Restoration of containment spray capability by operator action reduces estimated releases by a decade. Therefore, backfit requirements for subatmospheric containments are not supported for the same reasons cited above for PWR dry containments w/o SGFCs. PWR Ice Condenser Plants: As noted in Section 4.3 and Appendix E, PWR ice condenser plants are judged to be most prone to overpressure failure. However, this type of containment is limited to 10 plants of similar design, layout and recirculation flow features (see Section 4.3). For this group 'fo plants we have determined a separate value of sump failure probability based on known plant design parameters. The result is that the probability for sump blockage is judged to be lower, thus offseting the higher estimated conditional consequences from core melt. r l The calculated averted risk is 60 to 560 person-rem /Rx (similar in value to the PWR drys) and the calculated V/I ratios are 160 to 1400 person-rem /$M (low cost retrofit estimates); 40 to 380 person-rem /$M (high retrofit cost estimate). I

l l NUREG-0869, Revision 1B I' i

Thus, based on calculated averted risk and value/ impact noted above, a backfit requirement for PWR ice condenser plants is not indicated. BWR Mark I and Mark II Plants: NRC and industry PRA studies report an estimated BWR core melt frequency (attributable to all causes) of 2 x 10-5 to 3 x 10-4/Rx yr. The LOCA contribution to this total frequency is 1 to 10%, with large LOCAs having the lower value. Thus core melt frequency related to this safety issue is judged to be on the order of 10 ~7 to 10-6/Rx yr. If backfit actions are viewed from a reduction in core melt frequency perspective, the gain to be obtained (and this would apply only to BWRs where debris blockage was identified to significant) would be minimal. Therefore a backfit requirement is not indicated from this viewpoint. The calculated averted risk for this class of BWRs is 250 to 1250 person-rem /Rx (see Section 4.4) and is based on blockage probabilities derived from PWR studies. For reasons discussed in Section 4, these values are judged to be conservative. The corresponding V/I ratios derived were: 630 to 3,100 person rem /$M (low retrofit cost estimate) and 170 to 830 person-rem /$M (high retrofit cost estimate). The net result of the averted risk and value/ impact results is that an argument could be made to proceed with some type of plant assessments. However, this view must be balanced with the knowledge that only a few BWRs may fall into the higher V/I category. In addition (as for PWRs), it is the staff's opinion that the majority of BWRs will be determined to be at the lower end of the blockage frequency spectrum. An example is the analysis NUREG-0869, Revision 18 _ . __ __u submitted for Limerick-1* which resulted in adequate NPSH margin after analyzing per guidelines in the proposed RG 1.82, Revision 1. Another factor to be considered is suction realignment capabilities if blockage should occur. Mark I's can realign to alternate water sources (i.e., condensate storage tank). 23 of 41 BWRs have Mark I containments. Mark IIs do not have the option to align to the condensate storage tank, although realignment to the fuel storage pool is a possibility if such

            ' procedures were provided to the operators.

An additional significant factor is the ability for controlled venting. Controlled venting (which is an option that has received staff approval for submittal on a plant-specific basis) provides a means to maintain containment integrity by avoiding overpressurization loads. With wetwell venting, the calculated averted risk is reduced by a decade and the value/ impact cited above decreases accordingly. It is the staff's understanding that the BWR Owners Group is supporting implementation of controlled venting for all plants. i Therefore, given the above considerations, it is the staff's view that backfit requirements for Mark I's and Mark II's are of marginal significance and thereforr. are not proposed. Mark III Plants: The consequences (i.e., releases) associated with Mark III containments are estimated to be a decade lower than for Mark I's and II's due the channeling of fission products through the pool before release to the environment (assuming containment failure), with or without wetwell venting (see Appendix I E).

4/2/84, " Limerick Generating Station, Units 1 & 2, Containment Emergency Sump Performance", Philadelphia Electric Company SER submittal.

NUREG-0869, Revision 1B :

Therefore the calculated averted risk for Mark III's are 25 to 125 person-rem /Rx. The V/I ratios are: 60 to 310 person rem /$M (low retrofit cost estimate) or 17 to 80 person-rem /$M (high retrofit cost estimate). For the values shown above, a backfit requirement is not supportable for Mark III's. Option ; Option 3, which is a forward fit application of R.G. 1.82 Revision 1 to standard plant and new construction permit applications, will have no impact on current licensees and applicants. The incremental impact of requiring a new design to perform a sump blockage analysis is considered to be very small and therefore the value/ impact is believed to be very favorable. Also an important aspect for Option 3 is that all licensees and applicants will be informed as to the technical findings of A-43 and- the recommended guidance provided in the Regulatory Guide for performing a sump blockage analysis thus providing them with a basis for considerir.g analysis and corrective action as they deem necessary. Also this information provides each licensee or applicant with a better basis for considering changeout of insulation. The periodic changing of insulation is a conon practice at operating plants. Therefore because of its obvious high value and benefit with a very low impact we are recommending the adoption of Option 3. Option 4 Option 4 means that no overt means be used to inform the industry regarding the new information and understanding developed and that the existing incorrect NRC guidance with respect to sump blockage be left standing. Although this option has no impact it also has no value. The prospect of having incorrect guidance standing is unacceptable to the staff and this option was rejected. NUREG-0869, Revision 1B 5.3 Recommended Regulatory Action Based on the considerations discussed in Section 5.2, the staff recommends adoption of option 3 culminating in the the following actions for resolving USI A-43: (1) Issue the staff's technical findings (NUREG-0897, Revision 1A) for use as a technical information source. l (2) Issue SRP Section 6.2.2, Revision 4 and RG 1.82, Revision 1. These revisions reflect the staff's technical findings reported in NUREG-0897, > Revision 1A. This revised regulatory guidance would apply only to new construction permit applications and standard plant designs; implementation 4 would be required 6 months following issuance. (3) Issue a generic letter for information only to all holders of an operating license or construction permit outlining the safety concerns regarding potential debris blockage and recirculation failure due to inadequate NPSH.

A draft of such a generic letter is contained in Appendix H.

t 6 PLAN FOR IMPLEMENTATION I l The proposed resolution of USI A-43 would be accomplished through the i following actions: (1) Issue the staff's technical findings (NUREG-0897, Revision 1) for use as an information source by applicants, licensees, and the staff. (2) Issue Revision 1 of RG 1.82, to reflect the technical findings reported in NUREG-0897, Revision 1. In particular the 50% screen blockage criterion would be replaced by a plant-specific debris blot tage assessment. RG 1.82, Revision 1 provides specific guidance acceptable NUREG-0869, Revision 18 . - - . , ,- _ . - _-_ _ _ . . _ _ _

l to the staff for assessment of sump performance and RHR suction intakes including debris blockage effects. RG 1.82, Revision I would apply to new cps and standard plant designs only. (3) Issue the NRC Standard Review Plan (SRP, NUREG-0800) Section 6.2.2, i " Containment Heat Removal Systems," to reflect the guidance provided in j RG 1.82, Revision 1 and the technical findings in NUREG-0897, Revision

1. The revised SRP section would apply to new cps and standard plant designs only.

(4) Issue a Generic Letter to all holders of an operating license or

,                          construction permit outlining the potential safety concerns related to potential post-LOCA debris blockage and the inadequacy of the current 50% blockage criterion contained in Revision 0 of RG 1.82. This action is significant because replacement of plant insulation has become common practice in operating plants and licensees should be made aware of the

. potential for recirculation blockage from insulation following a LOCA. A draft Generic Letter is provided in Appendix H. The generic letter contains no requirements and no request for a response. The staff recommends that SRP Section 6.2.2, Revision 4, and RG 1.82, Revision 1, be issueu and become effective 6 months from date of issuance for implementation on new construction permits and standard plant designs. 7 STATUTORY CONSIDERATIONS 7.1 NRC Authority Because the proposed changes are revisions to RG 1.82 and SRP Section 6.2.2, these actions fall within the statutory authority of the NRC. Furthermore, , the recommendation to require applicants / licensees to demonstrate adequate NUREG-0869, Revision IB ,

  - s ._ .- ,-              - _ - . . .

__--m+.__ -

_e sump performance falls within the statutory authority of the NRC to regulate and ensure the safe operation of nuclear power plants.

7. 2 Need for National Environmental Policy Act (NEPA) Statement The proposed changes and potential plant backfit do not warrant a NEPA statement.

8 BIBLIOGRAPHY The following U.S. Nuclear Regulatory Commission doc ments were used in the preparation of this report: NUREG-75/014, " Reactor Safety Study," 1975 (formerly WASH-1400). NUREG-0800, " Standard Review Plan," July 1081; revised Standard Review Plan - Section 6.2.2, " Containment Heat Removal Systems," draft available from the NRC Division of Technical Information and Document Control,1717 H Street, NW, Washington, DC 20555. NUREG-0897, Revision 1, " Containment Emergency Sump Performance, Technical Findings Related to USI A-43," November 1984. NUREG/CP-0933, SAND 82-1659, " Proceedings of the Workshop on Containment j Integrity, Volume II of II, October 1982. l l NUREG/CR-2403: see Reyer. l NUREG/CR-2403, Supplement No.1: see Kolbe. l l l l l NUREG-0869, Revision 1B l \ i

NUREG/CR-2759: see Argonne. NUREG/CR-2760: see Padmanabhan and Hecker, June 1982. NUREG/CR-2761: see Padmanabhan, September 1981. NUREG/CR-2772: see Padmanabhan, June 1982. NUREG/CR-2982, see Brocard, July 1983. NUREG/CR-2791: see Rysocki. NUREG/CR-2792: see Kamath. NUREG/CR-3394: see Rysocki, July 1983. NUREG/CR-3616: see Brocard, December 1983. Regulatory Guide 1.82, Revision No.1 " Water Sources for Long-Term Recirculation Cooling Following a Loss-of-Coolant Accident," July 1985. Other documents used in the production of this report were: Argonne National Laboratory, "Results of Vertical Outlet Sump Tests," joint ARL/Sandia National Laboratory report, ARL47-82/ SAND 82-1286/NUREG/CR-2759, September 1982. l Brocard D., " Buoyancy, Transport and Head Loss of Fibrous Reactor Insulation," U.S. Nuclear Regulatory Commission report, NUREG/CR-2982, Revision 1, July 1983. NUREG-0869, Revision 18

                                                                                       .. AM
                                                                                             /

Brocard, D., " Transport and Screen Blockage Characteristics of Reflective Metallic Insulation Materials," U.S. Nuclear Regulatory Commission report, NUREG/CR-3616, December 1983. Ferrell, W. L. et al., "Probabilistic Assessment of USI A-43," Science Applications, Inc. report, September 1982. Kamath, P., T. Tantillo, and W. Swift, "An Assessment of Residual Heat Removal and Containment Spray Pump Performance Under Air and Debris Ingesting Conditions," Creare, Inc., Hanover, NH, U.S. Nuclear Regulatory Commission report, NUREG/CR-2792, September 1982. Kolbe, R. and E. Gahan, " Survey of Insulation Used in Nuclear Power Plants and the Potential for Debris Generation," Burns and Roe, Inc. , Oradell, NJ, U.S. Nuclear Regulatory Commission report, NUREG/CR-2403, Supplement 1, May 1982. Padmanabhan, M., "Results of Vortex Suppressor Tests, Single Outlet Sump Test and Miscellaneous Sensitivity Tests," Alden Research Laboratory, Worcester Polytechnic Institute, Holden, MA, U.S. Nuclear Regulatory Commission report, NUREG/CR-2761, September 1982. Padmanabhan, M., " Hydraulic Performance of Pump Suction Inlet for Emergency Core Cooling Systems in Boiling Water Reactors," Alden Research Laboratory, Worcester Polytechnic Institute, Holden, MA, U.S. Nuclear Regulatory Commission report, NUREG/CR-2772, June 1982. l-r Padmanabhan, M., and G. E. Hecker, " Assessment of Scale Effects on Vortexing, l Swirl, and Inlet Losses in Large Scale Sump Models," Alden Research Laboratory, Wrocester Polytechnic Institute, Holden, MA, U.S. Nuclear Regulatory Commission report, NUREG/CR-2760, June 1982. I NUREG-0869, Revision IB L_

                                                                                      ..s rq Reyer, R. , et al. , " Survey of Insulation Used in Nuclear Power Plants and the Potential for Debt is Generation," Burns and Roe, Inc. , Oradell, NJ, U.S.

Nuclear Regulatory Commission report, NUREG/CR-2403, Supplement 1, October 1981. Wysocki, J. J., " Methodology for Evaluation of Insulation Debris," Burns and Roe, Inc., Oradell, NJ, U.S. Nuclear Regulatory Commission report, NUREG/CR-2791, September 1982. Wysocki, J. J., "Probabilistic Assessment of Recirculation Sump Blockage Due to Loss-of-Coolant Accidents, U.S. Nuclear Regulatory Commission report, NUREG/CR-3394, Volumes 1 and 2, July 1983. l l l 4 k I i i NUREG-0869, Revision 18 i \ i

                                                                      - ...--...s APPENDIX A

SUMMARY

OF PUBLIC COMENTS RECEIVED AND ACTIONS TAKEN (REF. USI A-43) NUREG-0869, Revision 1

APPENDIX A

SUMMARY

OF PUBLIC COMMENTS RECEIVED AND ACTIONS TAKEN 1 INTRODUCTION The technical findings related to Unresolved Safety Issue (USI) A-43 were published for comment in May 1983. Notice of the publication was placed in the Federal Register on May 9,1983. The official comment period lasted for 60 days and ended on July 11, 1983. However, comments were received into September 1983, with followup comments received into November 1983. A listing of those who responded during the period and afterwards is shown in Table 1. Copies of the comment letters are on file in the NRC Public Document Room, 1717 H Street, NW, Washington, DC. A public meeting was held on June 1 and 2,1983, at Bethesda, Maryland, to offer additional opportunity for public comments; however, attendance was very small. Followup discussions were held with respondees to clarify issues raised at this meeting and in the written comments. An overview of the comments received is provided in Section 2 below. Section 3 contains summaries of significant comments and the actions planned to resolve them. 2 OVERVIEW 0F COMMENTS RECEIVED The major written comments received addressed seven specific subject areas. The comment categories and connentors are listed in Table 2 below. The commentors are identified in Table 2 as follows: Alden Research Laboratory (ARL); Atomic Industrial Forum (AIF); BWR Owners Group (BWR); Commonwealth Edison (CEd); Consumers Power Co. (CPC); Creare Research and Development (CRD); Diamond Power Co. (DPC); General Electric (GE); Gibbs and Hill, Inc. (GH); Northeast Utilities (NE); and Owens-Corning Fiberglass, Inc. (0CF). NUREG-0869, Revision 1 A-1

                                                                                                          . _ _ _ _ . . .                         - ,s-Table 1 Persons who commented on the technical findings related to USI A-43*

Alden Research Laboratory (ARL), M. Padmanabhan, letter to A. Serkiz (NRC),

                         " Comments on NUREG-0897 and 0869," June 13, 1983.

ARL, M. Padmanabhan, letter to A. Serkiz (NRC), " Revision to Table A-3 in NUREG-0869," June 22, 1983. Atomic Industrial Forum, R. Szalay, letter to the Secretary of the Commission, "NRC's Proposed Resolution of Unresolved Safety Issue A-43, Containment Emergency Sump Performance, Contained in NUREG-0869," July 22, i 1983. ' Atomic Industrial Forum, J. Cook, letter to R. Furple (NRC) and enclosure

                         " Examples of Staff Review Going Beyond Approved Regulatory Criteria,"

June 4, 1984. BWR Owners Group, T. J. Dente, letter to T. P. Speis (NRC), "BWR Owners' Group Comments on Proposed Revision to Regulatory Guide 1.82, Rev. 1," October 18, 1983. BWR Owners Group, D. R. Helwig, letter to V. Stello (NRC), BWR Owners' Group comments on Regulatory Guide 1.82, Revision 1, July 16, 1984. Commonwealth Edison, D. L. Farrar, letter to the Secretary of the Commission, "NUREG-0897, Containment Emergency Sump Performance; Standard Review Plan Section 6.2.2, Rev. 4, Containment Emergency Heat Removal Systems; and NUREG-0869, USI A-43 Resolution Positions (48FR2089; May 9, 1983)," July 13, 1983. Consumers Power, D. M. Budzik, letter to the Secretary of the Commission,

                        '* Comments Concerning Regulatory Guide 1.82, Proposed Revision 1 (File 0485.1, 0911.1.5, Serial: 23206)," July 15,1983.

Creare, W. L. Swift, letter to P. Strom (SNL), " Comments on Figure 3-6 of NUREG-0897 and Table A-9 of NUREG-0869," June 13, 1983. Diamond Power Company, R. E. Ziegler and B. D. Ziels, letter to K. Kniel (NRC),

                        " Containment Emergency Sump Performance, USI A-43," July 11,1983.

Diamond Power Specialty Company, B. D. Ziels, letter to A. Serkiz (NRC),

                       "HDR Test Result Summary, MIRROR Insulation Performance During LOCA Conditions," December 6, 1984.

General Electric (GE), J. F. Quirk, letter to K. Kniel (NRC), " Comments on Emergency Sump Documents," July 11, 1983. i

Including comments on NUREG-0869, NUREG-0897, proposed Revision 1 to Regulatory Guida 1.82, and proposed Revision 4 to Section 6.2.2 of the Standard Review Plan (SRP, NUREG-0800).

NUREG-0869. Revision 1 A-2

Table 1 (Continued) GE, J. F. Quirk, letter to T. P. Speis (NRC), " Comments on Proposed Regulatory Guide 1.82, Rev. 1," October 17, 1983. Gibbs and Hill, Inc., M. A. Vivirito, letter to the Secretary of the Commission, " Comments on Proposed Revision No.1 to RG 1.82," July 11,1983. Northeast Utilities, W. G. Counsil, letter to K. Kniel (NRC), "Haddam Neck, Millstone Nuclear Power Station, Unit Nos. 1, 2, and 3, Comments on NUREG 3897, SRP Section 6.2.2 and NUREG-0869," September 2, 1983. Owens Corning Fiberglass (OCF), G. H. Hart, letter to A. Serkiz (NRC),

       " Comments on NUREG-0897 and NUREG-0869," June 23, 1983.

OCF, G. H. Hart, letter to A. Serkiz (NRC), " Updated Comments on NUREG-0897 and NUREG-0869," July 14,1983. OCF, G. P. Pinsky, letter to K. Kniel (NRC), " Comments on NUREG-0879 and -0896," July 14, 1983. OCF, G. H. Hart, transmittal to A. Serkiz (NRC), "HDR Blowdown Tests with NUK0N In ulation Blankets," February 18, 1985. Power Component Systems, Inc. , D. A. Leach, letter to A. Serkiz (NRC),

      " Nuclear Grade Blanket Insulation," November 8, 1984.

Table 2 Categories addressed in major written comments ! Comment Category ARL AIF BWR CED CPC CRD DPC GE GH NE OCF (1) Survey of insulation used is X X not current or complete. (2) Cost estimates are low. X X 1 (3) Estimates of sump blockage X X X X l probabilities are high. (4) Value-impact analysis questioned. X X X X (5) BWRs should be exempt; A-43 is a X X X PWR issue. (6) Insulation material definitions and X X i descriptions need revision for l clarity and completeness. (7) Technical comments on and X X X X X X X X clarifications of subject matter l in NUREG-0897 and NUREG-0869. 1 ( NUREG-0869, Revision 1 A-3

By category, the actions taken in response to these comments are as follows: Categories 1 and 6: Tables have been added to NUREG-0897, Revision 1 and NUREG-0869, Revision 1 to include tne additional plant insulation information provided auring the public comment period. The text of the NUREGs has been revised to reflect recommended insulation definitions and the need to evaluate the specific insulation employed. Categories 2 and 4: The cost estimates provided by different industry groups have varied over a wide range. With the exception of Diamond Power Company, respondees claimed that the cost estimates in value/ impact analysis were too low. The revised value/ impact analysis reflects an averaged value derived from costs provided. Category 3: A detailed sump blockage probability analysis has been performed and is reported in NUREG/CR-3394. The results were used in the revised value/ impact analysis. These results show a sump blockage probability range for pressurized water reactors (PWRs) of 10 8 to 5 x 10 5/Rx yr and a strong dependence on plant design. Category 5: NUREG-0869 and Regulatory Guide 1.82 have been revised to specifically identify areas of concern for boiling water reactors (BWRs) and for PWRs. Category 7: Technical corrections and clarifications have been made in the appropriate sections of NUREG-0897 and NUREG-0869. The NRC staff greatly appreciates the review and comments provided by the respondees. The time and effort they have taken to review USI A-43 has resulted in an improved report that will reflect current findings and a balanced position with respect to this safety issue. NUREG-0869, Revision 1 A-4

3 C0FNENTS RECEIVED AND PROPOSED ACTION (OR RESPCNSE) ACTIONS The NRC staff has given complete and careful consideration to all coments received on USI A-43. Sumaries of significant coments and the actions taken by the NRC staff in response are provided in Table 3. Coments are presented in alphabetical order, based on the name of the comenting institution. 4 'l l NUREG-0869, Revision 1 A-5

                                ,e--..-m_,. , ..,-.--.,-,-e,   ,,,, - - - - . , -.n-,. ----_-.,,e ,,------.e ,.n--_ . , . . , . ..

E Table 3 Comments received on USI A-43 and NRC staff response '? c3 $ Comment NRC Staff Response ?

o Alden Research Laboratory 1,

v 8 ARL noted typographical errors and proposed These corrections and clarifications have been [ technical clarification to several tables incorporated into NUREG-0897 and NUREG-0869. - Atomic Industrial Forum The cost impact of $550,000/ plant used in Costs impacts were re-evaluated based on cost estimate value/ impact analysirs is low by at least information received from AIF and other respondees a factor of 2. Economic considerations related to reduced The essence of a value/ impact analysis is that it > probability of plant damage should be excluded attempts to identify, organize, relate, and make En from the cost-benefit balancing. Decisions " visible" all the significant elements of value expected should be based primarily on the value/ impact to be derived from a proposed regulatory action as well ratio. as all significant elements of impact. The net values are compared with the net impacts. Thus if a proposed safety improvement is accompanied by an adverse side effect, the impairment is subtracted from the improvement to arrive at a net safety value for consideration in the value/ impact assessment. Similarly, when the immediate and prospective cost impacts are summed, they should include all elements of economic impact on licensees, such as costs to design, plan, install, test, operate, maintain, etc. Plant downtime or decreased plant availability is included when applicable. The summed impacts, however, should be net impacts, for compariston with net values. Thus, any reductions in operating costs, improvements ir. plant availability, or reductions in the probability of plant damage are properly a factor in determining net adverse economic impact. Future economic costs and savings are appropriately discounted. ,

5 Table 3 (Continued) '? Comment NRC Staff Response h Qualitative differences among impact elements are 7 respected, and distinctive elements of impact (of which " averted plant-damage probability, as a favorable rather " than adverse impact, is a prominent example) are separately identified, for appropriate consideration in regulatory decision making. The ratio of avoided public dose to the gross cost of implementation is ordinarily a major decision factor. However, it is not by itself always a good guide to a sound regulatory decision. The issues involved are often too complex for a decision on this criterion alone. Other > factors that enter, often in important ways, may include 4 any economic benefits that reduce a net adverse economic impact, the safety importance of the issue, and values and impacts that cannot or cannot readily be quantified; for

                           .                        example, jeopardy to a defense layer in the defense-in-depth concept or expected reductions in plant availability that can be foreseen but not precisely estimated.

A sound regulatory decision rests on adequate consideration of all significant factors. An overly simple approach can mislead.if it simplifies away complexities that are the essence of the issue at hand. The assumption that sump failure will occur in A detailed sump blockage probability analysis has been 50% cases of the large LOCAs should be performed and is reported in NUREG/CR-3394. The results justified. show a wide range ci sump blockage failure probabilities (i.e., 3E-b to SE-5/ reactor year) and a high dependency on

                         ,                          plant design and operational requirements. These results are reflected in a revised value impact analysis utilizing a range of sump failure probabilities.

The use of PWR release categories from The containment failure probabilities and release WASH-1400 is too conservative. Containment categories used in the regulatory analysis for USI A-43 failure probabilitics used in WA5H-1460 were based on information presented in WASH-1400, and

E Table 3 (Continued) M

'?

o

 $        Comment                                                 NRC Staff Response
 ?

h are inadequate to describe the nuclear also on other considerations. The comments presented by 7 industry's present knowledge in this field. an AIF subcommittee regarding the validity of continued Releases due to " vessel steam explosion" use of WASH-1400 assumptions, etc. are being evaluated are unrealistic and should not be considered. through other activities such as: reevaluation of source terms, SASA studies, etc. USI A-43 regulatory analyses were based on the following considerations and for the reasons noted: (1) WASH-1400 assumptions were utilized to provide a common baseline calculations for reference plants and were used to estimate increases in releases due to a postulated loss of recirculation flow capacity. Until

m revised failure mechanisms and new source terms are determined, this approach provides a consistent set of calculations. (2) Although using a small containment failure probability associated with steam explosion would be more appropriate, release category PWR-1 (which includes steam explosion) was not a dominant contributor to release. Release categories PWR-2, -4, and -6 were the dominant pathways contributing to increases releases due to a failed sump for the plants analyzed. (3) Basing release effects on the assumption of simultaneous failure of core cooling and loss of containment sprays is conservative. If containment were not lost (as would be the situation for PWRs that have dry containments with safety grade fan cooler systems), the LOCA energy could be disspated without containment overpressurization and failure. Thus releases associated with PWR-2 and -4 categories could be discounted and PWR-6 releases only used. Such considerations have been incorporated into this - revised regulatory analysis. I i

g , Table 3 (Continued) $l T h Coment NRC Staff Response [ (4) Other factors--such as containment structural design

margins that argue against gross containment g failures (as postulated in WASH-1400), realignment to alternate water sources, controlled venting for BWRs, etc.--have also been considered this revised regulatory analysis.

The use of the CRAC Code and a "no-evacuation," The 50-mile radius reflects a substantial part (though 50-mile-radius model to develop public doses not all) of the total population dose, and is thus a is unrealistic. reasonable index of the radiological effect on the public. Standardization of calculations to that radius is helpful in comparing risks associated with different issues and ? average such risks for use with the 1000 person-rem /$H

criteria.

Evacuation of people is not considered because calculations suggest that, although it may sometimes be important for people directly affected, the effect of evacuation on the total population dose is likely to be small. URC should utilize information developed more Possible changes in the source terms are being considered recently (i.e., NUREG-0772) to reassess and by the special task force established by the Commission reduce the source terms, rather than continue to review the source-term issue. Changes would be to use the PWR-2 and PWR-3 release categories premature before this group completes its evaluation from WASH-1400. and the new values are accepted by all parties involved. NRC should utilize the " leak before break" Elastic-plastic fracture mechanics analysis techniq'ues to - concept in evaluating the safety significance analyze pipe break potential has been used in USI A-2, with of A-43. a limited number of PWRs being analyzed. For USI A-2, the submittal of such analyses for specific break locations (on a plant-specific basis) will require obtaining an exemption from the requirements of CDC 4. Submittal of such analyses to address the USI A-43 debris blockage issues would be reviewed by staff on a plant-specific basis, should a licensee or applicant elect to utilize this approach. L,

EE Table 3 (Continued) in

  'P o

8? Comment NRC Staff Response i

n

   @ BWR Owners Group T After quick review of the proposed revision to          The requirement for long-term decay heat removal is the regulatory guide, the BWR Owners Group and         applicable to light-water reactors, both BWRs and PWRs.

GE maintain that USI A-43 is not a generic issue for BWRs. The revisions to RG 1.82, which now proposes All types of insulation should be evaluated for the specific criteria for BWRs, should apply potential of debris generation, transport, and suction only to light-water reactors that have any strainer blockage. The wide variation in plant designs potential for harmful debris generation (i.e., and insulation employed does not support a generic light water reactors that extensively use statement. fibrous insulation). c' These comments and any future comments by RG 1.82, Revision 1 (along with NOREG-0897, the BWR Owners Group should not substitute NUREG-0869 and SRP 6.2.2, Revision 4) was issued for the normal notice and comment procedure "for comment" in May 1983. Only 14 respon=es were that allows potentially affected licensees received as of September 1983. Some of these comments to respond to proposed regulatory guide (in particular GE's July 11, 1983 letter) cited a need changes. to specifically address BWR-related concerns in the RG. This was done and copies were sent to GE and the BWR Owners Group. Given the previous extensive distribution of "for comment" reports and regulatory positions and the rather small number of responses, the staff does not plan to reissue RG 1.82, Revision 1 for comment. The NRC staff will incorporate additional valid technical points received from the BWR Owners Group and GE. C The most recent input from the BWR Owners Group (July 16, 1984) does not provide new significant findings; rather this input re-expresses concerns previously voiced and stresses possible misinterpretations of wording i RG 1.82, Revision 1.

1 b

ll Table 3 (Continued) a l Comment NRC Staff Response o

   @   Commonwealth Edison                                                                                                    ,

u', E7 The Commission has not sufficiently justified the A-43 resolution does not mandate retrofits; rather, need to impose retrofit requirements on either applicants are requested to assess long-term operating or near-term operating license units. recirculation capability utilizing RG 1.82, Revision 1 and to then determine what corrective actions may be needed. The use of an information bulletin to the majority of the plants does not constitute imposition of a retrofit. Cost estimates for surveys, design reviews, and The A-43 value/ impact evaluation has been revised retrofitting are questionable. based on detailed sump blockage probability studies 2 (NUREG/CR-3394) and cost estimates received from j; industry responses. The proposed RG 1.82 is overly conservative. The NRC staff acknowledges that conservatisms exist However, given the need for assurance that the in RG 1.82, Revision 1. However, such conservatisms recirculation sump remains a reliable source are prompted by the limited amount of available information of cooling water, the ccamentor agrees that an regarding insulation destruction due to high pressure evaluation of sump designs, potential for debris, jets and attendant debris generation, and the wide air ingestion, and adequate net positive suction variability of plant designs and types of insulation v head (NPSH) is fully justified. used. The commentor questions the assumption that 50% A detailed sump failure probability analysis was of LO.As lead to sump loss, the value/ impact ratio performed and is reported in NUREG/CR-3394. The given uncertainties in estimated costs, the basis " averaged" sump failure probability was 2E-5/ reactor-for assuming 23 years remaining plant life, etc. year with a range of 3E-6 to SE-5/ reactor year. Consumers Power Regarding the proposed Revision 1 to RG 1.82, the Appendix A of proposed RG 1.82, Revision 1 was commentor stated (1) that Appendix A should be always intended to provide additional information clearly delineated as being an information and and/or guidance, not design requirements. Appendix A guidance source, not as presenting design require- has been clearly labeled as such. ments, and (2) that consistency is needed with respect to NPSH terminology. l'

g Table 3 (Continued) N 'P 8! Comment NRC Staff Response H? Regarding the value/ impact analysis, the commentor That 50% of LOCAs lead to sump blockage has been lL questioned the assumption that 50% of the loss-of- reevaluated (see NUREG/CR-3394), and the results TL coolant accidents (LOCAs) lead to sump blockage and of that detailed study have been used in revising the 8 cites a sump failure frequency of 2 x 10 4 per A-43 release estimates. _. demand from another probabilistic risk analysis. The commentor questioned the direct application of The calculation of avoided accidents costs, core melt frequency reduction for computing avoided loss-of plant costs, etc., are consistent with current accident cost. The commentor disagrees with taking NRC staff evaluation practices. Recalculation of the credit for loss of plant cost. Rather, the parameters previously used will be carried out with commentor states that loss-of plant costs should be the revised blockage frequencies. deducted from avoided accident costs. T* Creare g A The beta factor used to predict a pump's Efforts were made to obtain the original data required NPSH in an air / water mixture is based tapes and calculate the data's scatter; however, this on data whose scatter was not reported. The information was not readily available. The suggested NUREG should note this and caution the applicant cautionary note has been added to NUREG-0897. and reviewer to carefully consider the adequacy of the NPSH margin if it is marginal. The use of an arbitrary minimum allowable NPSH NUREG-0897 and RG 1.82, Revision 1 no longer recommend margin, either as a fixed value (i.e., 1 foot) a minimum allowable NPSH margin. Instead, they or as a percentage value (i.e. , 0.5 x margin with note that whatever NPSH margin is available (after no screen blockage), is not justifiable. It should accounting for hydraulic and screen blockage be recognized that what constitutes a safe NPSH effects) should be evaluated with respect to each margin is a plant-specific judgment. plant's long-term recirculation requirements. Diamond Power Company NUREG-0897 resolves a significant safety problem in The NRC staff concurs. a thorough and equitable manner. f!' t' I'

_. Table 3 (Continued) a o A

$g        Comment                                                      NRC Staff Response o  The commentor provides recommendations regarding         The proposed classifications have been combined with 2    the classification of various insulating materials,      other similar proposals to revise and clarify the LT   particulcrly on the need to distinguish between          insulation classification and descriptions used in EI totally encapsulated insulation and jacketed             NUREG-0897.

insulation. The commentor provides listings of the types of The information has been added to NUREG-0897 and insulations purchased since 1980 and the types NUREG-0869, along with data received from other # of insulations used in recent retrofittings. manufacturers. Thc commentor states that the costs in the This cost information has been reflected in the value/ impact analysis are in agreement with its revised value/ impact analysis (NUREG-0869), along costs and provides the following figures: with other industry cost figures. Y g; Cost of MIRROR" reflective metallic insulation = $40/ft2 for material alone. Installation cost, excluding material

          = $25/ hour.

Productivity = 1.24 hours /fta of insulation. t,

l me Table 3 (Continued) E5 E.l 83 Comment NRC Staff Response EP Reflective metallic insulation is not the Information supplied by Owens-Corning Fiberglass Co. li predominant type of insulation ustd in newer and the Diamond Power Co. regarding types of insulation 5L plants. Recently insulated p! ants mainly used in existing and future reactors has been added to 8 use fiberglass insulation.* NUREG-0897 and NUREG-0869. These reports have been revi' sed to reflect this new information. The trend appears to be toward a greater use of fibrous insulations. A report on HOR test results on MIRROR

This report has been included as Appendix E in insulation performance during LOCA conditions NUREG-0897, Rev. 1. The results of this report was submitted to provide additional information do not support a hypothesis which postulates for the existing data base used in resolution free and undamaged inner foils being available of USI A-43.** to transport at low velocities and to cause 2, blockage. However, the limited data base

2. precludes developing a detailed debris generation model.

Letter of July 11, 1983.

   ** Letter of December 6, 1984.
                                                                                                                   *   *i

t = Table 3 (Continued) E E 6 Comment NRC Staff Response ? m General Electric Company 2 7 SRP 6.2.2 and RG 1.82, Revision 1 make no RG 1.82, Revision 1 and SRP 6.2.2 have been modified & distinction between BWRs and PWRs; regulatory to identify PWR- and BWR-related concerns. Regulatory criteria should differentiate between various Guide 1.82, Revision 1, has been renamed " Water Sources plant designs.* for Long-Term Recircuiation Cooling Following a Loss-of-Coolant Accident." Reference should be made to technical findings 3ased on the responses received, the A-43 technical that imply that A-43 concerns do not pose a findings will be revised to reflect (1) that there is serious problem for BWRs.* a more extensive use of fibrous insulations (i.e., NUKON") than previously identified and (2) that BWks are reinsulating with NUK0N". NUREG-0897 will reflect 1" current findings and identify both PWR- and BWR-5; related concerns. The value impact analysis utilizes a PWR for GE's point on utilizing a PWR probabilistic risk the risk assessment and PWR-oriented industry assessment for drawing conclusions for a BWR is impacts and, as such, is not directly applicable acknowledged. Similar assessments have been made to BWRs.* for BWRs and those results have been utilized in preparing this revised regulatory analysis. General Electric has reviewed the proposed The requirement for long-term decay heat removal is revisions and has concluded that the design applicable to both BWRs and PWRs. RG 1.82, requirements proposed in RG 1.82, Revision 1 Revision 1, Appendix A contains a series of tables (or are excessively prescriptive and not generically guidelines) that have been derived from extensive applicable to the BWR."* tests and analytical studies. This information is provided for ease of referral and can, or need not, be used--at the user's option. RG 1.82, Revision 1 is general, and not prescriptive. The applicant has the responsibility for design submittal and justification of the safety aspects thereof. Letter of July 11, 1983.

   ** Letter of October 17, 1983.                                                                                     i 5

g: Table 3 (Continued) El ? 8? Comment NRC Staff Response 8 g7 The proposed RG should be revised so that no The technical findings in 1983 (versus earlier findings)

$L  further requirements are imposed on designs that        are considerably different, particularly with
!1 have already included precautions that preclude          respect to insulation employed currently and the g   air ingestion into, or blocking of, suction lines       transport characteristics of insulation debris. The

_, used for long-term decay heat removal.* air ingestion potential has been experimentally quantified and found to be small. However the 50% blockage criterion in the current RG 1.82 permitted applicants to essentially bypass the debris blockage question. For those plants where design precautions can be clearly demonstrated, further actions (retrofits) are not necessary. 2, In addition, the proposed RG should be further The licensee and/or applicant always has the option 2. m revised to provide for altcrnative means of to propose alternate means to deal with a particular ensuring that long-term heat removal is not lost design or safety problem. as a result of suction blocking or air ingestion." In the SER for GESSAR, the NRC indicated that At the time the SER for GESSAR II was written, A-43 USI A-43 posed no problem for the Mark III concerns relative to BWRs were still under evaluation. containment configuration." The staff's SER cited several elements of the GESSAR II design that tended to reduce the probability for blockage of the RHR suction inlets due to LOCA-generated debris. The staff concluded that plants referencing the GESSAR II design could proceed, pending resolution of USI A-43, without endangering the health and safety of the public while the staff completed its evaluation of GESSAR. The unique aspects of each Mark III plant design should be evaluated during plant-specific reviews of A-43 concerns.

Letter of October 17, 1983. -

se Table 3 (Continued) E E E3 Comment NRC Staff Response a? The tests performed by Alden Research Laboratory The comment is partially correct, because BWR RHR fi (see NUREG/CR-3616) may even be very conservative suction inlets are located at some elevated distance El for BWRs since it appears the tests utilized sump above the wetwell or suppression pool floor. However, 8 screens directly on the sump floor.* the insulation debris transport characteristics (see _, NUREG/CR-2982, Rev. 1) showed that low velocities (i.e., 0.2 - 0.3 ft/sec) can transport fragmented debris and are applicable to both BWRs and PWRs. The proposed regulatory guide should be revised RG 1.82, Revision 1 states: "This regulatory guide to include criteria that will allow alternative has been developcd from an extensive experimental and measures for precluding loss of long-term analytical data base. The applicant is free to decay heat removal due to air ingestion or select alternate calculation methods which are based 2, blockage.* on substantiating experiments and/or limiting analytical 2. considerations." Thus, the applicant is free to select alternate methods or measures for precluding loss of long-term decay heat removal. Earlier surveys on the use of insulation in light As stated above, current findings do not support the water reactors have concluded that most BWRs utilize earlier surveys or conclusions. NUREG-0897 is being metallic insulation, which minimizes the potential revised to incorporate findings from public comments for formation and subsequent transport of debris received (particularly with respect to insulations to the sump screens.* currently used and the change to fibrous insulation from previously used reflective metallic insulations). Recent tests on the transport of thin stainless steel foils show that this material can be transported at low velocities (i.e., 0.2 to 0.3 ft/sec).

Letter of October 17, 1983.

I t

EE

??

T) Table 3 (Continued) 8 S

                         $o         Comment                                                        NRC Staff Response E

7 3 r Gibbs and Hill, Inc. Section B does not discuss the fact that sump Appendix A (page 19) has wording very similar to the configurations that differ significantly from commentor's suggested wording. the criteria of Appendix A may be equally acceptable. Gibbs and Hill recommends adding the following concluding paragraph to Section 8:

                              "If the sump design differs significantly from the guidelines presented in Appendix A, similar data from full-scale or reduced-scale tests, or in-plant tests can be used to verify adequate sump hydraulic 3,   performance."

oo Tables A-1 and A-3 are inconsistent and Table A-2 The inconsistencies have been corrected. has inconsistencies in water level noted. Northeast Utilities Tests show that gratings are as effective as Gratings were very effective in reducing air solid cover plate in suppressing vortices. ingestion to essentially zero. The procedure in Appendix B is too prescriptive. Appendix B in NUREG-0897 presents the staff's The NRC should allow licensees to define and technical findings for A-43. Appendix B was included develop their own evaluation methods. to illustrate major considerations. RG 1.82, Revision 1 is the regulatory document. 84

                                                                                                                                             . U k

E ' El c> 23 Table 3 (Continued) f Comment NRC Staff Response 1 i 8 Credit should be given for top screen area if for those plant designs and calculated plant _. it is deep enough to reduce the pctential for conditions where this point could be unconditionally clogging (RG 1.82, Revision 1, Section C, Item 7). substantiated, credit would be given. The licensee should be free to determine methods Regulatory Guide 1.82, Revision I has been revised of inspection and access requirements (RG 1.82, accordingly. Revision 1, Section C, Item 14). RG 1.82, Revision I will be used to evaluate sumps The need for backfitting will be based on plant-in operating plants. This may require backfitting specific analyses that will reveal the need for at substantial costs. and the extent of backfitting that might be required. The cost of backfit should be weighed against core

?*                                                          melt costs.

2 Appendix A to RG 1.82, Revision 1 requires obtaining Appendix A states: "If the sump design deviates performance data if sump design deviates significantly from the design boundaries noted, then significantly from the guidelines provided. similar performance data should be obtained for verifi-For operating plants, this may result in costly cation of adequate sump hydraulic performance." sump testing. NRC estimates for man-rem costs associated The value impact analysis has been revised based on with insulation replacement are grossly cost data received during "for comment" period. underestimated. The value impact analysis addresses only PWRs. The value impact analysis revision clearly addresses If the NRC has concluded that this issue only BWR and PWR concerns. applies to PWRs, the document should reflect this. The commentor concurs with the comments The AIF comments are addressed separately; see above. submitted separately on this document by the AIF.

e Table 3 (Continued) E 2 El Comment NRC Staff Response SS af Owens* Corning 5[ Detailed comments addressed the wide variation Detailed comments received on insulation types; E of insulations employed, descriptions, suggested descriptions, etc. have been used to revise ' terminology, etc.* NUREG-0897. Comments recommended including transport and head Data from NUKON" tests have been referenced ar.d loss data for NUKON" fiberglass tests.* major findings summarized in the revised NUREG-0897. The commentor questioned Table B-1, Criterion 2, Transport tests on reflective metallic foils that reflective metallic insulation foil debris revealed that they can be transported at low would not be transported at velocities less than velocities (0.2 - 0.5 ft/sec). 2 2.0 ft/sec.* r'o The commentor questioned the concept that if there Inputs received have been used in revising NUREG-0869. is all reflective metallic insulation there is no problem.* The commentor recommended changes to various Inputs received have been used in revising NUREG-0869. ' tables as discussed at the June 1 and 2, 1983, public meeting.* The commentor suggested word changes that would Inputs received have been used in revising NUREG-0869. minimize singling out fibrous type insulations as the screen blockage concern without considering blockages due to reflective metallic insulation materials.* ' The commentor addiessed the recommended revision Inputs received have been used in revising NUREG-0869. to reflect current status of insulations employed in nuclear power plants.*

Letter of June 23, 1983. i J

== Table 3 (Continued) E B. Comment NRC Staff Response ll E The potential for screen blockage by reflective A set of experiments to determine transport velocities fi metallic debris has not been adequately addressed. (similar to those performed on fibrous insulations) El In particular, the water velocities required has been cos:pleted by Alden Research Laboratory. 8 to transport debris and hold it against the sump The results are summarized in NUREG-0897 and used in _. screen have not been studied.* RG 1.82. The assumption that all fibrous blankets and The 7 L/D criterion is based on experimental studies pillows within 7 L/D of a break are destroyed is of representative samples of fibrous pillows exposed overly conservative. Different designs of pillows to high pressure water jets. These small water jet have varying resistances to destruction by water studies showed that increasing pressure (40-60 psia) jets.* results in destruction of pillow covers and release of core material. Furthermore, blowdown experiments 2 in the German HDR facility showed that fiberglass d2 "" insulations (even when jacketed) were destroyed within 6 to 12 feet of the break, and distributed throughout containment as very fine particles. Unless conclusive experimental evidence is obtained that accurately replicates the variety of conditions that may exist in a LOCA, it is prudent to retain the conservative 7 L/D criterion. The 7 L/D envelope is a significant reduction from the previously proposed 0.5 psia stagnation pressure destruction criterion in NUREG/CR-2791 (September 1982) anc (in general) limits the zone of maximum destruction to the primary system piping and lower portions of the steam generators.

    " Letter of July 14, 1983.

(i li

g Table 3 (Continued) E!

'?

8 Comment NRC Staff Response The commentor stated that estimated costs for OCF cost data are utilized in revisions to the

5. insulation installation and replacement are value/ impact analysis.

too low. OCF cost estimates were* S _. Cost of NUKON" = $90/ft2 for material (as fabricated) Cost of reflective metalli : = $100/f t2 for material (as fabricated) Installation cost = $112/ft2 #or labor and related support

 ?
 %  The commentor provided recommendatior.s for               Descriptive classifications provided for insulation classification of various insulating materials,           types have been combined with similar classifications stressing differences between NUKON" (an OCF              obtained from Diamond Power Company for inclusion in product) and other fiberglass and mineral                 NUREG-0897, Revision 1 and NUREG-0869, Revision 1.

wc,ol materials. The commentor also noted the differences between NUKON" and high density fiberglass.* The commentor identified 14 nuclear power OCF plant information has been utilized, along with plants that have been reinsulated with NUKON", information from Diamond Power Company, to develop are in the process of installing NUKON", or a current picture of insulation utilization in may install NUKON".* nuclear power plants. The major finding is that the e number of plants using or planning to use fibrous insulation is larger than previously estimated. For example, the Diamond Power list reveals that 25 of 130 operating and projected plants are utilizing fibrous insulation on primary system components. l

    " Letter of July 14, 1983.

I

Table 3 (Continued)

E 8 c3 Comment NRC Staff Response EP The commentor recommended inspection surveys of The recommendation for physical plant surveys (or

 $L plants to identify actual insulations employed and         inspection to identify types and quantities of
 $L recommended the modification of a draft generic           insulations employed) is a good one. However, the 8   letter to include this requirement.*                     generic letter for plant specific evaluation based on actual types and quantities of insulation employed will address this concern.

A report on "HDR Blowdown Tests with NUKON This report has been included as Appendix F in Insulation Blankets" was submitted to NUREG-0897, Rev. 1. The tests demonstrated that support the capability of NUKON* insulation jacketed and unjacketed NUKON* blankets within 7 L/D to withstand the impact of a high pressure will be nearly totally destroyed. NUK0N* blankets steam-water blast.** enclosed in standard NUK0N* stainless steel jackets 2 withstood the blast better; not enough of these A2 tests were performed to allov conclusions to be drawn for similar insulations. Power Component Systems, Inc. A report on " Buoyancy, Transport and Head Loss The formula provided for fibrous debris blocakge Characteristics of Nuclear Grade Insulation head loss is included in Section 5 of NUREG-0897, l Blanke,ts," was submitted to describe relative Rev. 1. I efforts in the area of fibrous insulations.*** l l l l l l Letter of July 14, 1983. Letter of February 18, 1985

     *** Letter of November 8,1984 i

(

                                                                          .- 4h5 s

APPENDIX B BACKGROUND AND

SUMMARY

OF MINUTES OF MEETINGS OF COPMITTEE TO REVIEW GENERIC REQUIREMENTS (CRGR) REGARDING UNRESOLVED SAFETY ISSUE (USI) A-43 RESOLUTION (CRGR MEETINGS NOS. 26, 28, AND 66) ] i 1 i i i NUREG-0869. Revision 1

A BACKGROUND AND

SUMMARY

OF MINUTES OF MEETINGS OF COMMITTEE TO REVIEW GENERIC REQUIREMENTS REGARDING UNRESOLVED SAFETY ISSUE A-43 RESOLUTION CRGR MEETING NOS. 26, 28, AND 66) BACKGROUND The staff's proposed resolution of Unresolved Safety Issue (USI) A-43,

 " Containment Emergency Sump Performance," was sent to the Committee to Review Generic Requirements (CRGR) on October 27, 1982 and was discussed in meetings with CRGR on November 24, 1982 and December 21, 1982. The December 21, 1982 CRGR minutes state that CRGR agreed with the staff's findings and proposed changes to Standard Review Plan Section 6.2.2, " Containment Heat Removal Systems," and Regulatory Guide 1.82, " Sump for Emergency Core Cooling and Containment Spray Systems." However, CRGR agrees only with " forward fit" implementation. The CRGR minutes cite the Deputy Executive Director for Regional Operations and Generic Requirements (DEDROGR) staff analyses. These question four key assumptions in the Office of Nuclear Reactor Regulation (NRR) calculations of averted public dose and state that the DEDROGR staff feels that the dose is high by a factor of 100. In conclusion, the CRGR recommended that the NRR staff review these risk-reduction calculations, re-affirm or revise the proposed lackfit actions, and then meet again with CRGR.

In response to the CRGR recommendations, the staff made additional calculations to estimate the frequency of large loss-of-coolant accidents. These calculations were based on a detailed piping and break probability analysis and estimates of the percentage of these breaks that could lead to sump screen blockage. The results of these calculations are in NUREG/CR-3394, which was published in July 1983. These findings, along with public comments received during the for comment period for USI A-43 (May-June 1983), were used in revising NUREG-0897 and NUREG-0869. NUREG-0869, Revision 1 B-1

l i A third meeting was held with the CRGR on July 11, 1984. The sumary minutes of Meeting No. 66 pertaining to USI A-43 are those noted as Enclosure 3 to the minutes for CRGR Meeting No. 66, which are included in this appendix. CRGR's views are as noted in this enclosure. After the July 11, 1984 meeting, the staff again revised the proposed resolution of USI A-43. t 1 r NUREG-0869 Revision 1 B-2 !

                                                                                                    . u-

SUMMARY

OF CRGR MEETING NO. 26 (November 24,1982)* The CRGR met on Wednesday, November 24, 1982, from 1:00 - 6:00 p.m. S. Hanauer, NRR presented for CRGR review the NRR recomendations to resolve USI A-43, Containment Emergency Sump Performance. The overall safety concern embodied in USI A-43 is related to the capability of the containment emergency sunp to provide an adequate water source to sustain long-tenn recirculation cooling following a large LOCA. The problem can be subdivided into (a) sump hydraulic performance, (b) LOCA-generated debris effects, and (c) recirculation pump performance under

post-LOCA conditions. Each has b2en studied by NRR and the technical findings are reported in NUREG-0897 and associated references. With this view, NRR proposed the following actions: (1) Revise the NRC Standard Review Plan (SRP) Section 6.2.2, " Containment Heat Removal Systems," and Section 6.3, " Emergency Core Cooling Systems." Issuance of the proposed revisions to the SRP is needed to correct previous sump review criteria that are not supported by current findings from full-scale sump tests and generic plant studies (i.e., judgment of sump hydraulic acceptability principally on vortex formation). (2) Revise Regulatory Guide (RG) 1,82 to reflect the findings in NUREG-0897, " Containment Emergency Sump Performance," to incorporate the results of 2 years of sump testing and generic plant studies and to correct deficiencies such as the 50% screen blockage criterion. Generic plant calculations addressing LOCA-generated debris effects have shown that the 50% blockage criterion can be excessive in some plants and nonconservative in other plants. Continued use, witnout revision, of this regulatory guidance would permit the sump designer to bypass the

Verbatim copy.

NUREG-0869, Revision 1 8-3

n.w need to assess debris blockage effects and to continue to show that a 50% blocked screen does not result in excessive head loss. Appendix A has been included in the proposed revision to RG 1.82 to provide guidance and criteria for assessing sump hydraulic performance, LOCA-induced debris effects and pump performance under adverse conditions. A combined consideration of these three aspects is necessary to determine overall sump perfomance and acceptability with respect to assurance that adequate pump NPSH margin will exist. (3) Operating plants should assess the extent of debris blockage potential and, based on the outcome of plant assessments, action should be taken ~ to modify the sump screens or to replace all fibrous insulation with encapsulated insulation. The Committee cocinended the staff for the thorough technical analysis described in NUREG-0897 and agreed with recomendations (1) and (2) above, which reduce requirements on future OL applicants. In support of recommend-ation (3), NRR presented cost-benefit analyses which showed the benefits (using $1000 per person-rem averted), outweighed the costs of the proposed requirements in (3) for operating plants. The Committee suggested that the

 . benefits (reduction in core melt probability) appeared to be overstated by at least a factor of 10, and perhaps 100, and that the costs appeared to be understated. Thus, it was not clear to the Comittee that recomendation (3) could be justified on a cost-benefit basis, even though it was acknowledged to be good engineering practice to replace unencapsulated fibrous insulation with encapsulated insulation.

NUREG-0869, Revision 1 B-4

In response to a question whether the staff has considered the effects of paint debris on sump performance, NRR said they had not considered it in the context of USI A-43, but they agreed to review what consideration had been given to paint debris in previous staff reviews. The Committee decided to discuss USI A-43 in a subsequent meeting after information on the potential effects of paint debris has been received from NRR. NUREG-0869, Revision 1 8-5

t l l

SUMMARY

OF CRGR MEETING NO. 28 (December ?1,1982)* The CRGR met with respresentatives of NRR to further pursue questions , regarding USI A-43 Containment Emergency Sump Perfonnance. The CRGR, during Meeting No. 26, had questioned the potential for sump blockage due to paint

removed from containment surfaces during a LOCA. The question of the potential for sump blockage due to paint rernoval and transport to the sumps was addressed in a memorandum from H. Denton to V. Stello dated December 16, l

1982. The NRR position on the paint blockage issue was that: l l

l (1) Analyses indicate that there 1: not a basis for concern as a generic safety issue; I (2) The issue will be further evaluated within established NRR procedures for treating proposed new generic issues, to detennine the priority for further evaluation; (3) The possible issue of paint removal therefore should not delay obtaining industry and public coment on the defined A-43 issue. l The CRGR accepted the NRR position on the paint blockage issue. THe CRGR addressed the level of risk reduction, or benefit, to be obtained l from the analyses and potential modifications proposed to be required of the ! several licensees that might be found to have combined insulation / sump designs that could lead to failure of long-term recirculation cooling. The Comittee (as reflected in the minutes of CRGR Meeting No. 26 November 24,1982) has agreed with the forward-fit aspects of the NRR proposed requirements. A revised Standard Review Plan Section 6.2.2 and a revised

Verbatim copy.

NUREG-0869. Revision 1 8-6

                                                                                    .ax tv Regulatory Guide 1.82 would incorporate changes in design criteria that would provide greater assurance of sump performance, but would be imposed only on Operating License and Construction Permit applicants filing Final or Preliminary Safety Analysis Reports at some time after the effective dates of the revised Standard Review Plan Section and the revised Regulatory Guide.

To support the proposed backfit requirements, NRR provided a generic value/ impact assessment comprised of a probabilistic risk analysis of the effects of loss of sump function and estimated costs of the backfit requirements proposed for licensees to reduce the risks of such loss. The prcbabilistic risk analyses resulted in an expected value of offsite public dose (person- ~ rems) that could be averted from the estimated six to ten plants that are expected to need modifications. Key assumptions in this NRR analysis are: (1) The expected value of large LOCA (greater than 6" diameter pipe) incidence is 10'4 per reactor-year. (2) For those plants having fibrous insulation that could potentially result in sump blockage, it is assumed that 50% of all LOCAs in piping greater than 6" diameter will result in complete failure to pump any water from any containment surp. (3) The assumed failure of recirculation flow (from sump) is assumed to conditionally fail both reactor building spray and emergency core cooling, thereby leading to a core melt with containment failure by overpressure. No credit was given for potential beneficial operator action to prevent sump blockage by throttling the emergency core cooling system pump or to utilize alternate water s'ources and systems to prevent either core melt or loss of containment function. Thus, for the class of plants above, the NRR analysis assumed the core melt frequency for this LOCA sequence is 5 x 10-5/Rx-Yr. NUREG-0869 Revision 1 8-7

(4) The offsite consequence model used to predict expected values of population dose assumed an average site, a 50-mile radius, and no evacuation of population during the accident. An analysis by the OEDROGR staff indicated that each of the assumptions above was probably too conservative and that the NRR predicted value of averted public dose of about 65 person-rems per plant per year was too high by a factor of at least 100. If this were indeed the case, the proposed implementation plan actions would not appear to be justified. The CRGR recomended that NRR review its risk reduction analysis in light of the analysis performed by the OEDROGR staff with the objective of developing the most realistic assessment of averted radiological dose. NRR should then reaffirm or revise the proposed backfit actions, and discuss with CRGR again if they believe the cost benefit analysis justifies the proposed backfit actions. NUREG-0869, Revision 1 B-8

( , SupmARY OF CRGR MEETING NO.66 (July 11, 1984) Enclosure 3 to the Minutes for CRGR 14eeting No. 66 CRGR REVIEW 0F THE PROPOSED RESOLUTION TO UNRE50LVED 5AFETT 155UE A-43

                             **CONTAINMtNT EMERGENCY SUMP PERr0RMANCE" l             -

l The NRR Division of Safety Technology representatives T. Speis, F. Schroeder, K. Kniel and A. Serkiz presented the proposed resolution for CRGR review. The package submitted for review was transmitted by a memorandum dated June 14, 1984 from H. Denton to V. Stello, Jr; it included the following documents:

1. Sumaries of USI A-43 Peferences.

2. NUREG 0897, Revision 1. March 1984, describing the technical findings of the effort.

3. Regulatory ' Guide 1.82, Rev.1. May 1984, " Sump Design and Water Sources fcr Emergency Core Cooling."

l 4 Standard Review Plan Section No. 6.2.2, Revision 4. " Containment Heat l Removal Systems." t

5. NUREG 0869, Rev. 1. USI A-43 Regulatory Analysis, containing a value/ impact analyses, sumary of public coments received and action taken, and a proposed generic letter for implementation of R.G.1.82, Rev. 1.

6. Draft Generic Letter, subject: " Assessment of Available NTSH Marpn for Long Term Cooling."

7. Note to A. Serkiz from B. Shields, April 3, 1984,

Subject:

Generic Letter on Containment Emergency Sump Perfonnance. l The NRR presenters at the CRGR meeting also provided a handout titled "US! A-43, Containment Emergency Sump Performance," which is attached. CRGR was requested to recorrvnend to the E00 approval of the following final actions:

1. Issue NUREG-0897, Revision 1 as the technical findings for resolving (fSI A 43;

2. Issue Regulatory Guide (RG) 1.82, Revision 1 and SRP Section 6.2.2, Revision 4 for guidance and use in the OL and CP review cycle as part of the normal review process.

3. Issue a generic letter to all LWR licensees and applicants pursuant to 10 CFR 50.54(f). This letter would request a plant-specific essessment of NUREG-0869 Revision I g, ,

(__

    . post-LOCA debris bl.ockage effects (using guidance prov4ded in Appendix A of RG 1.82, Revision 1) on net positive suction head (NPSH) margin, and would request the responders to submit the calculated NPSH margin available, and a description of any plant modifications shown to be necessary by this assessment.

The staff did not propose a generic solutic.n to the variety of potential deficiencies that were postulated to exist among licensees. The proposed generic backfit requirement was for licensees to complete an assessment and report to NRC using the staff specified criteria. Further staff / licensee interaction to define and approve acceptable solutions to design / operational deficiencies would be pursuec on a plant-specific basis. Five to 20 plants were expected to be in this category. - The safety rationale for the proposed requirements was that the containment recirculation mode of long tenn decay heat removal following a LOCA must be assured. The proposed backfit would not provide substantial additional protection over that thought to exist, but would assure a level of safety previously thought to exist. The backfit could in most cases be limited to analysis to verify the sump / pump perfonnance. Staff review and followup actions would be limited to those plants with identified problems. CP and OL applicants would be required to demonstrate adequate design margin for NPSH during the licensing review, in accordance with the SRP and Regulatory Guide revisions presented in the CRGR package. The staff proposal originally presented to the CRGR at two meetings in Neverber and December 1982 was issued for public cocinent in early 1983. The current proposal includes revisions made as a result of public cocinent. In addition, a more sophisticated analysis was ccmpleted to assess the likelihood of sump blockage than was presented in support of the initial proposal in late 1982. The current proposal was presented as based on (1) A review of expected LOCA probability as a function of pipe size, weld type, and joint configuration. (2) An examination of the Salem-1 plant containment layout and selection of 238 locations in pipe as those expected to represent all LOCAs significant with respect to then current staff guidelines for selection of postulated breaks in high energy pipe within containment (SRP Sections 6.2). (3) A mathematical model describing the volume of insulation removed by LOCA as a function of break size and location relative to adjacent insulated pipes and vessels. (4) Investigations of the likelihood of transport of stripped insulation to the sump. Five operating plant layouts were modeled analytically to evaluate the water velocity through expected pathways for transport of insulation. NUREG-0869 Revision 1 8-10

                                                                                                                 .. u

' (5) minimum debris. water velocities necessary to transport (6) Calculations of pressure loss across typical sump screens due to insulation deposition on the screens. A value-impact analysis was presented based on the calculations of sump probability

          .public   coment  and   cost data, period.         much of which was supplied by industry during the Core melt probability was i

sump failure probability, based on an SA! report tyssumed 1982. to be identical with completed in September ! To calculate net safety benefit from reducing this risk, the assumption was of 1. also made that given a core melt, the containment fails with a proba ' releases comensurate with the PWR 2 releas,e cateThis containm Calculation of offsite cose commitment in per_ son gory of WASH 1400. rem was cone using the CRAC l code all and a site population and meteorology selected to represent an averag plants. l The value-impact results produced by the staff are sumarized on page 9 of the attachment to this enclosure! The staff pointed out that the value/ impact ratio was only moderate and that for most plants no risk reduction was anticipated. bases in the drafoft generic 10 CFR 50.46 (b)(5) mandated the confimatory an letter. An additional concern voiced by the staff was the observed that are more recent likelytendency to block aof utilities to change insulation materials to types sump.

I The following major points were made in the discussion of the staff propesal at this meeting: 1. The application of the initiating event probabil:ty data in Table 1 Appendix B, page B-4, HUREG 0869, was considered unclear by the CRGR. The application of the data was called into question because the specification of 238 postulated break locations seemed to be a low number of locations relative1.0CA. pressure to the total length of piping in containment subject to a high The tabulated probabilities as a function of pipe size were intended to represent the population of all pipe in containment subject to LOCA failures. Further, the Comittee was concerned that the selected break locations were such that the probability of pipe displacement to give the postulated deutle ended or even single ended flow model used to predict insulation removal was much lower than the probability of pipe rupture given in the 1/ Ferreli, W.L.

~~

l Applications lnc., September 1982et al, "Probabilistic Assessment of US! A-43", Scie Table on Page 9 of a.ttachment is duplicated on Page B-14 NU9EG-0869. Revision 1 B-Il

                                          . tabulated data. The tabulated data is representative of a data base that includes all kinds of disruptive failures in pipe, not just the sudden explosive rupture with deflection at the break necessary to provide a full flow area blowdown.

2. The CRGR believed that there was considerable but undescribed conservatism in the model for determining, given a pipe break of full flow area, how much insulation is torn off a designated target area or volume. Two such conservatisms would be:

a. The' reduction in (initiating event) probability due to the selection of pipe break direction such that the maximum amount of insulation is targeted.

b. The use of a hemispherical volume of radius 7 x the pipe diameter oriented to contain the most target insulation, and the assumption that all insulation within that volume is stripped from the component

                                                 ,even if, as in the case of a steam generator, the "back" side of the component is not imoinged directly by the break jet flow.

3. The CRGR noted that there was considerable conservatism in the conclusion drawn from the analyses of five plants that all insulation debris torn off comoonents would be transported to the sump. NUREG/CR 2791, where the analyses are reported, reports that in four plants insulation debris arriving at the sump would not approach amounts necessary to significantly degrade sump operation. In one plant, Maine Yankee, material might adversely affect the sump based on the assumption that all material completes the trip to the sump, a scenario which seemed unlikely to the report's authors.

4 The CRGR noted further conservatisms in considering the effects of material at the sump, even given that material arrives at the sump in quantities necessary to cause sufficient screen blockage to unacceptably reduce the net positive suction head (NpSH) at the recirculation pump. Two such conservatisms were noted: (a) The assumption that insulation debris distributes unifo mly over the sump screens, (b) The assumption that recirculation flow in the range 6 to 10 thousand gallons per minute is the appropriate design flow recuirewnt for the sump screens. It was stated in the meeting that most plants have flow instruments in ECCS recirculation lines, so operators could be expected to reduce pump flow to the minimum retuired to cool the reactor core af ter vessel refill imediately following blowdown af ter a large LOCA. This flow requirement is likely to be in the range of 300 to 1000 gallens per minute, considerably less than the flows postulated in the NRR proposal. Since sumo blockage according to the NRR presentation is not expected in less than 1 2 hours, the les>er cooling flow requirements are highly likely. NUREG-0869. Revision 1 8-12

, s .

5. The CRGR noted prob.able conservatism in the postulated WASH 1400 release l categories used in the analysis. It was suggested that the release source terms associated with an early, energetic, above ground containment

! failure such as the WASH 1400 PWR-2 category would be excessively i conservative. Evidence pointed to at most a delayed core meltdown with eventual core melt through the containment base mat, resulting in releases no higher than that associated with a PWR S or 6 release.

6. The CRGR overall consensus was that given the questions raised about the assumptions and/or levels of conservatism of the analyses, NRR's position that the proposals based on the analysis are of only moderate benefit l

versus cost importance, the pecbability that further clarification or resolution of CRGR concerns may suggest a lower safety benefit than does the current proposal, and the estimated applicability of the proposed i solution to only a few plants, all combine to argue against approval of a b4ckfit requiring all licensees to expend resources to demonstrate that their designs are acceptable. Conclusions The Comittee concluded the following as a result of its review of the material transmitted and the meeting discussion:

1. The proposed requirements package was rated of medium importance by NRR.

CRGR review of the infomation presented on the safety benefit to be achieved resulted in uncertainty about the validity of the analyses, with the possibility that the proposed potential risk reductions may be far overestimated and would apply to relatively few plants.

2. The proposed backfit requirement on operating reactors to analyze containment sump perfomance and report to the NRC should not be promulgated at this time due to the uncertainties raised in item 1 above.

3. The CRGR would be willing to reconsider this proposal or a modified proposal by NRR at a regularly scheduled meeting af ter receipt of responses t,n the issues raised at the meeting and discussed in the meeting suma ry.

I NUREG-0869. Revision 1 8 13

                                                                                         %s VALUE/ IMPACT OVERVIEW PLANTS W/0 (CATEG,0RY A):

ANALYSIS COST = $45K/ PLANT AVERTED RELEASE = 0 MAN-REM /Rx PLANTS W/SOME BACKFITS (CATEGORY B): ANALYSIS COST = $ 65K/ PLANT BACKFIT COST = $300K/ PLANT

                                    $365K/ PLANT PLANTS W/ INSULATION REPLACEMENT (CATEGORY C):

ANALYSIS COST = $ 85K/ PLANT BACKFIT COST = $820K/ PLANT

                                    $905K/ PLANT AVERTED RELEASE =       650 MAN-REM /RX INDUSTRY DISTRIBUTION:

AVERTED ASSUMED NO. COST RELEASE V-1 RATIO CATEGORY OF PLANTS ism) (MAN-FEM) (MAN-REM /SM) A 90 4.1 0 0 : B 15 5.5 9,750 1,770 C __5 _gii 3,250 720 TOTALS: 110 14.1 13,000 920 l l l NUREG-0869. Revision I g.14

l . . l t l l REFERENCES H. R. Denton to V. Stello, Jr., memorandum dated October 27, 1982, "CRGR Review of Proposed Revisions to SRP Section 6.2.2 and RG 1.82 and the Supporting Technical Infonnation Document NUREG-0897, as related to USI A-43, 'Containnent Emergency Sump Performance.' l V. Stollo, Jr. to W. J. Dircks, memorandum dated December 10, 1982, l " Minutes of CRGR Meeting Nurber 26." l l l H. R. Denton to V. Stello, Jr., memorandun dated December 16, 1982, l " Potential Sump Screen Blockage Due to ' Paint Sheets' (Ref. CRGR Heeting l of 11/24/82 on USI A-43)." V. Stello, Jr. to W. J. Dircks, memorandum dated January 11, 1983,

                                                               " Minutes of CRG9 Neeting Number 28."

H. R. Centen to V. Stello, Jr., memorandun dated February 28, 1983, l " Response to CRGR Conenents on USI A-43." l V. Stello, Jr. to W. J. Otrcks, memorandum dated July 24, 1984, " Minutes of CRGR Neeting Number 66." H. R. Denton to W. J. Dircks, memorandum dated August 20, 1984, " Feedback and Closure: CRGR Meeting Number 66 (RE: PropoeJd Resolution of US! A 43)." i i I I i l 1 f4UREG 0869 Revision 1 0 1$ l

            .                                                                                                       .~

I l APPEN0!X C ESTIMATION OF PIPING FAILURE PROBASILITY l l l l l l l l l I l i l r NUREG 0869. Revision 1 1 I

a . . A. 1 l .

 !                            ESTIMATICN OF P! PING FAILURE PR08A81LITY                     !
)                                                                                            -

1

)        For US! A-43 evaluations, it was necessary to consider a large number of l        potential break locations over a wide range of piping sizes used for an actual reactor coolant system design. This, in turn, necessitated assigning l                                                                                           f piping failure probabilities to each of the potential break locations and

{ j pipe sizes considered. The assigned pipe failure probabilities shown in j ] Table 1 were used as basic inputs for subsequent computerized calculations ( l that weighted these values by the number of welds and the number of piping f l sections, for a particular diameter size category and for the plant piping I oesign utilized (see NUREG/CR 3394). l The pipe failure (rupture) probabilities shown in Table 1 were derived from j the assessments presented by Dr. S. H. Bush in his October 1977 paper

!        entitlec, " Reliability of Piping in Light Water Reactors," IAEA-SM-218/11.
)

This paper included assessment of the validity of piping failure prob- , t

;        abilities cited from various world-wide literature sources and their               l i         applicability to nuclear systess. Bush also evaluated the safety signi-             '

l ficance of reported failures in nuclear piping systems and presented in- l formation that would a110w an estimate to be made of the relative probability ( j of severance because of (1) general faults (e.g., inadequate piping flen- [ j ibility) and (2) various components (e.g., straight piping runs, joints, { tees, and elbows), depending on their size and potential crack orientation, j ! Using those world wide sources and failure emperiences deemed relevant to j nuclear piping systems, Bush reached the following conclusions: ! (1) The failure probabilities for larger sizes of nuclear piping are ! considered to be in the range of 10 4 to 10

per reactor year, exclusive of Intergranular stress corrosion cracking (IGSCC).

i j (2) Small pipe sizes of lesser safety significance have much higher l failure rates. I i t i t j NUREG 0669, Revision 1 C1 j 4 I

I

Table 1 Piping Failure Probability Estimates , Pipe Failure Weld Failure Probability Olstribution Size Diameter Probability Weld Weld Weld (inches) Class (1/Rx-Yr) Type 1 Type 2 Type 3 0 J P W W W y n a e 2 to <6 1 3E-4 0.7 0.15 0.15 6 to < 10 2 4E-5 0.5 0.30 0.20 10 to <16 3 3E 5 0.5 0.30 0.20 16 to 28 4 3E-6 0.5 0.30 0.20 328 5 JE-6 0.5 0.30 0.20 IPj = 3.76E-4/Rx yr l l Weld Type 1 = fabricated and non-standard joints Weld Type 2 = high restraint joints and tees with joints Weld Type 3 = elbows, reducers, and straight piping runs with joints I NUREG 0869 Revision 1 C-2 l

t l l l (3) In boiling water reactors (BWRs), IGSCC can cause failurt rates much l higher than 10 4 per reactor year (10- per reactor year) in piping 4 to 10 inches (102 to 205 mm) in diameter. l (4) Suggested failure mechanisms apply in most instances exclusive of IGSCC. l (5) Catastrophic failure would appear more likely from operator error or l design and construction errors (water hammer, improper handling of ( dynamic loads, and undetected fabrication defects) rather than conventional flaw initiation and growth or fatigue. l 1 l utilizing the failure information that Bush deemed relevant to nuclear plants, an overall piping failure probability of 3E-4/ reactor year can be derived from his paper for pipes greater than or equal to 3 inch diameter. This value is consistent with pipe failure estimates assigned in WASH-1400 for piping in the size range of more than 2 inches to less than 6 inches in diameter. The Bush paper can also be used to develop a failure probability distribution as a function of pipe diameter of about 4E-4/ reactor year frequency, with the relative distributions (as an approximate percentage va? v) being 6 inches to < 10 inches diameter 13%

                                 ,                       10 inchas to < 16 inches diameter 1M 16 inches to 28 inches diameter 1%

Furthermore, the Bush paper indicates that circumferential cracks (if they exist) would be espected to be of greater significance (by about a factor of

2) than axial cracks relative to the rupture probability. High restraint, fabricated joints would also be espected to make a higher contribution to the overall rupture probability than would straight runs of pipe. Therefore, the above percentages can be further redefined, this time as a function of piping joints,etc. The results of these types of considerations are reflected in the piping and joint failure multipliers shown in Table 1, which were utilized in the U5! A 43 sump blockage assessments (see Appendix 0).

l NUR[G 0869. Revision 1 C-3 l

m i . . I I l l In his paper Bush also observed that a well planned program of periodic inspection should dramatically reduce the probability of catastrophic failure l of piping, and he cites various studies that have 5.;gested that inspection benefits could result in reducing estimated failure f requencies by as much as 1 to 3 orders of magnitudes. l Experimental and analytical work based on mechanistic fracture mechantes that l has been done af ter the 1977 assessment by Bush (see, for example, NUREG-1061, Vol 1. August 1984) also indicates that the rupture probability of large-size ductile piping (unaffected by !GSCC) could also be significantly less than the assigned values of Table 1, perhaps by about several orders of l magnitude. The term " leak before break" has been coined from this later work. The sump blockage assessment performed toward resolution of U5! A-43 did not, however, give any empiteit probabilistic credit for this " leak before break" concept. References l l Bush, S. H. , " Reliability of Piping in Light-Water Reactors,' International Atomic Energy Agency, IAEA SM-218/11, October 1377. U. 5. Atomic Energy Commission, WASH-1400, " Reactor Safety Study " October l , 1975 (also issued by NRC as NUREG-75/018). U. 5. Nuclear Regulatory Commission, NUREG-1061, " Report of the USNRC Piping Review Committee, Investigation and Evaluation of Stress Corrosion Cracking in Piping of Bolling Water Reactor Plants," Vol 1, August 1984. NUREG-0869, Revision 1 C4 1

l APPENDIX 0 ESTIMATION OF PWR SUMP FAILURE PROBA8ILITY l l l 1 1 l l NUREG-0069, Revision 1 1 l l I

i . . ( l

 !                                                                                            l i                                                                                            i i

r 1 l l ESTIMATION OF PWR SUMP FAILURE PR08A81LITY l SUMP FAILURE ! l ' i I Sump fallare is defined as a loss of pressurized water reactor (PWR) sump (or i boiling water reactor (8WR) suction intake) capability to provide an adequate I water source and not positive suction head (NPSH) margin to the residual heat L removal (RHR) and containment spray system (CSS) pumps during the period i af ter a loss-of-coolant accident (LOCA) because of the effects of debris '

}          blockage. Stated another way:       does the head loss across a debris-blocked
!          screen or suction strainer exceed the NPSH margin available under zero blockage conditions?

SUMP FAILURE PR08A81LITY ! The sump failure probability (because of debris effects) is a function of: ' (1) The probability of a pipe break or weld failure, because the LOCA is the

initiating event that can destroy insulation. >

l 1 (2) The potential break sizes and locations within containment with respect l to other piping and insulated primary system components (e.g. , steam T

,                generators, pressurizer, pumps, safety injection tanks) because the          i expanding jet will destroy insulation that falls within the jet              l expansion envelope.                                                          j r

{ (3) The types and v entities of insulation employed, because blockage ! I effects will vary with insulation type (i.e., fibrous debris versus l l metallic insulation debris) and the location of such insulation. t i l i (4) The containment layout and sump (or suction inlet) location, which can ; j control debris transport. (Can the sump be directly targeted by the l

break jet, resulting in prompt transport, or does the debris transport occur later because of recirculation flow drag?). l i

(5) The size of debris screens (or suction inlet strainers), because larger ! screens can accommodate larger quantities of debris without incurring . large head losses.

,          (6) Post-LOCA recirculation flow and pump NPSH requirements, which determine       -

I whether a blocked screen situation will result in loss of NPSH margin [ and pumping capacity. ' I Thus, arriving at an estimated sump failure probability becomes a complex and , plant-specific evaluation based on: (1) probabilistic estimates (i.e., pipe l failure probabilities), (2) plant design features, and (3) a deterministic analysis of debris generated, potential transport to the sump, and potential attendant blockage which could lead to loss of NPSH. Such an evaluation ' i I l \ \ i i NUREG 0869 Revision 1 0-1 l i

begins with an estimation of pipe failure probabilities (vhich are a function of pipe size and weld type), followed by an estimate of the volume of debris that can be generated by any break postulated (which is a function of break size, break-to-target locations and possible combinations, and break jet model); debris transport potential; and blocked screen head loss (which is a function of the quantity of debris transported, the available debris screen area, and the post-LOCA recirculation flow rate requirements). The examples that follow are provided to illustrate such evaluations. ESTIMATING PIPE WELD FAILURE PROBA8ILITIES The first step is to estimate the probability of pipe (or weld) failure to calculate the initiating event probability (i.e., LOCA probability). The probabilities shown in Table 1 were estimated by M. Taylor (DEDROGR staff based on his review of Reliability of Piping in Light Water Reactors" by 5. Bush (IAEA-SM-218/11, October 1977); see also Appendix C of this report. They represent the estimated failure probabilities for all piping in a typical nuclear plant for the diameter classes shown. For example, the estimated pipe failure probability of any pipe in the 10- to 16-inch diameter range is 3E-5 per Rx yr, and the failure probability of a fabricated or non-standard weld in this diameter range is 1.5E-5 per Rx yr. To estimate pipe failure probabilities as a function of pipe diameter size and the type of weld (the assumption being that failure would occur at the weld joint), the data shown in Table 1 can be used to calculate a weld failure probability (Pwk), as follows: (Pj )(N n jn + N,W), + N,W),) p , (q ( ) (Xn *jn * *a*ja * *e*je) Where: Pwk = probability of a weld failure in diameter class "k" pipe, weld type weighted P j = probability of pipe break in any pipe in diameter class "j" N n

                                                            = number of welds of type "n" and diameter "k" X

n

                                                            = number of welds of type "n" in diameter class "j" W jn = probability weighting factor for type "n" welds N, a number of welds of type "a" and diameter "k" X, = number of welds of type "a" in diameter class "j" W

p = probability weighting factor for type "a" welds N, e number of welds of type "e" and diameter "k" X, a number of welds of type "e" in diameter class "j" W j, a probability weighting factor for type "e" welds NUREG 0869, Revision 1 0-2

                                                                                           .. p -

i . , i I Table 1 Piping Failure Probability Estimates - ' Pipe Failure Weld Failure Probability Distribution size Diameter Probability Weld Weld Weld (inches) Class (1/Rx-Yr) Type 1 Type 2 Type 3 0 J P W 3 n a "e 2 to <6 1 3E-4 0.7 0.15 0.15 6 to <10 2 4E-5 0.5 0.30 0.20 10 to <16 3 3E-5 0.5 0.30 0.20 16 to 28 4 3E-6 0.5 0.30 0.20

           >28           5           3E-6            0.5            0.30           0.20 IP) = 3.76E-4/Rx yr Weld Type 1 = fabricated and non-standard joints
     ,     Weld Type 2 = high restraint joints and teos with joints Weld Type 3 = elbows, reducers, and straight piping runs with joints NUREG 0469, Revision 1              0-3

The weld failure probabilities that were derived from the failure assumptions shown above and treated algebraically as described in Equation I were used to estimate the LOCA probability, using the Salem 1 plant and piping layout. NUREG/CR-3394 details those analyses. The weld sizes and distributions derived from a typical Selem 1 plant primary cooling piping loop are shown in Table 2. The breaks were assumed to occur at weld locations (following the criteria in Section 3.6.2 of the NRC Standard Review Plan (NUREG-0800)]. The loop analyzed contained 238 welds associated with piping that can be classified as LOCA-sensitive piping. Because the estimated pipe failure probabilities in Table 1 must be distributed on a per weld basis, the first step is to apportion (or redistribute) the total probability for a diametric size class to all existing pipe sizes. Table 3 has been constructed to illustrate how Equation 1 was used to develop such a distribution. If weld type is ignored in the first step, the per weld failure probabilities distribute as a function of the fraction of the number of welds (of a given pipe size) to the total number of welds in a particular diametric class (that is, j = 1, 2, 3, 4, and 5, as shown in Table 3). The summed totals (for a particular j class) must always sum to the total probability for that diameter class extracted from Table 1. Table 3 illustrates this process. In addition, Table 3 compares this type of fractional distribution with the weld type weighted values from NUREG/CR-3394. The variability within a particular diameter class (j = 1, 2, 3, 4, and 5) results from the distribution of weld types in the Salem 1 plant piping layout analyzed (see again Table 2). NUREG/CR-3394 analyses were based on both weld type and pipe segment probability distributions. The initiating event probability (P,) attributable to all 237 welds was calculated to be 3.7E-4/Rx yr and is in agreement with the IP) value shown in Table 1. Thus this type of single loop analysis is applicable to all loops. Had the distribution methodology discussed above been applied to four loops. the same initiating break probability would have been obtained, because the overall probability cannot exceed the summation noted in Table 1. Because satisfactory operation of the sump is essential when breaks occur in the primary coolant system LOCA-sensitive pressure boundary, the overall probability (P )o discussed above should be reduced (for purposes of estimating sump blockage probabilities) to account only for such breaks in the primary coolant system. In many PWRs.(such as Salem 1) such piping is located within the crane wall region and that break probability is designated P.g Therefore, elimination of secondary system piping weld locations and piping outside the crane wall resulted in a calculated Pg of 1.84E-4/Rx yr for Salem 1. These break locations (associated with Pg ) were further analyzed for debris generation as discussed below. NUREG-0869, Revision 1 0-4

Table 2 Break size distribution for Salem plant analysis System Weld Distribution System No. of No. of Welds 9 pipe designation System welds Diameter (inches) i 1 Hot Leg 8 8 9 34 2 Cold Leg 6 6 9 36.3 3 Cross Cver 6 6 9 32.3 4 Safety injection / 41 15 9 10, 4 9 11 cold leg 1106,1102 5 Safety injection / 33 696,2792 hot leg 6 Chemical volume and 13 5 9 16, 8 9 14 control 7 Feedwater 95 22 9 4,703, 66 9 2 i 8 Main Steam 20 19 9 30, 1 9 32 9 Pressurizer 16 16 9 14 Subtotal Welds 238 for loop analyzed

                                                                                                                                                                                                                        -16 for pressurizer
                                                                                                                                                                                                                        -13 for chemical volume and control
                                                                                                                                                                                                                        -33 for safety injection / hot leg 176 x 4 loops             = 704 Pressurizer loop welds                                                                                      = 16 2 Chemical Volume and Control loop welds (13)                                                               = 26 2 Safety injection and hot leg penetrations welds (33) = 66 Total welds                                                                                                                                                                              = 812 for 4 loops Diameter                                                                                   No. of l                                                          (inches)                                                                                  welds 34                                                                                         8 hot leg

36.3 6 cold leg 40 welds for 16 in.- 34 in, pipe 32.3 6 crossover (P = 3E-6) 1

! f 9 Mainsteam . 16 5 ',__4, 29 welds for 10 in.- 16 in. pipe 14 24 ' (P = 3E-5) 10 15 L 19 Selds for 6 in. -10 in pipe 8 4J (PO = 4E-5)

;                                                               6                                                                                   17 '

4 22 , 150 welds for 2 in. 6 in. pipe

3 7 2 104 ,

(P* = 3E-4) l Source: NUREG/CR-3394, Vol 2. Table A.1-1. NUREG-0869, Revision 1 D-5 i

                                                                                                                                                                                                                                                      . 0 DEBRIS GENERATION Estimating debris generation is a function of break size, jet expansion model, and the break versus target locations. For the analyses reported in NUREG/CR-3394, a hemispherical jet expansion region model was selected with the zones of influence that are shown in Figure 1, because the decompression pressure field for a high pressure, subcooled jet can be approximated with a hemispherical model. The energy levels (stagnation pressure level) within this expanding jet are also a function of distance from the break (or length to diameter, L/0). Calculational models that have been correlated with experiments show that at L/D = 3, the jet stagnation pressure is very nearly the same as the stagnation pressure within the jet emanating from the break location. For a PWR, this means pressures on the order of 2200 psi and extreme insulation destruction would take place, particularly for fibrous insulation. At L/D = 7, the PWR subcooled break jet stagnation pressure has been calculated to reduce to 20 to 40 psi. Although this is still a very high velocity field, experiments have shown (see NUREG/CR-3170) that the fiberglass covering shreds rather than totally destructing at these reduced pressures.

For analysis purposes, three L/D ratios of 3, 5, and 7 were selected to assess debris generation effects parametrically. The lower bound (L/D = 3) represents the highest jet intensity, because the expanding jet dynamic pressure at that distance is nearly that at the break jet exit plane and no conservatism exists for survival. The outer bound (L/D = 7) represents an axial distance where jet stagnation pressures have decreased to 20 to 40 psia, and at this distance the assumption that total destruction takes place does carry some conservatism. (See NUREG-0897, Revision 1 for further discussion on the selection of these L/D ratios to represent the different zones of insulation destruction.) In addition, blowdown tests in the HOR facility have demonstrated the highly destructive capabilities of LOCA jets. The weight of evidence is strongly against conceptualizing a high pressure pipe break as a simple water jet model. Such a break can be better termed as "an explosion." These flow rates investigated parametrically represent emergency core cooling system flow rates given in Final Safety Analysis Reports (FSARs) submitted by applicants for operating licenses. The range of debris screen areas evaluated is representative of PWR sump designs. Because there is no standard sump design or ECCS flow requirement, this range was used to scope the range of sump blockage probabilities for PWRs; it is discussed in detail in NUREG/CR-3394. The break location (that is, weld locations) versus target combinations for Salem 1 were systematically evaluated utilizing the plant insulation distribution. (The insulation was approximately half reflective metallic and half encapsulated fibrous insulation.) Table 4 shows the level of detail employed to evaluate all possible break-to-target combinations. The debris volumes associated with 14-inch pipe breaks are used for illustration, because these medium size breaks contribute significantly to the calculated overall sump blockage probability. NUREG-0869, Revision 1 0-6

Table 3 Probability distributions as a function of diameter * (

Reference:

238 welds / loop, Salem-1) (1) for j = 1, Pj = 3E-4 wn wa we k = 6", Nk6 " 1 .

                                                  .7 .15    .15    = 4",

Nk4 = 2R

                                                                   = 3",

Nk3

7

                                                                   = 2",

Nk2 = 104 Nk = 150 welds for j=1 w/o Weld Specification NUREG/CR-3394 Values Pk =2" = (3E-4)(104/150) = 2.07E-4 2.07E-4 P k=3,, = (3E-4)(7/150) = 0.14E-4 0.13E-4 k=4., = (3E-4)(22/150) = 0. 44E-4 0.41E-4 P Pb6" = (3E-4)(17/150) = 0.34E-4 0.38E-4 P3 ,7 = 2.99E-4 2.99E-4 (2) for j = 2: k = 8" and 10", Nk8 = 4 and Nkl0 = 15 w/o Weld Specification NUREG/CR-3394 Values P k=8., = (4E-5)(4/19) = 0.84E-5 0.71E-5 Pk =10" = (4E-5)(15/19) ==3.16E-5 3.29E-5 Pj =2 4.00E-5 4.00E-5 (3) for j = 3: k = 14" and 16", Nk14 = 24 and Nk16 = 5 w/o Weld Specification NUREG/CR-3394 Values Ppg 4. = (3E-5)(24/29) = 2.48E-5 2.47E-5 Pk =16" = (3E-5)(5/29) = 0.52E-5 0.53E-5 P3 ,3 = 3.00E-5 3.00E-5 (4) for j = 4: There was no piping in this diameter range for Sales 1 (5) for j = 5: k = 28", Nk32 = 7, Nk34 = 8, Nk36 ' 0 w/o Weld Specification NUREG/CR-3394 Values Pk =30" = (3E-6)(19/40) = 1.42E-6 0.93E-6 Pk =32" = ( E-6)(7/40) = 0.53E-6 0.88E-6 Pk =34" = (3E-6)(8/40) = 0.60E-6 0.72E-6 Pk =36.. = (3E-6)(6/40) = 0.45E-6 0.47E-6 P),$ a 3.00E-6 3.00E-6

     " Based on 238 welds / loop, as in Salem 1.

NUREG-0869, Revision 1 0-7

g gx v s 1 g

                                                                  /#          \
                                                                                 ). ,.s .
           ,            . tl1I                                                                                         ,
                                                                                                              .             5
                                                                                                                       . o c
                                                                          /.V              ,

v / s ~. J'e/'. r L I

                                                             .[5e       .

W, 7 's 1r L

b" H n P c S Q- o y i n C t a

                                                                                 "V                                                               r
                                                                                                         .'                                       e n

e g s i r

                                                              ,                                                                                  b
                                                                .                                                                                 e
                     *                                                                                                    ,                      d t                  .                                                                                                                      r
                                                 \

l l o

s. ' - - f w \ j \ d I e4

         .                           N'c           G              /,/ ,                                          .

z9 i 3 e

                                                  'Ss s        ~'                                                                                 l 3 i -

3 s , tR P ~ uC C - o

                                                                                              -       p
                                                                                                                -                                    /

4 eG R s v u , . cE nR

D eU ( uN

                                                                                                                   .                            l
                      \

f n

t. ni

                    '                                                                                                                           i
                      -                                           y           -

t.. .

                                                                                                                            )     t.         -       s f e P                              .              ,           L

( .m e [ (RQ,* v

t. 0 g h.

1 ot sm ei a e

i. .

                                                                                                                .            .e .

Tt nt os Z e

                                                                                                 . t.        .            o S

e e. . t y r .. c .b . 1

                                                                                                 . c . ,. u .4     . c
                                                                                                                             .d      .

e r u

t. t. t.v i , g

                                                                                                       . .           t. ti             .

i s s t

                                                                                                                     ~ OD            c          F v               ,           D o

e s s o c. 'a

L c r. w w t gyoRe, =3Eo[ gm

UNACCEPTALLE DEBRIS GENERATION Unacceptable debris generation is defined as that volume of debris" that, if transported to the sump screen, would result in a blockage head loss greater than NPSH requirements. Thus, the debris volume calculated through use of the break-to-target zones (or L/0 ratios) shown in Figure 1 must be compared to the volume that could generate unacceptable screen blockage head losses that exceed the NPSH margins. This effect was analyzed parametrically and is reported in detail in NUREG/CR-3394. Experiments have shown that the head loss for forced flow through fragmented fibrous insulations deposited on a debris screen can be expressed as an exponential relationship (NUREG/CR-2982, Rev.1) and that such head losses are highly dependent on material type. The following empirical relationships for head loss (HL) have been obtained:* HL = 1653 (Q/A)1.84 (V/A)1.54 for high density fiberglass HL = 68.3 (Q/A)1*79 (V/A)1.07 for NUKON HL = 123 (Q/A)1.51 (yjg)1.36 for mineral wool i

.                             In addition, these experiments have shown that shredded fibrous insulation materials distrioute uniformly over debris screens. The data for NUKON were l                             submitted by the Owens Corning Fiberglass Corporation during the For Comment 4

period for USI A-43. The analyses reported in NUREG/CR-3394 have parametrically investigated the following plant design and operational variables: Variable Range investigated t Pipe break-to-target distance 3 to 7 pipe diameters Sump flow rate 6,000 to 10,000 gpm Debris screen area 50 to 200 ft2

                              *Q/A is the approach velocity as calculated from volumetric flow (ft3/sec) divided by debris screen area (ft2) and the equivalent debris thickness (V/A). V/A comes from the transported debris volume (V) divided by the debris screen area (A).

i NUREG-0869, Revision 1 0-9 1

e e

   *eetsu faatti .. Tasett gasvt.                      InsW1. ,evai ,Insut..           _lsatst,_ ,faaett I ss lut . _ , t utuk             g,g. gesus, musst e      etaM.       LLNGlH InjCa.              eTFE          htP. TOtWes            bl&M.       t$m0fM IMICa.             fvPE
      ..                        ft        mCw           **                 ; rs     .t                                                     o.f.f.. TC rt %.:U.u t
                 .I .n .   , . .     .   .s....a.s     .. ..         ..3.                 ...i ss et
                                                                                                .g. ,_ ,.... .,,.I nC  .. H,_

a .s,.... ,,... .... . 9 to te.ee t.se 3.00 st 3 f.t to.48 2.st- 3.es st 3 3.3 _ . Its.ee .. 89 83 . a.se... ,s; _ z estet _ s t ,_. se.se s.ee .,a.ee _ ,,st_ 7 g!._ hee si its.ee is.as -s.se st ,_ 5,_ ,.3:,t,s e.i ,_ jp.sg t .. . . a . : .. . se ie -1Thi I I A'.*ee st- ~~ 's' *-

s . s ~~~ ~

              .14..eee d "*.e. s . ,_d i"-*1..eii*--' t sta         .$       >>e..                                                                                                     9
                                                                                                               .                                             . . . _ . _ _ .           g so       so.se           e.ts       3.ee          st          a            t.s       so.ee          e.se         s.ee        st     a             no.e its.et          to or        3.go          it          3        sine                                                                                                t    Q
                      .                                                                         .                                                                                 A    w

_tep _ to.ee- 19 13 _ 1.to it 2 ' L.f to e* t.te 1 ee it 1 t.e e a e . e. e f.to a.ee st Jo .is.a 4

 -'iii-*ii.ei-*i.iT                        3:ii- II                  a            e.e                                                                                                  b C
  " e 4'a '~~ ii .e 4' ' *""~ i . il        T.ns         !!           a          ti.e n

s e i'~~ ~ i e . e e ~~ e . 6 3~~ 3.bi "**it' a s.1 Q n

                                                                                                                                                       ~ ~ ~ ~ ~ ~

ese ie.ee e ee s.ee ~s t a ei:n 3

ses "* " s e.ee*~~~'i.33-* 5.66 ~ ~ it - a vi se.e i *.

i es "-* i s .ee ' e .e 6 * ~3. ei'** il a s.e ie.ee t.sa .a.ee at a e.i' r"p

                         .                                                                                                                                                            3
                                                                                                                                                                            - w ser        to.ee'           e.se      s.ee           st         :            e.,     sis.ee          ee.et         s.se        st     a           ase.e i.
   ._s e e . _. s . n            nt.._ L.se                ie          n           .i

_see e .ee _.._u s _.nse si_ _ _as .i t.se 2. __ne e i t___; > .i Table 4 Illustration of break vs. target combinations and debris volumes calculated Ref: Table 8.5-3e NUREG/CR-3394 Volume 2 NUREG-0869e Revision 1 0-10

.~

d J 9 Table 5 Pipe Break and Sump Blockage Probabilities vs. Pipe Diameter

                                                                                        .     .. nett e. .                 ..          .

Pi .u. .ef im., t t t.s. 0 in30.i e,. fe. u.it f fsu. ., Scut

                                                            - , . , n.ne,f
                                                                      .       90.,ts 09.

lu ti e.st a s. t. i te 4000, s. 9

                                                 . tote.st ,. .u..ue .-.. St.Caa.( f ue-...moCaset sue... .toCuce P90 ... StoCiast ren.. SteCanet.P.00
       . .tia . PC                 Pi .

L Ovte 0 9 L evte 8 . f L OVit S . 3 L Ovte 0 . S L 0918 9.f

18. L Ovte 9 3 PO GA 91 94 PO #E P4 DE PO Da Pl SE

           -                  -..~

og 9ts P _ t G.Fe. . .O,fWPO , P t ef W......,..~.sm.......-..,-o-..w-_- PO Of4 Pt Off ( 0.0 0.0 4.0 0.0 0.0 0.0 0.0 0.4

3. 3.018-04 f.Get-41 0.0 0.0 0.0 0.0 e.0 9.9. -.0.8 0.0 0.4 a.4 Ost S.A - 4.0

3. 0.0 0.0 0.0 . -0.0 0.0 0.0 8.0 0.0 0.0 0.0 0.0 0.0 0.8

3. 6.344649 0.8 0.0 0.0 0.0 0.0 (

0.0 0.0 0,e e,0 0.0 g.e , e,0 0,0 0.0 0.0 0.e e.0

e. 0.0 0.0 e.0 0.0 = 0. 0 e.0 0. & 4.0 0.0 0.0 a." 8.0 0.E 8

          .4      4.50$=0S 0.0                                                                     0.0            0.0          0.0         0.0         0.4       0.0      0.0           0.0 S. 0.0            0.0         . 0.0            0.0          0.0        0.0                                                    0,0         9.It=0S 1.It*09 I.4t-05 9.et el

8. 3.eSt es 3.0$t=09 0.0 0.' 8 3.f 4t+e9 f.fet*et 3.f tt+04 3.718 09 0.0 J.8 1.0 t*04 3.Ste06

           . e. f.I t t=04 f. Il t*04 0.0 . -                            0.0*       8.0            * .50ts04 3.90Leel.0.*                  4.0         0.0 0.0                                                                                        e. state 0.9t*40 0.01-04 9.9t*06 le. 3.394-09 3.39t=0S 6.358 48 t.39t-et 3.70t*01             0.0 3.70t+0 0.0 B  3.f S.0   ttee   t 3.f 0.0   tt.e4  4.4t=08 0.0        4.44-90 0.0         0.0       0.0       0.0          0.0 St. 0.0            9.0            0.0           0.0
           .8 4. 3.ef t et .3.30t*05. 3.33t*e t. 3. Set.at A.334et t.3.93Ee&L 3.004 0 L.S.f tt=0L. e J Lae4.4.et ma A - L t.e et.9.et.e44 A.f f =4% A.

le. S.34t*e4 3.70L*00 0.00t+44 1.00te64 0.0 9.04(404 0.0 f.006604 0.0 f. Sue 0.0 00 f.00t*00 0.0 0.3t*00 0.0 30.0 78-00 S.38-04 0.0 3.1t*000.0 0.0 9.3t*00 3.f t ce te. 0.0 0.0 0.0 0.0 A .A ** 8.t. a.O a.0- 4.E i

         " 30. 8.0     w   e.0          ..$.0. - e.0                  Ls         * . 0-      -. 4.0W .a.-.-                                                   0.0       0.0          e.0 0.0            0.0           0.0         0.0        0.0

33. 0.0 0.0 0.0 0.0 0.0 0.0 . 0.0 0.0 0.0 0.0 24 0.0 0.0 0.0 0.1 0.0 0.0 0.8 0.0 0.0 0.0 .- e .0 -.e . "

0.0 e .4 *4. *0. 4.* 8.* 8. W .0

30. 0. 0... 0. 9. .- . 4.e . 4. 0 0.0 e.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 9.34*0f 0.e 0.36*07 0.0

a. 26. 0.0 0.048600 0.0l f t e e t.0.004600 0.0 0.0n*00 0.0 0.36*0f 0.0 0 30. 9.30t*0f 0.8 S .80t+0 4.4.816 4 4.Ae0 Le e l.4,t a t.e l.0.3 L.01.2.e ta01 S. 3 t v e f 3.u.48. 0. S t.e t.3. 0 E st*

W

           .33. 8.70t*47 0.38t*ef 0.004-66.4                                                                                    3.et-et 3.0t*ef 3.et*ef 3.44-47 3.84 87 3.8E 07

34. f.34t-et 4.004

30. 4.00t-et f.34t**f ef 9.048-44

3. Set.04 3.3089.04t.*et 9.00t*el 0.074 00 0.e78-48 S.044-4 00 8 S.eet-el 0.00t*0f.f

0.-.

tteel f.f e t=09 3.04-07 t h . W .0 0.03.St+et

                                                                                                                                                                           * . S3.8- 8-470.* 3.t t-ef 3.St*4f 3.S 0.0          e.0        8.*    --

A 4 . 0.0

           .30. 0.0          0.0 .-.         0.0                                                                                                      0.0       0.0      0.0 0.0            0.0          0.9         0.0 et. 0.0           0.0            0.0           0.0          0.0        0.0 0.0         e.0         0.0       g.0      0.0           0.0 e2. 0.0          0.0             0.0           g,g          0,0        0.0            e.0*           0,0 9.a . . a,e        e. g.,,,,     a_^

e,s m.0 *m.= 0.0 .0,b -a *

           .44. 4.g.          e.e            0.e .         g.0 0.0                      0.0         0.0       0.0       0.0          0.0

48. 8.0 0.0 0.0 0;9 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.8 4.0 0.0 0.0 0.0 0.0 0.8 4.* ^^

de. 0.0 . 0.0 . ,

                                                                                                    ^4.            * .sk       -e.0 - .e.0 w .6                  A A.-                      _
           .00     0. 8 -       0. 0 . .. 0.0             L 0.0           4a         '_^

0.0 e.s e.9 0.0 0.0 0.0 0.0 0.0 e.e 4.0 0.0 0.0 0.0 St. 0.0 0.0 0.0 0.0 0.0 0.0 3.06-e9 t.St.e5 3.St-OS 3.38-41 3.0E-OS 4.lt-00 101 3.734.e4 9.704-e4 0.e Be3mtmCLAfuet 0vtsatt Pe00aetLiff 0F tulf, tytet OCCuestesCt 084 lulfl Afim0 tytet elauttle*lNC.sts pg ,, 88 .Pe00.* Intf. Evtuf.es ta. 5755-4m.Stamt WA'? 06ScaaCt.JeE0 0Af te.08 4statt.4awlise-eLOCm" St0Cuact Pe00-P90eatttiff 0F LasaCCEPfattt . 70 70f AL twistslutleHitD tallS3 (CCS 30m# St0Caaet L Ovia S . flectf etStassCt 80 Getas 01amette Safl0 8W.attOCs Act. F0tSUtNCT. utis f fPt.wei0854e6 4A01S.--a8b .0becaA44.e ete 9EGada5F0s 4saaef pelasa lae4ienf Ce Pect 4me.Oa$1* PO 70s Ovteatt pece Asse PI F0s Palm In Ce PeOS y F04 OftestL P90s Ame el 94 +0tOCEA06 PeOS PO FDe OvttaLL PB F0e Pelm Ill (W . . . . . . . , Source: NUREG/CR-3394, wh'ich contains the complete calculated data sets . i NUREG-0869, Revision 1 0-11

The analyses reported in NUREG/CR-3394 (as illustrated in Figure 4) considered each singular break-target combination to determine debris generation volumes. Each possible combination was considered on a singular weld / target basis for determining if that particular break probability should be retained for estimating overall sump blockage probabilities. The criteria for maintaining a particular break were based on a calculated head loss of 1, 2, or 5 feet of water because these evaluations were performed parametrically. Table 5 shows the probability distribution as a function of pipe diameter for the following: sump screen area of 50 ft2; recirculation flow rate of 8000 gpm; and an allowable head loss of 1 foot of water. The estimated sump blockage probabilities are 1.8E-5, 3.3E-5, and 4.5E-5/rx yr for L/D = 3, 5, and 7. The calculations reported in NUREG/CR-3394 considered insulation with debris associated both with piping and targeted compartments. They are admittedly , conservative because shadowing effects attributable to piping supports and other structures are not included. To illustrate the sensitivity of such an assumption, a series of calculations have been performed to assess the d significance of insulation only on failed piping, and as a function of diameter size, thereby ignoring other targets. These results are shown in Table 6. These estimated insulation volumes and associated head losses show that smaller diameter primary coolant system piping (6- to 10-inches), for the flow and debris screen area assumptions shown in Table 6, will not generate sufficient quantities of fibrous debris to result in head losses exceeding 1 to 2 feet of water, thus they support the conclusions that can be derived from Table 5. These findings would be most applicable to plants having small NPSH margins and small screen areas. The analyses reported in NUREG/CR-3394 were run with the high density fiberglass head loss correlation, and it was assumed that the insulation (for L/D = 3, 5, and 7) was totally destroyed, was transported to, and was deposited on the sump screen. The assumption of total transport is an imbedded conservatism for plants that have recirculation flow velocities less than 0.2 ft/sec within the containment and for those plants where intervening structures would inhibit transport. On the other hand, there are PWRs where the primary coolant system piping and the sump location are not isolated by intervening structures. SUMP FAILURE PROBABILITY When the calculational methods described above were applied to the plant parameters investigated they produced the range of estimated sump failure probabilities shown in Table 7. (More detailed tables are provided in NUREG/CR-3394.) For high flow rates (10,000 gpm), small debris screen area (50 f t2), L/D = 7, and low allowed head loss (1 ft of water), the calculated sump failure probability was 5.4E-5/Rx yr. For values of 6000 gpm,100 f ta, and L/D = 3, a NUREG-0869, Revision 1 0-12

r a Table 6 Headloss versus debris volume versus pipe size PIPE INSUL VOLUME SCdEEN RECIRC HEAD (1) HEAD (2) DIAPETER THICKNESS L/D DESTROYED AREA FLOW LOSS LOSS Inches Inches Cu Ft Sq Ft gpm Ft Water Ft Water 36.30 3.50 3.00 27.58 50.00 10000.00 i49.45 8.50 36.30

3.50 5.00 45.97 50.00 10000.00 328.21 14.69 34.30 3.50 7.00 64.35 50.00 10000.00 551.04 21.06 34.00 3.50 3.00 24.34 50.00 10000.00 123.29 7.44 34.00 3.50 5.00 40.57 50.00 10000.00 270.74 12.85
34.00 3.50 7.00 56.79 50.00 10000.00 454.56 18.42 32.30 3.50 3.00 , 22.07 50.00 10000.00 106.07 6.70 32.30 3.50 5.00 36.79 50.00 10000.00 232.93 11.58 32.30 3.50 7.00 51.51 .50.00 10000.00 391.08 16.59 16.00 3.00 3.00 4.97 50.00 10000.00 10.69 1.36 16.00 '3.00 5.00 8.29 50.00 10000.00 23.47 2.35 16.00 3.00 7.00 11.61 50.00 10000.00 39.41 3.37 14.00 3.00 3.00 3.89 50.00 10000.00 7.33 1.05 14.00 3.00 5.00 6.49 50.00 10000.00 16.10 1.81 14.00 3.00 7.00 9.09 50.00 10000.00 27.04 . 2.59 10.00 3.00 '3.00 2.13 50.00 10000.00 2.89 .55 10.00 3.00 5.00 3.55 50.00 10000.00 6.31 .95 10.00 3.00 7.00 4.96 50.00 10000.00 10.65 1.36 8.00 3.00 3.00 1.44 50.00 10000.00 1.58 .36 8.00 3.00 5.00 2.40 50,00 10000.00 3.48 .62 8.00 3.00 . 7.00 3.36 50.00 10000.00 5.84 .89 4.00 ~3. 00 3 00 .88 50.00 10000.00 .75 .21 6.00 3.00 5.00 1.47 50.00 10000.00 1.64 .37 6.,00 3.00 7.00 2.06 50.00 10000.00 2.75 .53 6.00 1.50 3.00 .37 50.00 10000.00 .19 .08 6.00 1.50 5.00 .61 50.00 10000.00 .+3 .14 6.00 ,1.50 7.00 .86 50.00 10000.00 .72 .21 -

2.00 1.50 3.00 .06 50.00 10000.00 .01 .01 2.00 1.50 5.00 .10 50.00

10000.00 .02 .02 2.00 1.50 7.00 .13 50.00 10000.00 , .04 .03 (1) H= 1653 ( (Q/ A) ^1. 84 8 (V/ A) ^ 1.54 (2) High Densi tyFiberglass, Low Density Fiberglass, H=68.3((Q/Al ^1.79t (V/A) ^1.07 NUREG-0869, Revision 1 0-13 L- -- _ - -. _ _ _

Table 7 Sumary of the probability of sump failure RECIRC SCKEEN ALLOWED CALCULATED PftCBABILITIES FLDW RATE AREA HEAD LOSS (BASED ON WELD TYPES) (GP11) (SQ FT) -(FT Hg0) L/D=3 L/D=5 L/D=7 6000 50 1.0 1.le-5 1.7e-5 2.8e-5 6000 75

1.0 5.7e-6 1.2e-5 1.8e-5 6000 100 1.0 3.te-6 6.7e-6 1.Se-5 6000 200 ,

                                              , 1. 0       2.9e-6   5.9e-6     6.9e-6 6000        50    . 2. 0       9.9e-6    1.2e-5   2.4e-5 6000        75         2. 0      3.te-6    9.2e-6     1.8e-5 6000      100        .2.0        3.Ie-6    5.9e-6     1.4e-5 6000      200          2.0       2.9e-6    5.8e-6    6.9e-6
   . . c.                6000        50         5.0       4.7e-6    1.2e-5    1.Ge-5 6000        75         5.0       3.1e-6    5.9e-6.. 1.5e-5               -
            .            6000      100         5.0        3.le-6    5.9e-6    7.7e-6 6QOO     200          5.0              0   5.3e-6    6.9e-6            ..

8000 50 1.0 1.6e-5 3.3e-5 4.5e-5 . 8000

75

                                  .             1.0       9.le-6   .1.2 e- 5  2.4e-5 8000       100          1.0       3.9e-6    1.2e-5    1.Be-5                             -

8000 '200 1.0 2.9e-6 5.9e-6 7.7e-6 8000 50 2.0 1.te-5 1.7e-5 2.7e-5 8000 75 2.0 4.7e-6 1.2e-5 1.Be-5 8000 .100 2.0, 3.te-6 6.7e-6 1.6e-5 8000 200 2.0 2.9e-6 5.9e-6 . 6.9e-6 8000 '50 5.0 7.4e-6 1.2e-5 1.9e-5 8000 75 5.0 3.te-6 6.7e-6 1.7e-5 8000 100 5.0 3.1e-6 5.9e-6 1.32-5 - 8000 100 5.0 2.9e-6 5.7e-6 6.9e-6 . 10000 50 1.0 1.6e-5 4.le-5 5.4e-5

10000 75 1.0 9.9e-6 1.2e-5 2.4e-5 10000 100 1.0 4.8e-6 1.2e-5 1.Ge-5 10000 200 1.0 3.1e-6 5.9e-6 9.4e-6 l 10000
50 2.0 1.le-5 1.Ge-5 4.3e-5 10000 75 2.0 'J.4e-6 1.2e-5 1.8e-5 l 10000' 100 2.0 3.te-6 9.2e-6 1.Be-5 l

10000 200 2.0 2.9e-6 5.9e-6 6.9e-6 10000 . 50 5.0 9.9e-6 1.2e-5 2.4e-5

 .                    10000         75        5.0       3.le-6     9.2e-6    1.Be-5 10000      100          5.0      3.le-6      5.9e-6    1.4e-5 10000      200-         5.0      2.9.e-6     5.8e-6    6.9e-6 e

NUREG-0869, Revision 1 0-14

.A

. a blockage probability of 2.9E-6/Rx yr is calculated. Because insulation change out is an ongoing plant activity, the staff does not specifically know the types and quantities of insulation employed. Therefore, calculation of a generic value is not possible. References Bush, S., " Reliability of Piping in Light Water Reactors," International Atomic Energy Agency, IAEA-SM-218/11, October 1977. U. S. Nuclear Regulatory Commission, NUREG-0800, " Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants," July 1981.

      -- , NUREG/CR-2982, Revision '1, " Buoyancy, Transport, and Head Loss of Fibrous Reactor Insulation," D. N. Brocard, Alden Research Laboratory, July 1983 (also Sandia National Laboratory, SAND-82-7205).
      -- , NUREG/CR-3170, "The Susceptibility of Fibrous Insulation Pillows to Debris Formation Under Exposure to Energetic Jet Flows," W. W. Durgin, and J. Noreika, Alden Research Laboratory, January 1983 (also Sandia National Laboratory, SAND-83-7008).
      -- , NUREG/CR-3394, "Probabilistic Assessment of Recirculation Sump Blockage Due to Loss of Coolant Accident," J. J. Wysocki, Burns and Roe Inc.,

July 1983 (also Sandia National Laboratory, SAND-83-7116). NUREG-0869, Revision 1 0-15

P .* APPENDIX E CONSEQUENCES OF LOSS OF RECIRCULATION CAPABILITY

CONSEQUENCES OF LOSS OF RECIRCULATION CAPABILITY 4 Risk analyses were utilized to assess the public health consequences associated with a loss of recirculation flow capability in the period after a loss-of-coolant accident (post-LOCA period) as a result of the effects of debris blockage. Blockage of a sump in pressurized water reactors (PWRs) and blockage of residual heat removal (RHR) suction inlets in a boiling water reactor (BWR) have similar consequences since loss of the redundant recirculation systems can result in core uncovery, which will lead to core melt. In addition, containment sprays will fail if a loss of suction takes place and containment overpressurization can occur. Loss of containment i structural integrity leads to high public doses. To obtain initial estimates , of public exposure, we referred to probabilistic risk assessments of five plants: Surry, Calvert Cliffs, Crystal River-3, Peach Bottom and Grand Gulf. The accident sequence of interest for the PWRs is designated AHF; a large LOCA (A) followed by failure of recirculation to the core (H) and to the containment sprays (F). For the BWRs, analogous sequences designated AHI, AI and AGHI were used as the basis for our evaluation. i All five PRA's are based on the RSS methodology. Consequently the conditional probabilities for various containment failure modes were approximately the same for all three PWRs. A synopsis of conditional l probabilities of source terms is shown in Table 1. Table 1 shows that 70 to 80% of the AHF sequences lead to release category PWR-6 (basemat j meltthrough), and 20 to 30% lead to PWR-2 (early overpressure failure). There are also small conditional probabilities of PWR-1 (steam explosions) i and PWR-4 (failure to isolate containment).

For the two BWR plants (using RSSMAP methodology), about 1% of the events are expected to result in BWR-1 releases (steam explosions), and about 80% would 4

lead to BWR-3 (radiation release to the reactor building). For Peach Bottom, j - 1 j l i NUREG-0869, Revision 1 E-1

                   -.                                                ._                              -.                                _                -            ._~           . _       ._.-_          . _.       - . .            _      . ___ _, .

there is a 16.5% probability of BWR-2 (direct release to the environment) and less than 1% chance of BWR-4 (failure to isolate the drywell). For Grand t Gulf, there is no chance of a BWR-2, but a 20% probability of containment

isolation failure reported in the RSSMAP study.

i i I The offsite doses associated with each release category were evaluated with l the CRAC code, under the following assumptions: a uniform population distribution of 340 peopic per square mile, meteorological conditions typical of the Byron site, no evacuation, and integration of conditional consequences

to a 50 mile radius from the plant. The resulting estimates of public exposJre are shown in Table 2. It should be noted that the conditional i

~ consequences are not monotonically decreasing with release category, as would ! be expected considering the release catego.ry definitions. This is most likely a result of neglecting the dose beyond 50 miles, the higher release energy for BWR-1 deposits of fission products at greater distances. i TABLE 1 ! Conditional probabilities of release categories for LBLOCA ; j with loss of ECCS recirculation, as estimated in plant-specific PRA's for the five reference plants. RSS/RSSMAP Methodology l PWR-1 PWR-2 PWR-4 PWR-6 ) Surry 0.01 0.207 --- 0.78 Crystal River 0.01 0.2 0.007 0.8 Calvert Cliffs 0.01 0.3 0.007 0.7 i BWR-1 BWR-2 BWR-3 BWR-4 Peach Bottom .008 .165 .824 .003 l Grand Gulf .01 ---

                                                                                                                                                                                         .79                      .2 l

4 e i ! NUREG-0869, Revision 1 E-2

  ._ . . - . _ . . _ - , , . . - , , - _ , , _ . . , _ _ . . . . _ . , , _ _ _ _ _ , . _ _ _ . . _ .    , . _ _ . . . _ _ _ _ _ _ _ , _ , , _ _ _ . _ . _ , , . . _ _ _ , _ _ _ __                 - - _ _ . . - _ _ _       ,,,,._.,.-,._m__               , __.

__. m l 1 TABLE 2 Calculated conditional public radiation exposures from each of the release categories for the A-43 reference site conditions. (RSS Methodology) Calculated Calculated Release Consequences Release Consequences Category (Person Rem) Category (Person Rem) ! PWR-1 5.4E+6 EWR-1 5.4E+6 l PWR-2 4.8E+6 BWR-2 7.1E+6 i PWR-3 5.4E+6 BWR-3 5.1E+6 PWR-4 2.7E+6 BWR-4 6.1E+5 i PWR-5 1.0E+6 l PWR-6 1.5E+5 l PWR-7 2.3E+4 l The conditional consequences of a core melt due to sump failure are obtained by weighting the calculated public exposure for each release category (Table 2) with its associated conditional probability (Table 1), and then summing over release categories. The resulting conditional consequences for the three reactors are listed in Table 3. The risk, in person-rem per reactor year, is the product of core melt frequency and conditional consequences. TABLE 3 Preliminary estimate of conditional consequences for loss of recirculation (based on WASH-1400 methods of assessing severe accident risks; RSS/RSSMAP Methodology). Conditional Consequences Plant (Person-rem) Surry 1.1 x 10 6 l Crystal River 1.1 x 10 6 Calvert Cliffs 1.6 x 10 Peach Bottom 5.4 x 10 66

                 . Grand Gulf             4.2 x 10 NUREG-0869, Revision 1                 E-3

( Since the publication of the WASH-1400 study in 1975, a great deal of progress has been made in two areas relating to the calculation of severe ) accident consequences; (1) containment response characteristics and (2) radiological release fractions. When the preliminary version of the A-43 resolution package was presented to the Committee for Review of Generic l Requirements (CRGR), members of the Committee inquired as to whether recent

research results related to containment response would substantively alter the consequence evaluation presented above. The staff has performed such an assessment, and in addition, we have examined the possible impact of the revised methodology for estimating radiological source terms.

) Reassessment of Containment Response The c.ontainment response assumed in estimating the consequences of sump ! failure in the reference PWRs may be characterized as follows: there is

about a 20% conditional probability of a serious radiological release to the i environment (PWR-2). and about an 80% probability of a benign release (PWR-6). If the estimates of societal risk are to be significantly reduced

, from those given above, then the probability of a serious release would have 4 to be significantly lower than 20%. To state with confidence that the release rate is that low would require a great deal of confidence in the , systems which prevent containment failure and reduce the resulting source ! term, i There is one containment type which meets this criterion, the PWR large dry t containment with safety grade fan coolers (a large majority of large dry containments safety grade have fan coolers). We have a great deal of confidence that overpressure failure of these containments can be prevented i j NUREG-0869 Revision 1 E-4

                                                                                   .. a 4 e either by operation of the fan coolers or by connecting the sprays to an alternative source of water. The probability that the fans will fail and the spray hookup to an alternative water source will not be made is judged to be small. Large dry containments are not susceptible to abrupt failure due to hydrogen burns and steam spikes, and the probability of steam explosions or failure to isolate containment is small.

For subatmospheric containments, prevention of eventual overpressure failure depends on connecting the sprays to an alternative source of water. Subatmospheric containments do not have adequately sized HVAC systems to meet the energy addition to containment associated with a large bresk LOCA. Non-condensible gas production due to core-concrete reactions could overpressurize containment in a day or more. Hydrogen burns associated with in vessel H2 production are not expected to fail containment, but late hydrogen burns with the containment at an elevated partial pressure of steam and/or non-condensible gases may threaten containment integrity. However, if inplant capability to detect flow degradation exists and corrective operational procedures have been developed, time exists to prevent eventual overpressure failure by operator action. In an Ice condenser containment, powered hydrogen ignitors would protect against a deflagration threat. However, because there would be water in the reactor cavity, containment failure due to steam production would occur at the time of vessel failure or within a few hours thereafter, depending on the amount of ice remaining at the time of core melt. The assumption is made in this analysis that refueling plugs have been removed in accordance with Technical Specifications and hence a wet cavity would exist at the time of vessel failure. NUREG-0869, Revision 1 E-5

Our base case assessment of BWR consequences is equivalent to a 100% pro-bability of severe radiological releases for Peach Bottom and an 80% probability for Grand Gulf. Our current understanding of containment response indicates that lower probabilities are possible. For the Browns Ferry BWR Mark I containment there is a specific design feature which allows the operator in the control room to connect the spray suction line to the condensate storage tank. Its existence for all other Mark I containments would have to be verified before generic credit for a radiological release reduction due to this feature is taken. It would also have to be verified that procedures exist to make the proper realignments and override the necessary interlocks within the available time in the event of failure of recirculation sprays. BWR Mark II and III designs do not have the option of taking spray suction from the condensate storage tank. It is possible to align spray suction to the fuel storage pool, but we are not aware of any 5 ..?dures. Furthermore, a source of makeup water to the fuel storage pool would have to be found. The physical plant layout for Mark III design has the advantage that, for the most likely mode of failure, the fission products will bubble through a subcooled suppressior pool, yielding a radionuclide release significantly lower than BWR-1. The resulting consequences (in terms of person-rem) would be reduced substantially. In addition, no additional credit is being given for scrubbling effects than assumed in prior BWR analyses. A similar mechanism for release reduction for Mark I and II containments would be obtained if the operator relieves pressure by venting the containment through existing wetwell penetrations to the atmosphere. NUREG-0869, Revision 1 E-6

                                                                                        . w_s 1

I Current BWR EPG's describe containment ventireg. Forthcoming BWR owners group EPG's uill discuss wetwell venting in greater depth with the emerging picture appearing to be that controlled venting will be adopted by all BWR owners. For each of the containment types, however, there are credible mechanisms which would allow fission products to bypass the suppression pool under certain circumstances. In the Mark I, drywell failure can occur either at the time of core melt due to overpressurization, or several hours later due to high temperature failure of penetration seals. In some Mark II designs the molten core would be retained on the diaphram floor and containment overpressure failure would occur in the drywell. Hence a direct path for some radionuclides could occur. In the Mark III design, direct leakage from the drywell to the wetwell can result either from existing measured leakage paths, or from failure to isolate the drywell. Reassessment of Radiological Releases A wealth of experimental and theoretical research on radiological releases in severe accidents has been performed in the past decade. The NRC office of research (RES) has developed a revised source term methodology based on that research and has documented the results of sample accident sequences for several reference plants (BMI-2104). A committee of the American Physical Society has reviewed the methods and results, and they have concluded, among other things, that there is considerable uncertainty in some aspects of the methodology. They further stated that the radiological release estimat,es for some sequences could be higbqr than predicted by WASH-1400. When the new results and their associated uncertainties are integrated into the analysis of specific sequences for specific plants, the calculated releases are generally lower than previously predicted, but the uncertainties are large. For the purpose of this assessment, we confine our consideration

NUREG-0869, Revision 1 E-7 1

i

                                                                                 -m to those cases in which the new source term methods unambiguously predict lower releases of radionuclides.

For sequences in which containment failure is predicted to occur long after the core has melted, the new methods predict considerable decreases in the suspended aerosol concentrations in containment, due to enhanced agglomeration and gravitional settling. During the early and intermediate time period (less than 10 hours) after core melt, there are competing mechanisms which would tend to somewhat offset this reduction; most notably, enhanced releases of refractory fission products during core-concrete interactions in some reactor cavity designs, and the possible revaporization of fission products deposited in the primary system. The latter mechanism is not a substantive consideration for A-43, because there is very little primary system retention of fission products in large LOCA sequences. For large dry containments with fan coolers, containment failure would occur more than a day after core melt, if at all. The revised source term methodology would predict significantly lower radiological releases and offsite consequences than we have assumed in this base case and prior A-43 analysis. Other large dry containments and subatmospherics fall into two categories: those in which the teactor cavity would be full of water following sump failure, and those for which it would be dry. In the wet cavity case, the core-concrete interaction would be suppressed, and the source term would be greatly reduced at the time when containment fails due to steam overpressurization (about 12 hours after core melt). For dry cavity designs, enhanced production of refractory metals during core-concrete interact. ion could occur, but the predicted containment failure time, if at all, is much later due to the absence of steam production. Consequently, we would also expect a greatly reduced source term for dry cavity cases. l l l NUREG-0869, Revision 1 E-8 l L

F _. an i Because the enhanced aerosol removal does not affect noble gases, organic iodine or gaseous fission products, we would limit the predicted reduction in offsite persnn-rem within 50 miles to about a factor of 10 for large dry and subatmospheric PWR containments. Because containment failure in an ice condenser is expected to occur early after core melt, we cannot be confident that the radiological releases are any lower than predicted in our base case. For BWRs, the contaiament failure times are generally much less than 10 hours, and the enhanced aerosol settling may not be as significant as estimated for PWR's. Furthermore, for early containment failure the releases of refractory metals can be significantly higher than we have previously assumed in the Mark I containment. Consequently, we do not expect significant reductions in the predicted fission product releases for recirculation failure in BWRs as a result of the revised NRC source term methods. Summary Based on our reassessment of containment performance and radiological releases, we have developed revised estimates of the offsite consequences for each reactor type. Because of the very approximate nature of the review, we quote the results with only order-of-magnitude accuracy. That is, our a:sessment determines whether the average consequences are about the same as a severe release (BWR-3 or PWR-3), an order of magnitude less (x 0.1), or two orders of magnitude less (x 0.01). The results for PWRs are shown in Table 4 and for BWRs in Table 5. I NUREG-0869 Revision 1 E-9

Large dry containments with safety grade fan coolers are not est4 mated to fail. Furthermore, in the event of a failure, radionuclide releases would be greatly reduced because of the long time to failure. For other large dry and subatmospheric containments, failure could occur within a day after core melt, but the reduction in source terms due to the enhanced gravitational settling would lead to an order of magnitude (x 0.1) reduction in consequences. A further order-of-magnitude reduction (x 0.01) would result if containment failure were prevented by connecting the sprays to an alternative water source. Ice condensers are expected to fail at the time of vessel failure, or a few hours thereafter, depending on the amount of ice remaining at the time of vessel failure. While source terms could be lower than the WASH-1400 categories, it would not, lead to an order-of-magnitude reduction in the predicted person-rem. Furthermore, recove'ry of spray operation before containment failure could not be assured, because containment failure may be rapid. Consequently, we conclude that the consequences of sump failure could be more severe then our estimates for other PWRs. TABLE 4 Approximate consequence reduction factors

for PWR containments based on reassessment of containment response and source terms.

Without Spray With Spray Recovery Recovery Large Dry Designs with Fan ---

                                                                      .01 Coolers Large Dry Designs without Fan               .1         .01 Coolers and Subatmospherics Ice Condensers                           1            ---

Because of the approximate nature of the revised consequence estimates, they are quoted with only order-of-magnitude accuracy.

               +     Revised consequence values can be obtained by multiplying the reduction factogs with the consequence estimates associated with PWR-3 (5.4 x 10 person-rem).

NUREG-0869, Revision 1 E-10

p :- I BWR Mark I and Mark II containments are expected to fail within a few l hours after core melt, and the radionuclide releases are expected to be on the same order as BWR-3 (Table 3). Failure could be later and releases could be lower for some Mark II plants, but variability of the Mark II designs and uncertainty of the phenomenology prevents us from accepting this as a general l conclusion. In both the Mark I and Mark II, successful venting of the I wetwell prior to containment failure could lead to substantial source term retention in the suppression pool, and a dramatic reduction in conditional consequences. A similar reduction would result if the drywell sprays could be connected to an alternative source of water. l We have limited our consequence reduction factor to one order-of-magnitude, l (x 0.1) because there is some possibility of bypassing the suppression pool due to early containment failure or leakage from the drywell. l The Mark III design fails in such a way as to channel the fission products j through the pool before release to the environment, with or without wetwell venting. As with the Mark I and Mark II plants, the reduction factor is l limited to 0.1 because of the possibility of pool bypass. 1 1 The revised estimates of offsite consequences are shown in Table 6. They are obtained by multiplying the reduction factors in Table 4 for PWRs and Table 5 for BWRs by the consequences associated with the PWR-3 and BWR-3 releases, respectively. Estimation of Offsite Releases Utilization of the estimated consequences for PWR ice condenser plants shown in Table 6, without consideration of plant-specific sump design features and recirculation flow reqairements could lead to the conclusion that consequences associated with a failed sump are a decade higher. Table 7 provides an overview of ice condenser plants; the significant factors are: NUREG-0869, Revision 1 E-11 l

   --       .--    .-    -      ~    _ ,        - -

  -  (1) All ice condenser plants utilize reflective metallic insulation on the primary coolant piping and major components (e.g., SGs, PER, RCP's, etc.);

thus debris blockage concerns associated with transport of fibrous insulation debris are not present (at least not in significant amounts). (2) For the majority of these plants, lower recirculation flow rates and larger debris screen areas prevail. The net effect is that approach velocities are less than 0.2 f t/sec, and therefore debris transport is not likely to occur. (3) NPSH margins are high, significantly higher for most plants than the 1-5 feet of head loss utilized in the sump failure probabilities discussed in Appendix 0. These factors all lead to the conclusion that sump failure probabilities developed for PWRs with a mix of fibrous and RM insulation and low NPSH margins should be reduced for applications to ice condenser plants. Interpolation within the values shown in Table 7 of Appendix 0 can be used to derive a sump failure probability (for ice condensers) in the range of 3 to 9

         -6 x 10 /Rx yr.      This value was used for estimating the consequences discussed below.

Utilization of the consequence values in Table 6 and estimated blockage probabilities (from Appendix 0) can be used to estimate averted risks. Table 8 contains such estimates and the blockage probabilities which were developed. These values were used to calculate the value-impact ratios discussed in Section 4.4. I i i l i i NUREG-0869, Revision 1 E-12 1 i

J.u w TABLE 5 Approximate consequence reduction factors

for BWR containments, based on reassessment of containment response and source terms.+

Without With Wetwell Wetwell Venting Venting Mark I 1 .1 Mark II 1 .1 Mark III .1 .1 Because of the approximate nature of the revised consequence estimates, they are quoted with only order-of-magnitude accuracy.

                 +    Revised consequence values can be obtained by multiplying the reduction factors with the consequengeestimatesassociatedwithBWR-3 (5.1 x 10 person-rem)

NUREG-0869, Revision 1 E-13

TABLE 6 Revised estimates of offsite consequences based on current understanding of containment response and radiological source terms. PWRs i i Conditional Consequences (person-rem) l Containment Type No Spray Recovery With Spray j Recovery I __

!        Large Dry (Safety Grade Fan Coolers)                                        4 5 x 10 1

Other Large Dry & Subatmospheric 5 x 10 5 x 10 4 j Ice condenser 5 x 10 6 ,,, i i BWRs - j Conditional Consequences (person-rem) i Without Venting or With Venting i or ) 1 Containment Type Spray Recovery Spray Recovery Mark I 5 x 10 6 5 x 10 5 l Mark II 5 x 10 6 5 x 10 5 i Mark III 5 x 10 5 5 x 10 5 i I I i t I i

NUREG-0869, Revision 1 E-14 i

1 Table 7, Overview of Ice Condenser Plant Design and i Operational Features ' l

;                                                      RHR       Debris               NPSH                                          Approach

Flow Screen Margin Insulation Velocity

. Plant {S2ml (So Ft) (Ft Water) Used (Ft/Sec) , l Catawba 1&2 6000 135 7.9 (RHR) Reflective .10 Metallic ,. DC Cook 1&2 6000 90 21.9 (RHR) Reflective .15 : Metallic ' McGuire 1&2 6800 120 6.9 (CSS) Reflective .13 16.9 (RHR) Metallic Sequoyah 1&2 9500 43 2.8 (RHR) Reflective .49 Metallic Watts Bar 1&2 8000 43 11.5 (RHR) Reflective .41 8000 164* Metallic .11

Trash rack area; this structure would intercept large size insulation debris before transport to the sump debris screen structure could occur.

l t i r I r I

!                                                                                                                                                                                       t 4

i

!                                                                                                                                                                                       I i

i i 1 1 l i I I l r NUREG-0869, Revision 1 E-15 L I l 1

E ? Table 8, Overview of Consequences Associated with Sump Blockage Estimated Estimated Assumed E Conditional Blockage Core Melt Estimated [- - Consequences Probalblity Conditional Risk Averted ** g Type Containment (person-res) (1/Rx yr) Probability * (AR, person-res/Rx)

s

- PWR Dry w/SGFCS -- 3 to 50 x 10 -6 0.5 -- PWR Ice Condenser 5 x 10 6 1 to 9 x 10-6 0.5 40 to 560 PWR Dry w/o SGFCs and Subatmospheric 5 x 10 5 3 to 50 x 10 -6 0.5 19 to 313 PWR Dry w/o SGFCs 5 x 104 3 to 50 x 10 -6 0.5 2 to 31 m and Subatmospheric, g w/ Spray Recovery Mark I and II 5 x 100 4 to 20 x 10 -6 0.5 250 to 1250 Mark III 5 x 10 6 4 to 20 x 10 -6 0.5 25 to 125 Mark I and II 5 x 10 5 4 to 20 x 10 -6 0.5 25 to 125 w/ venting or spray recovery

     *The assumption is made that 50% of the time that blockage occurs, core melt would occur. This assignment of a conditional core melt probability is realistic in view of potential operator dection and mitigating actions which could be taken.
    **An outstanding reactor life span of 25 years has been assumed.

1 1

_ ,;] 7 APPENDIX F CONTAINMENT SURVIVABILITY

                                                                             ~     .----aa 1

i CONTAINMENT SURVIVABILITY i The several different containment design concepts currently in use for the many operating plants can be grouped as follows: (1) The dry containment structures for pressurized water reactors (PWRs) must absorb all loads from accident conditions. The results are . characterized by large containment volumes (i.e. 2 x 108 ft3) and high design pressures (i.e. 60 psig) with considerable margin beyond the design point. - (2) Containment structures which incorporate a means of passive steam condensation have taken advantage of this approach to design smaller and lower pressure containment buildings. Boiling water reactor (BWR) containments utilize a water pool for condensing steam and PWR ice i condenser plants utilize an annular ice bed for condensing steam. Containment Structural Capabilities Containment failure modes and attendant releases were analyzed in WASH-1400 ' and related to a major loss of containment capability through: (1) steam explosion induced failures (a mode); (2) hydrogen burn induced failures (6 mode); (3) overpressurization of the containment building resulting from steam generation (molten core interacting with water) and noncondensible gases (molten core interacting with basemat) (y mode); and (4) basemat penetration (c mode). 'These WASH-1400 studies also assumed a loss of containment without assessing the significance of any design margin available in the different ' containment structures as currently designed, t Risk assessments performed in recent years indicate that risk from nuclear power plants is dominated by severe core melt accidents. Typical containment loading pressures and temperatures associated with whole core melt scenarios i are on the order of approximately 100 psia and approximately 300*F. More ' recently, mechanistic models for containment failure (see NUREG/CR-3653) 3 have also been included in these assessments of severe accident scenarios. , t These severe accident and containment structural capability studies provide the following insights: (1) Because of structural design margins, containments have inherent capabilities beyond their design basis. This provides a capability to contain or mitigate a wide spectrum of severe accidents. (2) Best estimate analyses of containment performance indicate that containments can retain structural integrity at pressures as high as 2.2 to 2.5 times the design pressure. " Extensive yielding" is the i term used in NUREG/CR-3653 to describe loss-of-structural integrity for these best estimate analyses. NUREG-0869, Revision 1 F-1

as 1 i (3) Although it is possible that leaks through penetrations could occur before loss of structural integrity, the risks frcm such leaks would be considerably smaller than from the gross containment failures assumed in WASH-1400 studies. Containment Heat Removal Systems Make-up water flow is needed for the energy released into the containment atmosphere or to the pools due to the decay heat in order to keep the core flooded, is small compared to the design basis recirculation flow. However, as the flow to the vessel is reduced, boil-off will occur. The steam released into the containment will tend to raise the pressure and temperature. This energy must be removed from the containment in order to maintain the containment pressure below its failure pressure. The Containment Heat Removal Systems (CHRS) are sized to remove the energy released into the containment due to decay heat. Depending on the containment type, any one or combination of the following systems might be used as CHRS: suppression pool cooling systems, containment atmosphere spray systems and air cooling systems. BWR plants consider the pressure suppression pool (Mark I, II and III containments), as the short term heat sink for both the blowdown energy and the decay heat. In the long term, the pool cooling system is the most important CHRS in limiting the maximum pool temperature (and pressure for Mark III). Reducing the flow to the suppression pool water cooling system due to debris breakage would result in increasing the pool temperature. If the reduction of the flow rate is small (some tens of percent), the temperature increase would be small because of heat exchanger capacity margins. However, a major reduction (factors of 5-10) in flow rate would very soon lead to a temperature increase in the pool beyond design limits. This is due to the strong connection between mass flow rate through the heat exchanger and overall heat transfer coefficient of the heat exchanger; of course, together with the reduced flow rate. Containment atmosphere spray systems are used in all types of containments for both fission p'oduct and energy removal from the atmosphere in accident situations. The relative importance of the spray system is dependent on the containment type. In ice condenser containments the spray system is the only means to remove the decay heat released as steam into the containment atmosphere after the ice beds have melted. A reduction of the containment spray flow rate would immediately lead to higher containment pressure and temperature in addition to a reduction in the atmospheric fission product removal effectiveness. This is due to several factors: 1) smaller total amount of water flow, 2) bigger droplet sizes because of smaller pressure drop over the spray nazzles, 3) shorter stay time in the containment atmosphere, 4) smaller total heat transfer area of the droplets, 5) loss of thermodynamic equilibrium between containment atmosphere and droplets, and

6) coverage by spray in the containment is smaller. It is expected that a reduction of containment spray flow rate would cause a significant reduction of spray cooling efficiency.

NUREG-0869, Revision 1 F-2

ww.. For Mark I and II designs, the spray systems are not required to mitigate the course of the transient. Mark III containments do however rely on sprays, primarily for the short term mitigation. As a result, loss of spray system effects can be considered to be of secondary importance for BWR designs. Dry containments usually employ both sprays and fan coolers. Each system is sized for 100 percent capacity. Therefore, the coolers could perform the same heat removal function without the sprays being operational. Safety grade air coolers, typical for dry containment designs, are not affected by a reduction of RHR-system flow rate and therefore would not be affected by loss of this system. However, the availability of containment air coolers, which are not safety grade and are not designed to operate during post-LOCA accident situations, cannot be independently relied upon. Due to the elevated pressure and steam-air mixture density in the containment, such fans would be expected to trip because of overload. Another factor would be, of course, the very questionable survivability of the related electrical components not qualified for the post-accident environment. Expected Containment Design Response Due to Loss of Recirculation Water Sources The influence of reduced RHR-system flow rate upon the overall containment behavior is a plant-unique matter, depending on the containment design (Mark I, II, III, Ice Condenser or Dry Containment). Mark I and II containments are both small and quite similar in their pressure responses following an accident. The peak pressure is reached very early; and after one hour the pressure is already low compared to the design pressure. However, the pool temperature is high and still increasing and it would probably be the first limiting design parameter associated with reducing the RHR-system flow rate being exceeded. This elevated water temperature to the RHR pumps could jeopardize their operation. Thus, a total loss of RHR-system, or a major reduction of the flow rate, would result in a slow overpressurization and failure of the containment. This failure could be expected in less than 12 hours, if the containment is not vented as recommended by EPG. A rupture of the containment can be expected, based on current engineering judgment, to occur in the range of 2-3 times the design pressure. In Mark III containment, the pressure and pool temperature are still hfgh and increasing after one hour. A reduction of RHR-system flow could hve the consequence that both the containment pressure and pool temperature would exceed the design value. This would result in the same consequences as with Mark I and II plants. Without venting, containment failure cculd -be expected during the first 24 hours following loss of recirculation flow capability. The ice of an ice condenser containment will melt in about'l-2 hours. After the ice is melted, the containment spray system is needed to keep the pressure below the design limit. A reduction of RHR-system flow rate could NUREG-0869, Revision 1 F-3

also cause the containment design pressure to be exceeded in this type plant. " Reduction of the RHR-system flow rate would occur in the same time period when the ice would be totally melted and would result in a rapid overpressurization and early containment failure. This could be expected in about 3 hours. The peak pressure in a dry containment (normal atmospheric or subatmospheric) is reached early in the transient, and, af ter one hour, the pressure is expected to be very low. The dry containments also have safety grade air coolers, which could maintain containment pressure and temperature below the design values and, therefore, a reduction of RHR-system flow will not adversely affect the dry containment. Thus, containment failure for large dry containments with SGFCs would not be expected. Severe Accident Study Insights For large dry PWR containments with safety grade fan coolers, we can conclude that the containment response assumed previously in the A-43 consequence analysis is conservative because the probability of containment i failure due to overpressure has been overestimated. Reaching a similar conclusion for BWR Mark I containments would be contingent on the licensee's demonstrating that there is a high likelihood of the operator successfully i exercising the option to align the containment sprays to take suction from the condensate storage tank. For PWR subatmospheric and ice condenser designs, and BWR Mark II and Mark III plants, we are not aware of a design option which allows the operator to readily switch spray suction to an alternative water source.

Although alternate water sources exist, and in some cases the piping appears to be in place, it remains to be demonstrated that the proper connection can be made, valves can be aligned, and interlocks overridden within a reasonable time under accident conditions. Without the sprays, the probability of overpressure (or overtemperature) failure for these containments could be higher than previously assumed. The PWR subatmospheric containment design has an advantage over the ice condensers, Mark II's and

, Mark III's insofar as its predicted failure time is much later than for the j others, thereby allowing mere time to recover the sprays. The BWR Mark III containment stands out insofar as its principal failure mode leads to significant scrubbing of fission products in the suppression pool. The reduction in conditional consequences that would result is offset somewhat by our conclusion that the probability of early failure for Mark III plants is higher than prior estimates. A similar mechanism for source term reductions for Mark I and Mark II containments would result if the operator can successfully relieve pressure by opening a wetwell vent. Without a plant specific analysis by the licensee, we cannot say with confidence that the containment response estimates developed for resolution of USI A-43 are i conservative for BWR containments (see also BMI-2401). i I NUREG-0869, Revision 1 F-4 I 4

   - - - - - - - - , , - - - - - ~ , . , , . _            m - , . _ , . - -

                                                                                                         .b y

Conclusions In summary, the following situations appear to exist: . (1) PWR dry containments (with safety grade. fan coolers) will likely survive a core melt situation, even with a loss of the containment emergency sump. Even without safety grade fan coolers, large dry containments will not overpressurize for 6 to 12 hours following a LOCA, leaving time for the operator to take corrective action to provide an alternate water source for containment spray (provided a loss-of-sump condition is detected and ccted upon). The availability of SGFC's to minimize containment pressure, thereby enhancing containment survival is d significant factor. (2) Mark I containments have a high containment structure survivability if alternate containment spray suction can be provided should loss of RHR suction occur. However, Mark I's are most susceptible to containment failures which lead to release consequences comparable to WASH-1400 releases.

(3) Controlled venting of any of the BWR containment designs will preserve structural integrity for all BWR containment designs, if such venting is implemented correctly. (4) Small dry PWR containments (without safety grade fan coolers) and PWR ice condenser plants are most susceptible to loss of integrity if the containment emergency sump is lost and core sprays cannot be recovered through use of alternate water sources. References U.S. Atomic Energy Commission, WASH-1400, " Reactor Safety Study", October 1975 (published by NRC as NUREG-75/014). U.S. Nuclear Regulatory Commission, NUREG/CR-3653, " Final Report, Containment Analysis Techniques a State-of-the-Art Summary", March 1984. J. A. Gieske, et al, BMI-2104 (Volumes 1-7) "Radionuclide Release Under Specific LWR Accident Conditions", Battelle' Memorial Institute, July 1984. ' NUREG-0869, Revision 1 F-5

 .-   _ _ , _ _ _ . ,                 4, _ _ 1 . _ _

p = APPENDIX G ESTI!!ATION OF COSTS TO REPLACE INSULATI0ft

                                                                                -. R ESTIMATION OF COSTS TO REPLACE INSULATION The monetary costs of insulation replacement are dependent on plant design, material, and preplanning efforts. Table 1 shows the insulation replacement costs and estimated exposures that were used in the For Connent version of the USI A-43 value-impact analysis (NUREG-0869), which was published in April 1983. Cost information received from the industry (Atomic Industrial Forum) and insulation vendors (0 wens-Corning Fiberglas and Diamond Power Company) is summarized in Table 2. This information was used to develop the reinsulation costs shown in Table 3. The variability in estimates from industry sources is large and further illustrates the variances in plants and installation experience. The composite average estimates of $153/f t2 (Table 3) for insulation replacement can be compared to the earlier USI /-43 estimate of

$75/ftr. Table 3 also shows an estimated insulation replacement cost range of $92/fte to $244/ft2 The most severe monetary impact would result from a decision to replace all of the problem insulation in a given plant. The PWR sump failure probability study showed (see NUREG/CR-3394) that insulation on the primary system piping and the lower one-third of the steam generator are the principal sources of debris that leads to unacceptable sump blockage conditions. Appendix 0 of this report discusses the significance of insulation on the primary system piping (_10-inch diameter or greater), steam generators, reactor coolant pumps, and pressurizer that could lead to unacceptable debris screen blockage. Table 4 illustrates the variability of fibrous debris as a function of pipe size and was derived from the Salem plant analysis (see NUREG/CR-3394). Plant insulation variability (as installed) is illustrated for the Salem 1 and Maine Yankee plants as shcwn in Table 5. Approximately 2400 ft3 (covering an area of 8200 ft2) of potentially troublesome insulations are identified. On the basis of these two illustrative plants, it is estimated that only 4400 f t2 to 5235 ft2 of insulation might have to be replaced rather than replacing all plant insulation. In the cost impacts developed 171 Section 2 of this report, these amounts are used as the upper bounds of the amount of insulation that would have to be replaced for determining insulation replacement costs for a typical PWR. NUREG-0869, Revision 1 G-1

a Table 1 Estimates of costs to replace insulation and associated exposures illustrating plant variability Unencapsulated Cost est* Cost est** Cost est+ Estimated ++ insulation 1 2 3 exposure Plant (ftt) ($ X 103) ($ X 103) ($ X 103) (person-rem) Salem 1 13,200 281 238 660 99 Maine Yankee 6,700 142 121 335 47 Ginna 1,000 21 18 50 8 Hillstone 2 1,300 28 23 65 10

*These costs are derived from Surry 1 and 2 steam generator removal and reinstallation data, and discussions with onsite staff. A per-unit rate of 0.85 hr/fta for replacing insulation was derived, and labor costs of
  $25.00/hr were assumed.

- Telephone estimates from New England Insulation Company (Maine Yankee has employed this firm) were: $3/ft2 to remove insulation, $11/f t2 to fabricate new panels, and $3 to $5/ft to install the new panel.

+ Telephone estimates of $35 to $50/f te for MIRROR _ insulation fabrication and installation were obtained from Diamond Power, which supplies such insulation.

The value of $50/f t2 was used. ++ Exposure data were derived from Surry 1 and 2 data. Discussions with the Surry site staff indicate that a 50-person-rem exposure level for insulation replacement is realistic if the job is adequately pre-planned. An equivalent dose of 7 x 10 3 person-rem / f t2 of insulation to be replaced can be derived. NUREG-0869. Revision 1 G-2

                                                                                       . l Table 2 Insulation fabrication and installation cost estimates received during the A-43 (NUREG-0897) for comment period Commentor                             Comment / Estimate Atomic Industrial Forum                Productivity:    Industry experience shows an average rate of 2.0 to 2.5 hr/ft2 for installation.

Labor: Industry labor costs are $30 to

                                          $45/hr.

Total: The averaged cost of $550,000 per plant is off by at least a factor of 2; material costs are not included. Diamond Power Company MIRROR *: The cost of material supplied at (manufacturer of MIRROR

Cooper was approximately $40/ft2 insulation)

Productivity: Installation rates averaged 1.24 hr/ft2 Labor: The estimate of $25/hr used previously seems raasonable. Owens Corning Fiberglass NUKON*: Materfil cost is estimated to Company (manufacturer of be $90/ft2, NUKON* insulation) Reflective: An assumption of $100/ft2 would be very reasonable. Labor: Current labor costs of $40 to

                                         $50/hr are common.

Productivity: Installation rates of 7 to 12 ft2/hr are typical. Total: This leads to a cost of $112/f ta for labor and $100/ft2 for material, for a total cost of $212/f t2 to remove existing insulation and replace it with reflective insulation. NUREG-0869, Revision 1 G-3

Table 3 Insulation replacement cost estimates developed from public comments received Commentor Comment /Es timate Atomic Industrial Forum Reinsulation Labor Costs: $60 to 5112/ft: Insulation Fabrication Cost: $75/ft Combined Total: $135 to $185/ft2 Average Combined Total: $160/ft: , Diamond Power Company Installation Labor Costs: $31/ft Insulation Fabrication Costs: $30 to 50/ft: Subtotal: to $81/ft: , Assumed Remt al Costs: $31/ft: Combined Total: $92 tc $112/ft: Average Combined Total: $102/ft: Owens-Corning Fiberglass Reinsulation Labor Costs: $60 to $144/ftr Company Insulation Fabrication Cost: $90 to $100/ft: Combined Total: $150 to $244/ft2 Averaged Combined Total: $197/ft: COMPOSITE AVERAGE: ($160 + $102 + $197)/3 =

                                                                                                                       $153/ft NUREG-0869, Revision 1                                                                     G-4

                                                                                  .3 Table 4 tiaximum LOCA-generated debris summarized by break size

(

Reference:

NUREG/CR-3394) Pipe dia- Total fibrous Total all meter (in.) debris (ft3) type (ft3) 2 1 1 6 2 22 8 2 3 10 4 31 14 227 227 16 270 270 32 144 295 34 315 726 36 118 402 Notes: (1) These values correspond to break locations in the primary system within the crane wall and represent the largest quantity of debris generated by a single break of a given pipe diameter. (2) The insulation types and distribution within containment are those used in Salem 1. All insulation within 7 L/Ds o' a break location is assured to be destroyed and released as fragmented debris. NUREG-0869, Revision 1 G-5

                                                                                    . _ . . ._N Table 5 fypical volumes of primary system insulation employed

Salem Maine Yankee Component Volume Type of Volume Type of (ft3) insulation (ft3) insulation Steam generator 1284 reflective metallic / 1144 calcium silicate /

fibrous fibrous Hot leg 160 reflective metallic 149 fibrous Cold leg 140 reflective metallic 156 fibrous Cross-over 60 reflective metallic 279 fibrous Pressurizer 160 reflective metallic 302 calcium silicate / fibrous Pressurizer surge 129 reflective metallic 57 calcium silicate / line fibrous Reactor coolant 570 reflective metallic 149 calcium silicate / fibrous Bypass N/A N/A 88 fibrous TOTAL ** 2507 2324 SUBTOTAL *** 1284 1527 (excluding reflec- (4402 !ta) (5234 ft2) tive metallic and calcium silicate)

  *This table is based on information provided by the operators in 1981. Plant changes since 1981 have made the data less iccurate for these specific reactors.

However, as representative data for reactors in general, the table is still valid.

**This volume includes all of the insulation that could be hit by a water jet from a LOCA pipe break (in pipes      10-inch diameter). If the volume were restricted to insulation within L7D = 7 of a break, it might be significantly smaller.

- - For conservatism, Salem's steam generator is assumed to be covered entirely with fibrous insulation; 50% of the insulation on Maine Yankee's steam generator, pressurizer, and reactor coolant pump is assumed to be fibrous.

NUREG-0869, Pevision 1 G-6

_.1

 ;               Using these revised insulation replacement costs and the revised estimates of the amount of insulation that must be replaced results in the cost estimates work sheet shown in Table 6. A cost range of $300,000 to
                 $1,300,000 (average cost = $750,000) is arrived at and is used in the value-impact calculations contained in Section 2.2 of this report.

Other estimated costs associated with plant evaluations and potential backfits are shown in Figure 1. These estimated costs were derived as follows: (1) Utilization of information provided in RG 1.82, Revision 1 and NUREG-0897, Revision 1 provides a rapid means for estimating sump air ingestion and debris potential. An evaluation impact of $10,000 (see 1 in Figure 1) is estimated for plants having high post-LOCA water levels (which prevent air ingestion) and low containment recirculation flow velocities (i.e., _,0.15 ft/sec) which precludes debris transport. (2) Should air ingestion pose a problem, the use of vortex suppressors has been shown experimentally to reduce air ingestion to zero; the relevant information is provided in RG 1.82, Revision 1 and flVREG-0897,

Revision 1. The design, fabrication, and installation of a vortex suppressor (consisting of connercially available floor grating materials, either installed horizontally or formed into a cage) are f

estimated to cost $35,000 to $50,000 (see 2 in Figure 1). i (3) Debris blockage problems can be assessed in two steps. The initial step--based on limiting calculations as described in RG 1.82, Appendix A and flVREG-0897, Revision 1--is estimated to cost $15,000. Should a second step--a detailed debris-generation, transport, and screen ( blockage analysis--be required, the cost would be higher. Plant-specific analyses in USI A-43 studies were on the order of $35,000 to

                                                 $50,000 per plant analyzed. Thus, this cost impact is estimated to range from $25,000 to $65,000 (see 3 in Figure 1).

j NUREG-0869, Revision 1 G-7

l Table 6 Estimated insulation replacement costs Fibrous Amount Estimated cost

insulation requiring ($ thousands)

Plant employed (ft2) replacement (ft2) High Avg Low Salem 2 13,200 4,400** 880 675 440 Maine Yankee 6,700 5,235** 1050 815 525 Millstone 2 1,300 1,300 (assumed) 260 200 130 AVERAGED COST 730 565 365 AVERAGED COST (w/o Millstone 2) 965 745 485 Estimated Cost Variance ($ thousands) Case 1 (averaged cost of three plants): 940 7 730(+30%) = 475 t 730 e 165 ; 365 P 475 a 365(-35%) = 240 Case 2 (averaged cost without Millstone) 1255 e 965(+30%) T 630 a 965 r 145 w485 0 630 4 485(-35%) ; 315 Case 3 (rounded values from Case 2. used for value-impact ratio calculations): 1300 1000 690 h y b 1000 e 750 1 - 500 F if y 650 500 300

Utilizes cost range shown in Table 3.

       **See Table 5.

NUREG-0869 Revision 1 G-8

                                                                                      '~
 ~

(4) Should the debris assessment calculations show a need for plant modifications, consideration logically should be given to alternatives that would be less costly than replacing large quantities of insulation. Because preservation of the net positive suction head margin is the key criterion, two ways the plant could be modified are Increasing debris screen area to reduce the impact of a loss of pump suction head caused by blockage. This could be done by enlarging the sump or intake screen. Along the same lines, use of . screens upstream of the currently installed screen would have a two-fold benefit: it would intercept undesirable debris at some distance from the sump location and it would reduce the impact of head loss because of the reduced approach velocities associated with the enlarged upstream screen area. Such a hardware backfit is estimated to cost $250,000 to $350,000 (see 4 in Figure 1). Re-examining the recirculation flow rates required for the long-term recirculation mode, possibly reducing the currently established flow rates (which are set by imediate post-LOCA flow requirements), and submitting the re-analysis of long-term recirculation needs. This option can be considered an analytical backfit, and the cost of such an analysis is estimated to range from

                 $25,000 to $65,000.

(5) The estimated costs for replacement of insulation are $300,000 to

           $1,300,000 (see 5 in Figure 1); these represent the major cost impact and are discussed at the beginning of this section.

References U. S. fluclear Regulatory Commission, flVREG/CR-3394. fiUREG-0869 Revision 1 G-9

E x 8 h PWR Vortex Suppressor 3 Required to p@ Reduce Air Ingestion N

Sump Hydraulics / Air Ingestion 127.

g Sump Design Check per RG 1.82, Rev. 1

 -       Assessment %

per RG 1.82, Initial Debris Screen Blockage Rev. 1 ~ ?.ssessment Effects Minimal per RG 1.82, Rev. I N (per limit analyses) Detailed Debris Acceptable c % Generation and NPSH Impact - @ 4, Blockage Analysis o Needs Estimated Costs Miniul

         $10,000
         $35,000 - $50,000 Retrofits, or Action
         $25,000 - $65,000
         $250,000 - $350,000 hardware costs
         $300,000 - $1,300,000 (for replacem nt of large                                       Replace quantities of insulatien)                                                 Problem      @

Insulation Figure 1 Actions required to determine sucp design acceptability l l l

I

                                                                                                                                                                                                                                                                              - eee M4 APPENDIX H DRAFT GENERIC LETTER                                                                                                                     ,

i l t I i i N i l t

l

o .

                                                                                                           ~L
    .*      \  g UNITED STATES NUCLEAR REGULATORY COMMISSION                                  ,

i

 ;;                                     WASHWOTON, D. C. 2tMS pg, j

S 4 o . TO ALL LICENSEES OF OPERATING REACTORS, APPLICANTS FOR OPERATING LICENSEES, AND HOLDERS OF CONSTRUCTION PERMITS Gentlemen:

SUBJECT:

POTENTIAL FOR LOSS OF POST-LOCA RECIRCULATION CAPABILITY DUE TO INSULAT:0N DEBRIS BLOCKAGE (Generic Letter 85- ) This letter is to inform you about a generic safety concern regarding LOCA - generated debris that could block PWR centainment emergency sump screens or BWR RHR suction strainers, thus resulting in a loss of recirculation or containment spray pump net positive suction head (NPSH) margin. The potential exists for a primary coolant pire break to damage themal insulation on the piping as well as that on nearby components. Insulation debris could be transported to water sources used for long-tenn post-LOCA recirculation and containment sprays (i.e., PWR containment emergency sumps ~ and BWR suction intakes in the suppression pools) and deposited on debris screens or suction strainers. This could reduce the NPSH margin below that required for recirculation pumps to maintain long-term cooling. This concern has been addressed as part of the efforts undertaken to resolve US! A-43, " Containment Emergency Sump Performance." The staff's technical findings contain the following main points:

Plant insulation surveys, developnent of methods for estimating debris generation and transport, d ebris transport experiments, and infomation provided as public corrents on the findings have shown that debris blockage effects are dependent on the types and quantities of insulation employed, the primary system layout within containment, post-LOCA recirculation patterns and velocities, and the post-LOCA recirculation flow rates. It was concluded that a single generic solution is not possible, but rather that debris blockage effects are governed by plant-specific design features and post-LOCA recirculation flow requirem ents.

The current 50% screen blockage assumption identified in Regulatory Guide (RG) 1.82, " Sumps for Emergency Core Cooling and Containment Spray Systems," should be replaced with a more comprehensive requirement to assess debris effects on a plant-specific basis. The 50% screen blockage assumption does not require a plant-specific evaluation of the debris-blockage potential and usually will result in a non-conservative analysis for screen blockage effects. NUREG-0869, Revision 16 H-1

The staff has revised Regulatory Guide (RI) 1.82 Revision 0, " Sumps for Emergency Core Cooling and Containment Spray Systems" and the Standard Review Plan Section 6.2.2 " Containment Heat Removal Systems," based on the above tech-nical findings. There is no requirement for analysis or modification of any operating plant or plant now under construction. These revised regulatory documents will be used on Standard Plants or new Construction Pennit applications only, effective six months from date of issuance. Although there are no new requirements as a result of the resolution of USI A-43, we reconsnend that RG 1.82, Revision 1 be used as guidance for conduct of the normally required 10 CFR 50.59 review for future insulation changes. RG 1.82, Revision 1 provides guidance for estinating potential debris blockage effects. It is expected that those plants with small debris screen areas (less than 100 ft8), high ECCS recirculatio,i psmping requirements (greater than 8000 gpm), and small NPSH margins (less than 1 to 2 ft of water) would benefit the most from this type of assessment in the event of a future insulation change. RG 1.82, Revision 0 with its 50% blockage criteria does . not adequately address this issue and is inconsistent with the technical findings developed for the resolution of US! A-43. This information letter along with enclosed copies of NUREG-0897, Revision 18 and RG 1.82, Revision 1, should be directed to the appropriate groups within your organization who are responsible for conducting 10 CFR 50.59 reviews. No written response or specific action is required by this letter. Therefore, no clearance from the Office of Management and Budget is required. If you have any questions on this matter, please contact your project manager. Hugh L. Thoripson, Jr., Director Division of Licensing

Enclosure:

NUREG-0897 R1B RG 1,82 al NUREG-0869, Revision IB H-2

ENCLOSURE 3 USI,A-43 CRGR PXG NUREG-0897 Revision IB l t June 5, 1985 CONTAIMENT EMERGENCY SUMP PERFORMANCE Technical Findings Related to Unresolved Safety Issue A-43

  . FOR CONMDif PURJCMICN CATL April 1983 PEM90N NO.1A CCWPLITED: Merch 1985 CATE F1,11USHED-A. W. Sahir, Took donoger DMeien of Sefety Technology Office of Nuclear Peoctor Regulation U. S. Nucteer Reg

tory Commesion Washington, D. C. 20S55

!       i           .

i i j i i i i l l ABSTRACT ! 1 1 l

\                                                                                                                  t

, This report summarizes key technical findings related to the Unresolved ; Safety Issue (USI) A-43, Containment imergency Sump Performance. Although t ! this issue was formulated considering pressurized water reactor (PWR) sumps, j the generic safety questions apply to both boiling water reactors (BWRs) and j PWRs. l Emergency core cooling systems require a clean and reliable water source to

maintain long-term recirculation following a loss of coolant accident
(LOCA). PWRs rely on the containment emergency sump to provide such a water

;                     supply to residual heat removal pumps and containment spray pumps. BWRs rely                 ,

i on pump suction intakes located in the suppression pool, or wet well, to ' i provide a water source to residual heat removal systems and core spray i systems. Thus, pumping performance under post-LOCA conditions must be evaluated. l The key safety questions relate to: (1) PWR sump or BWR suction intake ! j hydraulic performance (i.e., air ingestion potential); (2) potential sump screen or suction strainer blockage as a result of LOCA damage to insulation l 1 materials; and (3) pump performance under post-LOCA conditions where ingestion of air and debris particulates could occur. 1 { The technical findings presented in this report provide information relevant I t to assessing these safety concerns. These findings have been derived from i extensive experimental studies, generic plant studies, and assessment of pumps -

utilized for long-term cooling. Hydraulic results have revealed a less 1 severe potential for air ingestion than previously hypothesized. Debris blockage effects on NPSH margin should be dealt with on a plant-specific
basis because of the large uncertainty in quantifying the extent of debris: blockage. Therefore, these findings have been used to develop revisions to j Regulatory Guide 1.82 and Standard Review Plan Section 6.2.2 (NUREG-0800).

1 i T I t r i l i i 111 - ! l ! t i

i . i l TABLE OF CONTENTS PAGE l Abstract------------------------------------------------ 111 ! Foreword------------------------------------------------ xii l Acknowledgements---------------------------------------- xiii 1 INTRODUCTION-------------------------------------------- 1-1 i 1.1 Safety Significance--------------------------------- 1-1 j 1.2 Background------------------------------------------ 1-1 l 1.3 Technical Issues------------------------------------ 1-3 ) 1.4 Summary of Technical Findings----------------------- 1-3 l 2

SUMMARY

OF KEY FINDINGS--------------------------------- 2-1 i 2.1 Pump Performance------------------------------------ 2-1 2.2 Effects of Debris on Recirculation Capability------- 2-4 l 2.3 Sump Hydraulic Performance Findings----------------- 2-8 i 3 TECHNICAL FINDINGS-------------------------------------- 3-1 3.1 Introduction---------------------------------------- 3-1 3.2 Performance of Emergency Core Cooling System Pumps-- 3-2 l 3.2.1 Characteristics of Pumps used for Emergency . l Core Cooling Systems------------------------- 3-3 l 3.2.2 Effects of Cavitation, Air or Particulate l Ingestion, and Swirl on Pump Performance----- 3-10 l 3.2.3 Calculation of Pump Inlet Conditions--------- 3-21 1 3.3 Debris Assessment---------------------------- ------ 3-25 3.3.1 Overview------------------------------------- 3-26 3.3.2 Types of Insulations Employed---------------- 3-30 3.3.3 Insulation Debris Generation----------------- 3-41 3.3.4 Two-Phase Jet Loads Under LOCA Conditions---- 3-42 3.3.5 Transport and Screen Blockage Potential for Reflective Metallic Insulation Materials----- 3-67 1 3.3.6 Buoyancy, Transport, and Screen Blockage ! Characteristics of Mass Type Insulations----- 3-68 l 3.3.7 Effects of Combined Blockage (Reflective l Metallic and Mass Type Insulations)---------- 3-79 l I v

                                                                                                                                   . e 3.4 Sump Hydraulic Performance--------------------------           3-79 3.4.1 Envelope Analysis----------------------------         3-84 3.4.2 General PWR Sump Performance (All Tests)-----         3-85 3.4.3 PWR Sump Performance During Simulated Accident Conditions (Perturbed Flow)------------------      3-91 3.4.4 Geometric and Design Effects (Unperturbed Flow Tests)----------------------------------     3-92 3.4.5 Design or Operational Items of Special Concern in PWR ECCS Sumps----------------------------     3-92 3.4.6 BWR Suction Pipe Intakes---------------------         3-96 4       INDEPENDENT PROGRAM TECHNICAL          REVIEWS-------------------  4-1 4.1 Sump Hydraulic Performance Revi ew-------------------          4-1 4.2 Insulation Debris Effects          Review--------------------  4-3 5

SUMMARY

OF SUMP PERFORMANCE TECHNICAL FINDINGS---------- 5-1 5.1 General Overview------------------------------------ 5-1 5.2 Sump Hydraulic Performance-------------------------- 5-1

5. 3 Deb ri s A s s e s sme nt.s---------------------------------- 5-5 5.4 Pump Performance Under Adverse Conditions----------- 5-15

5. 5 Comb i ned E f f ec t',------------------------------------ 5-18 6 REFERENCES---------------------------------------------- 6-1 APPENDIX A -

SUMMARY

OF PUBLIC COMMENTS RECEIVED AND ACTIONS TAKEN APPENDIX B - PLANT SUMP DESIGNS AND CONTAINMENT LAYOUTS APPENDIX C - INSULATION DAMAGE EXPERIENCED IN THE HOR PROGRAM APPENDIX 0 - DETERMINATION OF RECIRCULATION VELOCITIES APPENDIX E - MIRRORe INSULATION PERFORMANCE OURING LCCA CONDITIONS APPENDIX F - HOR BLOWOOWN TESTS WITH NUKON INSULATION BLANKETS vi,

s LIST OF FIGURES P,A_G_E 3.1 Assembly schematic of centrifugal pump typical of those used for RHR or CSS service------------------------------- 3-6 3.2 Performance and NPSH curves for RHR pumps, head versus flow rate data normalized by individual best-efficiency-point values---------------------------------------------- 3-7 3.3 Performance and NPSH curves for CSS pumps, head versus flow rate data normalized by individual best efficiency-point values---------------------------------------------- 3-8 3.4 Assembly schematic of multistage pump used in BWR emergency cooling systems------------------------------------------- 3-11 3.5 Typical head degradation curves due to cavitation at four flow rates (Qg, Q2 ' 0 3, and Q4 )--------------------------- 3-13 3.6 Reduction in pump NPSH requirements as a function of Ilquid temperature----------------------------------------------- 3-14 3.7 Head degradation under air ingesting conditions as a function of inlet void fraction--------------------------- 3-16 3.8 Effect of air ingestion un NPSH requirements for a centrifugal pump------------------------------------------ 3-18 3.9 Schematic of suction systems for centrifugal pump--------- 3-22 3.10 As-fabricated, reflective metallic insulation components-- 3-32 3.11 Mass-type insulations------------------------------------- 3-33 3.12 Encapsulated insulation assemblies------------------------ 3-37 3.13 Jacketed insulation assemblies---------------------------- 3-38 3.14 Structural damage to railings and walls in the HOR facility following a blowdown experiment-------------------------- . 3-43 vil

i l l l l l 3.15 Erosion of reinforced concrete in the HDR facility due to direct break jet impingement------------------------------ 3-44 3.16 Blowdown damage to fiberglass insulation covering the HOR ' pressure vessel------------------------------------------- 3-45 , I l 3.17 Distribution of fiberglass insulation after an initial l l HOR blowdown test ------------------------------------- 3-46 ! 3.18 Blowdown damage to jacketed reinforced fiberglass in the i HOR blowdown compartment---------------------------------- 3-47 3.19 Schematic of jet impinging on target---------------------- 3-48 l 3.20 Centerline target pressure as a function of axial target l position for break stagnation conditions of 150 bars------ 3-51 3.21 Centerline target pressure as a function of axial position for break stagnation conditions of 80 bars---------------- 3-52 l 3.22 Composite target pressure contours as a function of target I L/D and target RADIUS /0 for stagnation conditions of 150 bars and 35 of subcooling-------------------------------- 3-53 3.23 Composite target pressure contours as a function of target L/0 and target RADIUS /D for stagnation conditions of 80 bars and saturated liquid--------------------------------- 3-54 - 3.24 Comparison of calculated target pressures with HOR , experiment V21.1------------------------------------------ 3-55 1 j 3.25 Multiple region insulation debris generation model-------- 3-56 i l 3.26 Possible variation of debris types and relative quantities l in regions of the three-region jet model 3-59 J.27 Zones of influence for debris generation------------------ 3-60 3.28 A half segment flipped onto screen------------------------ 3-69 i j 3.29 Uncrumpled foil sheet flipped vertically on screen (flow r velocity = 0.5 ft/sec)------------------------------------ 3-69 3.30 Crumpled foil sheet against screen (flow velocity

        = 0.3 ft/sec)---------------------------------------------                           3-70 3.31 Several foil sheets on screen (flow velocity
        = 0.7 ft/sec)---------------------------------------------                           3-71 I

viii

i

),

6 e ! I i t [ l i : i I

;      3.32 Approach flow perturbations and screen blockage schemes--- 3-82 3.33 Break and drain flow                impingement--------------------------                                        3-83 l       3.34 Void fraction as a function of Froude number; horizontal                                                                 '

intake configuration-------------------------------------- 3-86 ! l 3.35 Surface vortex type as a function of Froude number; horizontal intake configuration--------------------------- 3-87 l 1 ; 4 + 3.36 Swirl as a function of Froude number; horizontal intake I configuration--------------------------------------------- 3-87 I

!      3.37 Void fraction data for various Froude numbers; vertical
,           intake configuration---------------------                                         .

3-88 l

l I 3.38 Surface vortex type as a function of Froude number; i j vertical intake configuration----------------------------- 3-89 l
3.39 Swirl as a function of Froude number; vertical intake i configuration------ -------------------------------------- 3-89 ,

1 \ i 3.40 Vortex type classification-------------------------------- 3-90 l

j 3.41 BWR pipe inlet configurations as built in full-size : ) facility-------------------------------------------------- 3-97 l j 3.42 Perturbed flow schemes; schemes A, B and D used for BWR .l l inlet tests----------------------------------------------- 3-98 i 1

,      3.43 Test-average void fractions for tested BWR suction   '

t j intakes--------------------- 3-100 a i I 3.44 1-minute average void fraction for tested BWR suction I j intakes--------------------------------------------------- 3-101 i I 5.1 Technical considerations relevant to ECCS sump l t p e r f o rma n c e - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 2 4 ! ! 5.2 Flow chart for calculation of pump inlet conditions------- 5-16 1 l j 5.3 Combined technical considerations for sump performance---- 5-19 ! I ) { i i 4 i 4 l I

                                                                                                         . o.

5.4 Debris volume versus debris screen area, recirculation flow rate, and blocked screen head loss, for high density fiberglass------------------------------------------------ 5-23

5. 5 Debris volume versus debris screen area recirculation flow rate, and blocked screen head loss, for low density fiberglass------------------------------------------------ 5-24 5.6 Flow chart for the determination of insulation debris effects--------------------------------------------------- 5-25 m

a X

l l

1 l

l 1 , l ! LIST OF TABLES l PAGE l i 3.1 RHR and CSS pump data------------------------------------- '-4 l I l 3.2 RHR, CS and CI pump data for BWRS------------------------- 3-9 . l 1 i l 3.3 Types and percentages of insulation used within the l primary coolant system shield wall in plants surveyed----- 3-39 t 3.4 Insulation types used on nuclear plant components--------- 3-40 3.5 Maximum LOCA generated insulation debris summarized by break . size------------------------------------------------------ 3-61 i l 3.6 Typical volumes of primary system insulation employed----- 3-63 ( 3.7 Transport and screen blockage characteristics of reflective j insulation materials-------------------------------------- 3-64 l 3.8 Summary of transport and screen blockage characteristics ! of high density fiberglass-------------------------------- 3-77 l [ 5.1 Hydraulic design findings for zero air ingestion---------- 5-4 l

                                                                                 .t 5.2 Hydraulic design findings for air Ingestion 1    2%----------   5-6 5.3 Geometric design envelope guidelines for horizontal                     r suction outlets-------------------------------------------    5-7 5.4 Geometric design envelope guidelines for vertical                       :

suction outlets------------------------------------------- 5-8 P 5.5 Additional considerations related to sump size and placement------------------------------------------------- 5-9 j 5.6 Findings for selected vortex suppression devices----------' 5-10 i t 5.7 Screen, grate, and cover plate design findings------------ 5-11 5.8 Debris assessment considerations-------------------------- 5-13 , t 5.9 First round assessment of screen blockage potential------- 5-21 r xi i I t d

e FOREWORD This report has been prepared to provide a concise and self-contained reference that summarizes technical findings relevant to Unresolved Safety Issue A-43, Containment Emergency Sump Performance. This report was originally issued for pubite comment in May 1983; comments received were reviewed, and those of substantive technical or informational content have been incorporated into this Revision 1. It should also be clearly noted tnat this report is not a substitute for requirements set forth in General Design Criteria 16, 35, 36, 38, 40, and 50 in Appendix A of Title 10 of the Code of Federal Regulations Part 50, nor is it a substitute for guidelines set forth in NRC's Standard Review Plan (SRP, NUREG-0800), Regulatory Guides, or other regulatory directives. The information contained herein is of a technical nature and can be used as reference material relevant to the revised SRP Section 6.2.2, Revision 4, and Regulatory Guide 1.82, Revision 1. P xiii

ACKNOWLEDGEMENTS the technical findings relevant to Unresolved Safety Issue A-43, Containment Emergency Sump Performance, set forth in this report are the result of the combined efforts of the staff of the Nuclear Regulatory Commissicn (NRC), the (00E), Sandia National Laboratories (SNL), Alden Department of Energy Research Laboratory (ARL), Burns & Roe (88R), Inc., and Creare, Inc.The following persons deserve special mention for their participation and contributions: W. Butler, NRC/DSI G. Otey, SNL W. Durgin, ARL M. Padmanabhan, ARL G. Hecker, ARL P. Strom SNL D. Jaffee, NRC/DL W. Switt, Creare J. Kudrick, NRC/OSI G. Weigand, SNL A. Mullunzi, 00E F. Wind, HOR Project P. Norian, NRC/ DST J. Wysocki, B&R F. Orr, NRC/DSI In addition, acknowledgement is given to persons whose efforts are referenced herein, and to these persons who participated in peer reviews of the results obtained and the application of such results. Particular acknowledgement is given to Gilbert Weigand (SNL), who played a major role in developing investigative approaches and maintaining technical quality and continuity in those early efforts, and to G. Hecker (ARL) for his keen insight and questioning regarding the application of the results obtained, ~ particularly in the concluding phases of this USI. In addition, acknowledge-ment is given to the Ofamond Power Speciality Company and Owens Corning Fiberglass, Inc. for providing the technical data and other product line information that have been included in this report. Acknowledgement is also given to S. Khalid Shaukat (NRC/ DST), who rade valuable contribution to various sections of this report. Final thanks are given to Cindy Barnes (NRC/ DST), who persevered with me through all the concluding revisions to this document. A. W. Serkiz Task Manager

i, t

                                                                                 'A r 1

i 1 INTRODUCTION 1.1 Safety Sionificance After a loss-of-coolant accident (LOCA) in a pressurized water reactor (PWR), water discharged from the break will collect on the containment , floor and within the containment emergency sump. PWR emergency core cooling l systems (ECCS) and containment spray systems (CSS) initially draw water from the refueling water storage tank (RWST); long-term cooling is implemented by realignment of these ECCS pumps to the containment emergency sump. In boiling water reactors (BWRs), the break flow collects in the suppression pool (or torus), and the residual heat removal (RHR) and core spray (CS) i

systems take suction from intakes located in the suppression pool. Thus successful long-term recirculation depends un-the PWR sump design--or BWR suction intake design--to provide adequate, debris-free water to the RHR recirculation pumps for extended periods of 'ime. t -

The primary areas of safety concern addressed in this report are- as fellows: . (1) post-LOCA hydraulic effects (i.e., air ingestion potential) 1

(2) generation of insulation debris as a result of a LOCA, with subsequent l transport to PWR sump screens (or BWR suction strainers) and blockage I

thereof (3) the combined effects of (1) and (2) on the required recirculation i pumping capacity (i.e., impact on net positive suction head (NPSH) of , the recirculation pumps) 1.2 Backaround i t The importance of the ECCS sump and the safety considerations associated with its design were early considerations in PWil containment design. NPSH i 1 NUREG-0897, Revision 1 1-1 i

j . . t i i i l

                                                                                   .                  i i                                                                                                      t i

4 l requirements, operational verification, and sump design requirements are ( issues that have evolved and are currently addressed in the following i Nuclear Regulatory Commission (NRC) Regulatory Guides (RGs): l i RG 1.1 Net Positive Suction Head for Emergency Core Cooling l

and Containment Heat Removal Systems Pumps, 1970 ;

/ , ! RG 1.79 Preoperational Testing of Emergency Core Cooling Systems i for PWRs, 1974 1 l RG 1.82 Sumps for Emergency Cooling and Containment Sprays

!                            Systems, 1974                                                           ,
!                                                                                                    L Review of these Regulatory Guides reveals that the concerns of the NRC staff i regarding emergency sump performance evolved over time.                 Initially, in plant       l

tests were called for in RG 1.79. Then, there was a transition to [

containment and PWR sump model tests in the mid-1970s. At that time, ; f considerable emphasis was placed on " adequate" sump hydraulic performance ,f during these model tests, and vortex formation was identified as the key f f determinant. The staff's main concern was that formation of an air-core $ vortex would result in unacceptable levels of air ingestion and severely j degraded pump performance. There was also concern about sump damage or I blockage of the flow as a result of insulation debris generated by LOCAs, l j missiles, and break jet loads. These concerns led to the formulation of some }

of the guidelines sit forth in RG 1.82 (those relating to cover plates, .

r i debris screen, and a 507, screen blockage criterion), i ! e j In 1979, as a result of continued staff concern about the safe operation of l ECCS sumps, the Commission designated the issue as Unresolved Safety Issue [ l (USI) A-43, Containment Emergency Sump Performance. To assist In the ! ! resolution of this issue, the Department of Energy (00E) provided funding for i 1 construction of a full-scale sump hydraulic test facility at the Alden i l Research Laboratory (ARL) of Worcester Polytechnic Institute (WP!) (Durgin, l l I NUREG-0897, Revision 1 1-2 l

Padmanabhan, and Janik, 1980). 1 At about the same time an NRC Task Action Plan (TAP)A-43wasdevelopedtoaddressallaspept'softhissafetyissue. Potentidl oebris effects were investigated through plant insulation surveys, sample plant calculations, and supplemental experiments conducted at ARL to determine the transport characteristics of various types of insulation debris and attendant screen blockage head losses. 1.3 Technical Issues The principal concern is summarized in the following question: In the recirculation mode following a LOCA, will the pumps receive water sufficiently free of debris and air and at sufficient inlet pressure to satisfy NPSH requirements so that pump performance is not degraded to'the point that long-term recirculation requirements cannot be met? This concern can be divi 6dd into three areas for technical consideration: - sump (or suction intake) hydraulic design," insulation debris effects, and pump performance. The three areas are not independent, and certain combinations of effects must be considered as well. This report presents the technical findings derived from extensive, full-scale experimental measurements, generic plant surveys, sample plant calculations, assessment of the performance of residual heat removal pumps, and public comments received. Public Comments received and staff response to them are contained in Appendix A. These technical findings provide a basis for technically resolving USI A-43 and for developing revisions to RG 1.82 and Section 6.2.2, of the NRC Standard Review Plan (SRP, NUREG-0800).

1. 4 Summary of Technical Findings The following key determinations are derived from the technical findings presented in Section 3 below: -

NUREG-0897, Revision 1 1-3

 .. _;;..;.-- z.;; ; ;.__; ....:- :_ _      . . . . _ _

(1) Visual observations of vortex formation cannot be used to quantify levels of air ingestion. Full-scale PWR sump experiments and BWR suction inlet experiments have shown that levels of measured air ingestion were generally less than 2% under a wide range of simulated post-LOCA conditions. On the other hand, the absence of air-entraining vortices can be used to infer zero air ingestion. (2) Air ingestion levels have been correlated with the Froude number (Fr) that embodies suction submergence level and suction inlet flow velocity. Full-scale experiments have shown zero air ingestion in PWR sumps for Fr 1 0.2 and zero air ingestion for BWR suction inlet designs up to Fr 1 0.8. Envelope, or bounding, plots for estimating air ingestion levels as a function of Froude number are presented in Section 3.4 (3) Excessive air ingestion levels (i.e., > 2 to 4 volume %) can lead to degradation of pumping capacity (see Section 3.2). Use of vortex suppressors (fabricated from floor grating materials) can effectively , reduce air ingestion to 0% (see Section 3.4). For BWR suction inlets, the inlet strainer appears to act as a vortex suppressor and retardant to air ingestion. (4) RHR recirculation pump operation can be assessed using the findings and methods provided in Section 3.2. As noted above, low levels of air ingestion can be tolerated. However, pumping performance should be based on calculated pump inlet conditions for the postulated LOCA including adjustment of the net positive suction head requirements (NPSHR) for low levels of air ingestion (see Section 3.2). (5) Ingestion of small particulates does not appear to pose a pumping problem as a result of erosion for the post-LOCA circulating pumps in either PWR or BWR plants because of the materials of construction used in the impellers and casings. Pump seal systems should be reviewed from l NUREG-0897, Revision 1 1-4

T' the viewpoint of possible clogging. Catastrophic failure of shaft seals (as a result of debris generation) is unlikely because of the safety bushings built into pump seal assemblies. If water-lubricated bearings are specified or used in any of the post-LOCA circulating pumps (e.g. , in multistage RHR, reactor core isolation cooling (RCIC), high pressure coolant injection (HPCI) or high pressure core spray (HPCS) pumps in some BWRs), the seal system should be carefully reviewed. Particulate ingestion may be sufficient to cause seal failure and/or bearing seizure in these cases. (6) Surveys of plant insulation materials have shown a wide variability in the types and quantities of insulations employed in nuclear power plants (see Section 3.3). Furthermore, feedback received during the "for comment" period on USI A-43 has shown that the types and quantities of insulation have changed over time and with replacement changes made in operating plants. Thus, because of the nature and quantities of insulation materials used, debris blockage assessments become very plant specific and time dependent. (7) Estimating the effects of debris blockage requires an estimation of (a) the quantity of debris that might be generated by a LOCA, (b) the transport of such insulation debris to the PWR sump screen (or BWR suction strainer), and (c) the potential blockage as a result of flow entrainment of debris to the screen (or strainer) surface. Plant-specific studies have shown that there is a strong dependance on plant layout (which affects migration of debris) and on PWR sump design features (or BWR suction intake design). Appendix B provides illustrative sump designs and containment layout. , (8) The destructive power of a LOCA jet has been demonstrated in HDR* blowdown experiments, particularly from the viewpoint of destruction of

        *The Heissdampfreaktor or superheated steam reactor, in the Federal Republic

( of Germany; see Appendix C. NUREG-0897, Revision 1 1-5

fibrous insulation materials. Because finely shredded insulation can be transported at low recirculation flow velocities (i.e., 0.2 ft/sec) and distribute uniformly over debris screens, or suction strainers, such insulations warrent a close review for estimating post-LOCA blockage effects on pumo NPSH margin. Experiments have also shown that reflective metallic insulations can suffer severe damage from LOCA jets (see Appendix E) and that undeformed thin foils (such as those used internally in reflective metallic assemblies) can be transported at low velocities (e.g., 0.2 to 0.4 ft/sec). Information on the transport characteristics of simulated insulation debris and debris generation is contained in Section 3.3. (9) Sample plant analyses and experiments have shown that the uniform 50% blockage criterion in RG 1.82 is not adequate for the reasons noted above. Sump screen blockage (or suction strainer blockage) should be evaluated on a plant-specific basis on the basis of the insulation materials employed, and a plant-specific' assessment of potential debris t'ransport and debris screen blockage should be made. Therefore, RG 1.82 , has been revised accordingly. (10) The technical findings in Section 3 have been further refined to develop PWR sump and BWR suction inlet evaluation guidelines. These guidelines are in Section 5. (11) Methods for estimation of debris generation and transport developed in NUREG/CR-2791 (published in September 1982) are superseded by those outlined in Sections 3.3 and 5.3 of this NUREG. NUREG/CR-2791 (published in September 1982) should be reviewed from the viewpoint of later information (such as those contained in Section 3.3 and 5.3 of this NUREG). Certain assumptions previously made in NUREG/CR-2791 (i.e., insulation damage effects extending outward to a stagnation pressure level of 0.5 psi) are not supported by more recent evaluations. NUREG-0897, Revision 1 1-6

2

SUMMARY

OF KEY FINDINGS 2.1 Pump Performance Sustained operation of PWR RHR and CSS pumps, or BWR RHR pumps, in the recirculating mode presents two principal areas of concern: (1) possible degradation of the hydraulic performance of the pump (inability of the pump to maintain su'ffcient recirculation flow as a result of sump screen blockage, cavitation or air ingestion effects) (2) possible degradation of pump performance over the long- or short-term because of mechanical problems (material erosion due to particulates or severe cavitation, shaft or bearing failure due to unbalanced loads, and shaft or impeller seizure due to particulates) Pumps used in RHR and CSS systems in PWRs are primarily single-stage centrifugal designs of low specific speed. PWR CSS pumps are generally _ rated at flows of about 1500 gpm, with heads of 400 feet, and require about 20 feet of NPSH at their inlet; PWR RHR pumps are generally rated at about 3000 gpm, with heads of 300 feet, and require about 20 feet NPSH at maximum flow. Rating points and submergence requirements for the pumps are plant specific. Pump impeller materials are generally highly resistant to erosion, corrosion, and cavitation damage. Experimental results show that under normal flow conditions and in the absence of cavitation effects, pumping performance is only slightly degraded when air ingestion is less than 2%. This value would be a conservative estimate for acceptable performance and is dependent on many variables. However, air ingestion greater than 15% almost completely degrades the performance of pumps of this type. NUREG-0897, Revision 1 2-1

Submergence or NPSHR for RHR and CSS pumps (routinely determined by manufacturers' tests) are established by percent of degradation in pump output pressure. Individual pump specifications determine that NPSH required be set according to a 1% or 3% degradation criterion. No industry standard exists for the percent degradation criterion, nor for the margin between available NPSH and that required in setting RHR and CSS pump submergence criteria. Air ingestion affects NPSHR for pumps. Test data on the combined effects of air ingestion and cavitation are limited, but the combined effects of both increase the NPSH required. A value of 3% degradation in pump output pressure for the combined effects of air ingestion and cavitation appears to be realistic for assessing recirculation pump performance. The types and quantities of debris small enough to pass through screens (or suction strainers) and reach the pump impeller should not impair long-term hydraulic performance. In pumps with mechanical shaft seals, accumulated quantities of soft or abrasive debris in the seal flow passages may result in clogging or excessive wear, both of which m&y lead to increased seal leakage. Catastrophic failure of a shaft seal in the post-LOCA circulation pumps in , either PWR or BWR systems as a result of debris ingestion is considered unlikely. In the event of complete failure of shaft seals, pump leakage would be restricted by the throttle or safety bushing incorporated in these seals. There is a much broader spectrum of both design features and rated performance values for centrifugal pumps used in BWR safety systems than for those used in PWR systems. Although there is a wider variation in BWR pumping capacities, the pumps in BWR systems are also low to medium specific speed designs. They have performance characteristics very similar to those used in PWRs. Pumps in BWRs should be subject to the same technical considerations regarding hydraulic performance as those for PWR pumps (i.e. , the criteria used in calculation of NPSH and in considering the quantities of air will apply directly to the BWR pumps). NUREG-0897, Revision 1 2-2

The main bearings for BWR safety pumps are similar in construction and protection details to those of their PWR equivalents. That is, the main bearings are rolling element or ball bearings, either grease or oil lubricated. These bearings are generally protected from damage as a result of pump leakage by mechanical shaft seals equipped with safety bushings and, in some cases, downstream deflectors. This is true for multistage pumps as well as conventional single-stage pumps. As is the case for comparable PWR pumps, even a complete mechanical seal failure produces only a limited amount of leakage. The outboard ball bearings for these pumps 're protected by disaster bushings and deflector disks, and, therefore, total failure of these bearings is not likely. The BWR pumps are distinguished from PWR safety system pumps principally by the fact that multistage pumps are frequency used in BWR safety systems. When multistage pumps are used, one should be concerned about the effects of particulates and debris on the interstage bushings. In multistage pumps, interstage bushings are generally cooled and lubricated _ by the pumped fluid. For plants where it has been determined that significant amounts of abrasive particulates or fiberous debris may be transmitted from the pump inlet screen into the pumps themselves, the interstage bushing systems should be evaluated to determine whether external pressurized cooling or flushing is needed to prevent damage as a result of wear or clogging. Plant operational experience (based on periodic start-up and verification of safety system (s) operation) has shown no problems with interstage bushing assemblies even though the suppression pool water quality is less than that used for reactor recirculation . 1 NUREG-0897, Revision 1 2-3

2.2 Effects of Debris on Recirculation Capability The safety concerns related to the effects of LOCA generated insulation debris on RHR recirculation requirements can be viewed as dependent on the following: (1) the types and quantities of insulation employed (dependent on plant design.and installation) (2) the potential for a high pressure system break to severely damage or destroy large quantities of insulation (dependent on plant layout and insulation distribution, and on break-targeted insulations) (3) the potential for LOCA generated insulation debris to be transported to the PWR sump screen or BWR suction strainer (dependent on plant layout and recirculation velocity) (4) the extent to which such transported debris would result in blockage of , the sump screen or suction strainer (dependent on screen design and size) (5) the blocked screen head loss impact on RHR recirculation pump available NPSH (dependent on the material and blockage characteristics of the debris transported to the screen) The variability of plant layout, sump design, insulations employed, and recirculation requirements make debris assessments very plant specific. The results of debris considerations studied can be summarized as follows: j (1) Types of insulation vary from plant to plant and are subject to change l with time (i.e., replacement insulation may be different from the original i installation). l l NUREG-0897, Revision 1 2-4 I

(2) Generally speaking, insulations can be categorized as (a) reflective metallic insulation (both stainless steel and aluminum are utilized) (b) encapsulated, by metallic or other types of coverings, but with various core materials; typical core materials are calcium silicate, fiberglass, mineral wool Cerablanket", and Unibestos" (c) nonencapsulated insulations, which are typically fabricated as

              " blankets" or " pillows" and in which the core materials noted in (b) are used, with varying methods of attachment (d) molded insul'ations with closed-cell structure (i.e. , foam glass)

(e) antisweat insulations (typically fiberglass, urethane and polyurethane foams, and closed-cell rubber) Although encapsulation can afford protection from high pressure jet loads and missile impacts, encapsulated structures must be reviewed to assess the real degree of protection that is afforded. The characterization " totally encapsulated" can be misleading because of the variability of encapsulations ,. and attachment mechanisms provided. Thus assessment should be made to determine whether the insulation is totally encapsulated or semi-encapsulated. lesulation surveys conducted in 1982 (see Section 3.3) indicated a decreasing trend in the use of insulations such as fiberglass, mineral wool, and calciua silicate, etc., with licensees of newer plants appearing to elect to install reflective metallic insulation. However, feedback received during the "frer comment" period (June-July 1983) reversed this finding. More recently, some licensees of operating plants have elected to replace old insulation with fiberglass, and applicants for plants in the operating license (OL) review stage also have selected fiberglass. The more extensive use of fiberglass should be reviewed on a plant-specific basis to assess the screen blockage impact. NUREG-0897, Revision 1 2-5

LOCA jets are capable of high levels of insulation destruction, as evidenced by the HDR blowdown experiments (see Appendix C). In these HDR experiments, all glass fiber insulation, within 2 to 4 meters of the break nozzle of diameters up to 450 mm was destroyed and distributed throughout the containment as very fine particles. In addition, Sandia National Laboratory (SNL) has analyzed two-dimensional-break jet expansion phenomena and target pressure loads. SNL calculations correlate well with HDR data and show that significant jet loads occur within 3 to 5 L/D's* of the pipe break location. More recent HDR experiments (see Appendix E) illustrate the level of damage that than be incurred by reflective metallic insulation. These experiments revealed severe damage near the break location and much less damage at 7 L/D's from the break. Debris generation is discussed in Section 3.3.3. Insulation debris transport tests at Alden Research Laboratory (ARL) show that severely damaged or fragmented insulation can be transported at low velocities (0.2 to 0.5 ft/sec). Both fiberglass shreds and thin (0.0025 to 0.004-inch) metallic foils (if undeformed) can be transported at these low velocities. Therefore, the level of damage near the postulated break location (s) becomes a dominant consideration in assessing the type and volume of debris generated as well as in estimating transport probability. Larger , or intact pieces require much higher transport velocities (> 1.0 ft/sec). Thus determination of recirculation flow velocities within containment is an important factor in assessing debris transport (See Appendix D). In PWR containments, recirculation flow velocities on the order of 0.2 to 0.6 ft/sec can be calculated; hence, the transport of large pieces of debris is less likely. However, because the types of insulation used, levels of damage, available recirculation paths and sump locations versus break are controlling considerations, such assessments become highly plant dependent.

Hore L is the centerline axial distance from the break to the target and D is the pipe break diameter.

NUREG-0897, Revision 1 2-6

 . .                                                                                  l l

Assessment of the probabilities for PWR sump failure (NUREG/CR-3394) has also revealed that: (1) Principal attention should be given to insulation on the primary coolant system piping and lower half of the steam generators, because insulation on these components is the major source of potential debris, based on postulated break locations and possible break jet targets. , (2) Piping less than 10 inches in diameter is of secondary importance because smaller diameter breaks generate lower quantities of debris. The jet envelope and target area are reduced for these sizes. Although these findings should not be applied unilaterally, these trends are applicable to PWRs for initial debris assessments and thus provide a means to scope the magnitude of the debris generation potential. Low density insulations with a closed cell structure will float and are not likely to impede flow through the sump screens, except where the screens are not totally submerged. Low density hygroscopic insulation with submerged , densities greater than water require a plant-specific assessment of screen blockage effects. Nonencapsulated insulation (particularly mineral fiber, fiberglass, or mineral wool blanket) requires a plant-specific evaluation to determine the potential for sump screen blockage. If reflective metallic insulation is damaged to the extent of releasing interior foils, transport and potential screen blockage must be assessed on a plant-specific basis. In summary, all insulations should receive a plant-specific evaluation. Conservative methods have been developed for estimating quantities of debris, break sources, transport mechanisms, and blockage effects based on the findings summarized above. These methods are detailed in Section 3.3 and summarized in Section 5.3. NUREG-0897, Revision 1 2-7

l

2. 3 Sump Hydraulic Performance Findings Data obtained from full-scale sump tests provide a sound base for assessing sump hydraulic performance. Both side-suction and bottom-suction designs were tested over a wide range of design parameters, and the effects of elevated water temperatures were also assessed. Scaling experiments (1:4,1:2,1:1) were also conducted to provide a means for assessing the validity of previous scaled-model tests. The effectiveness of certain vortex suppression devices was also evaluated. For completeness, plant-specific and LOCA-introduced effects (ice condenser drain flow, break flow impingement, large swirl and sump circulation effects, and sump screen blockage) were evaluated experimentally at full scale. In addition, a limited number of BWR suction tests were performed. The results of this test program can be summarized as follows:

(1) The broad data base from the sump studies resulted in the development of envelope curves for reliably quantifying the expected upper bound for the hydraulic performance of any given sump whose essential features fall approximately within the flow and geometric ranges tested. , (2) Vortices are unstable, randomly formed, and, for cases where air ingestion occurs, cannot be used to quantify air ingestion levels, suction inlet losses, or intake pipe fluid swirl. The full-scale tests show that at water submergences deeper than 9 feet and inlet water velocities of less than 4 ft/sec, significant vortex activity disappears. Correspondingly, air ingestion is negligible or non-existant. NUREG-0897, Revision 1 2-8

4 (3) Based on void fraction measurements, air ingestion was found to be less than 2% in most cases. A few test conditions resulted in higher air ingestion, 2% to 8%, with or w;thout perturbations of the approach flow. Maximum air ingestion of 8% to 15% were recorded for only short time periods with deliberately induced adverse approach flow conditions of severely blocked screens. These tests revealed the importance of measuring void fraction and demonstrated the ineffectiveness of visual obs'ervations of vortices as a means of quantitatively evaluating air entrainment. (4) Swirl angles in suction pipes were generally found to have decreased to about 4* at a distance 14 pipe diameters from inlets. Swirl angles of up to 7* at a distance 14 pipe diameters from inlets were observed in some sump tests at low submergence with induced flow perturbations. (5) Hydraulic grade line measurements for all experiments revealed that the sump intake loss coefficient was insensitive to overall sump design variation. Loss coefficients are basically a function of local intake . geometry, and the measured values are consistent with those obtained from standard hydraulic handbooks. (6) Testing over the temperature range of 70* to 165*F revealed that water temperature (or previously hypothesized Reynolds number effects) had no measurable effect on surface vortexing, air ingestion, pipe swirl, or loss coefficient. (7) Vortex suppressor testing for PWR applications revealed that cage-type and submerged grid-type designs generally (a) reduce surface vortexing from a full air-core vortex to surface swirl only; (b) reduced air ingestion to essentially zero; (c) reduced pipe swirl to less than 5*; and (d) had no significant effect on the loss coefficient. These vortex suppression structures were fabricated from floor grating materials typically used for walkways. NUREG-0897, Revision 1 2-9

(8) There were no major differences between the hydraulic performance of vertical outlet sumps and that of horizontal outlet sumps of similar design geometry and similar flow conditions. (9) Comparison of the results of different scale models showed that scale modeling down to 1:4 scale using Froude number similitude adequately predicted the sump hydraulic performance variables (void fraction, vortex type, swirl, and loss coefficient) of full-scale tests. Tests on 1:4 ,1:2 , and 1:1-scale versions of the same sump under comparable operating conditions showed no significant scale effects in the modeling of air withdrawal because of surface vortices or in free-surface vortex behavior. Additionally, model tests accurately predicted swirl and inlet losses if specified Reynolds number criteria were maintained. (10) A parametric assessment of nonuniform approach flow into the sump as a result of specific structural features did not reveal any significant adverse effects (see also Section 3.4). (11) Orain flow impingement on the sump water surface resulted in extensive turbulence that tended to reduce vortexing and did not lead to increased air ingestion. (12) Break flow impingement tests produced considerable air entrainment at the water surface, but void fractions of the pipe flow were generally small, lest than 1%. In one case, a considerably higher void fraction was recorded, 6%, because of a change in approach flow to the sump caused by the break flow. NUREG-0897, Revision 1 2-10

(13) PWR sump screen blockage tests sometimes revealed slight increases in air ingestion and some degradation of the hydraulic performance of the sump, depending on the sump configuration and test conditions. However, no significant changes were noted. In each case where air-core vortices were generated, the use of a vortex suppressor eliminated the air-core vortex and reduced the air ingestion to zero or negligible levels. Thus, the effectiveness of vortex suppressors (such as submerged floor grating designs) has been demonstrated. (14) BWR suction intake tests (see Section 3.4.6) revealed that air ingestion was essentially zero for Freude numbers less than 0.6. The suction strainers typically utilized in BWR installations appear to act as vortex suppressors, thereby inhibiting air ingestion (even though air core vortices were observed at lower Froude numbers). Thus the full-scale sump hydraulic test program conducted at ARL has resulted in an extensive data base that has broad applicability and can be used in lieu of model tests or in plant tests (if the sump design being evaluated falls within the design and flow envelope investigated). Sump hydraulic design guidelines and criteria for assessing air ingestion potential are in Section 5. l l l l l l l NUREG-0897, Revision 1 2-11

. _ _ . . . _ . - . _ . . _ . . . .. . __. ._._..~......._..&_ 3 TECHNICAL FINDINGS 3.1 Introduction Before a plan for the resolution of Unresolved Safety Issue A-43 was developed, the following key safety questions were identified: (1) What are the performance capabilities of pumps used in containment recirculation systems, and how tolerant are such pumps to air entrainment, cavitation, and the potential ingestion of debris and particulates that may pass through screens? (2) Were a LOCA to occur, would the amount and type of debris generated from containment insulation (and its subsequent transport within containment) cause significant sump screen blockage and, if so, would such blockage be of sufficient magnitude to reduce the NPSH available below the NPSH required? (3) Can geometric and hydraulic sump system designs be established for which acceptable sump performance can be ensured? It was recognized that resolution of USI A-43 depended upon the responses to these questions. The effort to resolve these questions was undertaken in three parallel tasks, each designed to respond to one of the key safety questions.

The first question was addressed through an evaluation of the general physical and performance characteristics of RHR and CSS pumps used in existing plants. Conditions likely to cause degraded performance or damage to pumps performance were evaluated. The investigation of pump cavitation, air ingestion, particulate ingestion, and swirl is reported in NUREG/CR-2792 and Creare Technical Memorandum 962. It is summarized in Section 3.2 below.

NUREG-0897, Revision 1 3-1

q To address the second question, 19 power reactor plants were surveyed concerning the quantity, types, and location of insulation used within containment (see NUREG/CR-2403 and its Supplement 1). Then, calculational methods were developed for estimating (1) the quantities and sources of debris that could be generated during a LOCA, (2) the transport of such debris, (3) the quantities and properties of insulation debris that could potentially be transported t,o sump screens, and (4) head losses as a result of debris buildup on sump screens (NUREG/CR-2791). Many of the methods for tre assessment of oebris blockage in NUREG/CR-2791 are superseded by those described in this report. Experiments were conducted to estimate the onset of jet erosion damage to fibrous insulations (NUREG/CR-3170) and to determine the transport and screen blockage head losses associated with fibrous insulations (NUREG/CR-2982, Rev. 1). Tae transport and blockage charac-teristics of reflective metallic insulations are reported in NUREG/CR-3616. The third key safety question was addressed in an investigation of the behavior of ECCS sumps'under diverse flow conditions that might occur during a LOCA. The test program was designed to cover a broad range of geometric , and flow variables representative of emergency sump designs. The results are reported in NUREG/CR-2758, NUREG/CR-2759, NUREG/CR-2/60, NUREG/CR-2761, and NUREG/CR-2772. 3.2 Performance of Emergency Core Cooling System Pumps This section summarizes the general physical and performance characteristics of RHR and CSS pumps used in PWRs and RHR, CS and CI pumps used in BWRs. The summary characteristics are based on information from 12 PWRs and 7 BWRs that were sampled in the study. Effects likely to cause degraded performance or damage are identified, and the results of an analysis of these effects on pump performance are presented. NUREG-0897, Revision 1 3-2

c..,_ 3.2.1 Characteristics of Pumps Used for Emergency Core Cooling Systems The pumps used in PWR and BWR systems have different characteristics. 3.2.1.1 RHR and CSS Pumps Used in PWRs A study of pumps used in 12 PWR plants has shown that although individual pump details are plant specific, the pumps used in RHR and CSS services are similar in type, mechanical construction, and performance. Similarities in the types of pumps are shown in Table 3.1; the table lists the manufacturer, model number, and rated conditions for each of the pumps used in the plants surveyed. The column labeled " Specific Speed" provides a parameter conventionally used by pump manufacturers to specify hydraulic characteristics and, hence, the overall design configuration of a pump. As the table shows, all pumps are relatively high-speed, centrifugal pumps and are in the specific speed range of 800 to 1600 rpm, with specific speed defined as Ns = (speed) (volumetric flow)t/2/(head)3'4 _ The pumps used for RHR and CSS service have the following similarities in mechanical construction:

     ,    (1)   Impellers and casings are usually austenitic stainless steel, highly resistant to damage by cavitation.

(2) Impellers are shrouded with wear rings to minimize leakage. (3) Shaft seals are the mechanical type. (4) Bearings are grease- or oil-lubricated ball type. NUREG-0897, Revision 1 3-3

Table 3.1 RHR and CSS pump data Manutacturer*/ModeL - - Aated CondatAons-- RER CSS (RPM) (M) (CPM) Specific Plant Speed Need Flow Speed Arkansas Unit 82 1-R 6 23 WD 1800 350 3100 123e t-R 8 20 WD 1900 525 2200 851 Calvert Cliffs I-R en21 AL 1780 360 3000 1205 142 S&W 6xem11 MSMJ 3500 375 1350 1544 Crystal River 83 W eHN-104 1780 350 3000 1205 W6MND=134 1550 450 1500 1407 Cinna Pac 6* SVC 1770 200 1560 1016 Raddos Neck Pac 3" 1.X 1770 300 2200 1152 Pac e' t.Z 1770 300 2200 1952 Eawaunee 5-J 6x10xte VDSM 1710 260 2000 1222 1-3 4xit AM 3550 475 1300 1257

         ' McGaire 162          1-R 9 20 up                                        1780    37 5   3000      1144 t'-R Gu20 WD         1700    300    3400      1205 Midland 82           asw 10x12 21 Asset                                 1780    370    3000      1156 saw 6sem135 put      3550    387    1300      1467 Millstone Unit 2     I-R (No Model el                              ,

1770 350 3000 1998 C3736-4m6-13DV 3560 477 1400 1370 I-R es21 AL 1700 360 3000 1180 IOconee83 T-R da ti A 3550 460 1490 1390 Prairie Island 5-J 6x10xte VDSM 1770 285 2000 1141 T-R 4x11 AN 3550 500 1300 1210 Prairie Island 5-J 6x10ste VDSM I-R 4x11 AN 1780 200 2000 1956 162 1550 510 1100 1210 Sales et 1-R es20w 1780 350 3000 1205 C 3415 ento-22 1700 450 2600 929 !

Pac -- Pacific 1-R d Zagersoll-Rand W -- Worthington G - Gould 56W -= Babcock & Wilcom 5-J -- Syron Jackson Specific Speed is defined as N, = Speed (Flowi t/2 /(Mead)I/4 In this definitions Speed is in rpm. flow in go and head Aa f t.

NUREG-0897, Revision 1 3-4

s . _ . . _ . . . . . . . . . -. .. . . A pump assembly typical of pumps used for RHR and CSS service is shown in cross-section in Figure 3.1. Similarities in the performance of pumps used in RHR and CSS service are shown in Figures 3.2 and 3.3. Performance and cavitation data from each of the pumps listed in Table 3.1 have been plotted for comparison. Performance data are given in terms of normalized head versus normalized flow rate where the best-efficiency point head and flow are used for the reference values. Cavitation data are given in terms of NPSH required. 3.2.1.2 RHR, CS and CI Pumps Used in BWRs There is a wider variation in rating conditions for pumps used in BWR safety systems than for their counterparts in PWRs. Table 3.2 lists rating points, pump types and specific speeds for a sample of seven BWR plants. Flow rates and rated heads for the BWR pumps are in many cases significantly larger than those conditions for PWR pumps discussed in Section 3.2.1.1. In spite of these plant-specific differences, the pumps are all low to medium , specific speed designs with performance characteristics similar to those used in PWRs. Many of the pumps used in BWR ECC systems are multistage designs. Both the single stage and multistage design pumps used in BWR systems have many construction features similar to those for PWR pumps: (1) Impellers are usually austenitic stainless steel with high resistance to damage from cavitation. (2) Impellers are shrouded with wear rings to minimize leakage. (3) External shaft seals are mechanical. (4) Main bearing; may be grease or oil-lubricated ball types or oil-lubricated sleeve bearings. In the multistage designs, internal sleeve bushings may be used between stages to provide additional support to the shaft. NUREG-0897, Revision 1 3-5

E Aa j

.~

y8 6009 i r n i F-1 )

                 .. 1  -

[ Q K ' 2 NQ 0' e . me@e' 8

                      %$              ;              h    .          ;-., rt;         , ;1                 ,
                         ;__ .(

[b -

                                                                   .x
               ._(                               ._    _    .. _

q_ .

                                                                                           ... . = = .

3n WW Ml Tk ~;QQ if

                                                                             )h Jgg                            i 1;

[NS(s h

                                 ---                                        c z        s        .q                                            -

rm 1 y x. ww 9p d O 8 0888 @ apwxas se e se m: M EEik _E 7 .

                                                    ..   =:'t' Figtre 3.1 Assembly schematic of centrifugal pump typical of those used for RHR or CSS service                        .

RHR Pumps l.5 , , , , , , , 0 .- Arkansas unit a2 g b-calvert Cliffs 152 (4 _ g C-Crystal River 8 3 , j d-Cinna j e-Haddam Neck ( f-Kewaunee 1, k ,d 9-accuir l'2 g,3 h-Midland 62 . h,C, f i-Millstone Unit 2

               \                                                                I 3-oconee 83 k-Prairie Island 162 h 1-Salem #1 1.2 Li   -                                                                                                               -

3 2 Perfonnance E LO - - 50 0 W m 0.3 - 40 0 m 3 E s i

                                                                                                                                         =
     =    18   -

g 30 0 g NPSH C 0 g . 0.7 -

                                            ,/        g, j                                                                     -

20 0

w Os - - CO b,e

                                                                   ~
                                                                                                     \              k,f,d
                         ,            i             e                           i               ,        i                i g3                                                                                                                        ,,

to 12 14 46 0.1 LO 12 L4 LE NORWALIZED FLot RATE, (0/Qwp) Figure 3.2 Performance and NPSH curves for RHR pumps, head vs. flow rate data normalized by individual best-efficiency point values NUREG-0897, Revision 1 3-7

CSS Pumpo , 4 4 4 4 4 4 4 a- Arkansas tinit 82 g b-Calvert Clif f s 1&2 c-Crystal River #3 I.4 - d-c. anna

                                                                                                   -a**da' "*ck b,i,h                 f-Kewaunee g- necaire 162 h""' eland e2 C'f            i-M211 stone t? nit 2 1.3    *
                                                                /
                                                                 /                    j          3-Oconee e3 k-Prairie Island 1&2 d,g,l1-Salesel
                   -                                                                      0

f. 2 _

k 1.1 1 - I h

    -       LO
                   -                   Performance ] /                                                                             -

50 0 I E - 40 0 o M 03 - d

a \

     =
     =      0.3     -                                                   b,g                                                        -

30 0 - 6 NPSH m , g , '

                                                                                                                                                  =
                                           )

10 0 01 - v e 10 0 06 - 0 j,k f,I 40 15 L4 16 46 S.8 1.8 12 te 42 0.4

         '                                   NORMALIIED FLOW RATE, (0/0b9)

Figure 3.3 Performance and NPSH curves for CSS pumps, head vs flow rate data normalized by individual best-efficiency point values NUREG-0897, Revision 1 3-8

   . . _ - ._ a__ _;               _.c.,     _ w _ _ n     -
                                                                                                        .c .

n - - - - - h TABLE 3.2 4 RHR, CS AND CI PUMP DATA 'Rs > RATED CONDITIONS ECCS PLHP SPEED HEAD FLOW SPECIFIC PLANT MODE TYPE * (rpm) (ft) (gpm) SPEED Cooper CS VSS LPCI VSS 1760 420 7800 1675 HPCI STD Dresden (2) CS VSS 3560 585 4700 2052 : LPCI VSS 3560 570 2700 1585 l HPCI STD Edwin Hatch CS VMS 1780 670 4700 982 (1&2) RER VMS 1780 420 7700 1684 HPCI STD LaSalle (1&2) LPCS VMS 1780 725 6350 1015 HPCS VMS 1780 1569 6942 595 RHR VMS 1780 280 7450 2244 Limerick CS VMS 1780 668 3175 763 (1&2) EHR VMS 1180 525 10000 1076 _ Susquehana CS VMS 1780 668 3175 763 (182) RHR VMS 1180 600 10000 973 , HPCI STD Varies 525/ 5070 770 2940

Zimmer (1) LPCS VMS 1780 690 4750 911 HPCS VMS 1780 1347 5142 574 RHR VMS 1780 270 5050 1900 j L

STD - Steam Turbine Drive VSS - Vertical Single Stage VMS - Vertical Multistage r

4 4 NUREG-0897, Revision 1 3-9

, . . . - . .. . ~. . The technical considerations relative to hydraulic performance (i.e., cavitation, air ingestion) are the same for single stafe or multistage designs. However, because of the differences in construction details between the two types of pumps, the effects of particulates may be significantly dif ferent for each design. Figure 3.4 illustrates the main features of a multistage design typical of those found in BWR emergency cooling systems. These pumps use interstage shaft bushings which are lubricated by the pumped water and are therefore subject to wear or clogging from debris. 3.2.2 Effects of Cavitation, Air or Particulate Ingestion, and 5wirl on Pump Perforraance Several items have been identified as potential causes of long- or short-term degradation of emergency cooling pumps in PWRs and BWRs. They are (1) cavitation, which may cause head degradation and damage to impellers (2) air ingestion, which may cause head degradation , (3) particulate ingestion, which may cause damage to internal parts (4) swirl at the pump inlet, which may cause head degradation All of these effects also have the potential for inducing hydraulically or mechanically unbalanced loads. They are discussed below. 3.2.2.1 Cavitaticn Net positive suction head (NPSH) is defined as the total pressure at the pump inlet above vapor pressure at the liquid temperature, expressed in terms of liquid head (pressure / specific weight); it is equivalent to the amount of subcooling at the pump inlet. If the NPSH available at the pump is less than the NPSH required, some degree of cavitation is ensured and some degradation of performance and perhaps material erosion are likely. NUREG-0897, Revision 1 3-10

. . ~ . . . - . - - - - - e e e e _. . e ' '.6 83 7 ~~ 36 2 34-- j ,a ^4'. ,, 4C 358 E a L \ ' - 228' 2eI A~ 238A ,- f fy, M i - 36. ' 2 A 2- ~ ,-"' m / ,_ _ g.. g 3., 3 60 - ~ W ' ' d' II k "I

                                                ,       -t' 23-~

[, . BJ-asea - --e 79 a I i j1 4, M i?e.- w m, -

                                        ' T f 2
                                       ,,,                 ,ig
                                                                             'i f

kW w me

                                                                                                                                                               ?.T.'
                                                                      .                            \                               $           I                42M 3 13SC -- --               , ,                                 t      1                    -              4                42302 44--                 .m                                    .                              . g                      . e4 2

{e

                                                                                 ,l                     ,'                      a,
                                                                                                                                          -{             ---42 3A 2 3g, i,                                                                             ***
                                                                                   ;                       ;     i,t                           .
                                                                                                                                                                ,o s
                                                                                 , I,                                                          -

42 3 A- 3 i s r.. s 6358-- i l 22'9- M i l ' II4' i ---. 2 4 A l

MS-- - 9 :; ; m si g-4568- p- 423A3 I, 423a.4 24 a-- . -

                                                                                    .f*e        I                                              i 42Ja-S
                                                                                                '                                                               M79 25ta -
                                       ' 70 0 ---        s,$)" '             '                  l                                                               ,33 3 tS S A-- -          ,                    j                                                     ,,,
                                                                                                                                         , ,{                  423s..
                                       ,    6-                                 f        e '                                                    ',- ast a                    ,
                                                                                                                                                       -- 9 t

ll' Ye*

                                     ;l A, ]             g         4 - -32
                                                                                                                                                               ~

j :. aM j'

                                                                                                             >                          _ _ _.e.

28 -~ .I ) / g _, ,

1. _g, 2s -- _4 i y p \g- n;., - 2, .

                                                                                                                                                      .a
                                                                                                                                    -.                 --g 35 9                                 .                     ,,                           ' ~ I '.
                                                                                                                                   - I - - 4564 gyss .
                                                                                                           &          ,                   e-   *~             -24s0
                                                         - l                                                                                       ? - 2*+C O'                                          ta D              4)                        ,,
                                                                                                                                                   ,    - ,r.C lA -                                   % -340
                                                                        ,-               !                                                             -m yl              -              n                        [
                                                                                                                                                          --gw     ns ol.                       ,                                             ms Figure 3.4 Assembly schematic of multistage pump used in BWR emergency cooling systems NUREG-0897, Revision 1                                                       3-11

E 5 There is no standaro for identifying the NPSH required for a 'given pump. Unless there is a stipulation in the specifications, manufacturers have used some percentage (1% to 3%) in head degradation as the criterion for establishing the NPSH required at some flow condition. Tnese are empirically established values for which very rapid degradation occurs (see Figure 3.4) and when cavitation occurs severe erosion is likely to happen. Figure 3.5 illustrates the changes in pump performance at several flow rates as a function of NPSH; these curves are typical of those provided by pump, manufacturers to define the NPSH required for their pumps. As NPSH is reduced for each flow rate shown (Q1-Q4), a point is reached below,the ~3% limit at which substantial degradation begins. Fluid system designers may choose to apply some margin to the NPSH requirements for a pump when , designing emergency core cooling systems, but currently no standard margin between NPSH required and NPSH available has been established by NRC' regulations. Some conservatism may be introduced in the calculation of NPSH following guidelines established in RG 1.1 where no credit is allowed for increased containment pressure. However, RG 1.1 does not address subatmospheric i conditions in containment with respect to NPSH. The cavitation behavior of pumps changes at elevated liquid temperatures. Figure 3.6, which is extracted from the Hydraulic Institute Standards (Hydraulic Institute, 1975) shows that as liquid temperatures increase less NPSH is required by the pump. As a result, increases in liquid temperatura have two effects on NPSH: (1) the vapor pressure increases, which reduces NPSH available, and (2) the NPSH required is reduced by an amount, as given in Figure 3.6. 1 i NUREG-0897, Revision 1 3-12

3 (8 '

s, - 3%

                                   !                   \              0l lh                                             3%

l

l\ l I 02 INCREdSING FLOW RATES

                                  \g             3%

03 HEAD s y

   =            -

l \" l \ l3'A 04 l l l , l 4 NPSHR VALUES J AT 01-04 l NPSH Figure 3.5 Typical head degradation curves due to cavitation at four flow rates (Ql, Q2, Q3, and Q4)

E

8 E 1 i g 4.0 i g i [ l l ' w - { _ t 3.0 - Q g - a - W . y - g 2.0 __ 5 z - 5 - g - - { 1.0 - g _ . g - 0 ' i I

                                                                                           >                     I       I     i 100            _   150             200            250         300 TEMPERATURE ('P)

Figure 3.6 Reduction in pump NPSH requirements as a function of liquid temperature (Murakami and Minemura, 1977) ~ i h

I The austenitic stainless steels specified for impellers and casings in these pumps are highly resistant to erosion damage caused by cavitation. Erosion rates for extended operation are not significant as long as the NPSH available exceeds the NPSH requirement of the pump. 3.2.2.2 Air Ingestion The key findings derived for emergency cooling pumps with respect to air ingestion are based primarily on data from carefully conducted tests in air / water mixtures on pumps of a scale and specific speed range comparable to emergency cooling pumps.* Test data from independent programs on different pumps have been plotted in Figure 3.7 to illustrate the degradation in head at different levels of air ingestion (percent by volume). Performance degradation is indicated by the ratio of the two phase (air / water) pressure rise to the single phase (water) pressure rise.

             *All relevant test data were gathered through reviews of technical papers                   -

and interviews with pump manufacturers. Manufacturers' test data on air / water performance of pumps are sparse, and apply primarily to the development of commercial pumps for the paper industry. Although these pumps are similar to those used for emergency cooling service, test methods and results are generally poorly documented. Therefore, manufacturers' data have not been used to establish the air / water performance characteristics of pumps in this report. (Manufacturers' data and testimonials do, however, corroborate published data.) Only :ources of information meeting the following criteria were used:

Pumps must be low specific speed (Ns = 800 to 2000 rpm).
Pumps must be of reasonable design (with efficiencies > 60% and impellers diameter > 6-inch).
It should also be noted that the quantities of water recirculated in BWRs are significantly larger than those in PWRs.
Reasonable care must have been used in experimental techniques and in the documentation of results.

Test results meeting these criteria were then reduced to common, normalizing parameters and plotted for comparison. NUREG-0897, Revision 1 3-15

                    2 i      i    i     I               1     I      i  i l

Merrv a O rturakamissinemura y 1T . Q Florjancic -

                 !                                      %'~
                 =

o C 0.3 - Oh - E ct G M h 0.6 - E - E

h= 0.4 - G-C

                            -     Syrtol       9/98 .e
                $           Open Syrtols            0.6                                      -
                @           Closed Syrtols         0.8              I 0.2  -                                                          _

h C Half-Closed 1.0 Syr.cols Dashed Lines - Density Effect f I f I l I I i 1 0 4 8 12 16 20 PUMP INLET VOID TRAC"'ICN - % l l l { Figure 3.7 Head degradation under air ingesting conditions as a function of inlet void fraction (% of total flow rate by volume) t l NUREG-0897, Revision 1 3-16 t

~.-..-. . . . Figure 3.7 shows that for low levels of air ingestion, the degradation in pump head follows the curve (dashed line) predicted by the change in average fluid density due to the air content. Above 2% void fraction, the data depart from this theoretical line, and the rate of degradation increases. The data in the figure are shown for tests on single stage pumps. Similar tests show that multistage pumps degrade less in performance for comparable quantities of air. Above void fractions of about 15%, pump performance is almost totally degraded. The degradation process between 2% and 15% void fraction is dependent on operating conditions, pump design, and other unidentified variables. These findings closely approximate the guidelines empirically established by pump manufacturers: at air ingestion levels of less than 3%, degradation is generally not a concern; for air ingestion levels of approximately 5%, performance is pump and site dependent; for air ingestion greater than 15%, the performance of most centrifugal pumps is fully degraded. For emergency cooling pump operation at very low flow rates (< about 25% of best efficiency point), even small quantities of air may accumulate, resulting in air binding and complete degradation of pump performance. 3.2.2.3 Combined Effects of Cavitation and Air Ingestion Few data on the combined effects of cavitation and air ingestion are available. Figure 3.8, which uses test results from Merry, (1976), shows that as the air ingestion rate increases, the NPSH requirement for a pump also increases. The curves for this particular pump show that air ingestion levels of about 2% result in a 60% increase in the NPSH required (allowed head degradation based upon 3% degradation from the liquid head performance). NUREG-0897, Revision 1 3-17

E h 8 l.2

                                                                              ;           i                           i 7                                              SPEED   2940 RPN                                 SOLID LINES - TEST RESULTS

FLOW 480 GPN BROKEN LINE-INTERPOLATED

{ mi,i _ HEAD 368 FT. _

~                       =

s

                        $                                           3% HEAD DEGRADATION a 0%

g I - u j' g .0 p _

                                                                                   -                         N a 3.3%

O

a l l Y =0 .9 - l x - t E a - E.S %

                        $                                                I                  l E  0.8              -

NPSH REQUIRED l NPSH REQUIRED __ i l 2 % AIR (ZERO % AIR) I I I 0.7 I I l i 0 10 20 30 NPSH -FT Figure 3.8 Effect of air ingestion on flPSH requirements for a centrifugal pump , 1

3.2.2.4 Particulate Ingestion The assessment of pump performance under particulate-ingesting conditions is based on estimates of the type and concentrations of debris likely to be transported through the screens to the pump inlet. In the absense of comprehensive test data to quantify types and concentrations of debris that will reach the pumps, it ha,s been estimated that concentrations of fine, abrasive precipitated hydroxides are of the order of 0.1% by mass and concentrations of fibrous debris are of the order of L% by volume." The effects of particulates in these quantities have been assessed on the basis of known behavior of this type pump under similar operating circumstances. Ingestion of particulates through pumps is not likely to cause performance degradation for the quantities and types of debris estimated above. Because of the upstream screens, particulates likely to reach the pumps should be small enough to pass directly through the minimum cross-section passages of the pumps. Because of generally low pipe velocities on the pump suction side, particulates reaching tha pumps should be of near-neutral buoyancy and, , therefore, behave like the cump fluid. Manufacturers' tests and experience with these types of pumps have shown that abrasive slurry mixtures up to concentratiens of 1% by mass should cause no serious degradation in performance. Tests on single stage pumps similar in construction to those used in RHR service have shown that quantities uo to 4% of fiber paper stock by mass could be hendled without appreciable degradation.

   *The concentration for abrasive A10(H) was obtained from Niyogi and Lunt, 1981, in which it was estimated that 3000 pounds of precipitate would develop in 30 days and recirculate with 3.7 million pounds of water. The 1% by volume concentration of fibrous debris is based on the quantity of fibrous insulation reaching the sump screens typical of a PWR (see Table 3.4), mixing with 200,000 gallons from the refueling water storage tank and being recirculated through the pumps.

NUREG-0897, Revision 1 3-19

A major concern regarding the effects of particulates on pump parformance l and operability has been the effects of fibrous or other debris (such as paint chips) on pump seal and bearing systems. Porting within cyclone separators and the flush ports for mechanical shaft seals or wa.er-lubricated bearings may become clogged with debris. In such an event, seal or bearing

failure is likely. In the PWR plants that were reviewed, pumps used oil-lubricated or permanently lubricated bearings and mechanical shaft seals.

For these configurations, the seals may be subject to failure because of clogging, but the bearings are not. The construction of mechanical face seals used in these pumps is such that complete pump degradation or failure is not likely, even in the event of seal failure. In many cf the applications in BWRs, multistage pumps incorporate interstage bushings which are lubricated by the pumped fluid. In these applications, it is possible that excessive wear or clogging due to the presence of particulates or debris >

;  may cause bearing failure.

i 3.2.2.5 Swirl The effects on pump performance resulting from swirl due to sump vortices

!  are negligible if the pumps are located at significant distances from sumps.

Test results discussed in Section 3.4 indicate that swirl angles in the l suction pipe were typically 4* in PWR sump configurations (measured at 14 I pipe diameters from the sump outlet) and 0 to 7* in BWR configurations. RHR ! and CSS pumps are generally preceded by valves, elbows, and piping with { characteristic lengths on the order of 40 or more pipe diameters. This i system of piping components is more likely to determine the flow distribu-tions (swirl) at the pump inlet than the swirl caused by sump hydraulics. However, for swirl angle > 10* it should be noted that swirl induced by the

sump causes a higher friction loss than is the case with nonswirling flow. ;

For pumps with inlet bells directly in the sumps, vortices and accompanying swirl in the inlet bell can cause severe problems, because of asymmetric hydraulic loads in the impeller. Hence, this type of installation should be i avoided, j

NUREG-0897, Revision 1 3-20

[ , . . _ . .., 3.2.3 Calculation of Pump Inlet Conditions The steps given below delineate the calculational procedure for assessing the inlet conditions to the pump, based on the findings noted above. The procedure follows routine calculation methods used for estimating NPSH available, except that the procedures incorporate steps to allow for air ingestion effects. Figure 3.9 shows a schematic of the pump suction system with appropriate nomenclature. The procedure is as follows: (1) Determine the hydrostatic water pressure (gage), P,g, at the sump suction inlet centerline, accounting for temperature dependency and minimum sump water level. An important factor to include in determining the maximum sump water level is pressure head loss across the sump screen (see Section 3.3). (2) Based on the sump hydraulic assessment, determine the potential level of air ingestion at the sump suction pipe, a , as discussed in Section 3 5.2. . (3) Calculata the pressure losses in the suction pipe between the sump and the pump inlet flange. Pressure losses are calculated for each suction piping e?ement (inlet loss, elbow loss, valves, pipe friction) using the average velocity through each element, Vg , and a loss coefficient, K g, for each element. The total pressure losses are then N Pg = (y/144) I K V 2/2g 9 9 i=1 where y is the specific weight of water (lb/ft3), 144 is the conversion from psf to psi and N is the number of elements. ' The loss coefficients are defined as h gj K= j V /2g NUREG-0897, Revision 1 3-21

E = '? 8 4 i t 1 n p T,- Water Temperature Pc- Containment Absolute Pressure p ,

                                                                                                                               - Hydrostatic Pressure of Sump

( P 3g Suction inlet

                                                  /                   Tw                                                  Z - Water Elevation in Sump w
                                                                                '/
                                                                                /    /                                      Zs- Sump Suction Center Line y                                                                 Z,                77f                                            Elevation
                                                               / -- -- sg      r        Sump Suction                         p - Pump Inlet Flange Center Line Elevation p;p, f                                                         , yReducer Sump Pump Suction Valve I         " "

_ Pump

                                                                                                                             '/////

Figure 3.9 Schematic of suction systems for centrifugal pumn , 5

where h gj is the head loss in ft of water in element i g is the acceleration due to gravity V is the average velocity in element i in fps 9 Loss coefficients can be found in standard hydraulic data references such as Hydraulic Institute Standards (1975). (4) Calculate a value for Ppthat will be used to correct the volumetric flow rate of air at the sump suction pipe for density changes (If air ingestion is zero, Steps 4, 5, and 6 can be ignored): i P p

                                                  =P sa  -Pg+Ph         -P d where l                                             P sa
                                                  = the total absolute pressure at the sump suction pipe centerline, which is the sum of the hydrostatic pressure, Psg, and the containment absolute pressure, Pc (determined in                       -:

accordance with RG 1.1 and 1.82 for NPSH determination) P g = the pressure loss determined in Step 3, j P h

                                                  = the hydrostatic pressure due to the elevation difference between the sump suction pipe centerline, Z,, and the pump inlet flange centerline, Z, i

P h * (Y/144) IIs -Z)p P d

                                                  = the dynamic pressure at the pump inlet flange using the average velocity at the pump suction flange, V p Y (V,)2 Pd" 144                         2g l

NUREG-0897, Revision 1 3-23 _ _ _ - _ - , _ _ . . _ . _ . _ _ _ _ _ . . _ _ . _ _ - _ _ _ _ _ _ _ _ _ - - _ . . _ _ _ . . ~ - . .

(5) Calculate the corrected air volume flow rate at the pump inlet flange, ap, based on perfect gas, isothermal process op= (Psa/Pp)as (6) If a p is greater than 2%, inlet conditions are not acceptable. (7) Calculate NPSH at the pump inlet flange, taking into account the requirements of RG 1.1 and 1.82, as follows: NPSH = (P c 2*

                                 ~         ~

sg h vp) (144/Y) I where P yp

                    = the vapor pressure of the water at evaluation temperature and the other terms are as defined in Steps 1, 3, and 4 above.

(8) If air ingestion is not zero, the NPSH required from the pump , manufacturer's curves must be modified to account for air ingestion as follows: p = 0.50 (a ) + 1.0 p where o = the air ingestion level percent by volume at the pump inlet p flange. Then NPSH required (air / water) = Dx (NPSH required for water) The expression for p is empirical. It has been selected because it provides a reasonable amount of conservatism in predicting NPSH requirements in the presence of less than 2% air ingestion at the pump inlet. However, the data on which this conclusion is based are NUREG-0897, Revision 1 3-24

limited mainly to the tests of Merry, (1976), and the test data scatter mentioned in the published report are not quantified. Therefore, it is important that good judgment be used in the application of the correct factor p to plant calculations. In particular, the conservatisms in assumptions for calculating the pump inlet conditions should be weighed carefully if the calculated NPSH available for air / water operation is marginal wich respect to the required NPSH. (9) If NPSH available from Step 7 is greater than NPSH required from Step 8, pump inlet conditions should be satisfactory. 3.3 Debris Assessment The safety concerns related to the generation of thermal insulation debris as the result of a LOCA and the potential for sump screen blockage were addressed generically as follows: (1) Nineteen reactor power plants were surveyed in 1982 to identify _ insulation types used, quantities and distribution of insulation, methods of attachment, components and piping insulated, variability of plant layouts, and sump designs and locations. Additional information was contributed during a public comment period in 1983. (2) Experiments were conducted to establish the pressure conditions leading to the onset of damage to typical nonencapsulated mineral wool and fiberglass insulations, and attendant debris generation. The buoyancy and transport characteristics of both fibrous and reflective metallic insulations were investigated, along with screen blockage and head loss. NUREG-0897, Revision 1 3-25

3.3.1. Overview Assessing LOCA generated insulation debris requires consideration of the following elements: (1) The type and quantities of insulation employed. These are important because the potential for transport and blockage depends upon the insulaticn material employed. Identification of insulations employed and their distribution on piping and major components is important, as is the identification of methods of attachment. (2) Long-term cooling. For both PWRs and BWRs, the maintenance of long-term recirculation cooling is the underlying safety concern and breaks (or LOCAs) requiring long-term cooling must be assessed. For PWRs, breaks in the primary coolant system are of principal concern, and evaluations of potential break locations (and size) should be the basis for estimating quantities of debris generated. For BWRs, potential breaks in the feedwater and recirculation loop piping and steamline breaks _ constitute the LOCAs that necessitate long-term cooling. SRP Section 3.6.2, " Determination of Rupture Locations and Dynamic Effects Associated with the Postulated Rupture of Piping," should be used to identify potential break locations. (3) Possible break-target combinations. On the basis of the break locations identified in Step 2, possible break-target combinations must be assessed. Sections 3.3.3 and 3.3.4 provide guidance for defining the break jet envelope. Analyses should consider the effects in close proximity of the break (within 1 7 L/O's of the break) where insulation destruction will be highest. Beyond 7 L/O's, insulation could be dislodged in the as-fabricated state, depending on the methods of attachment. NUREG 0897, Revision 1 3-26

c. (4) Level of insulation damage and volume of LOCA-generated insulation j debris. The level of damage can be severe, partly damaged or dislogement of as-fabricated insulation segments. Insights regarding ! potential levels of destruction can be derived from the HOR ! (Heissdampfreaktor or superheated steam reactor) experiments (see Appendix C). In those experiments, destruction of insulation (particularly fiberglass insulation material) within 2 to 4 meters of the break was very severe. l Analytical studies (see Section 3.3.4) of expanding two-phase Jets also show very high stagnation pressures near the break location (within 3 to 5 L/D's). The insulations and coverings within this region will be subjected to stagnation pressures on the order of 10 to 50 bars. Small-scale experimental studies on some typical fibercloth-jacketed insulation pillows (see Section 3.3.3) revealed that the onset of I destruction (the start of tearing of the fibercloth Jacket) occurred at j stagnation pressures of 20 to 35 psi. - i Thus the estimation of debris generation is complex and material dependent. Sections 3.3.3 and 3.3.4 provide means for making such estimates. (5) Transport Characte-istics. The transport of LOCA-generated insulation debris will be controlled initially by the blowdown phase (when the jet forces will distribute debris). Long-term transport will occur during the recirculation phase when containment-flow forces (or velocities) control the transport of debris. This 1cng-term transport depends on the type of insulation, level of damage and flow velocity. Both fibrous insulation and RMI debris fragments transport at icw velocities (0.2 to 0.5 ft/sec). RMI debris generally accumulate at the lower portion of debris screen while fibrous insulation debris builds up uniformly on the screen. Thus, highly damaged insulation debris will exhibit transport characteristics significantly different (i.e., transport can occur at l low velocities) from the as-fabricated insulation segments. NUREG-0897, Revision 1 3-27 L

f i i 1 l I The plant layout, particularly for PWRs, is an important consideration - ) in the initial transport (or blowdown) phase. If the sump and break I

locations are such that the break jet can target the sump region directly, direct transport to the vicinity of the sump screen can be !

j postulated immediately. Moreover, if the break jet can target the sump l screen, screen survivability relative to jet loads should be assessed. I I (6) Screen blockage (or suction strainer blockage). This blockage is r l dependent on the material characteristics of the debris transported to 4

!            the screen and on the local velocities, which can pull such debris to                  i i

l the screen, as well as on the findings obtained for the transport of l { fibrous and metallic materials and as-fabricated sections of typical insulation materials. l There are two parts to this element: l' l (a) Will the debris be transported? Transport is dependent on !

!                    recirculation flow velocities within containment.                            .

i j (b) Will blockage occur? Blockage is dependent on the approach i velocities near the screen or suction strainer, and the approaca ! l, velocity will establish the blockage patterns that will occur. I l Shredded fibrous debris is transported at near-neutral buoyancy l conditions and is deposited (in a general sense) uniformly across a screen structure. Metallic foils (such as those used internally in I reflective metallic insulations) exhibit transport characteristics and l screen blockage patterns that are a function of foil thickness (or l j rigidity) and screen- approach velocities. Development of a blockage i model for foils is more difficult than it is for fibrous debris. t $ L (7) Head loss as a result of the estimated screen blockage. The results of j Step (6) dictate the estimating methods applicable. Results of , j experiments have shown that blockage losses for fibrous insulation j t NUREG-0897, Revision 1 3-28

1

m . - .- _...___ - .. . ._ _. _ , ~ . , _ . _ . _ . _ , _ . _ . , , - . _ , - . . , materials can be described as a power function such as AH = a Ubc t where a, b, and c are coefficients that should be derived from experimental data t(thickness) = volume of debris / effective screen area U = approach velocity Head losses that result from impervious materials (such as metallic sheets) are dependcnt on the potential blockage patterns resulting from the plant-specific reviews. For example, a PWR sump with a horizontal debris screen will incur a different type of blockage than will a sump with high vertical debris screens. Sections 3.3.5 and 5.3 provide additional information relative to these considerations. (8) Accurate predictions of recirculation flow velocities within the containment during the long-term cooling mode. These are as important as the experimentally derived debris transport velocities discussed abov;. If predicted recirculation velocities exceed transport velocities, debris will move toward the sump. An analytic method that permits estimation of velocities within containment is reported in NUREG/CR-2791. However, because of simplifications inherent in that modelling technique, a more refined analysis may be warranted if the predicted fluid velocities are within a factor of two of the transport velocity determined experimentally for each of the insulation types. That is to say, although the recirculation flow velocities discussed in Appendix 0 would predict one-half of the critical transport velocity * (thereby indicating zero transport), transport might actually occur because of flow field variabilities within containment that are not accounted for. NUREG-0807, Revision 1 3-29

3.3.2 Types of Insulations Employed Insulations utilized in nuclear power plants can ba categorized in two major groups, as discussed in the following paragraphs. (1) Reflective Metallic Insulation fhis is an all-metallic insulation design based on the concept of utilizing a series of highly reflective fails to retard heat transfer. reflective metallic insulation (RMI) is generally constructed from stainless steel, although aluminum interior foils have been used in conjunction with stainless steel inner and outer liners. Figure 3.10 provides details for typical, as-fabricated RMI segments and details of the internal foil construction. Generally RMI is manufactured in half-shell segments or other geometric shapes that are prefabricated to fit piping or other major components (reactor vessels, steam generators, and the like) and that use snap-on latching for attachment. There are currently at least four different manufacturers of RMI: Diamond Power Speciality Company, TRANSCO, Johns-Manville, and ROMET. All vendor designs vary. Some designs have open ends; others have sides sealed with foils. Interior foils range in thickness from 0.0025 inch to 0.010 inch. Inner and outer liners are generally thicker (on the order of 0.030 inch to 0.040 inch) and may be flat, corrugated, or dimpled. (2) Conventional or Mass Type Insulation Mass type insulation is an indus+.ry-derived term that encompasses a wide range of insulation materials anc differentiates them from RMI. NUREG-0897, Revision 1 3-30

V . . . . . . . . - . . - . . . . , . . . . ., - e . In mass type insulation, the materials used as the insulation filler are from one of two broad categories, fibrous and others. Fibrous insulations include. Calcium Silicate Molded Block Calcium silicate molded block insulation is a molded, high-temperature pipe and block insulation composed of hydrous calcium silicate. It is light weight, has low thermal conductivity and high structural strength, and is insoluble in water. Its density (dry) is 13 to 14 lb. per cubic foot. Its compressive strength (based on 1-1/2 inch thickness) is 60 to 250 psi. The molded blocks are provided in thicknesses of up to 4 inches and lengths of up to 3 ft. Expanded Perlite Molded Block Expanded perlite molded block insulation is composed of expanded perlite _ with reinforced mineral fiber and inorganic binders. It is an insulating material with properties similar to those of calcium silicate insulation. The average maximum density is 14 lb. per cubic foot. Its flexural strength should be not less than 35 psi, and its compressive strength dry is 60 psi and wet is 25 psi. Fiberalass Molded Block Fiberglass molded block insulation is comoosed of glass that has been foamed or cellulated under molten conditions, annealed, and set to form a rigid incombustible material with hermetically sealed cells. The density is between*7.0 and 9.5 lb. per cubic foot. Its flexural strength is 60 psi, and compressive strength is 75 psi. NUREG-0897, Revision 1 3-31

l 1 4p..,

                                                                                                                                                                                                       , .                     ..        .         ,                            - r.1                                -
                                                                                                                                                                                                                                                                      *:'ay',;;!-

aj"u g g :

                                                                                                                                                                                      ?!t.           .,                                                    -
                                                                                                                                                                                                                                                                           . y

o -

m - - i ;j..

    ?
                                                                                                                                                                                                                     -2'  .-.-                   -
                                                                                                                                                                                                                                                                           . ;y, O
                                                                                                                                                                                                                         . I y.,M i '                                 g . ;(-

e s

                                                                                                                                                '                                                                               '""1i h us7r                                                                          i

\

W ' s

    =
                                                                                                                                                                                                                                                                                                                   'i i

HALF S!!LLL (OUT-SIDE) TYPICAL OUTER SliELL w l , ~m '

                                                               - , , <                                                                                                                                                                                                                       (

[ . i g , ., ---?-'.g

                                                                                                                                                                                                                                  ,.       . . . .                      ~M                                           ,
                                                                                                                                      ~
                                              -amn                                                                                                                                                                                                                    '

g- '. ,

                                                              .. ,..,....-..,                                                                                                V                        " *-

i p ,, . . %

: v .w s , - ' 6a . < * * -

i ' AT ::.a:,';pI ^..t@- * -

i Y7.lINM -~ ~

4 3 ; i INNER F0ILS (.0025 inches thick) ! HALF SitELL (IN-SIDE) . I l Figtre 3.10 As-fabricated, reflective metallic insulation components

  . . _ . _ _    _ _ _          _ _ ~           _ _ _ _                    _ __ _ ._ .                            . _ _ _ _ _ _ _ _ _ _ _ _                            . _ _ _ _ . _ _ _ _

S

                                            .ii  .

e.-4 b5 k h.5xy.' - -

                                                                              ,m:%.g[n .h *,[.' hl , '

i I

                                                                                  ' ' # = s ggdy* p[.'f .-/                                                                                 MINERALWOOL e 9,: -
                                                                                                                                             ,/

l l I i i \ l j t 1 1

                         .x                                                                                                                                r.L ,,

t< j 1 i k .._ er. % . , CALCIUM SILICATE i

                                                                                                                                                                   -     u. i

m-~ - ~4 L c~+- w ;,us> . .., .

m::. .' .

                                                                                                                                                             ...ra..J                                         >

\ . swr 2 {

                                                                                                                                                    ,.[0~d

!i Y :t y ,g u

                                                                .. .s ._r..                                                                      . . . . .

t,- y;rt c . . . j .s. - _a.... .

                                                                  ,               f         .,                               .: %

1 "s.' .~f,

                                                              . ~. .- .....a.           ,.                   'N                         -   . :     .
                                                ~L. , J                     -
                                                                                              -) */ . $...:}. '.a**.'f
                                                                                                                                            . a+s. a                                                          \

l

                                            /            .  .  .I .    )     .        ,   . , . . ". f                . .' . T.; -                                                         NEEDLED FIBERGLASS I

f.yf ,a. f .,y,/-[q ;'s,.*

/.?

(E - TYPE)

                                                                                                         .t.V                                                                                                 !
                                                                                                         *,~. 4j[f ...r ;,                                                                                    >

i

-sG e l

                                 ,,.v,c,.

p ;, -' ' I Figure 3.11 Mass-type insulations 1 4 I NUREG-0897, Revision 1 3-33 . I i

2 i 1 f Nukon Fiberglass Blankets ! t The leading manufacturer of this type insulation is Owens-Corning which ! makes thermal insulation system called NUKON for use in the containment { areas of light water nuclear power plants. NUKON* is a blanket [' j insulation consisting of fiberglass insulating wool reinforced with fiberglass. scrim and sewn with fiberglass thread. The blanket may have {

!          secondary holding straps attached to it and wrapped completely around it. This material has a low density (i.e., 2 to 4 lbs. per cubic foot).                        i i

Figure 3.13 shows this type of insulation as fiberglass core material. ! l >

;          Mineral Wool Fiber Block Mineral wool fiber block insulation is made of a mineral substance, such 3           as rock, slag, or glass processed from a molten state into fibrous form.                          l l

The density, depending on kind, ranges from 10 to 20 lb. per cubic foot. i j The strength varies considerably with the classes of insulation. The ! ] moisture is less than 1.0 percent by volume. . 1 r

}          Other insulations include.

l j Cerabianket i

!          Cerablanket", manufactured by Johns-Manville, is a ceramic fibrous                                ,

insulation material with a density of 6 lb. per cubic foot. The l i Cerablanket is enclosed in 0.006 inch metal foil and then encapsulated

!          in a reflective insulation structure.

i i i ! 4

!                                                                                                            I i                                                                                                           i I;                                                                                                            t i

4 : i NUREG-0897, Revision 1 3-34 l l ' '

O

Unibestos Unibestos" insulation is composed of lime and diatomaceous silica taken from natural deposits. These basic ingredients are bonded with asbestos fiber possessing the tensile strength of piano wire. This composition is then encased in stainless steel sheet.

Figure 3.11 illustrates a variety of materials of this type of insulation. (NUREG/CR-2403 provides a more extensive description of insulations employed, particularly those used in older plants.) Any of the above described mass type insulations can sometimes be enclosed in an outer shell or jacket or cloth covers. The following categories are currently being used by the industry: Totally Encapsulated or Semi-Encapsulated Insulation Internal insulation in this category can be mass type materials that act as the principal heat barrier. The outer shell is generally made of . sheet metal and in some cases the ends are closed. The encapsulation is being utilized to contain the mass insulation and to ease installation and removal. Caution is recommended in assessing encapsulated insulation because of the generalized use of this category and wide variability of designs procured and installed in plants. Figure 3.12 illustrates some encapsulated insulations. Survivability under break jet loads requires assessment of the specific insulation employed and the structural capability of the encapsulation provided. The construction of semi-encapsulated insulation modules is exactly the same as that of totally encapsulated ones, except that semi-encapsulated modules are assembled in the field and clamped, not welded, together. NOREG-0397, Revision 1 3-35

Jacketed Insulations In this category the principal heat barrier (internal insulation) is the same as it is for mass type insulation. The jacket (which is usually a separate outer metal cover such as a stainless steel sheet, asbestos cloth, fiberglass cloth or aluminum) is simply an outer cover to protect the core material. Thus jacketed insulations are an intermediate arrangement between encapsulated and nonencapsulated insulation. Generally banding or latching mechanisms are employed for jacketed insulations such as shown on Figure 3.13. Urethane and polyurethane foam antisweat is another jacketed type insulation. It is a rigid cellular foam plastic that combines light weight and strength with exceptional thermal insulating efficiency. The foam is a vast cross-linked netwook of closed cells; each cell is a tiny bubble full of gas that accounts for 90 percent of its volume. Its density ranges from 1.8 to 4.0 lb. Der cubic foot. The insulation is sealed with a vapor barrier of aluminum foil or a metal jacket. . Regardless of the type of insulation employed, the assessment of debris effects must focus on types and quantities of materials present and their survivability during a LOCA, as discussed in Sections 3.3.3 and 3.3.4. Plant insulation surveys were performed in 1981 and 1982, and the results are summarized in Table 3.3. (The details associated with these surveys are in NUREG/CR-2403 and its Supplement 1.) These surveys showed that there was a wide variability in types of insulations employed, but that the newer plants were electing to utilize RMI. Moreover, based on the two BWRs surveyed, the trend appeared to be total use of RMI or totally encapsulated insulation. NUREG 0897, Revision 1 3 36

l~ 1- ~ - . . - . . ._ . . . .. . . . . . . . , a . l L I . l I i I l

~-

l " . . . , ..- ) j l i j I \ t i l , I ) I

                             ' ' ' 'j,'h , "
                                      ,                                      ,3            ' :.rg; l

y er. l ENCAPSULATED FIBERGLASS

f w t l l l , . -y. l w '. , l 1 ENCAPSULATED REFLECTIVE METALLIC l l l l i ,I Figure 3.12 Encapsulated insulation assemblies NUREG-0897, Revision 1 3-37

 ~.                                                                              _

[ l i i i I

i l

l i ) . ..... ! q.. *

                                                                                                                           .. . r                     '1                   (
                                          + )(,;-r , / -                                      -
                                                                                                                               ;#                                          l 41                             ..i.;                                     , -

1

                                                        \                         .g o                                                                                     ;

i '

                                                                                 .i n .

r ,..,c. i 4 4, % . ,'O ) l . .!

                                                                                      ,11 k l

l 3 e( * . i +. N f

                                                                                                                                  .. ;*_ fl-        .g FIBERGLASS CORE MATERIAL                                                 JACKET AND LATCH                                                  {

t i ! s 1

                                                                                                         ~4 h
                                                                                                            . I                                                          i l

l i A l 1 i i I l JACKETED ASSEMBLY I Figure 3.13 Jacketed insulation assemblies NUREG-0897, Revision 1 3-38

Tabla 3.3 Types and percentages ef insulction used t:ithis the ! primary coolant system shield Wall in plants surveyed -Ii E

s !

m )' i o e ----

                                                                                                                  -Types of Insulation and Percentage *                                            - - -     - - - - - - - - - - - -

I e y Mineral Calcium Reflective Totally Fiber / Wool Silicate Unibestos ! si Plant Metallic Ehcapsulated Blanket Block Block Fiberglass .I 1 ' i 1 Oconee Unit 3 98 - - -- - 2 . S Crystal alver Unit 3 94 5 1 - - -- w Midland Unit 2 78 -- -- -- -- 22 , Neddam Neck l 3 - -- -- 959 1 , l Robert E. Ginna - -- 5 80 10 -- l M. 3. Robinson I l -- - -- 15 85 -- Prairie Island Units 1 & 2 98 - -- -- -- 2 ; Newaunee 61 -- - -- 39 -- solem Unit 1 (

                                                              - 39                                                8                                         53**      --        --                                       -                       5 McGuire Units 1 & 2                         100                                            --                                           -          --       --                                        --                     !

t.a Sequoyah Unit 2 100 -- -- -- -- -- c'a maine Yankee 13 - 4a 25 13 1

l l Millstone Unit 2 25 35 5 30 -- --

St. Lucie Unit 1 { 10 - -- 90 -- -- 41 59 Calvert Cliffa Unita 1 & 2 -- -- -- -- Arkansas Unit 2 46 53 - - -- 1 Waterford Unit 3 15 85 -- -- -- -- Cc. twr 30 70 - -- -- -- WPPSS Unit 2 100 -- -- -- -- --

                    *Toler. co is + 20 percent t
                   **Both totally and semi-encapsulated cerahlanket is used, however, inside containment only totally                                                                                                                             !

encaps.utated is employed. '

                    %nibestos is currently being replaced by calcium Silicate. Ilowever, both tygws of insulation have the same sump blockage char.acteristics.                                                                                                                                                                                     I
                 -                                                                                                                                                                                                                                l i

l i i t t -

                                                                                                           .                                                                                                                                       o

 .... .J.; ; . . . ' ~ ~  ~E          :___
                                                                           ~
                                 .                                               - .          . . _ . . , m_ n.  -.

Table 3.4 Insulation types used on nuclear plant. components

Coolant Coolant S. G. S. G. Pressurizers Vessel Piping Pumps (less bottom Bottom head)

EIE.E. . Haddcm Neck Rm C C C C C IP-2 & IP-3 Rm C C C C C Maino Yankee Rm C C C C C Mil 10 tone-3 Rm C C C C C YCnk0e C C C C C C Palisades Rm C C C C C W31f Creek Rm C C C C C Ft. Calhoun Rm C C C C C C311cway Rm C C C C C R0binson-2 Rm C C C C C TurkCy Pt-3 Rm C C Rm Rm C Turk 0y Pt-4 Rn C C C C C St. Lucie-2 Rm C C C C C Watcrford-3 Rm E E E E E S;uth Texas 1&2 Rm C C C C C _ 500 Onofre-1 Rm C C C C C Ginna Rm C R:n Rm C C Marble Hill Rm Rm Rm C Rm Rm .& C , Afl0-2 Rm Rm - Rm C Rm Rm MiE.E. . Licorick 1&2 Rm C C N/A N/A N/A Fitzpatrick Rm C C N/A N/A N/A PCrry 1&2 Rm C C N/A N/A N/A Monticello Rm C C N/A N/A N/A Hatch-1 Rm C C N/A N/A N/A , insulation Lecend: RJ - Reflective Metallic Insulation C - Conventional Insulations (e.g., fibrous & mass materials) E - Encapsulated Insulation

Based on material obtained during a public comment period; may be obtained by writing to Generic Issues Branch, NRC, Washington, DC 20555.

NUREG-0897, Revision 1 3-40

_ . . . . . ~ J _. L . _ _ . . ~ ._ _. -c. :- - - - - - - -- ----- -------c----- . ;;wr ;s _. .;z 7.w However, comments received during the public "for comment" period associated j with USI A-43 (June-July 1983) presented a changing picture (see Table 3.4). j Some older operating plants (e.g. , Monticello) have been reinsulated with I fibrous insulation. Newer BWRs (e.g., Limerick) are being insulated with fiberglass, and the increasing use of fiberglass is evident. Replacement of selective insulation also occurs during, or following, inservice inspections. These recent observations re emphasize the large varibility of insulations employed, the plant-specific aspects associated with insulations used (plants

!                  handle insulation on a site-specific basis and changes need not be reported),

I and the time dependency factor. As new insulation products are developed new materials are being introduced into nuclear plants. 3.3.3 Insulation Debris Generation Jet impingement forces are the dominant insulation debris generator. Other contributors, such as pipe whip and impact, have been studied and shown to be of secondary importance (NUREG/CR-2791). The criteria for defining break or rupture locations should be consistent with ! the requirements of SRP Section 3.6.2, which provides guidance for selecting the number, orientation, and location of postulated ruptures within a containment. The safety concerns associated with debris relate to ensuring long-term recirculation capability. Therefore, for PWRs, the postulated breaks of concern are those in the primary coolant system and in components (or other systems) that are connected to the primary coolant system. For BWRs, the i postulated breaks of concern are in the feedwater and recirculation systems and in the steam lines. The destructive nature of high pressure break jets has been experimentally a demonstrated in blowdown experiments conducted in the HOR facility (see l Appendix C). Figures 3.14 and 3.15 show damage to reinforced concrete structures"in the HOR. Figures 3.16, 3.17, and 3.18 show the damage to insulation and insulated components in the HDR. NUREG-0897, Revision 1 3-41 '

These blowdown tests (blowdown was from 110 bars and 280 C to 315*C, under steam and subcooled water conditions) revealed that all glass fiber insulation was destroyed within 2 meters of the break nozzle and distributed throughout the HDR containment as very fine particles. In addition iron wrappers were thrown away from vessels within 4 to 6 meters of the break nozzle, with glass fiber untouched. With enforced shieldings (steel bandages) around the vessels, the damage was reduced. Mineral wool insulation that was encapsulated in iron plate, withstood the rough blowdown conditions well. Break sizes 200-mm, 350-mm, and 430-mm diameter have been investigated. 3.3.4 Two-Phase Jet loads Under LOCA Conditions Determination of the extent of potential damage requires estimation of pressure and flew field forces resulting from the expanding jet. On the other hand, the flow field for a two phase jet is extremely complicated and multidimensional. The jet impingement model discussed in this section is based on a study of HDR experinental data by Sandia National Laboratory. This model is under peer review by the ANS-58.2 Committee on Pipe Rupture _ and has not yet been incorporated in SRP 3.6.2 as an endorsed approach. Sandia National Laboratory has analytically studied two phase jet impingement on targets over a range of pressures and temperatures representative of postulated LOCAs for BWRs and PWRs. Those results are reported in NUREG/ CR-2913. In the expanding jet flow field, there are three natural divisions.of the j field (see Figure 3.19). There is a nozzle (or break) region where the flow i chokes. In this region, there is a core at choked flow thermodynamic properties that projects downstream of the nozzle at distances that depend on the degree of subcooling. Downstream of this region there is the free jet region. Here the jet expands almost as a free, isentropic expansion; the flow is supersonic throughout this entire region. The free jet rejion terminates at a stationary shock wave near the target. This shock wave arises because the target propagates pressure waves l l NUREG-0897, Revision 1 3-42 w _

t I l I 1

f. ~:

t, z - * - 4 w \ 'El f i -

                                                         *~l<.Q}.g,&

y. '

                                                            ' g}.                 ' *)! Mt                                                                         '

l

a ' z1 dh'g( i

                                                                                    .*                                                                               f
                                                                                                                                                                     \
                                       ;                          i;.:, "                                                         '
                                                                             +g;
                                                                  - .= -a                                                                                         --
                                                               &=S. . . . .           .                            %                                                 l
                                                          . - . .~ en.s:             -

p.g;9:. , . ,_

                                                                                                                                                                     \
                                                                                                                                                                     \

l l Figure 3.14 Structural damage to railings and walls in the HDR facility i following a blowdown experiment l l i l i NUREG-0897, Revision 1 3-43

i l

                                                     ,o                  ,,
                                 .C.*-,
                                ., r                                                                                        .-

l%g - L (-

l

j %..- . ) 4< .

                                                                 .e
                                                                                                                           $ [. .   '

I l % i f ' _.. -1 !" l l ! % b' , l v* ,

                                                                                                                                    ,s                                                  ;
                                       ']

f '.

50

                            '~ ' ., . r.
                                                      %~

N gf 49 b 's a L. -

                                                                                                                        *e T,, s e .
                          . , . , . . . ,_                                ,,I          s            ,

j ' 17 f* O

                                                                                        ,-   ,4 a

r;'.. ,.>!i g ,'<'.

1 . s l

                                                            . . , $)
                                                                                .    #. x                                                                                               ,
                                      -         3  .-                 ..

l l Figure 3.15 Erosion of reinforced concrete in the HDR facility due to direct break jet impingement l NUREG-0897, Revision 1 3-44

.l                                                                                                                             .

2 l I I I I 4 4 , ; h ,b. M , .' , if I i ; , ;+ . O{t?f.

^^

t 'y l I

                                      ;f T * **h       .'.**'                                                            /
                                                                                                                                   'l I                                      g      ,j
                                              " J?-                                                                                  '

o y ,. i ! Iw.

                                                              *P t.

l

94. .. . .

                                                                                                                .                    J j

e Figure 3.16 Blowdown damage to fiberglass insulation covering the HDR pressure vessel i NUREG-0897, Revision 1 3-45 l

e e j . I l i i 'l

                                                                                                                                                                                                                                                           *~
                                                                                                                                                              ~k ' ' f, f                        4h,.b.,., 'h,$}                                                   4             ,- {[               I,
                                                                                                                                                                     )n_- '

i i i,g

                                                                                              % .. ., ,' n, i 'g.1 .,                                      *
                                                                                                                                                                                                                                                                      .x e                                                                                  o .s                                                                                                          g, . , . ' ,. n'^-..k.f-

[ ,, * '# ^ -.

                                                                                                                               - s                  '#
                                                                                                                                                           ,    ,$jj f[,, ' $                  grga,.., .9 )' h ' I' .['           I
                                                                                                                                                                                                                                                               .                     ..f,,-'.,..                      .        g!
                                                                                  ;.;fyy....?,            _k,.;;J,                       .q,   ')fh'1.*.
                                                                                ? ,,. H '.7                                4               L .                   ),. .,i.lI:~m 5-
                                                                                                                                                                                       ..)*               ; ..;; p' , .?"M.,$                      / i.                   p: .;: '% c :.v;Jnf.

8

                                                                                                                                                                                                                                                                                       .5     i ...
                                                                                        ';,,., , .-                          N9              .                .-                        ..                     *, ,                       . 45 , ' , ' g"- **[                                                     ' :.'              ',._.
                                                                                                                                                     '"; ^                                                                 ;     *"'            g: .',

f. [. .,5 .

                                                                                'l..,,.                                                                                                                                                                                                                     ,,
                                                                                                                    ' . [. ., (. ,c ~ ... , ..e
                                                                                                                                                                                                                                         , j.. 4 g * ., . J#4. ;'.,*,.                                                                '    /
                                                                                             ..                                                                                                                     ,, .. ... .. .. . - ~                                                                                  . ..                  .
                                                                                            .O.. .          r.,.' A ,[.. y. .0 $ ( ' \x ,s p j#
                                                                                                                                                                                                                         ?'.%                                                                             ;6 ,                              .
                                                                                             ..,.             .' ' s,, ,'7 M .,,g k , ,>
                                                                                                                                       .,.,..'.,"e',,                                          -                       -                            '
                                                                                                                                                                                                                                                         ' ' ,, , 6 . .*. . , - ,$

x .,. '

                                                                                         .u .                   -                                                                                                        -

7

                                                                                                                                                                                                                                                                              . .f-                     .,                           ,-
                                                                                                                                                                                                                                                   .' '. ='                                                                      b 'it/

T i y *-..s *.. s

                                                                                                                                                .",\g                                    . ,              7 g ..J.                                                                          ,g                   -?                                                           !
                                                                                             ;r
                                                                                                               - 7; . 4.                w f L..                              ..--                                    ""' --'                                        . . ,; . . . *F                                      :.4 9.

s 4 . . . ( z y.. a. ;. . ).. :?,:,.,. . . ;,..:.x,,; .

                                                                                                                                                                                                                                                                                                    ..             p ' .,g;y::

n .v t .. ,. ., . . . s= g . , - a . .. . . v. . _ _ . , ,a.. . ; .g ,. m,

;(

f.' .

                                                                              ;                 {, ;.                                                                        , .-
                                                                                                                                   ;:.l.,- : 4,-Q;,.', , % ; _.                                                               ;. ;y;.          . . jip.g                      :.' ;. jy(; . jf.y.7                                        '.
                                                                                   . .,                      ~- <
                                                                                                                                                  . . _ .                                                                                                                                                                        n ,.,.;
                                                                                                                                  - s :. . .                                                      .
                                                                                                                                                                                                                              . . .! .,Vl                ar                                             tn .n m J. ,
                                                                                                                                            . t . i, ;g e;                . .,                                                                                       5:,,;.*+.\=               .

l 2,..

                                                                                                                                 '-.                                                                                         , , . , s. . ,- -,.                                                                                                                               l
                                                                                                                                               .;;h.                 ;;      '.                                                                                                                               ,

3...^. as- - , s.' .-. i;4 q$g" . ,',- . . % .. J

 \
                                                                                                                                                                                                               .,'           MV.e. :? ,A . - .        .                                             ' -

, .. su c ~ l l .f: l ~; l.; ' *. -

                                                                                                                                     ., ,                                                                                      ' ly. f  ? {                                                                               _ , , '6           ,                                l
                                                                                                                                 ;.- '. y.                       ..

t 3; . . , .,,

                                                                                                                                                                                                                                                                                                            .;S,,y Q,                                                                      f' :                                       , .I                r                                                          ~f                       ,

7. h ;

                                                                                                                           ' 1' .;.'=. .*

z;y n . >- . n ,.

                                                                                                                                                                                                                                                                                                                             .,,,.                                              t a

y . , 1 s.s .. 4.. -. ... ,

                                                                                     . v ._ ~.                                    . .g .                                                                  ,

w. . ;- .

t i __o.. - .

                                                                                                                                                                                                                                                                                          ~
                                                                                                                                                                                                                                                                                                      .s e~-                                              i
                                                                                                                                                                                                                                                                 +.. n d.a ; .                                        . . s                                                    ;

9 . x' ' ^r..

^.

d

                                                                                                                                                                                                                    ..g',.
                                                                                                                                                                                                                                                             ' : es '.,J.6                           ;

4e

                                                                                                                                                                                                                                                                                                                            .y                                              .,

i

                                                                                                      ,; .g              _.,t_                                                                                     .. =,. } : - r j                                                                                                                                                                                                ~

l, ,,

                                                                                                                                                                                                                                                                                                   };                        ,

l

                                                                                   ' ;[ > % .;; } [, ,y.,                                                                                                                                                                                                                  ' &

Q Pia ~ 1 - ,; wf)&.. ' h. h l e

                                                                                                                                                                                                                                                                                                                                   ,....,1
                                                                                                                                                                                                    . g ).:.i.                                       9vt- -

r.g. -

                                                                                                               . ; _ '7. --       -

p y, . , .

3 l

                                                                                              . .            . ...                                                                                                              .                                                                .                                                   n                         .
                                                                                                                                                                                                      . q . . Q, *.-
                                                                                       .          a. .. g . . 4 -, .;.            . .,. ;., :. ,; , , :_,                                                                                                                                                       -(                                   $

t

m. .

                                                                                                                                       ...             g,                  ..                                   .                                                                                                       .
                                                                                                    ,-t-. . . . . , ; ..

t c ... . . . .

                                                                                                                                                                                           ..v          ,
                                                                                                                                                                                                                             ~,,.                                                                           .. + -

l

                                                                                ,        3:                                ,
                                                                                                                                       #. ,         ,                '..j,.                   .
                                                                                                                                                                                               .          .;:qn                                                                 . , y. . t. v u                                           , , .

l f, . .i Q.Q .. ..p. .t ., [ ,' ' ; .Iy- ,' [,..h.. h. .p; ) h. R.: > p ., <

.s.

m . , y.. . ,s ,._.,..t . . ;.

                                                                                                                                                                                                                                                     . ..                          r-              g y I

Figure 3.17 Distribution of fiberglass insulation after an initial HDR blowdown test NUREG-0897, Revision 1 3-46 [ l ' l l l l ,

1 d l l 3 . g ute-. *

                                                                                                                     ,                                            t*ygg 4
                                                                                                              "' k,*..
                                                                                                , _,, e9
                                                              ,                                                                                     ."[' ,' (.NI '

y- / < 9 4 -

                                                                                                                                                                          .,*    .O*
;                                                                                            y                                                     Tg-s ..;, ;

a. .

                                                  . f. ~
                                      \                                                                                                              '
                                    \

l. .. , , ,

                                                ./c y~
                                                                                       / ' '-                                      ..                   .

4. a-

K ;f, .

p[}

k. lg, -

x - W._ g, d 'Jg. y ..y

                                                                         *[..
                                                .--* >f           83D                      i
                                                                                                                                                . gp%.

I i I l ,! Figure 3.18 Blowdown damage to jacketed (sheet metal cover) reinforced i (with wire mesh) fiberglass in the HDR blowdown compartment i ) .i i i I I 1 NUREG-0897, Revision 1 3-47 l

TARGET STAGNATION PRESSURE, PO TARGET 2RESSURE OISTRIBUTICN. PT E Q ADIUS MEASURED AT TARGET SUBSONIC FLOW N\ TARGET

                                  -                                                        S^

ISENTROPIC FLOW L SHOCK (SUPERSONIC) WAVE JET CORE - STAGNATION POINT Y O'.

                                                               '                         BREAK NOZZLE LCHOKED)                      _

j

                                                               ~' W        PIPE DIAMETER LARGE VE3SEL        -
                                             . . // Y /// Y/ Y               /Y Pg = Stagnation pressure at break ATg = Subcooling of stagnation temperature at break X, = Stagnation quality at break Figure 3.19 Schematic of jet impinging on target NUREG-0897, Revision 1                3-48

upstream and, thus, produces a pressure gradient that will direct the fluid around the target. Downstream of the shock is the target region where the local flow field imposes a pressure loading on the target. Depending upon the upstream flow conditions and the L/D's of the target, there may be a substantial total pressure loss across the shock wave. Thi; loss arises because of the irreversible physics that characterize the shock. The pressure loss across the shock and radial velocity components can lead to negative pressure loads across the target, which can lift away materials (such as insulation segments) from targeted components. The HDR tests revealed evidence of such loadings. NUREG/CR-2913 addresses the centerline behavior of two-phase jets and the radial loading for axisymmetric impinging two-phase Jets. The method developed for calculating centerline behavior indicates that the jet stagnation pressure at a given target distance from the break (in terms of L/D) is a function of the stagnation pressure and steam quality or the degree of subcooling in the vessel. This functional dependence (on pressure and subcooling) largely disappears at about 5 L/D's from the break. At - approximately 7 L/D's downstrean of the jet origin along the centerline of the jet, stagnation pressure falls to roughly 20 psig regardless of the break thermodynamic conditions. Two-dimensional pressure distributions were calculated and are reported in NUREG/CR-2913. These results indicate that the region targeted by an impinging two-phase jet is highly dependent on the thermodynamic conditions at the break. The constant pressure contours (as a function of target L/D) NUREG-0897, Revision 1 3-49

form complex shapes in space. Figures 3.20 through 3.23, which are i reproduced from NUREG/CR-2913, illustrate axial and radial pressure distributions of an expanding jet representative of PWR and BWR blowdown conditions. Figure 3.24 is a comparison of Sandia calculations (taken from NUREG-2913) with HDR experiment V21.1. The significant findings to be derived from the calculations contained in NUREG/CR-2913 are (1) Target pressure loadings increase asymptotically at L/D's less than 3.0 to break exit pressures. At L/D's less than 3, survivability of insulation materials is highly unlikely. (2) At L/Ds from 5 to 7, the centerline stagnation pressure becomes essentially constant at approximately 2 1 1 bars. (3) The multidimension pressure field loads the target over a large region (see Figures 3.22 and 3.23); this region may be approximated by a 90* _ jet cone expansion model. A hemispherical expansion model could be another approximation for this expanding pressure field. These two-dimensional calculations do not support the use of the Moody jet model (a narrow jet cone) for target close to the break locations. The two phase jet modelling results and the levels of insulation damage I evidenced by the HOR experiments lead to the development of a three-region jet-debris generation model which is shown in Figure 3.25. Region I (1 3 L/D from the break) is where extremely high levels of destruction would occur due to the very high break jet pressures (see also Figures 3.20 and 3.21) and total destruction can be assumed to occur. Region II (3 < L/D <

7) is a zone where high levels of damage (or destruction) are possible; but with the recognition that the types of insulation employed (e.g., reflective NUREG-0897, Revision 1 3-50

Pn = 150 BARS I.I i . . . . . . . . 1.0 ^'o " 7 50 AT, = AT, = 35

           *90    -                  '                                            o AT,
                                                                                           = 0
                                                                                               'S                  -

x, = 0.333

                                                                                     *o    "          75
           .80    -                                                                                               -
           .70    -
                            /
                              /                                                                                   -

o .60 - - c. N

           .50   -                                                                                                -

N Q.

           .40   -                                                                                               -
           .30   -                                                                                               -
           .20   -                                                                                               -
           .10   -                                                                                               -

0.0 O.0 2.0 4.0 6.0 8.0 10. L/D Figure 3.20 Centerline target pressure as a function of axial target position (L/0) for break stagnation conditions of 150 bars and various subcoolings and qualities. L is the target position, O is the pipe diameter P z is the centerline pressure, and P, is the stagnation pressure at the break. NUREG-0897, Revision 1 3-51

P = 80 BARS 1.1 . . . . . . . . . 1=0 #'*

AT, = 50 of = 35

            .90   -
                                                                      $     lO             -

x, = 0.333

                                          ..                            x,  =   0.750     ,
            .70   -                                                                        -

o .60 - Q.

     '      .50   -

N Q.

            .40   -                                                                       -
            .30   -                   -
            .20   -                                                                       -
            .10   -                                                                        -

0.0 O.0 2.0 4.0 6.0 8.0 10. L/D Figure 3.21 Centerline target pressure as a function of axial position (L/D) for break stagnation conditions of 80 bars and various subcoolings and qualities. L is the target position, D is the pipe diameter, P is the centerline pressure, and P g is the stagnation presIure at the break. NUREG-0897, Revision 1 3-52

- _ . . . . . . . _. _ . _ . .~ . . 7- - tBARSI A= 1.0 __ B= 2.5 C= 5.0 D= 10.0 6- - Em 15.0 Fa 20.0 __. G= 25.0 Hs 30.0

                                                                                                                                         !=   35.0 5-     -            =                                              =

a= 40.0 K= 45.0 L= 50.0 M= 55.0 N= 60.0 4-- * *
Os 65.0 P= 70.0

                                   --                                         .                       .                       .          0=   75.0 C                                                                                                              R=   80.0 s 3-        -             =                                 =                *                      -

a -- . . . . . 2- - - = * < = * - 1- a = * - - q c

                                                  ~ ' = ' .s
                                  ~

P o=150 AT o= 35 .L c/D= 1.07 0- l l l l l l l l l l l 4 I 1 2 3 4 5 6 . RADIUS / D Figure 3.22 Composite target pressure contours as a function of target length / jet diameter (L/D) and target radius / jet diameter (RADIUS /D) for stagnation conditions of P = 150 bars and 35 degrees of subcooling. Smoothlin0sconnectinglike alphabetic letters form an approximate pressure contour corresponding, in bars, to the pressure versus alphabetic letter key. This contour is approximate and is only informational. NUREG-0897, Revision 1 3-53

 ._   . --                 -           -        . ~ .       . . .            - . - . . . . _ . . . . . . . .       .   ..     . . . . .        . ..   .. ..

7- - (BARSI A= 1.0 8= 2.5

~~

C= 5.0 O= 10.0 6- - E= 15.0 F= 20.0

~~

G= 25.0 H= 30.0 1= 35.0 5- - .Ja <0.0 K= 45.0 L= 50.0 Ms 55.0 N= 60.0 4- - =

On 65.0 P= 70.0

                         ..                                   e                                       a                              Q= 75.0 O                                                                                                                    R=    10.0 3-         -

s e e J .. e e 2 . e e : e . e e a

                         ..          e
                                    ,s        e          e           e                   .

1 .....e . . .

1 80 AT o= 0 L c/D= .50 P o=

                                    "                                                           l         l     l    l    l        l 0                        l         l         l         l          l Al         1                  2                    3                    4           5            6 RADIUS                          /       D Figure 3.23 Composite target pressure contours as a function of target length / jet diameter (L/D) and target radius / jet diameter (RADIUS /D) for stagnation conditions of P = 80 bars and saturated liquid.                       Smooth lines connectin0 like alphabetic letters form an approximate pressure contour corresponding, in bars, to the pressure versus alphabetic letter key.                                                     This contour is approximate and is only informational.

NUREG-0897, Revision 1 3-54 _ t . _-

Measured Pressures 12 21 Bars 7- 9 1 Bars 3 ! 1 Bars 6 1.510.5 Bars o g 5 - a: e 5 E

                  .!!! 4 0

5 3o HDR Force Plate Location 3 -

2 -a

                             \                    t 4

Calculated Pressures for

1 Pg= 80 Bars Er ATg = 0*K J , 15 10 5 2.5 1.0 Bars 0 ' ' ' ' ' ' ' ' ' ' ' ' 8 0 1 2 3 4 5 6 7 Radius / Jet Diameter Ratio

                   "      O Figure 3.24 Comparison of calculated target pressures with HOR experiinent V21.1 NUREG-0897, Revision 1                    3-55

e . I I I I I l Boundary of a Right I l Circular Cylinder Originating at the l Postulated Pipe Break I I 1 i I

                               #                 i REGION I             REGION 11            REGION lli g                                             l Total           High Love's of    l Dislodging in l Destruction      Damage Possible         "as-fabricated" I Materials and     i       Pieces or     l I       Attachment             Segments       I
         + 3 L/ D's              Dependent       l l

Note: 2 1 4 8' L/D' l Pressure Isobars l f 25 15 10 l Shown Are Calculated Target Pressures for i l Break Conditions of l 2.5 150 Bars and 35'K I l Subcooling 2- l (- l l

   "'  ~

4-VN l \ l % l 8 R = Radius of Circular

       -                                         l                     1         Flat Plate Target 6"

1 Bar L = Distance From Break l to Target

7 L/D's D = Diameter of Broken f l Pipe l

l Pstag = 0.5 poi, or _s Mejor Wall Boundary \ ~3 1 I

                                                                'N       \

l Figure 3.25 Multiple region insulation debris generation model NUREG-0897, Revision 1 3-56

                                                           ;. =..                     . - - . .

u metallic, fibrous, foamglass, etc), methods of attachment, whether the materials are er.capsulated,etc. are factors which should be considered in estimating the types and volumes of debris generated in Region II. Region III (L/D > 7) is a zone where destruction (or damage) will likely be dislodgement of insulation in the as-fabricated mode, or as modules. Beyond 7 L/0, break jet pressures have decayed to 1 to 2 bars. It should also be noted that superimposed pressure field on Figure 3.25, is representative of a PWR primary coolant system break. BWR jet expansion fields decay more rapidly (see pressures in Figure 3.21 versus Figure 3.20). i t Despite the calculational simplification afforded by a three-region model, determination of the types and quantities of insulation debris will always

be material (or type) dependent. Figure 3.26 has been constructed to illustrate the possible variation of debris types as a function of distance

]                   from the break jet and the relative quantities of different types of possible debris. A quantified debris distribution model would require extensive experiments designed to develop s'uch data; these do not exiat. On
  !                 the other hand, results from HDR experiments (see Appendices C,E, and F) do                                      _.

provide insights regarding debris generation and were used to construct Figure 3.26. 4 First of all, the assumption of severe damage or total fragmentation within 3 L/D's is supported by experiments and is applicable to both reflective metallic insulation (RMI) assemblies and fibrous insulation assemblies. However, the hypothesis of " exploded" RMI assemblies releasing free, or undamaged interior foils (which can transport at very low velocities), is not supported by the experimental evidence reported in Appendix E. Pursuing those potential levels of damage expected in Region II (see Figure 1 3.25) it appears that the RMI debris could consist of damaged inner foils and damaged assembly or components that were the result of further LOCA damage. Experimental data available for fibrous insulations indicate that shredding ar.d damage can extend into Region II, with such damage decreasing with t distance from the jet. However, if the " inner core" of fibrous insulation f NUREG-0897, Revision 1 3-57

is exposed to the break jet (such as would occur if the cover blanket was breached), blowdown transport of this material would be expected to extend for distances much greater than 7 L/D's. Jacketing of fibrous insulations does appear to provide some protection provided such jackets are not blown away by the initial blowdown jet forces, as demonstrated by HDR blowdown tests (Appendix F) where unjacketed fibrous insulations or insulations covered by a metal mesh are nearly totally destroyed within 3L/D's with some damaged and partiali;' destroyed segments within 7 L/O's. But the same blankets enclosed in stainless steel jacket withstand the blast better (see Appendix F). Figure 3.26 illustrates examples of debris generation for RMI, fibrous jacketed and fibrous non-jacketed insulation materials. Thus, debris generation in Region II can be very complex, and generic conclusions should not be drawn nor extrapolated to cover different materials or conditions. The specific materials and products used as insulation should be carefully reviewed in light of the data base available as results of tests (see Appendices C, E, and F). The assessment for the volume of debris generation, transport and screen blockage should be made a: a plant-specific basis. If such a determination shows that estimated blockage head losses do not exceed the NPSH margin, a conservative safety assessment has been made. - The size of the third volume (Region III) was established using the Moody jet analysis as modified and discussed in NUREG/CR-2791. It begins at L/0 = 7 and extends to an axial position in the jet where the jet thrust (as calculated by the Moody jet expansion model) would be equal to 0.5 psig when calculated for a flat axisymmetric target. The Moody-type jet expansion model was selected for establishing the outer boundary of Region III because it always results in a larger L/0 value for the boundary than the two phase jet analysis in NUREG/CR-2913, thus ensuring that the effects of debris modeling uncertainties are mitigated by a conservative outer boundary selection. Break location (s) and insulation (s) targeted by the break jet are the key factors in estimating debris generation. This is illustrated in Figure 3.27 for a typical PWR where the influence of an expanding jet is shown. A break NUREG-0897, Revision 1 3-58

Totally frogmented **As-febricated** eegment] . 100 % - 2 5 0 e 1 % Demoged & partially destroyed segmente D UD's 0 0 3 7 Example for non-jocketed fibrous insulation meteriese. Totally frogmented "As-febricated eegmente

                                                                                            /        ..

2 5 i a

1 t

2 E

O' W U D's 0 3 7 Exemple for jacketed fibrous ineutetion meteriele.

Severe damage end destruction of RMI seeemblies, ejection of inner meteriale 100% = =

                                                     /         %                         'As-febricated** segmente
                        !                        /               \

k \ Possibio demoge to MMI seeemblies

                       !                 /                                      pave blowdown of demoged Innes
                                     /                                          components from Region I, 2

k 0 V \J U D's 0 3 7 Exempi. for r.n tev. met.m. ine.i.tson m.t.ri.e.. Figure 3.26 Possible variation of debris types and relative quantities in regions of the three-region jet model (see Figure 3.25) \ NUREG-0897, Revision 1 3-59

ii!! ;!i!I + ! t '~ ;- li ;! > {

                                                                       \l j                           \\

j

l j

                                                                                        /
                                                                                                      \        7 5
                                                                                                    ,           , =

T G NfN D r S Ih& t# i

i r U ia L h H P C

                                                                            ]F R

RPv I n i o t a r e n e g l la isr kl ! b W e e / d n r C a r

                                                -O G

S

                                                               / /                                                                     f o

C / e

                                                       \l                                                                                c n

e P 3 5 7 l u C = = = f n R g i

                                                           .                         D D D                                             f
                                                            .                        U U U                                               o
                                        \                                                                                                s   t
                                                        '                                                                                e
                                                       '                                                                                 n o

V Z

                                                           -   P     -                                                                 7 R                                                                       2
                                                                                                                        )              3 D

(- ( r t e e r u g le e iF s p ) m s La e m (i V u tD P e e r o g k r t n r a u t a a ae s r TrB s e lo e r n o to y r P e C eb e r G r g cd v o o e ne o t m t L s c a c ad ti s a e a t sv o e t e o ;i r R S R H DD C V G P L D O P R S CR H U C r e e h W 2E Y oco

N< Eo3 ~ Y$

I

Table 3.5 Maximum LOCA generated insulation debris summarized by break size Pipe Total fibrous Total all diameter (inches) debris (ft3) types (ft3) 2 1 1 6 2 22 8 2 3 10 4 31

14 227 227 16 270 270 32 144 295 34 315 726 36 118 408 Notes: (1) These values correspond to break locations in the primary system within the crane wall and represent the largest quantity of debris generated by a single break of a given pipe diameter. (2) The insulation types and distribution within containment are those used in Salem-1. All insulation within 7 L/0 of a break location is assumed to be destroyed and released as fragmented debris. (3) For reference see NUREG/CR-3394. NUREG-0897, Revision 1 3-61

in the primary coolant system piping will target large quantities of insulations located in the lower portions of the steam generators. Although break locations are identified in SRP Section 3.6.2, the reviewer (or analyst) should determine which breaks are most significant and estimate the extent (or volume) of insulation debris generation. Such a detailed break evaluation was carried out for a reference PWR (Salem Unit 1) and is reported in NUREG/CR-3394. Although this study was primarily directed at estimating the probability of sump blockage, the analyses revealed that breaks in large diameter piping (> 10-inch diameter) were the dominant contributors to debris generation (see Table 3.5). This finding can be used by the analyst in scoping the extent of LOCA debris generation. Table 3.6, which illustrates typical volumes of insulation employed (for two typical PWRs) on the primary coolant system and related components, provides an insight regarding volumes of insulations employed and their distribution on the PWR primary coolant system and components. Although a generic conclusion cannot be drawn from these studies, because of plant variabilities, the results do indicate that PWR debris assessments should concentrate on the primary coolant system insulation within the crane wall region and for pipe breaks of pipe diameter > 10 inches. Because such a detailed break study has not been done for BWRs, the reviewer should consider debris generation as occurring for breaks postulated in the BWR feedwater and recirculation piping and for postulated breaks in BWR main steamlines. NUREG-0897, Revision 1 3-62

_ _ _ _ _ . _ . -. - - - - ~ ~~ ~ '~~~~ Table 3.6 Typical volumes of primary system insulation employed (1) Salem Maine Yankee Volume Type of Volume Type of Component (ft3) Insulation (ft3) Insulation Steam Generator 1284 reflective metallic / 1144 calcium silicate / fibrous fibrous Hot Leg 160 reflective metallic 149 fibrous Cold Leg 144 reflective metallic 156 fibrous Cross-Over 60 reflective metallic 279 fibrous Pressurizer 160 reflective metallic 302 calcium silicate / fibrous Press. Surge Line 129 reflective metallic 57 calcium silicate / fibrous RCP 570 reflective metallic 149 calcium silicate / fibrous Bypass N/A N/A 88 fibrous TOTAL (2) 2507 2324 SU8 TOTAL I3) . 1284(=4402 ft2) 1527(=5234 ft2) (excluding reflective metallic and calcium silicate) (1)This table is based on information provided by the operators in 1981. Plant changes since 1981 have made the data less accurate for these two specific reactors. However, as representative data for reactors in general, the table is still valid. (2)This volume includes all of the insulation that could be hit by a water jet from a LOCA pipe break (in pipes >10" diameters). If the volume was restricted to only insulation within L/0 = 7 of a break, it might be significantly smaller. (3)In order to be conservative, Salem's steam generator is assumed to be covered entirely with fibrous insulation. 50% of the insulation of Maine Yankee's steam generator, pressurizer, and reactor coolant pump is assumed to be fibrous. NUREG-0897, Revision 1 3-63

I Table 3.7 Transport and blockage characteristics of reflective metallic ' insulation materials (see also NUREG/CR-3616) Velocity to Velocity to initiate transport i Sample motion to screen Description (ft/sec) (ft/sec) Comments Undamaged half jacket normal to flow concave side up 1. 0 1. 0 Either flipped on screen (see Figure 3.28) or got stuck partially flipped concave side down above 2.2 Never moved, i Outside Cover (0.037" thick diameter = 19" concave side up 0.7 0.8 Same blockage mode as undamaged half jackets. concave side down above 1.8 Inside Cover (0.015" thick diameter = 13") . concave side up 0.7 0.8 With both initial positions, concave side down 1.1 1. 6 covers flipped against the screen on arrival and got flattened against it by the flow force. , End Covers above 2 Never moved. t Single sheet : Inner Foil 0.35 0.5 Moves in folding and tumbling l (0.0025" thick mode. Flips vertically i 36" x 25") against screen as soon as it

uncrumpled reaches it. (Figure 3.29) with and without May be folded on screen, separating crimp i.e., not cover full sheet l area. i Never covered screen higher than maximum sheet dimension, -

even for flow velocity of 2 f t/sec, and water depth of 60 inches. NUREG-0897, Revision 1 3-64

Table 3.7 continued Velocity to Velocity to initiate transport Sample motion to screen Description (ft/sec) (ft/sec) Comments Single sheet 0.20 0.25 Moves in folding and tumbling . Inner Foil mode. Flips against screen i (0.0025" thick as soon as it reaches it. 36" x 25") Gets flattened on screen by current. Four sheets 0.25 0.4 to 1.8 When numerous foil sheets inner foil are used they tend to jam (0.0025" thick up in piles that may need 36" x 25") high velocity to unjam, two crumpled Significant overlapping on two uncrumpled screen. Single cut-up sheet inner foil (0.0025" thick 24" x 21") uncrumpled 0.20 0.25 Folding and tumbling transport mode. Flip veritically on screen - upon arrival, sometimes folded, crumpled 0.20 0.25 Flip veritically on screen upon arrival, sometimes folded. (See Fig. 3.30) Several cut-up sheets inner foil (0.0025" thick 8" x 8") uncrumpled 0.5 1. 2 Pieces not folded by flow as larger ones. Sliding transport mode. One piece reached screen at 0.5 ft/sec - all flipped vertically on arrival to screen. (See Fig. 3.31) crumpled 0.5 1. 2 One piece reached screen at 0.9 ft/sec - all flipped vertically on arrival to screen. - NUREG-0897, Revision 1 3-65

Table 3.7 continued Velocity to Velocity to initiate transport Sample motion to screen Description (ft/sec) (ft/sec) Comments Several cut-up sheets inner foil (0.0025" thick 3" x 3") uncrumpled 0.8 2.0 Pieces not folded by flow as larger ones. Sliding transport mode. Several cut-up sheets inner foil (0.0025" thick 3" x 3") (continued) crumpled 0.6 1. 0 Pieces flip vertically on screen unless a corner gets trapped under screen bottom, in which case the piece stays flat on bottom. NUREG-0897, Revision 1 3-66

3.3.5 Transport and Screen Blockage Potential for Reflective Metallic Insulation Materials A limited amount of testing has been conducted with reflective metallic insulation components to gain an insight into the transport and possible screen blockage configurations. The results are reported in NUREG/CR-3616. The thrust of these tests was to determine ve,locity levels that would transport various components, particularly thin foils that are used internally. As might be expected, intact units were not transported until flow velocities exceeded 1 ft/sec. On the other hand, very thin, stainless steel foil (0.0025 inch thick) materials were transported at low velocities (0.2 to 0.5 ft/sec) if such foils were in an uncrumpled and intact state. Table 3.7 summarizes experimental findings. In these tests, as the foil material became more rigid (increased thickness), the foil type debris was transported by sliding along the floor, rather than in a tumbling mode, and higher velocities were required to flip the material into a vertical orientation against the debris screen. Of more significance are the screen blockage patterns observed during these transport tests. Intact shells (or halves) can flip against a debris screen if velocities exceed 1 ft/sec (see Figure 3.28). On the other hand, free thin foil sheets tend to crumple resulting in the blockage configurations shown in Figures 3.29 and 3.30. Multiple foil sheets can form a blockage pattern such as shown in Figure 3.31. Generally blockages occurred at the lower portion of the debris screen. Although enough sheet material to totally block the screen was introduced into the transport flume, total blockage did not occur (see Figure 3.29). The very thin foil material (when in large sheets) is transported with a tumbling, lifting-type motion; however, lack of structural rigidity results in transport deformations, as shown in Figure 3.29. Another significant finding was that none of the foil samples tested became water borne. This is particularly important in BWR considerations because the RHR suction intakes are generally 6 to 8 feet above the suppression pool floor. NUREG-0897, Revision 1 3-67

l Thus transport of metallic insulation debris at fairl,y low velocities cannot be discounted and therefore plant-specific assessments should be made for those plants employing this type of insulation. The transpor and blockage findingsdiscussedabovecanbeusedtoestimatelevelsofpbtential blockage. Of equal importance is the severity of LOCA induced damage (see Section 3.3.4) and types of RMI debris generated (see Appendix E). The HDR tests discussed in Appendix E do not support a debris generation model consisting of free, undamaged interior foil materials being available for transport. 3.3.6 Buoyancy, Transport, and Screen Blockage Characteristics of Mass Type Insulations The buoyancy and transport characteristics of fibrous insulation' materials are important because long-term screen blockage is a function of whether, and how, such debris material would be transported. Information regarding transport of shredded mineral wool insulation is provided in the Finnish tests conducted in the late 1970s (see Imatran Voima Oy, "Model tests of the , Loviisa Emergency Core Cooling System and Model Tests of Containment Sumps of the Emergency Core Cooling System"). These tests showed that shredded mineral wool would be transported at low velocities and build up uniformly on a debris screen, and thus could result in high head losses. , l j NUREG-0897, Revision 1 3-68

1 _ _ iJ..".- ,...~- {}E,.k [.. .

                                                                                                                                                                                                                                                                                                                                                                                             # .w. n gw .fwea. w - a .+ . .
                                                                                                                                                                                                                                                                                                                                                                                                             $f... . ..                               cm. f.e . _- . ;,.

h. .,.

                                                                                                                                                                                                                                                                                                                                                                                                                ?
                                                                                                                                                                                                                                                                                                                                                                                                               ,y,. e ,k.m. , ..
                                                                                                                                                                                                                                                                                                                                                                                             ,.                   . . . . - ;, ,w . ..

s. .

                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      ., , n~
                                                                                                                                                                                                                                                                                                                                                                                                                              . a, y                       v", = ., . ,'a.a. + e r
                                                                                                     .*.a. .h.                                                                                                                                                                                                                                                                                               ., ,, , . . -.lcy.

e. -,.

m

                                                                                          '., . ., q . ,                   .

p

                                                                                                                                                                                                                                                                                                                                                                                          -n                , ., ..                     , .,      +. . e; .v
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                          %r. ;

( .n.. a . .: ,. . . g: , ;.. . .;_ ; ,, ' ,

                                                                                                                                                                                                                                                                                                                                                                                                                               . ,;                       .Q.,...,....          .; e
                                                                                                                                                                                      .       p',          g
                                                                                                                                                                                                               , c ... .
                                                                                                                                                                                                                                                                                                                                            ,      .       y             ,        -{        y       ,54, A,.f; _ ;,.p ; '.;,' ; #,y :
                                                                                                                             ,..              ' . ,7
                                                                                                                       - ~                                                                                                                                                  ,   ,
                                                                                                                                                                                                                                                                                   '      ,    .             n.

s

                                                                                                                                                                                                                    ",,s[,,                                                       P_                               y               ,        .           ,      ..,                    '.y c .. ..                                    . .. ..
                                                                                           ~
                                                                                                  ....i.,....:       [ .. ,; ..                             .~        Jf.9,d[.~                                                                                           '        G                       ,        h.

r,..,..

                                                                                                                                                                                                                                                                                                                                    .6,         . F      .  .,   ..,..   .;.'{.[            [y.                       .. ."

t, . , [

. , r. . ' .-. e. + , . ,.,; . p .

_ J.+ - ,

                                                                                                                                                                                                                                                                                                                                                           .w                          '

u i l ?

                                                                                  .,J '& : p                                                                                                                                                                                              a, y. e.                                            - *;,.>
                                                                                                                                                  ..,J'.
                                                                                                                                                                                                           #.y ;$ . ,                                                       ..                                                                               ..
                                                                                                                                                                                                                                                                                                                                                                                                                     . .                               .. .. i                             . ;; ;

y. :. . . y ;~- c. a

                                                                                                                                                                                                                                                                                                                                                                                                        .'.-(:'T . .*. - 9 :q ' ? ,.
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                    ~
                                                                                  . :.. M + . ,l, i.,,.4                       ,.                    .....

l.. p

                                                                                                                                                                                                                 . u ?. 4.                                           *e,                         ,t,7:                                                   '.f,.,. . ,. .

1 \ ,

                                                                                                                                                                                                                            ...,..                                                                                                                                                                                              s.
                                                                                                           .i .,:                                     .
                                                                                                                                                                                                        ,v,.                                                                         ..

\

                                                                                  ,3 p
                                                                                                                                                                                                                                  . . , . -3                                                               .,                           . ....     . . ... ,. .   .
                                                                                                                                                                                                                                                                                                                                                                                             . . .-.                     -$           .. s
                                                                                   . + . , . - 3',.e                      .. <w                                                                                                                                                                                                                                                                                                                                                                  ..g   ,.,

s .

                                                                                 ;;',. ; .._ , - ' . ' f. . [.

g . ._..,* *'_ A ;, - ..,, .e, ' .'.,,, * ,~. . 'f . , , , ,

\ r .

a. :

                                                                                                                           .s                                                                         .,    ..

y

a.,.,,.'.J e .

                                                                                                                                                                                                                                                                                                                                                                                                                                                      .: * .p'                             ,
                                                                                  .,g             ,     .                                      ,.                 ,s                                                                                                      .
                                                                                                                                                                                                                                                                                           ,p:                                        .: . q...-                                                          . _ . :s ,, y                                  ;         <       .            . .;

l .,,... n ., '

                                                                                                                                                                                                                                                                        . y .:..                             ; %., .. -                                               --                                                      . , :u: .
                                                                                  %, yl.V jY '. g. ~                                                                                                         .              . - .                                            +

_ q . .s ( ,, . '

                                                                                                                                                                                                                                                                                                                                                                                           - : .,                                                      .-*                       -,f .; _

l '[ .) ' ' ._ %f ,} ., .f ' y: "4 ' l - }. ; ' q, -}} . '*', .; . , ll_._;,}, .. , , ; g * '. .- 4y r. en . .s I o. + -

                                                                                                                                                                                                                    .e,: e. s,,.. . .s.,                                                                      . , , 3. t , t.. . A: . .. . w. _ ;; ..- , ,..? ,,,. n
                                                                                                                                                                                                   ..o,,,fg.;v 3

I s

                                                                                                                              . ., n e                                                             a                                                                                                                                                            .

2;.v ,. ..- .c.+.,

                                                                                                                                                                                                                                                                                                                                                !                     .y                                          . .*, . . , .                                                            .* ;,.

v

                                                                                                                                                                                                                                          ., ..i+,.'*... y J .sg ' . ' : ,(3,,r s5., 3. . .. ' s.,                                                                                                                                     . ? , f. y . .~

d, u s,

                                                                                                                                                                                                                    . ' . ' 5. . ' J . :
                                                                                                      .f!-- r t.(W"
                                                                                                          . ,.                 ~                  3' e..*3..., ,y .' K, '_,l;.                           ,

t Figure 3.28 A half segment flipped onto screen g.q.[l 2g< ' .s

                                                                                                                                                                  .,2
                                                                                                                                                                                                                                                                                                                                            ' '.. g.         x-e p, M . ,. .] ~ ~ .. T] g q 5 y ,                                                                                                                                                                                                                                                                                     ,_
                                                                                                                                          . ' ;k
                                                                                                                                             .                    a
                                                                                                                                                                                    ;;.:Er.     .:. s a i c-            .
                                                                                                                                                                                                                                                                                                                                                                                         ~

gl. .;, .,. -.

                                                                                                                                                                                                                                                                                                                                                                                                             ; - < .' , m.

e. ; . c,, p. , '. Ny

o. n. ,w.

                                                                                                                                          +.                .: -
                                                                                                                                                                                                                                                                                                                                                                                                                                - ? -                          ,        .                    ~p
                                                                                                           *s;                                                                                                                                                                                                                                                                                                       a n . ~ : . f. ... .yv. ;
                                                                                                                           . n : , r :s ., . , f '. ..' . '.'
                                                                                                                   . .-e . < r. .. ,a ..g 3,                                                                                                                     .

gr* . .,. . . . , .,, *

.g*

l

                                                                                                                                                                                                                                            ... ~ . , .

4 ..

                                                                                                                                                                                                                                                                                                                                                                                                                     ,    >..i.
                                                                                                                                                                                                                                                                                                                                                                                                                            , c..,.:,.,. ..                           s 2..
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                             . w.
                                                                                                                 . ..- . , ,                                                                .<.. 3, , ., - s . ;-

e.

                                                                                                                                                                                                                                                        <,.e.. 1.,                                                   .:.

s. > . .

y

                                                                                                                                                                      .,                                                4. . , .                                                                             --

l .

                                                                                                                                                                      .+.-                            ,I y.'.:                                .
                                                                                                                                                                                                                                                                              -- . ,_, _ . -%                                       > . ;;.-.y ' la'., g.s                                                                         .- .,                           .. . ..
                                                                                                                }
                                                                                                                                                                                   , . '. . ' .,-..9.[!... -' , ; r
                                                                                                                                                                                                                                                                                                                       '.                                                                                                                                                  3..- .,

p . r ;. - , . ."

                                                                                                                                                                                              .                                                                                                    . . , ,                                                                                                                                                                                        4
                                                                                                                                                                                                                                                                                                                                                                                                                                 - t, **.%.          / ., :

97 , a en..3 -~.m... .,.'. 's 2 ..,

                                                                                                                                                                                                                                                                                                                                                                                                                                                              *v..ts.
                                                                                                                               ;e,               .                                                                                 ,                                                                                   t                .,                   w,,,,                                                         rzo          et
                                                                                                                                                                                                                                                                                                                                                                                                                          ?
                                                                                                                 ,p~ : .
                                                                                                                                                    ,                                                                                                                .                                                                                                                             w
                                                                                                                                                                                                                                                  ,w . .                                                               ...                                                                                     {...                             . .e
                                                                                                                                                                                                                                                                                                                                                                                                                              . ~.b.n..,

e Q . ... 4

                                                                                                                                                                  *..<                                                                  .;                                                                                                                                         . . m
                                                                                         .r -                                    ..
                                                                                                                                                    ..,c.._..

. F,ag, :, .p .

                                                                                                                   . ,                     +-                                                                                 , ,.. .- ,_.                                                                                               :.
                                                                                                                                                                                                                                                                                                                                                                                                    ,.. . . . .. ;W. .

e

                                                                                                                                                                                                                                                                                                                                                                                                                                                                                          ~5
                                                                                                                                                                                                               . . ' . r ;. : - , ^ :L.y.
                                                                                                                                        * ~ . , , ..                                                                                                                                                                                                                                                                                                                                          ..

C. i.,.p .: W .T

                                                                                                                                                                                                             ,                                                                                                                                                                                           .r Yl 4 k.' f z.l l .?) ,W . .l .:' . +.:                                                                                                                             _,.

A

                                                                                                                                                                                                                                                                                                                                                   ,'.'                                                                      hl-                . f ;- hQ e,
                                                                                                                                                                                                                                                                                                                                                                                                                                                                                  ,e
                                                                                                                                                                                                                                                                                                                                                                                  % . ._ . ,>-.m<>%,                                                    - r:

1

                                                                                                                           .-y~_.,..--... -
                                                                                                                                                                                                                                                                                                                        ~
                                                                                                                                                                                                                                                                                                                                              %. . g. .*          -

w. w .

                                                                                                                                                                                                                                                                                                                                                                                                                                                                                             ~.
                                                                                                                                                                                                                                                                                                                                                                                           ,"'*'aumme .                                                         % ',
                                                                                                                                                                                                                                                                    !                                                                                                                                                         4                                                              e g_                      '
                                                                                                                                                                                                                                                                                                       .                                                                                                                       a
                                                                                                                                                                                                                                                                                                                                                                                                                                                          .s l                                                                                                          y                                             :

4 4

                                                                                                  .v                                                       . , , . . .
                                                                                                                                                                                                                                                                                                                                                                                                                         -A
                                                                                                                                                                                                                                                                                                                                                                                                                                         ~a                                 ~                     s 4
                                                                                                                                                                                                                                                                                                                                                        ~N 2              .
                                                                                                                                                                                                                                                                                                                                                                                                                             -4
                                                                                                                                                                                                                                                                                                                                                                                                                                          ~ -                                                        4 Figure 3.29 Uncrumpled foil sheet flipped vertically on screen (flow velocity = 0.5 ft/sec)

NUREG-0897, Revision 1 3-69

i

.l i,

k a l i

                              . _.                                   ,             .- =
                                 .;;.                                            -.-             =~

: ~.-- -.

                                                                                                    -                                 t
                                                                             ~~                  ~

1 amme= d~ . - 2 j l 4

'4 i,

N l i - 4 .,

!                                                                                                                                          a I                                                                                                                                        CT.a~,

i I

                         -/..
                                                                                                                                        ~:i j
                                 ' .d,*

p , i i l Figure 3.30 Crumpled foil sheet : p , ;creen (flow velocity = 0.3 f t/sec) t i

l l i.

NUREG-0897, Revision 1 3-70 l

l 1 l t I i I l I I ', l F

 ,                           .                 .                                                                                                                                                                    m-                                                                      _

i '. '_ - .: . nu mm.ummmmmummmm une

                                                                                                                                                                                                                                               ~.                                              __
                                                                                                                                                                                                                        , - . .. . . .n .
                                                                                                    --                                                                                 r oi
                                                                                                                                                                                                                                                                                               .                 1
                  ..                                                                                                                                                                                                                                                  . mmmum a

f -

                                                                                                                                                                                                                         -.T                                                                                     i m-            __ . __ _. .                               =='.usumu.ma

_, m . , ._ . 7 .. u .

                                                                                                                                                                                     . , ,.                         ,               z.,....,,,.,9,,...                                                           ;
                            . . w. - + %: .                                ,

r- -r' .-

                                                                                                                                                                                                                                                        ..a                n . a; l

i 1 ._ U',~ t p8ll!,f sigth f .

                                                                                                                                  '.               ? ~. :      . . .        .73.Q                                     .(..l . . :e".i .. .p ,g.7. ,; ;
                                                                                                                                                                                                                                                                                  ....,e
                                                                                                                                                                                                     >~g.,,q:.';..              ..,.
                                                                                                                                                                                                                                           - . ,y]q,c           c ~. .,

1 x :e- - n,.

                                                                                                                                                                                                                                                                                   , :a 1            -
                                     ".c..       %... . .      ..%. p                                                                                                                                                                                                                         .

l 4 g 9. y:% i.s . ., Q,.,% W4,:;.;.# .

                                                                                                                                                              . 3A.y. g,9.g:.;y._y,...

y m;., . .ar.':. s . . t r -. . ;i y, .,, .- v

                                                                                                          ;y                          c                      ,..,.                   m. -                                 - ..      .:                     q%,;,u .y~:                     .+        g.

1

                     '1m..a 1                                            ,.
                                                                                 - { y,,
                                                                                                                . : . :.'m.a. .?. .
                                                                                                                                                                                <4   g .**        , - .. 4      - r e.3..        Q..                           ,- 4;                                             '
                                                                                                                                                                                                                                       .. y'h.' T e ' 1.q:                                        ;

> 4p.9p- -

                                                       ,A..                      <.                              .. -av.                                                                                                                                                            ..
                                                                                        .;                                        .g .                                                                                               .%

> .- - 1

                                                                                                                                                                                                                                                                                       . m. .
                                                                                                                                                                                                      . ..e ... .

J b.. . T..n.; f.n.

                                               . t   .                                              .

i. q - u....

                                                                                                                                                                   .1..j             ..
                                                                                                                                                                                                                                                        - 1 . , . . ;<.      .,,,.

f,fly . ;; e e n. amuy .' . n

      ~:                                                                                                                                                                                            i l
                                                               -h.kS . ig:, m, , >%
                                                                . ,            ..~..<                                h.:::..            ..y                                         .:
                                                                                                                                                                                                                       '.           .  "}..;         }J_'kj,$.(    . 'N .y'. .&, s
                                                                                                                                                                                                                                                . x. ... ~ . u. , h, ?. : .-

h.*- .tb >'+g. ' , ~ . Q, e .)-. y.l

c. -

gb. }

                                                                                  .                                                  - ". . . - z. '                                    f>.                                         . : . ' * <,. - r . > . . %
                                                                                                                  "- .,) ,'..-;.                                                                   'y-                                 . '; y.3 % - . #

g, . n; s , k, ~.. cu ,r s',.%'. , ' ', y aqf -g .. l.6 y ,:

s.M(;~ ' * #

                                                                                                                                                                                                     -                 _.,5
                                                                                                                                                                                                                                                           '.'# 3.-l ,u.                            %.s d

g .. -

                                                                                            .e                                                                                                                         X             g                              .

W h.; ..*,; l.,t.. ....'h(.' ' Y,.n ...v ' ).n.,. .

&.,w... -

                                                                                                                                                                                                                                                                  ' ft. : .: , e.

v, s.e.. m.+i..%. .+ - 4 w,:. ;g.. f. ,. .r<: . , . , .-. 4

                    . ... Q.. ,.         ' . : . a t.%- s, i.g. '. . .. ..4.,p.c w.. 'f.,.,                                                                .;                                                                                                                        .

j , .,.- . 6,c .t '4 2, * *.r. ::. . ,.. T . - . 4.; g - .- . .,7J. , , , _4. - ;

                                                                                                                                                   ;       .>      u     '      :

j .g,

                                                                                                    ., ,,                                ,."..;,.,,,., 3.,. :!; e
                                                                                                                                                               +
                                                                                                                                                                                        ,,                     ,   ,< .. .Q  , , , ;y,,.l e   - ' %          ,-* ,       :j.x  ,,;), .:. )..g ',

r. , . . . ' >+:>+ , ;,; :g ;> . . a .; >:w ~,,,...;. - x ,. * ., ;. 4 w . .: '. w , , x y 1;o-. . . . .. e< - ;, r ;,m. .: ,,k~ . ..o

                                                                                                                                ..        ;,.r;. :

y d %.,p e..Q. .. g .; y './.7,..< '. .. ._x h ij k, ) l. ., .... p

 \                   . . .                     .

i , :- ca 4- x. . r:

                                                                               . . ;.x,?        , , f h.. {,f

_.. , ,6f,y,4:

                                                                                                                                                              .. ': 4 . .                     ~ .g ,;          -l^?l{h. me s.'
                                                                                                                                                                                                                           \. . .
                                                                                                                                                                                                                                       ,.;,l< r~ p.;                ,.

l8 h-].eU -f ,i f 'l f i Figure 3.31 Several foil sheets on screen (flow velocity = 0.7 f t/sec) - 1 t j NUREG-0897, Revision 1 3-71 i I f I

           -.  .   .=_                                 .                -.-   .-. _ .   . -

1 Similar tests were conducted under NRC sponsorship at the Alden Research Laboratory and are reported in NUREG/CR-2982, Revision 1. The results of those tests are summarized below. , Buoyancy, transport, and head loss experiments were conducted with three types of as-fabricated insulation panels and with fragmented fibrous insulations. The three types of as-fabricated insulation panels were Type 1: 4-inch mineral wool or refractory mineral fiber core mineral (6 lb density), covered with Uniroyal 6555 asbestos cloth coated with 1/2-mil Mylar. Type 2: 4-inch Burlglass 1200, or 4 layers of 1-inch-thick Filomat 0 (fiberglass) core material, an inner covering of knitted stainless steel mesh, and an outer covering of Alpha Maritex silicone aluminum cloth, product 2619. Type 3: Same insulation core materials as Type 2, but with an inner and outer covering of 18-ounce Alpha Maritex cloth, product 7371. - The fiberglass core material in Types 2 and 3 is a high density fiberglass (310 lb/ft3). Various types of fiberglass insulation are employed in nuclear plants, and, as evidenced by the data reported (Durgin and Noreika, September 1983) for the Owens Corning Fiberglass product NUKON", they can exhibit different characteristics. Therefore, evaluations should be based on the l actual material (s) utilized in a given plant. i NUREG-0897, Revision 1 3-72

 .                  . t       . .                                     -               -

4 The buoyancy tests revealed (1) In general, the time needed for both mineral wool and fiberglass insulation to sink was less at higher water temperatures. (2) Mineral wool (Type 1) does not readily absorb water and can remain , afloat for several days. (3) Fiberglass insulation (Types 2 and 3) readily absorbs water, particularly hot water, and sinks rapidly (from 20 seconds to 30 seconds in 120*F water). (4) Undamaged fiberglass pillows of Type 3 (and possibly also of Type 2) can trap air inside their covers and remain afloat for several days. (5) Based on the observed sinking rates, it may be concluded that mineral wool pillows and some undamaged fiberglass pillows (those that trap air inside their cover) will remain afloat after activation of the , containment recirculation system (approximately 20 minutes after the beginning of LOCA). Those floating pillows will move at any water velocity and can be transported to the sump before activation of the recirculation system. The transportation tests revealed (1) Water velocities needed to initiate the motion of insulation are on the order of 0.2 ft/sec for individual shreds, 0.5 to 0.7 ft/sec for individual small pieces (up to 4 inches on the side), and 0.9 to 1.5 f t/sec for individual large pieces (up to 2 feet on the side). NUREG-0897, Revision 1 3-73 l

(2) For whole sunken pillows to flip vertically onto the screen, flow velocities of 1.1 ft/sec for Type 1 (mineral wool) and 1.6 to 2.4 ft/sec for Types 2 and 3 (fiberglass) are required. (3) Whole floating pillows require a water velocity in excess of 2.3 ft/sec to flip vertically against the screen. (4) Insulation pillows broken up in finite size sunken fragments tend to congregate near the bottom of the screen if there is no turbulence generator, and, depending on the water depth, unblocked space can remain near the top of the screen. With turbulence generators (vertical posts 2 feet upstream of the screen), some insulation fragments are lifted from the bottom and collect higher on the screen. (5) 'Once insulation shreds are in motion, they tend to become suspended in the water column and collect over the entire screen area. The head loss tests revealed _ (1) The measured head loss across a vertical screen in a flume as a result of blockage by insulation released upstream varies from 7 to 10 times the approach velocity head, U2/2g, for whole sunken pillows; from 13 to 36 times the approach velocity head as that for opened or broken up pillows; and more than 240 times the approach velocity head for shredded pillows. These results are for an equivalent volume for 50%

screen blockage with the undamaged pillows.

Opened pillows with separated, fragmented, or shredded insulation layers ! had enough area to block the entire screen. However, the screen was entirely (but not uniformly) covered only in the test with the shredded insulation. In the other tests, open space remained on the screen. , 1 I NUREG-0897, Revision 1 3-74

u.. ..-. . .- -- s , , For these conditions, the maximum measured head loss of 240 times the approach velocity head (for shredded pillows) would result in screen head losses of 0.15 foot to 0.60 foot for approach velocities of 0.2 ft/sec to 0.4 ft/sec. (2) Measured head losses through beds of accumulated fragments or shreds of mineral wool or fiberglass insulation varied nonlinearly with approach velocity and bed thickness. For mineral wool fragments, the larger head losses were observed for the tests of larger fragments (3 x 2 to 4 x 1/8 inch). For an original insulation thickness of 1 inch, the maximum head loss was 0.4 foot at 0.2 ft/sec and 1.4 feet at 0.4 ft/sec. For fiberglass insulation fragments and shreds, the larger head losses were observed for the shreds. For an original (as-fabricated) insulation thickness of 1 inch, the maximum head loss was 1.2 feet at 0.2 ft/sec and 6 feet at 0.4 ft/sec. , (3) The head loss through as-fabricated insulation material is higher, by a factor of up to 10, than that for accumulated , fragments. For example, with water at 105* to 120 F and with an approach velocity of 0.2 ft/sec, the head loss through 2 inches of undisturbed mineral wool is about 3.5 feet, and the head loss through 1 inch of undisturbed fiberglass is about 20 feet. These head losses are for insulation samples sealed to the walls to prevent leakage. The head loss would be less if leakage occurred around the sample. (4) In addition to the variables of insulation thickness and approach flow velocity, the actual head loss that may be expected across a sump screen depends critically on how the screen is blocked. If some unblocked screen area remains, or if water can flow between pieces of insulation, NUREG-0897, Revision 1 3-75

the head loss would be small; if the entire screen area is uniformly 1 covered with mats of undisturbed insulation or accumulated fibers, the head loss can be many feet, j (5) Best-fit expressions for the head loss through shredded fibrous I insulation, were derived as follows: l for mineral wool (Type 1): aH = 123U1 .51t l.36 1 for fiberglass (Types 2 and 3): aH = 1653U .84t l.54 l l where l U is the screen approach velocity (ft/sec) t is the original (as fabricated) insulation debris thickness (ft) AH is the head loss (ft H 2O) Table 3.8 summarizes these transport and head loss characteristics. _ The strong dependence on material characteristics cannot be overemphasized. Owens Corning Fiberglass conducted similar tests with fiberglass utilized in NUKON* (a low density fiberglass, 2 lb/ft3). The transport characteristics were similar to those reported in NUREG/CR-2982, Revision 1, in that the transport of fragments occurred in the 0.2 to 0.3 ft/sec range. However, the screen blockage head loss correlation for fragments (experimentally derived) was aH = 68.3Ul *79t l.07 This equation is significantly different from the two previous equations, and l these results are reported in ARL Report No. 110-83/M489F (Brocard, D. N., September 1983). Thus, the reviewer should base evaluations on the particular type of insulation material (s) employed in a given plant application. NUREG-0897, Revision 1 3-76

Table 3.8 Summary of transport and screen blockage characteristics of high density fiberglass Y Y Y F111ow i a v a Condition Type Ift/sec) (ft/sec) (ft/sec) (ft) h Comments Floating thole pillows 1 N/A N/A > 2. 3 Never flipped 2 N/A N/A N/A Sunk while against screens flipped vertical 3 N/A N/A N/A Sunk while against screens flipped vertical sunken 1 1.1 l'.1 1.1 0.13 only one pillow thole tested pillows 0.9 1.1 1.1 0.07 - Only one pillow

                                                     .                             tested folded in half on screen 2         1.2          1.8            2.0      0.44    7.1 1.4           1.6           2.4
                                                                                                       ~

3 1.5 1.7 2.0 0.60 9.4 1.1 1.6 1.6 0.33 83 Fillows on screens overlap by 2 inches sunken pillows 1 1.1 1.1 1.1 0.67 36.0 cith covers 0.9 1.5 0.96 27.5 Not all pieces removed but

vertical included and separated 2 or 3 1.1 1.6 insulation 0.9 1.2 1.2 0.71 32.0 liyers 8:nken pillows 1 1.0 1.9 1.4 25.0 Not all pieces cith covers 11 2.0 1.6 26.0 vertical cad insulation layers in 5 2 or 3 1.0 1.4 1.6 0.54 14.0 significant pieces (see overlap of Figure 2.6) pieces on screen

        *For details in the size and amount of the insulation materials utilized in these tests see NUREG/CR-2982, Revision 1.

NUREG-0897, Revision 1 3-77

Table 3.8 Continued Pillow Y Y Yv i s AN Condition Type (ft/sec) (ft/sec) (ft/sec) (ft) comments 79 sunken 1 0.4 1.4 1.6 1.35 34.0 Fragments pillows in collect on 4' x 4" x 1" fragments, botton 1 ft of screen Covers not included. 0.6 1.3 1.4 2.45 80.0 with turbulence generators. Fragments collect on botton 3 f t of screen 2 or 3 1.0 > 1.6 Not all pieces reached the screen. Collected near screen bottom, Figure 4.4 1.0 > 1. 6 0.72 18.1 with turbulence generators. Only about half the pieces on screen. Some pieces at mid-height. sunken 2 or 3 0.4 > 1. 3 N/A 3.7 240 Not all pieces pillows in shreds. for on screen. Covers not 1.0 Screen entirely fps but not uniformly included. covered. Sunken single Tests conducted fragments in 1 ft wide 4*x4*x1" fluse with 7 1 0.6 inch water depth 2 or 3 0.7 4*x1*x1" 1 0.3 2 or 3 0.5 shreds 1 0.3 2 or 3 0.2 NOTATIOus Vi = velocity needed to initiate action of at least one piece of insulation (not including covers when separated from pillows) vs

velocity needed to bring all material on screen -

v, a velocity needed to flip all pieces vertically on screen m = head less et v, (or v. te v, not given) NUREG-0897, Revision 1 3-78

l l l l l i In summary, the following consideration should be made for determination of fibrous insulation blockage effects: (1) Recirculation velocities and break jet loads must be evaluated to determine that they are high enough to transport debris to PWR sump screens or BWR suction strainers? (See Appendix 0.) If not, blockage , is not likely to occur. (2) If the material can be shredded by the break jet, transport can occur at low velocities and a determination of screen head losses must be made, provided recirculation velocities are high enough to result in transport of the fragmented insulation d(bris. 3.3.7 Effects of Combined Blockage (Reflective Metallic and Mass Type Insulations) l l Assessment of the effects of combined blockage, wherein both reflective metallic and mass type (fibrous) insulations are employed, is more ,

difficult. As described above, both types of insulations can be transported at low velocities and block debris screens. Because metallic-type debris does not become water borne, blockages that can be ascribed to metal foils l would occur at the lower (or bottom) portions of vertical screens. Fibrous insulation fragments can be trancoorted at near-neutral buoyancy and do l migrate to open flow passages. Therefore, a combined-effects model should be l applied. Unfortunately, not enough experimental data are available to allow for development of a combined generic blockage model. Plant-specific evaluations should also consider the potential for this type of combined debris blockage. l 3.4 Sump Hydraulic Performance l

To investigate ECCS sump behavior under flow conditions that might occur

during a LOCA, a test program was undertaken that covered a broad range of geometric and flow variables representative of PWR containment emergency sump designs. To avoid scaling uncertainties, a full-scale experimental facility NUREG-0897, Revision 1 3-79 t

at the Alden Research Laboratory was used. Scaling effects resulting from the use of reduced-scale hydraulic models were subsequently evaluated. The three broad areas of interest for ECCS sump design investigated were (1) fundamental behavior of the sump with reasonably uniform approach flow conditions (2) changes in the fundamental behavior of the sump as a result of potential accident conditions (screen blockage, break and drain flow, obstructions, nonuniform approach flow, etc.) that could cause degraded performance in the recirculation system (3) design and operational items of special concern in ECCS sumps Information from initial testing was used to plan or redirect later tests; hence, the tests were not necessarily conducted in the order listed below. The tests performed may be divided into six series as follows: . (1) Factorial Tests A fractional factorial matrix of tests was used to study primary sump flow and geometric variables. The factorial matrix provided a wide range of parameter variations and a method for effectively testing a large number of variables and determining their interdependencies. (2) Secondary Geometric Variable Sensitivity Tests The effects on sump performance of secondary geometric variables and design parameters of special concern in ECCS sumps were tested by holding all sump variables constant except one, for which several values were , tested. NUREG-0897, Revision 1 3-80

i l l (3) Severe Flow Perturbations Tests i i l i The behavior of selected sump geometries subjected to approach flow l perturbations was investigated. Major flow disturbances considered were screen blockage (uo to 75%), nonuniform approach velocity distribution, break-flow and drain-flow impingement, pump startup transients, and obstructions, as illustrated in Figure,s 3.32 and 3.33. (4) Vortex Suppression Tests The effectiveness of several types of vortex suppressors and inlet configurations was evaluated. l (5) Scale Tests Scaling effects in geometrically scaled models using Froude number similitude and pipe velocity similitude were tested, t

(6) BWR Suction Pipe Inlet Tests

The hydraulic performance of BWR suction pipe geometries typical of Mark I

I, II, and III RHR suction inlet designs was evaluated. i Data generated during the sump performance studies were analyzed using two j spproaches as follows: (1) Functional Correlations of Dependent Variables Correlations using response-surface regression analysis of nondimensional l empirical data fitting were developed. Because of the extremely small values of the dependent variables and the complex time-varying nature of the three-dimensional flows in the sump, the use of functional correlations showed no NUREG-0897, Revision 1 3-81 L

r I r I z - E SAL e l o NON-GNIAOWN A10AA WNO B3833N 8103l)VD3 S3H3N3S 1 v"T uu uu uu <J ,] enoes-nepsosaes dlom 8 r 0 ,", 1--

                                                                                          ~
                                                                                                       'r        -
                                                                                                                   ?

iN D g s a tsiom assisiseniow e103xwmt 7 l : [

\
                                   . c. u 3 = no s1 i

f I((

                                                                                            ?
                                                                                                 \
                                                                                                ~"r g

i S

                                                                                                                       } g""b(     
                                                                                                                                             '  ~l 5 f                %

5 l T r ii n iiii

^^

I--"' ni

e @ n G

                                                                                                                               = --.

O

                                                                                                                                                        =======i

- ,:, uu uu u .u uu l a H+ H" l-+ iH- " !! !!

                             -e_                                  =     _                     .
                                                                                              ;1
                                                                                                                                                     .'S+i"iim co 1                                                                                    . .

l , ,--~

                                - - _~ s                       > j f <.
                                                                                                =    _O                      O_ %                  ;.l                4i
                                                                                                                                           =-

l . . .n - i : i- -'- trmiiii 1:mii l O O O O Uu ,U,U, , UU UU

                                                                      .. .. is                                                ,,,,                     i i n i:,. . .

mo ri l ' E} i e CQ t-3 J

                                                                                                                                                   .m                 =.
                                                                                                        = ..                 u . - ..

l e e e e - i [ i

3g6naa E'EZ Vddace3y ; tom daagnagettous tup s32aau qtopeSa sayamas

[ l i - r

12" SUPPLY LINE P NO2ZLE

             ]            _ kg-Ni  Cl                                  l
                                                                              =

bll l l

                                               !                       I
                                                                                         =

a. Break Flow Jet Impingement

12" ORAIN FLOW I NOZ2LE 1

DRAIN FLOW JET ~

17' TO 23' j i

                                      'l SCREENS AND lg,,                                      GRATINGS
                                 -               W = 1 FT                ,
             "                      $k       t   !I r        -

(r(yi. l tJ

                                                                         ) !

SUCTION PIPE O

b. Drain Flow Jet Impingsment Figure 3.33 Break and drain flow impingement NUREG-0897, Revision 1 3-83

i l j consistent, or generally applicable, correlation between the dependent and independent variables. Thus, the hydraulic per 'ormance of a particular sump under given flow and submergence conditions could not be reliably predicted using this approach. I (2) Boundina Envelope Analysis The broad data base that resulted from the sump studies made possible the use of envelope analysis for reliably predicting the expected upper bound for the . hydraulic performance (void fraction, vortex type, swirl angle, and inlet loss coefficient) of any given sump whose flow and geometric features fall

!   approximately within the ranges tested. The data boundary curves generated j    indicate the maximum response of the data for each of the hydraulic l    performance parameters as a function of the sump flow variables, particularly when plotted as a function of Froude number. Thus, the ability to describe I

the performance of PWR ECCS sumps, with or without flow perturbations, using bounding envelope curves was the most significant result of the ARL test t program. The application of an envelope analysis to test data resulting from . l all the sump performance tests is discussed in Section 3.4.1. Findings of the sump performance tests are described in greater detail in subsequent

sections.

I 3.4.1 Envelope Analysis i The sump performance test program generated a data base covering a broad

range of ECCS geometric variables, flow ronditions (including potential j accident conditions), and design operations (horizontal or vertical inlets, I single or dual pipes, etc.). An envelope analysis applied to this broad l range of data resulted in boundary curves for vortex activity, swirl, and

! sump head loss as a function of key sump flow variables (Froude number, velocity, etc.). , 1 I l ! NUREG-0897, Revision 1 3-84 1 .

Figures 3.34, 3.35, and 3.36 show typical envelope analysis curves for air ingestion, surface vortex activity, and swirl in PWR sumps with dual horizontal pump suction intakes. Figures 3.37, 3.38, and 3.39 show typical envelope analysis curves for air ingestion, surface vortex activity, and swirl in PWR sumps with dual vertical intakes. 3.4.2 General PWR Sump Performance (All Tests) The following items were studied while testing for the sump performance. (1) Free Surface Vortices Vortex size and type (see Figure 3.40) resulting from a given geometric and flow condition are difficult to predict and are not reliable indicators of sump performance. Performance parameters (void fraction, pressure loss coefficient, and swirl angle) are not well correlated with observed vortex formations. (2) Air Ingestion Measured levels of air ingestion, even with air core vortices, were generally less than 2%. Maximum values of air ingestion with deliberately induced swirl and blockage conditions were less than 7% for horizontal inlets and 12% for vertical inlets. These high levels always occurred for high flow and low submergence (Froude number (Fr) generally greater than 1.0). For submergences of 8 feet or more, none of the configurations tested indicated air-drawing vortices ingesting more than 1% over the entire flow range, even with severe flow perturbations. (3) Swirl (measured at a distance 14 diameters from suction inlet) Flow swirl within the intake pipes, with or without flow perturbations, was

very low. In almost all cases, the swirl angle was less than 4 degrees, an ~ NUREG-0897, Revision 1 3-85

m- /

                                                                                                                   =
                              .-           ,.,      .                                                      p I
             =. .

L

        .i
                                                                                                                  =

l ms r .. . l *

                                                               "                    **                     .s i    == .

r=" l .. .

                                                 ..    .. s.                  2
                                                                                       -! d .                    -               .
                                         . L .a a 1 . L ." . $ 'r ". f 1a .=>^ *. *: ' ,,3,.
             .=

s. u i

a. a. u.

a u. u. u. u a

                                                                                    .vr
                                  ...~....           .
                                                                                              =

[ f

                                                                                ,i                                  .

l .' -

              .s .                                               .                 .                ,

i

g .. .

igu

                                                     ". J. * * (. . .*, , . .
                   .                                       .J.l ' * * " ,*                                *
                                        .. U..STJ.Eh.dd.&d..!I.i.:

e . .vr Figure 3.34 Void fraction (% by volume) as a function of Froude number; horizontal intake configuration; only data points indicating . nonzero void fraction are plotted l NUREG-0897, Revision 1 3-86

C I,.7. .

I"

i&TlT:WF-Ec. .. . . :: - Ii .e. . t. .% .k.>.4.,. ,..

                                                                                      ...n .. .Ag<. . :,. #-:. ...
                                              .l . .          .

i c . E

                                                                       $.. b                                                       W:.. "4. , ,         .

Iu

                                                                                         ,                                                    J1
                                                                    A                 ,

9 ,j,.9* (. . . l ..

                                                       ..    . fMBRWf.:.y_
                                                      .a.                :.
                                                                                                      .nsnee s. amen, vy-

c. c. c. c. c. i.

Figure 3.35 Surface vortex type as a function of Froude number; horizontal intake configuration (Type 1: surface swirl only Type 6: full air core to intake)

l. . ::=::"r '."' .

a 8 .. .

                                                    .I ..
                                                                                                                                                    ,        ... ,     .s
                                                                                          . 's ? . z .. .oght.4,.*'                                      ..

t,[ l, .IS[y[E9s.gS"?.l . F . mw - war-c: 1". $. ' .. Figure 3.36 Swirl as a function of Froude number; horizontal intake configuration - NUREG-0897, Revision 1 3-87

I a - 3

1 iss .

5

is s .

_b . . . E ~ v> ize .. - 8 j iss i as . [

7

Es . x 5 54.s .

                                                                                                          =
       .&                                                       x       =
                                                                            "   n                   **

a Ls . y 5 .. . sa

                             ,   ....a!"

s. ALM

s. . na 1s. W .

E M s'F

s. a s. .

E... '.oys. .,*f a. , g g FROUDE NUMBER. uA/g

       $ as W
       % its      .

E o UNIFORM APPROACH FLOW TESTS (EXCT.PT FOR c = 0 SUMPS) 2

UNIFORM APPROACH FLOW TESTS (FOR c IO)

E ' s, s s PERTURBED APPROACH FLOW TESTS - E O it. p . m 2 it s . 4

       $                                                        N 4
       / is. a    .

ft 8 & g as . 5-E as . O

4. s .

C e , g ts . , u W W

      $      g,             ,      , a .3 o "m

8 - = a h "31_em* e am s o.aI $% en

s. s a. a L. a, e s. s s. s s. a n. e s. s s. s a, e FROUDE NUMBER, uG g Figure 3.37 Void fraction data for various Froude numbers; vertical intake configuration; only data for nonzero void fraction plotted NUREG-0897, Revision 1 3-88

                                                                                          "                                  ~
                                                " - ~ 'i 5 '                                             -

W. '." I"~ f: .. s::s , ." -

1,, -

                                    /              ..< -e<t . . t
                      . . ..                   a ; .... .-                              .
                                    "< u "

g- -

. e ,< " ', . , ,

t , '. . e . .

                                    ,, c~:1.;. s.. y, ~a.                               ., t..

p. . . s ,u .

1 :,5u i t.c : 4 ,. - p .

                        "L.      6      a       a       i.         J.          a            J.           J.                a        as e.ousemm.sa,wfT-Figure 3.38 Surface vortex type as a function of Froude number; vertical intake configuration (Type 1:                          surface swirl only; Type 6:

full air core to intake) as

r. . .

t -

                                                            /                   -

s !q a. .

                                                                ,,,,,,,,,,,,,,,,g,,,,,,,,,,,,,,,,..

E , u ns. , wee em . m - q],,. . . ~ - I .

                 = f ...                             -

3a ,. g 5 ... , . . , , , . , ad ".st.." S g5 ,, g;, . ,s , .J , .

                                                                             ,.. s .,       .                                  ,
                 <eu.                ,
                                           .g                     4                               .                                 .

p- - s,yJ.y"g%i..... if

                                                                                                     .p.                    .
                         *I.

a 4e.=,us.s;s.

                                     .         " JJ"irm     , ,. 1 a, ay ....

a . t ,f.* a -a, u enews mas.ea. .y;T Figure 3.39 Swirl as a function of Froude number; vertical intake _ configuration . NUREG-0897, Revision 1 3-89

i , , l ! VOMTEX l TYPE t I 1 v, INCOHERENT SURFACE SWlRL , l 1 2 7 SURFACE DIMPLE: COHERENT SWIRL AT SURFACE l I 3 $ DYE CORE TO INTAKE; } (X)HERENT SWlRL THROUGHOUT WATER COLUMN ; l 4 VORTEX PULLING FLOATING TRASH. BUT NOT AIR t l l t l 5 -$ VORTEX PULLING AIR l ( h OU88LES TO INTAKE l l l o.........

                                                                                                                       '4 6

FULL AIR CORE h5 TOINT4xE l l l Figure 3.40 Vortex type classification . 1 l l I NUREG-0897, Revision 1 3-90

I

acceptable value for RHR and CSS pumps. The maximum value for severely perturbed flows was about 8 degrees and occurred during the screen blockage test series.

(4) Sump Head Losses The suction pipe intake pressure loss coefficient for most of the tests, with and without flow perturbations, was in the range of 0.8 1 0.2 and agreed with recommended values in standard hydraulic handbooks. 3.4.3 PWR Sump Performance Ouring Simulated Accident Conditions (Perturbed Flow) The following items were considered for sump performance under perturbed flow conditions. l (1) Screen Blockage Screen blockages up to 75% of the sump screen resulted in air ingestion , levels similar to those noted under 3.4.2(2) above. (2) Nonuniform Approach Flow Distributions Nonuniform approach flows, particularly streaming flow, generally increased surface vortexing and the associated void fraction. (3) Orain and Break Flow Drain and break flow effects were generally found not to cause any additional air ingestion. They reduced vortexing severities by surface wave action. (4) Obstructions - Obstructions 2 fee *,or less in cross-section had nn influence on vortexing, air withdrawals, swirl, or inlet losses. NUREG-0807, Nevision 1 3-91

(5) Transients Under transient startup conditions, momentary vortices were strong, but no air-core vortices giving withdrawals exceeding 5% void fraction (1- minute average) were observed. 3.4.4 Geometric and Design Effects (Unperturbed Flow Tests) In general, no consistent trends applicable for the entire range of tests were observed in the data between the hydraulic response of the sump (air withdrawal, swirl, etc.) and secondary geometric parameters. However, for some ranges of flow and submergence, the following observations are applicable: (1) Greater depth from containment floor to the pipe centerline reduces surface vortexing and swirl. (2) Lower approach flow depths with higher approach velocities may cause ' ,l increased turbulence levels serving to dissipate surface vortexing. ( (3) Suction pipe inlets located with less distance to the adjacent sump wall and greater pipe spacing reduces vortexing and swirl. , (4) There is no advantage in extending the suction pipe beyond 1 pipe diameter from the wall. 3.4.5 Design or Operational items of Special Concern in PWR ECCS Sumps , l (1) Pump Intake Orientation j t Comparison of vertical intake data to corresponding horizontal intake data . showed minor differences in hydraulic performance for sumps of the same geometry and flow conditions. Average vortex types agreed within i 1 (types range from 1, incoherent surface swirl, to 6. full air core to pump Intake); NUREG-0897, Revision 1 3-92

1 air withdrawals were somewhat higher for vertical intake sumps but usually within 1% (30-minute averages) to 4% (1- and 5-minute averages); swirl angles differed only within t 1 degree. Both vertical and horizontal intake sumps performed better under perturbed flow when the pipe inlets were closer to an adjacent wall rather than at the center of the sump. (2) Single Intake Sumps l Two sump configurationi (4 x 4 feet and 7 x 5 feet in plan, both 4.5 feet deep with 12-inch-diameter intakes) were tested under unperturbed (uniform) and perturbed approach flows with screen blockages up to 75% of the screen area. For both the configurations, unperturbed flow tests indicated air l withdrawals were always less than L% by volume for the entire range of tested flows and submergences (Fr = 0.3 to 1.6.). Even with perturbed flows, zero or near zero air withdrawals were measured in both sumps for Froude numbers less than 0.8, suggesting insignificant vortexing problems. For Froude numbers ! above 0.8, a few tests indicated significantly high air withdrawal (up to 17.4% air by volume; 1 minute average) especially for the smaller sized sump. , l Measured swirl values in the pipes were insignificant for both the tested sumps, in the range of 2 to 3 degrees, even with flow perturbations. The inlet loss coefficients for both sump configurations were in the expected ranges for such protruding inlets 0.8 2 0.2. l (3) Oual-Intake Sumps with Solid Partition Walls Four dual-intake sump configurations (one 20 x 10-foot sump with 24-inch i diameter intakes and three 8 x 10-foot sumps with 24-inch, 12-inch, and 6-inch intakes, respectively) were tested with solid partition walls in the sumps between the pipe inlets and with only one intake operational. None of ! the tests indicated any significant increases in vortexing, air withdrawal, swirl, or inlet losses compared to dual pipe operation without partition walls. Thus, a partition wall in a sump should not cause any additional problems when only one pipe is operating. l l l NUREG-0897, Revision 1 3-93

j . j (4) Bellmouths at Pipe Entrance t t Limited tests on a sump configuration were conducted with and without a bellmouth attachment to the 12-inch intakes. Adding bellmouths at the f pipe entrances did not result in any significant changes in the vortex - types, air withdrawals, and pipe swirl compared to those that otherwise l ! existed under the same hydraulic conditions. An expected redaction of up to , about 40% in inlet losses was noticed with the addition of a bellmouth, j (5) Cover Plate A solid top cover plate over the sump was effective in suppressing vortices j as long as the cover plate was submerged and proper venting of air from l j underneath was provided. No air drawing vortices were observed for the submerged cover plate tests, and no significant changes in swirl or loss , coefficients occurred. (6) Vortex Suppressors ,, .i Cage-shaped vortex suppressors made of floor grating in the form of cubes 3 } 1 and 4 feet on a side and single or multiple layers of horizontal floor )

gratings over the entire sump area were found to be effective in suppressing vortices and reducing air ingestion to zero. These suppressors were tested 4 in sump configurations using 12-inch-diameter intake pipes, and with the water levels ranging from 0.5 foot to 6.5 feet above the top of the suppressors. Adverse screen blockages were imposed on these sump j

configurations, which produced considerable air ingestion and strong ( vortexing without the suppressors; thus, the effectiveness of the suppressors was tested when hydraulic conditions were least desirable. The suppressors I } also reduced pipe swirl and did not cause any significant increase in inlet ! ! Iosses. Both the cage-shaped grating suppressors and the horizontal floor - l grates were made of standard 1.5-inch floor grates. I i I ! 1 i i l NUREG 0897 Revision 1 3-94

l , . l Tests on a cage-shaped suppressor less than 3 feet on a side indicated the existence of air-core vortices for certain ranges of flow and submergences, even though air withdrawals were found to be reduced to insignificant levels. Therefore, either properly sized cage shaped suppressors made of floor grating, or floor grating over the entire sump area, may be used to reduce I cir-ingestion to zero in cases where the sump design and/or approach flow creates otherwise undesirable vortexing and air-ingestion. (7) Scale Model Tests To evaluate the use of reduced scale hydraulic models to determine the performance of containment emergency sumps and to investigate, in particular, possible scale effects in modeling the hydraulic phenomenon of concern, a test program involving two reduced-scale models (1:2 and 1:4) of a full-size sump (1:1) was undertaken (NUREG/CR-2760). The test results show that the hydraulic models predicted the hydraulic , performance of the full-sized sump; namely, vortexing, air-ingestion from free surface vortices, pipe flow swirl, and the inlet loss coefficient. No scale effects on vortexing or air-withdrawals were apparent within the tested range for both models. However, an accurate prediction of pipe flow swirl and inlet loss coefficient was found to require that the approach flow Reynolds number and the pipe Reynolds number be above certain limits. Based on these results, it is concluded that properly designed and operated reduced scale hydraulic models of geometric scales 1:4 or larger could be used to properly evaluate the hydraulic performance of a sump design. Evaluations of sump hydraulle model studies conducted in the past can be derived from this series of tests,

r l l i NUREG-0897, Revision 1 3-95 l

(8) Pump Overspeed Tests Two 8 x 10 x 4.5 ft sumps (one with horizontal suction intakes and one with vertical suction intakes) were tested at higher flow rates to simulate pump overspeed or run out (to Froude number = 1.6) conditions. No strong air-core vortices were observed with air-withdrawals greater than 1 percent (1 minute or 30 minute averages). l M:ximum recorded pipe swirl angle was 0.9* (at 14.5 pipe diameters from entrance); inlet loss coefficients averaged 0.8 (NUREG/CR-2761). (9) High Temperature Tests A series of tests were performed on horizontal suction intake, and the conclusion was that changing water temperatures over the range from 40*F to 165*F had no significant ef fect on sump hydraulic performance parameters (see NUREG/CR-2758, Section 4.6.). 3.4.6 BWR Suction Pipe Intakes Because BWR plants do not have a sump or a floor depression with surrounding screens and gratings, typical residual heat removal system suction pipe inlet configurations applicable to Mark I, Mark II, and Mark III containment designs were investigated in full-scale flow experiments. Figure 3.41 shows the two inlet pipe and strainer configurations of the three designs under consideration. Key parameters of interest were air" ingestion levelt, vortex formation, suction pipe swirl, and the RHR inlet pressure loss coefficient. The tests were conducted with both perturbed and unperturbed approach flows to the inlets, as indicated in Figure 3.42. Flows ranged from 2000 to 12000 gpm per pipe, while submergences varied from 2 to 5 feet. The resulting Froude numbers ranged from about 0.2 to 1.1. NUREG-0897, Revision 1 3-96

s . . . . . __ NO. 8 MESH SCREEN 1/8" HOLE OVER 5/8" HOLE PERFORATED PLATE PERFORATED PLATE (40% OPEN AREA) (24% OPEN AREA) .

                                                                - 6"                 "      00v 24"                                   O
                                                                                            ????M                        t ggn     >ggggggyC                   3g" y      oooo r

u CLOSED END SOLID PLATE CLOSED END

36" : *- - 28 " ---*

DETAll A DETAIL B

                               *- 7 * --*
                                                                                                          -* + 6"
                           \-                                                                        %
                           '             bg CONICAL      STRAINER                                6 l'

[ FALSE WALL (SEE OETAIL Al I CONICAL f I I Y STRAINER

                                                                                               '            l     (SEE DETAll 8)

j O , .l_

                       -V                       24" PIPE                                              ,,,   24" PIPE PLAN VIEW                                                     PLAN VIEW 24" PIPE                                                    4          24" PlPE 4-s
                                                                                                                         " ].

k SUBMERGENCE ~ SUBMERGENCE F- ; ; .- O F- ,_. O t , f t :s , t 7' j-CONTAINMENT 7' l f-CDNTAINMENT { ..,.,.y

                          ;;                      / FLOOR                      [                ,,

r w i * . c r~ ~

                                                                                                                   / FLOOR
                                  .      . . p . . . .. s . .

NOT TO SCALE ELEVATION ELEVATION CONFIGURATION A CONFIGURATION 8 (MARK H AND MARK E DESIGNS) (MARK I DESIGN) Figure 3.41 BWR pipe inlet configurations as built in full-size facility NUREG-0897, Revision 1 3-97

E A JL a 4 a g - w w

   "   o                                  r                             -T                         n
                                              $                                                     d i
  ~

o

   =   -
                                         =
                                                \.                      _,,

7 -= b

               /
                                  )

b _

                                                         /               E j             k k---o.s LWl i I i i i i i i i SWIR L COUPLE DOUBLE SWIRL NON-UNIFORM FLOW (FLOW DISTRIBUTOR BLOCKAGE)

W-30 FEET L-60 FEET

1 Figure 3.42 Perturbed flow schemes; schemes A, B, and 0 used for BWR inlet tests

4 % f Figures 3.43 and 3.44 show the test-average (30-minute) and 1-minute void fractions for the two inlet configurations (A and B) and the various flow schemes examined. Essentially zero air withdrawal was measured for both configurations at Froude numbers less than or equal to 0.6 under all tested approach flows. For the double inlet or tee inlet design (Configuration A), maximum air withdrawal was less than 0.5% at all Froude numbers examined. For the single inlet design (Configuration B), air core vortices drawing up , to 4% air by volume were observed to form at a Froude number above 0.6 under perturbed approach flows. No air-core vortices were observed for either inlet configuration over the entire range of tested flows at submergences equal to or above 3.5 feet (Froude numbers less than 0.6). Swirl angle in the Configuration B inlet

            ' pipe ranged from 0 to 3 degrees, while the Configuration A pipe swirl angle fell between 2 and 7 degrees for the Froude numbers tested.
        ?

The measured inlet loss coefficients expressed in terms of suction pipe velocity head averaged to about 1.7 and 1.0 for Configurations A and B , respectively. The loss coefficients reflect entrance, strainer, and tee losses (if applicable). f NUREG-0897, Revision 1 3-99

 ~ . - . -      .                        -                                                                                     .

s. = ,

w E e 7*' -

9 ABOUT 15 PSI w

                   # s. s_

w

                   >-                         . umsonM AMac4CH Flow D                       NON-UNtrORM AMROACH PLOW
                   $ is        .              a SCwEME A 2                          =sCusMe e esewtus o o
                   "     4. s.

e n

                  .- 1 s.

z 9 D 2.s . m m

1. s_

a 5 as a_ _u _x _ i ,.

s. s s. 2 s. 4 s. s s. s 1. s i. 2 a. 4 1. s i.e 2. s FROUDE NUMBER. ug

a. CONFIGURATION A

s. s w

o e 78- 9 ABOUT 15 PSI w w -

                  >=
                                             . umronM AMacACH Flow N04-UNipOAM AMROACH FLOW D                           aSCwfME A E is,                       e KMW s 2                           e SCMGME O n        s .

Y %

                  ." 18.

2

9 l b ts.

a: m g 1. s. 6 ga _& _ _m _ a_ m .s e a e e i t Le s. 2 s. 4 s. 6 s. s 1. s !. 2 1. 4 1. 6 f. s 2. s FROUDE NUMBER. uA/gs

b. CONFIGURATION B Figure 3.43 Test-average void fractions for tested BWR suction intakes _.

NUREG-0897, Revision 1 3-100

                             &8
                                       . O ABOUT 15 PSI
                        < . 7. .  .

z-W

                        < EI-                . m,on. AmeACH etow W                pdOfe-UBelFOnes APPAoACH Flow h                    a SCHEME A
                                             . SCHEut a Z   SE .

g

SCHEWE D a *b E_

.~

z 3. 5_ 9 b

                       < 2. s.

E 4 1 .E . . 3

                                       *m  _ m           m_           .      ..     ,        ,     ,     ,

E S' E. 2 E4 8. 8 8. 0 1. 0 1. 2 1. 4 8. 8 f. 8 2. 8 FROUDE NUMBER, ug

a. CONFIGURATION A Es g..

                       < 7.
                                         . @ ABOUT 15 PSI z                    .

W

                      >                        . U woam AmoACH etow
                       < EI-               PsO40NtFonu APPROACH Flow W                        e SCHEME A
                      >=                       oSCHEu6 e                               -
                      @     s,,                e sCHEuf o 5
                        ,   4. s.
                      -                                                       S 6

z ss_ o 9 0

                      < 2. n .

a: 4

1. E.

                     >                                       g        A u       n          O  t              t     t La         2 - . L                              s                t

s. s e. 2 s. 4 s. e s. s t. s t. 2 1. 4 1. s i. e 2. s FROUDE NUMBER, ug

                                                           'b. CONFIGURATION B Figure 3.44 1-minute average void fraction for tested BWR suction intakes l

NUREG-0897, Revision 1 3-101

 . .. 2  . ..                      .       m-     . ......-..a...~...-.-     - . . . - . . .  ..   . - - . . . . . . . -

4 INDEPENDENT PROGRAM TECHNICAL REVIEWS f Independent program technical reviews were conducted before and during key phases of the work reported in Section 3 to solicit comments and technical views about the program's direction and goals from experts not connected with the implementation and execution of Task Action Plan (TAP) A-43. The reviewers were selected from among the foremost authorities in each of the areas reviewed. Two reviews were conducted: sump hydraulic performance and insulation debris calculational methods effects. 4.1 Sump Hydraulic Performance Review The sump hydraulic performance review consisted of two panel meetings,* held on March 17 and June 4, 1981. The primary purpose of the first meeting was to introduce in detail the program plan and initial test results. The second meeting was primarily for reviewer followup response and comment. Addition-ally, at both meetings the reviewers were provided with preliminary program redirections, and were asked to comment on results to date and give an . analysis of the proposed future program plan. Overall, the reviewers approved of the program, the experimental test plan, its conduct, and data analysis. They concluded that the program was appropriate for resolving the sump hydraulic performance issues.

Meetings were held on March 17, 1981, at Germantown, Maryland, and June 4, 1981, at Alden Research Laboratory of Worcester Polytechnic Institute, Holden, Massachusetts. Those attending and their affiliations were P. Tullis/

Utah State University; D. Simons/Simons, Li and Associates; R. Gardiner/ Western Canada Hydraulic Laboratories; D. Canup/ Duke Power Company; W. Butler /NRC; S. Vigander/ Tennessee Valley Authority (TVA); J. Kennedy / University of Iowa; and R. Letendre/ Combustion Engineering, Inc. (R. Letendre did not attend the meeting of June 4, 1981.) Those attending were asked to provide formal written responses and comments at the close of the second meeting. Copies of the responses are available through the Office of Light Water Safety Research, Department of Energy, Washington, DC. - NUREG-0897, Revision 1 4-1

Divergent opinions emerged during the review concerning the potential for pump performance degradation when the fluid temperature was near saturation. Some concerns were expressed regarding the possibility of degraded pump performance as a result of cavitation or the release of dissolved air into the water in the suction lines leading to the pumps. Other opinions suggested that pump performance should be satisfactory at coolant temperatures near saturation, because the (1) solubility of air in water is lov near saturation and, (2) if cavitation were not occurring in the pump, any voids would collapse as a result of the static pressure increase with depth in the sump. These collapsing bubbles would then form a turbulent environment and inhibit surface vortex activity. Although the pump issues raised by the-reviewers are indirectly pertinent to the sump hydraulics program, they are a part of USI A-43 and have been addressed (see Section 3.2). In direct response to reviewer comments, elevated temperature tests were performed immediately following the first 25 configurations, which was earlier in the program than originally planned. The experimental research _ program did not examine the effects on operation at temperatures near saturation conditions due to the operational limits of the experimental facility (about 165 F). However, up to that limit, no significant, or adverse, temperature effects on sump system performance were detected. An area of general peer review group agreement was that sump system

  ' performance with respect to air entrainment could be improved in most sump configurations by the addition of a vortex suppression device (s).                            One reviewer, however, commented that such a device (s) might be removed during NUREG-0897, Revision 1                    4-2

       ..     ---. . . . . . . . . - . - . . . - . ~       --_.-.a.....        . . . . . - . - .   ..

some phase of reactor operations and not be replaced. Such a possibility, in his judgment, was sufficient justification for an experimental research program that would allow the development of adequate sump design guidelines that were based upon justifiable physical criteria (in the absence of vortex suppressors). The results of the studies provided in Section 3.4 confirm the effectiveness of vortex suppressors to reduce air ingestion to zero and provide hydraulic results for developing acceptable sump design guidelines. The adequacy of recirculation SJmp pumps for performing reliably when ingesting air / water mixtures was a matter of some concern to the review group. These concerns have been resolved by the development of sump design guidelines that take into account pump performance specifications under such conditions. 4.2 Insulation Debris Effects Review The purpose of the insulation debris effects review was to determine the adequacy of metnods (described in Section 3.2 and in detail in NUREG/CR-2791) . to conservatively estimate quantities of insulation debris that might be produced in containment, its transport, and its potential for sump screen blockage. The review was conducted in two phases. In the initial phase, a draft report describing the methods was provided to peer panel and other reviewers

to solicit their comments. Reviewers provided highly useful criticisms and comments with recommendations for improvements in the physical basis and rigor of the development of the debris generation and transport models.

   *The peer panel reviewers and their affiliations were R. Gardiner/ Western Canada Hydraulic Laboratories; D. Simons/Simons, Li & Associates, Inc.;

D. Canup/ Duke Power Company; R. Mango / Combustion Engineering, Inc.; P. Tullis/ Utah State Unviersity; J. Kennedy / University of Iowa; W. Butler /NRC; and ' S. Vigander/TVA. Other reviewers included G. Weigand/Sandia and R. Bosnak, G. Mazetis, and T. Speis/NRC. Their written review comments are available - through the NRC Division of Safety Technology, NRC, Washington, DC 20555. NUREG-0897, Revision 1 4-3

 -              . . . . - . .~.. ... -. - . - - . .         . . . - . -

The draft document was then modified in response to the comments of the reviewers. The modified document was transmitted to the reviewers, who were then requested to prepare comments for a formal peer panel review, which was the second phase of the review process. Formal peer panel review took place at NRC Headquarters on March 31, 1982. Panelists Kennedy and Canup were unable to attend the meeting; however, a number of other persons, in addition to peer panel members, participated in the review.* Questions that were raised during the meeting and their disposition are given below. [t was observed that, under some circumstances, the amount of debris generated with the potential to migrate to the sump could be greater than that estimated in the draft report. This concern was resolved by determining that the report would require the selection of those pipe break locations and jet targets that would generate the maximum quantities of potentially transportable debris without regard to initial blowdown and transport direction. _ Questions were raised about (1) the applicability of the jet model used in the deoris generation portion of the report, (2) the assumption of uniform distribution of debris across the face of the jet and, (3) the use of a 0.5 psi stagnation pressure cutoff for debris generation. Resolution of (1) was arrived at by agreement that a modified Moody jet model (Moody, 1973) would be allowed to model the jet. It was agreed that the stripping of all insulation from plant and piping within the crane wall and within the jet represented a conservative treatment of insulation debris generation.

   *0ther' attendees were: S. Hanauer, K. Kniel, C. Liang, P. Norian, F. Orr, A. Serkiz, J. Shapaker/NRC; G. Hecker/Alden Research Laboratory; E. Gahan, J. Wysocki/ Burns and Roe; W. Swift /Creare, Inc.; and P. Strom and                         -

G. Weigand/Sandia. NUREG-0897, Revision 1 4-4

s _ _ . ..~.. _ .,.. _ _ . _ .._ _.._ _ .-_ - .-.~ _ .. . - - . . _ _ _ _ l I Discussions of (2) concluded that a definite probability existed that debris distribution across the face of the jet would not be uniform. It was agreed that a distribution of debris across the jet face would be provided that would represent the geometric distribution of insulation targeted by the jet in the containment. In addition, because of uncertainties in jet transport to walls, it was agreed that the quantities of debris estimated to exit through crane wall openings would be doubled. The use of a 0.5 psi stagnation pressure cut-off (item (3)) for insulation damage was questioned by a number of reviewers. Technical views were put forward by a Sandia staff member on the expected performance of jets under LOCA conditions. He stated that centerline stagnation pressures above 15 psig could be expected for at least five diameters downstream of high-energy, high pressure breaks. An Atomic Energy Commission report (Glasstone, 1981) was cited by Burns and Roe as the origin of the cut-off estimate for debris generation. Alden Research Laboratory personnel reported on preliminary experiments at ARL that have shown t'-t little insulation damage occurred to fibrous insulation assemblie up to 6.5 psi water jet pressures. It was agreed

that the 0.5 psi stagnation pressure represented a conservative treatment for the onset of insulation debris generation. It was further agreed that the assumption that all insulation within the jet cone would be transformed to insulation debris was conservative. This assumption was chosen to represent the volume within which insulation debris would be generated under the treatment provided in NUREG/CR-2791. The results of work performed subsequently on these issues are provided in Sections 3.3 and 5.3 of this report.

                        *This decision has been superseded by information discussed in Section 3.3.

NUREG-0897, Revision 1 4-5

Discussions were held on the physical accuracy of the model in representing pipe whip, pipe impact, and the direction of motion of dislodged insulation and its trajectory. It was first pointed out that the quantity of insulation generated by this mechanism would amount to 10% or less of that generated by jet forces. It was further pointed out that the use of the treatment in the report would conservatively estimate the quantities of insulation debris produced by a minor contributor to debris production and, as such, was satisfactory. Questions were raised on the treatment of long term transport following blowdown. These questions related to (1) recirculation flow velocities within containment (2) hydraulic lift provided to sunken debris (3) drawdown of floating debris onto less than fully submerged sump screens (ice-jam effect) (4) transport mechanisms of sunken debris, such as tumbling and sliding In the resolution of (1), agreement was reached to account for obstructions in flow paths and subsequent flow expansion (Appendix 0 and NUREG/CR-2791). Agreement was reached on (2) that for horizontal orientation, lift would be approximated by drag for horizontal debris, would be zero for vertically oriented debris, and would be disregarded for tumbling debris. NUREG-0897, Revision 1 4-6

  . m         _.._. . . . .      _.... _.                         ..-                 _     .. _ . .                                    .

8

Item (3) was recognized as a potentially important mechanism for screen blockage. It will be treated by established methods available as described in the literature, (Uzuner, July 1977; NUREG/CR-2791).

Tumbling and other transport mechanisms, as noted under (4), could I significantly affect the movement of debris towards screens. Panelists agreed to treatments that they considered to be conservative in dealing ! with debris transported by these mechanisms. Recent experiments at ARL have shown a wide variability of transport characteristics depending on the debris geometry (section 3.3; NUREG/CRs-2982 and -3616). Arguments were raised that a period of debris transport (intermediate-to short-term transport and long-term transport, as defined here) might exist. It was postulated that transport during such an interim period might seriously affect potential sump blockage. Because the report assumes that all floating debris reaches the sump, such an interim migration period would not affect the consequences of such transport. With respect to debris of density equal to or greater than unity and its . transport, discussions brought out views that the likelihood of a significant effect during such an interim period would be minor, flow patterns would show no preferential transport toward the sump, and entrainment would be higher in the recirculation mode than in the interim period. An issue that was not resolved concerned the behavior of fibrous insulation in its migration toward a sump and the potential for blockage by such material. Because this problem appears to exist at only a few plants, it is considered plant-specific. Nevertheless, it was an open issue at the time of the meetings. Following the meetings, experimental studies were conducted at ARL to estimate stagnation pressures required for the onset of debris generation for nonencapsulated mineral wool and fiberglass insulations (NUREG/CR-3170), the transport characteristics of such debris, and the pressure losses NUREG-0897, Revision 1 4-7

at sump screens caused by the accumulation of fibrous debris on screens (NUREG/CR-2982). These findings are reflected in the findings provided in Sections 3.3 and 5.3 of this report. All panelists, except S. Vigander of TVA, concluded that the use of the methods discussed would result in conservative estimates of sump screen blockage. Vigander commented that while he was of the opinion that the treatment would yield conservative, perhaps ultra-conservative, results, he could not with certainty arrive at that conclusion. He suggested that uncertainty analyses be conducted to establish the levels of conservatism (if any) that are provided in the development. Other panelists agreed that quantitative or qualitative error analyses would be desirable, although tne needs for such analyses were deemed not to be immediate or pressing. NUREG-0897, Revision 1 4-8

5

SUMMARY

OF SUMP PERFORMANCE TECHNICAL FINDINGS 5.1 General Overview Emergency core cooling systems require a clean and reliable water source for maintaining long-term recirculation following a LOCA. PWRs rely on the containment emergency sump to provide such a water supply to residual heat removal pumps and containment spray pumps. BWRs rely on pump suction intakes located in the suppression pool, or wet well, to provide a water source to residual heat removal pumps and core spray pumps. Thus, recirculation pump performance under post-LOCA conditions must be evaluated for both BWRs and PWRs. Typical technical considerations are shown in Figure 5.1. Each major area of concern- pump performance, sump hydraulics, and debris generation potential--can be assessed separately, but the combined effects of all three areas should then be assessed to determine the overall effect on both the available and required NPSH requirements of the pumps. The sections below , summarize technical findings and provide concise data sets. . 5.2 Sump Hydraulic Performance Full scale tests show that adequate PWR sump (or BWR RHR suction intake) hydraulic performance is principally a function of depth of water (the submergence level of the suction pipe) and the rate of pumping (suction inlet water velocity). These variables can be combined to form a dimensionless quantity defined as the Froude number Froude number = U/ 5 NUREG-0897, Revision 1 5-1

r ! 2 - g DEBRIS SUMPS PUMPS S

Types. Quantities, and Location
Location in Plant;

 $      of insulation Redundancy

Pump Design and Operating Characteristics x
Containment Layout and Broek
Geometric Parameters
NPSH Requirements (No Airl tj Locations

( o

s

Estimate Quantity of Debris Generated
I" C e e

                                                                                                                            ' 'Ptors (Racks. Screensl.

Sump and Suction Piping - Losses i

i l ' I f i f i

Effects of Air ingestion on
Short-Term Transport by
Hydraulic Characteristics NPSH Required Blowdown Jet -Water Level Above Sump Outlet

                                                                                                                      - Sump Outlet Velocity                  ,     ,       p      ,,,

Long-Term Transport by -Air Ingestion -Inlet Design Recirculation Velocities -Inlet Losses - Temperature Effects T

N t i

Effects of Particulate and Debris Ingestion 1 P 1 r
Potential for Interceptor Blockage
NPSH Required
Head Loss Across Interceptors
NPSH Avellable
1 f

                                                                                                       -            Is There Adequate NPSH Margin      -

Under All Post LOCA Conditions? ' Figure 5.1 Technical considerations relevant to ECCS , sump performance

0

where U = suction pipe mean velocity ,

1 s = submergence (water depth from surface to suction pipe centerline) g = acceleration due to gravity The extent of air ingestion is the principal parameter to be determined. Small amounts of air (less than 24 by volume) do not significantly degraua pumping capacity (Merry, 1976; Murakami and Minemura, 1977; and Florjancic, 1970). Generally speaking, full-scale tests revealed low levels of air ingestion (< 2%) over a wide range of Froude numbers despite the presence of air-core vortices. Other hydraulic effects, such as intake swirl, were found to be small, and inlet loss coefficients were in agreement with handbook values for similar intake geometries.

Section 3.4 summarizes the results of full-scale PWR sump hydraulic tests and BWR suction inlet tests. Figures 3.34 and 3.37 show typical void fraction data as a function of Froude number for PWR sumps; Figures 3.43 and 3.44 ,

show void fraction data for BWR suction inlets. More detailed results are provided in NUREG/CR-2758; NUREG/CR-2759; NUREG/CR-2760; NUREG/CR-2761; and NUREG/CR-2772. Generally, sump (or suction intake) design acceptability should be based upon < 2% air ingestion criteria. PWR sump hydraulic performance can, therefore, be assessed as follows: 1 (1) Table 5.1 summarizes the conditions for PWR type sump designs where .i negligible (or zero) air ingestion would exist. Adequate submergence and low intake velocities are the key parameters derived from ARL tests. i . NUREG-0897, Revision 1 5-3 _ - _ - _ _ ~ . _ -- _. . _ - - , - ._ . -__ - .-, . .__ _---. ,___ - _ - - __. , ,

l Table 5.1 Hydraulic design findings" for zero air ingestion Item Horizontal Outlets Vertical Outlets Minimum submergence, s (ft) 9 9 (m) 2.7 2.7 Maximum Froude Number, Fr 0.25 0.25 Maximum Pipe Velocity, U (ft/s) 4 4 (m/s) 1.2 1.2 Cover Plate

                                       #                                and i                              '

Debris Screen j! U Minimum Water 1' . ll n Level as gj in Determined ll

                                          During Design                             _

a MW ar. j. ~,9.el i+ c'"o i u P - - h.f h ..

*The hydraulic findings were established using experimental results from NUREG/CRs-2758, -2759, and -2760, and the variable ranges over which such data were taken for sump geometries which were of rectilinear design.

NUREG-0897, Revision 1 5-4

e . (2) If the adequacy of the sump geometric design and hydraulic performance is to be based on air ingestion levels of 1 2%, such assessments can be made using Tables 5.2, 5.3, 5.4, and 5.5. Under such conditions, sump design features should be comparable with those sump geometries tested at ARL and as noted in these tables. (3) Vortex suppressors provide a very effective means to achieve zero air ingestion. Vortex suppression devices such as those shown in Table 5.6 have been shown to reduce air ingestion measured levels to zero on PWR sump designs. (4) Table 5.7 provides additional information pertinent to screens and grates that could affect PWR sump hydraulic performance and represents the types tested at ARL. (5) Elevated water temperature has been shown to have negligible effect on sump hydraulic performance in full-scale tests conducted at temperatures up to 165 F. . BWR pump suction intake designs (employing suction strainers) that result in a Froude number of 1 0.6 were found to have insignificant air ingestion. NUREG/CR-2772 reports experimental findings for Mark I, Mark II, and III intake designs. 5.3 Debris Assessments Debris assessments should consider the initiating mechanisms (pipe break locations, orientations, and break jet energy content), the amount of debris that might be generated, short- and long-term transport, the potential for PWR sump screen or BWR suction strainer blockage, and head losses that could degrade available NPSH. In addition, an evaluation of the effects of small NUREG-0897, Revision 1 5-5

Table 5.2 Hydraulic design findings

for air ingestion < 2%

Air ingestion a is empirically calculated as a = a, + ay x Fr where a and a are coefficients derived from test resultsasgivkninthetablebelow Horizontal Outlets Vertical Outlets Item Dual Single Dual Singleam Coefficient u, -2.47 -4.75 -4.75 -9.14 Coefficient og 9.38 18.04 18.69 35.95 Minimum Submergence, s (ft) 7.5 8.0 7.5 10 (m) 2.3 2.4 2.3 3.1 Maximum Froude Number, Fr 0.5 0.4 0.4 0.3 Maximum Pipe Velocity, U (ft/s) 7.0 6.5 6.0 5.5 (m/s) 2.1 2.0 1. 8 1. 7 Maximum Screen Face Veloc;ity (blocked and minimum submergence) (ft/s) 3.0 3.0 3.0 3.0 (m/s) 0.9 0.9 0.9 0. 9

Maximum Approach Flow Velocity (ft/s) 0.36 0.36 0.36 0.36 (m/s) 0.11 0.11 0.11 0.11 Maximum Sump Outlet Coefficient Cg 1. 2 1. 2 1.2 1.2 c..., m. .

i , ~.... /cA-il

                                          " ,o $ $ '  g il i;d                                !" >9 u+-               l-
"See note on Table 5.1                 .

- These numbers are not from test data, but are extrapolated.
NUREG-0897, Revision 1 5-6

Table 5.3 Geometric d sign envelcpe guidelines for horizontal suction outlets a. . E

m a Size Sump Outlet Position

Screen m

e

 ."                                    Min.                      .

Min. Area m Sump Aspect Perimeter M Outlet Ratio (ft) (m) ey/d (B ey )/d c/d b/d f/d eX/d 2 (ft ) (m2 ) Dual 1 to 5 36 11 >4 75 7 8

 ~                                                      >1                 >3                > 1. 5     >l                   11.5 Single      1 to 5             16      4.9                                                                     -

35 3.3

Preferred location.

    ** Dimensions are always measured to pipe centerline.

T N Tresh neck and

                      ;                  L                 ;

Debele Screen

                  .f I                                                e o             e                         rb-II 11 lli l
                                                                                                              ,                  N ll                         ,,..

5 11 s mer= ll

rat ,'-

e ll ll m . _ . ll ll

                                                                                          ;- A_h - + -                  ,        'i.

ll . . ll ...

                                                                                                        'y-l    r'                      f1          11     "                    l                  .

d - lF

                                                                                                                   ~       ~

L --i ; ====iy-.Ji .

                                                                                                            *'~**'':

inJ tia -

                                                                                                       - .-     e, .--e.., -

_ . . , a- n ie - u. 5 ,'

Muurnum reatmeter = 2(L + SI ,

                                                                                                                 -      g      _

i

Table 5.4 Geometric design envelope guidelines 3 for vertical suction outlets ** A

8 e ,

 ."  I,                              Size                                    Sump Outlet Position

Screen

o W Min.

, 7 ' Sump Aspect Perineter Min. Area - T a Outlet Ratio (ft) (m) ey /d (B ey )/d c/d b/d f/d ex /d 2 (ft ) (,2) Dual 1 to 5 36 11 > _0 > _4 75 7 11 11 11 > 1. 5 Single 1 to 5 16 4.9 < 1. 5 - 35 3.3

Preferred location.

u, ** Dimensions are always measured to pipe centerline. cm Treeh Rock Detwe Screen I I: lI pl - = === = = = = = = _. !j] lli ll n r- N .. N. . il . lI , d d e ll b** 3 .- ,. lll 9 - ce ll L +- ll 1 f ;. ' ll . l e ,- .. li I l1l o o c- .t M

                                                                                                       ..-ry;                 ....u Lt_ _--     -._. _ _ _ _ _        - _j)                                     a         -

u . . I f

                          -: -j =         .-        4e: .                                                     ._. .         .-... .

Aspect Reelo = L/B 2 i : Minimum Perimetae = 2tL e Di

R :

e 0

4

Table 5.5 Additional considerations related to sump size and placement

1. The clearance between the trash rack and any wall or obstruction of length 1 equal to or greater than the length of the adjacent screen / grate (B s

                                   " 's) should be at least 4 ft (1.2 m).

2. A solid wall or large o'astruction may form the boundary of the sump on one side only, i.e., the sump must have three sides open to the -

approach flow.

3. These additional considerations are provided to ensure that the experimental data boundaries (upon which Tables 5.1, 5.2, 5.3, and 5.4 are based) resulting from the experimental studies at Alden Research Laboratory are noted.

( > L, u $ !N N Y ll// N N /Y /N /O D *

an " 's (mini c L : n re . _ _ _ _ _ _ _ _ _ _ _ _ _.q

                                '   it                                      i li                                      ll 11                                      Il
                                                                                                                                      ~

8, 8 Sump Pit I Il Tresh Rock I ,

                                "            b                   Il                       Debris screen L._ Jj n

[L___ 4 h,4__ , _ _lu) _ _ L a

2 L' I n) c L : o o G

                                "                                        ,'ll pl                                       11 l1                                       11 8 8 il 1

Sump Pit ll f>8 11 il 11 . 11 o !! r1 r1 l e => - e,==el 4 Y;== = =f = =:Y v Treeh Rock and oesase sereen NUREG-0897, Revision 1 5-9

Table 5.6 Findings for selected vortex suppression devices *

1. Cubic arrangement of standard 1-1/2-inc (38-mm) deep or deeper floor grating (or its equivalent) with a characteristic length, 2 , that is >

3 pipe diameters and with the top of the cube submerged at Yeast 6 inches (15.2 cm) below the mir.imum water level. Noncubic designs with 2 > 3 pipe diameters for the horizontal upper grate and satisfying the d5pth and distances to the minimum water level given for cubic designs are acceptable.

2. Standard 1-1/2-inch (38-mm) or deeper floor grating (or its equivalent) located horizontally over the entire sump and containment floor inside the screens and located between 3 inches (7.6 cm) and 12 inches (30 cm) below the minimum water level.

Tests on these types of vortex suppressors at alden Research Laboratory have demonstrated their capability to reduce air ingestion to zero even under the most adverse conditions simulated.

Design #1: soud Top cow., 7,esh neck S tandard jl end Floor Grating , 'gOebria Screen i _______________f i ll g

                !              '                                                                                  ~

Flo e Grat ng 4-

                '                                 W.0                 !          i   / \i                 p e!     't               j             ll;             l t          W      l

{A s@1LljiQJl

                                                                                                       !Q
                       ~

Tresh Rock W

end Dobrie Screen Solid Top Cove, Tresh Rock 9"
hh rf Minimum i Dobrie reen j' Water Level j'

Standard I t Floor pg4

                                      ,"         , ,         c,. ting
                                                $$$$$1$$

NUREG-0897, Revision 1 5-10

Table 5.7 Screen, trash rack, and cover plate design findings *

1. Minimum plane face screen area should be obtained from Tables 5.3 and 5.4.

2. Minimum height of open screen (debris interceptors) should be 2 feet (0.61 m).

3. Distance from sump side to screens, g s, may be any reasonable value.

4. Screen mesh should be 1/4 inch (6.4 mm) or finer.

5. Trash racks should be vertically oriented 1- to 1-1/2-inch (25- to 38-mm) standard floor grate or equivalent.

6. The distance between the screens and trash racks should be 6 inches (15.2 cm) or less.

7. A solid cover plate should be mounted above the sump and should fully cover the trash rack. The cover plate should be designed to ensure _

the release of air trapped below the plate (a cover plate located below the minimum water level is preferable).

     "These design findings are based on full-scale tests conducted at the Alden Research Laboratory.

l 50u0 COVEN PLATE

                        .          !5il0F l                  j,             _

y!.9 '

                       . ' i..

l! Nk'$ Trash Rack

                                        .,                                        q f       .,,,,@

l i.- )

                                                                            '    ;f
                                    $$$l (min)$j d3[ify:@f'}[.fi, ,
                         ~
                                                . .s          . ,p.-      . f. -     .
                                                                                         ,.g.-
                                            ,,        ..e :":. ...w.M.
                                                                                  ~.-

6" % U),Debria 2k. , %" Mesh Screen (max) '* (max) NUREG-0897, Revision 1 5-11 l

s I

debris (or particulates) that can pass through screens or strainers should be made. Particulate effects on bearing and seal systems should be evaluated. Table 5.8 outlines key considerations requiring evaluation. Evaluation of potential debris effects requires the following information: - (1) Identification of major break locations (per SRP 3.6.2) and jet energy . levels. (2) Types and quantities of insulations employed, and methods of fabrication i and installation (i.e., mechanical attachments). Material characteristics of the insulations utilized are important for determination of transport and head loss characteristics. The primary and secondary system piping, reactor pressure vessel, and major components (PWR steam generators, reactor coolant pumps, pressurizer, tanks, etc.) that can become targets of expanding jet (s) identified under Item (1) are of importance in assessing debris generation. For BWRs, the feedwater and recirculation piping and the steamlines are of . importance in assessing potential debris generation. (3) Containment plan and elevation drawings showing high energy line piping runs, system components, and the piping that are sources of insulation debris should be reviewed. Structures and system equipment that become obstructions to debris transport, and sump location (s) are important. Orawings showing PWR sump design and debris screen details are needed; for BWRs, downcomer inlet design (from drywell to wetwell), RHR suction inlet and debris strainer design details are needed. (4) Expected containment water levels and recirculation velocities during the I post-LOCA recirculation period are needed to assess debris transport and NPSH effects (see Appendix 0). i NUREG-0897, Revision 1 5-12

Table 5.8 Debris assessment considerations

CONSIDERATION EVALUATE (1) Debris generator
Major Pipe Breaks and Location (pipe breaks and location
Pipe Whip and Pipe Impact as identified in SRP
Break Jet Expansion Envelope (the ~

Section 3.6.2) (major debris generators) (2) Expanding jets

Jet Expansion Envelope
Piping and Plant Components Targeted (i.e., steam generators)
Jet Forces on Insulation
Insulation That Can Be Destroyed or Dislodged by Blowdown Jets.
Sump and Suction Structures (i.e.,

screens), Survivability Under Jet Loading (3) Short-term debris transport

Jet / Equipment Interaction (by blowdown jet forces)
Jet / Crane Wall Interaction
Sump Location Relative to Expanding Break Jet (4) Long-term debris transport
Containment Layout and Sump (or (transport to the sump during Suction) Locations the recirculation phase)
Debris Physical Characteristics
Recirculation Velocity -
Debris Transport Velocity (5) Screen (or suction intake)
Screen (or suction strainer) Area blockage effects (impairment of
Water Level Under Post-LOCA Conditions flow and/or NPSH margin)
Recirculation Flow Requirements
Head Loss Across Blocked Screen or Suction Intakes P

Key elements for assessment

Estimated Amount and Type of Debris of debris effects That Can Reach Sump
Predicted Screen (or Suction) Blockage
AP Across Blocked Screens or Suction Intakes
NPSH Required vs. NPSH Available

     *Per debris estimation methods described in Section 3.3                                                        .

Y NUREG-0897, Revision 1 5-13

Generic findings regarding debris that might be generated, transported, and lodged against sump screens (and the plant-specific dependence of these phenomena) are discussed in Section 3.3. The following paragraphs summarize the findings. Break locations, type and size of breaks, and break jet targets are major factors to con, sider in the estimation of potential quantities of debris generated. The break jet is a high-energy, two phase expansion that is capable of shredding insulation and insulation coverings into small pieces or fibers by producing high-impingement pressures and large jet loads. If the PWR sump location can be directly targeted by an expanding break jet, a close examination should be made of possible jet load damage to such insulations at that location and their possible prompt transport to the sump; jet loads on sump screens, etc., also should be evaluated. Low-density insulations, such as calcium silicate and Unibestos, that have closed cell structures can float. Thus, they are unlikely to impede flow , through screens if water levels are above screen height. Partially submerged screens should, however, be evaluated for pulldown of floating debris Uzuner, July 1977). Low-density hygroscopic insulations that, upon being wetted, have submerged densities greater than water require plant-specific determinations of screen (or strainers) blockage effects. Fibrous insulations (such as mineral wool and fiberglass materials) that are transported at low velocities have been shown to present the possibility for total screen blockages (NUREG/CR-2982). Even if these materials are deposited onto screens in layers of relatively small thickness (on the order of an inch or less), high pressure drops can result. The potential for NUREG-0897, Revision 1 5-14

screen blockage can be calculated using the methods provided in Sections 3.3.5, 3.3.6, and 3.3.7. The methods for debris assessment noted above should also be reviewed in light of the appendices E and F. Appendix E provides information received - from Diamond Power Company about HDR test results on MIRROR insulation performance during LOCA conditions. Appendix F provides information received from Owens-Corning about HDR blowdown tests with NUKON insulation blankets. The NRC staff response to above mentioned information is included in Appendix A. 5.4 Pump Performance Under Adverse Conditions The pump industry historically has determined NPSH requirements for pumps on the basis of a percentage of degradation in performance. The percentage is arbitrary, but generally is 1% or 3%. A 2% limit on allcwed air ingestion was selected in this review because data shcw that air ingestion levels exceeding 4 2% have the potential to produce significant hea'd degradation. Either the 2% - limit in air ingestion or the NPSH requirement to limit cavitation may be used independently when the two effects act independently. However, air ingestion levels less than 2% will affect NPSH requirements. In determining these combined effects, the effects of air ingestion on NPSH required must be taken into account. A calculational method for assessing pump inlet conditions is shown in Figure 5.2. For a given sump design, the following procedure can be followed: (1) Determine the static water pressure at the sump suction pipe after debris blockage effects have been evaluated (see Section 5.3.). For PWRs, the water level in the sump should not be so low that a ' limiting critical water depth occurs at the sump edge in a way that flow is restricted into the sump. NUREG-0897, Revision 1 5-15

SUMP GE0 METRY 8 LOCKAGE, FLOW RATE I SU;4P WATER LEVEL , Ps c DETERMINE AIR INGESTION a,- % . CALCULATE PIPING LOSSES , P, . SUMP TO PUMP ELEVATION CHANGE,Ph AIR . INGESTION ? YES & a3> 0 CALCULATE ~ NO PUMP INLET V STATIC PRESSURE 4 Pp j CORRECT a CALCULATE FOR DENSITY CHANGE, a p AV LABLE 4 CALCULATE S,NPSHR NPSHR FROM COM PAR E PUMP CURVE t NPSH AVAll l AND NPSHR l Figure 5.2 Flow chart for calculation of pump inlet conditions NUREG-0897, Revision 1 5-16

(2) Assess the potential level of air ingestion (see Table 5.2) using the criteria in Section 5.2. (3) Determine pressure losses between suction pipe inlet and pump inlet flange for the required RHR and CSS flows. If the pump inlet is located , less than 14 pipe diameters from the suction pipe inlet, the effect of sump-induced swirl should be evaluated (see Section 3.4). (4) Calculate the static pressure at the pump inlet flange. Static pressure is equal to containment atmospheric pressure plus the hydrostatic pressure due to pump elevation relative to sump or suppression pool surface level, less pressure losses and the dynamic pressure due to velocity. Note that no credit is allowed for containment overpressure, per SRP Section 6.2.2. (5) Calculate the air density at the pump inlet, then calculate the air-volume flow rate at the pump inlet, incorporating the density difference from suction pipe to the pump. , (6) If the calculated air ingestion is found to be less than, or equal to 2%, proceed to Step 7. If the calculated air ingestion is greater than 2%, reassess the sump design and operation per Section 5.1. (7) Calculate the NPSH available. , (8) If air ingestion is indicated, correct the NPSH requirement from the manufacturer's pump curves by the following relationship: HPSH required (air / water) = NPSHrequired(water) x NUREG-0897, Revision 1 5-17

e t

where

 ,.                  p = 1 + 0.5 a p
              .and a p

is the air ingestion rate (in percent by volume) at the pump _ inlet flange. (9) If the NPSH available from Step 7 is greater than the NPSH requirement from Step 8, inlet considerations will be satisfied. If the above review procedure leads to the conclusion that an inadequate NPSH margin exists, further plant-specific discussions must be undertaken with the applicant / licensee for resolution of differences, uncertainties in The lack of credit for containment calculations, plant layout details, etc. overpressure should be recognized as a conservatism that should be assessed on a plant-specific basis. f In addition, an evaluation of small particulate (or debris) ingestion should ; be made to assess pump bearing and seal design effects. Small particulates (which can pass throdch PWR screens or BWR suction strainers) should be assessed for adverse impacts on pump operation and pump bearings.

5. 5 Combined Effects The findings summarized in Sections 5.2, 5.3, and 5.4 can be combined as shown in. Figure 5.3 for determination of adequate sump performance. This sequence.is straight forward; it begins with assessing air ingestion potential, followed by assessing debris blockage effects on NPSH margins, and concluding with pump performance under post-LOCA conditions.

To facilitate first round, or scoping evaluations, the following guidance is provided: s NUREG-0897, Revision 1 5-18

ECCS SUMP DESl2N SUMP oES404

Geometry D Leesteen Redesign 1
b A*'* "*9"8'*8 T Redesegn
Men 6 mum Water Leves
Poet LOCA Cenetiene
Ineviettente) Used I

oESA:s CoNSeoERATioMS MYDRAULIC CONSiOERATIONS

Types Quentteles & Leest6en
Air ingesteen, e of inevietsental Empeeved
Sump outest Cendt6ene.
Estemeted Veaume end Type P T,U, etc. of Detr6e
Menemum _1 _ ,_ _; Level -

Racersuest6en Flow Petteme .
W Hydrevas > and Vessest6es in Centeenment
Typse 4 yeaume of Detrte Transported
Sereen Steenage
Esthneted Head Lees Aereen ym b _

e'**n*d i"**' 'ON8I l e,>2%7 - 1 r 1 P i b CORRECT DErtC3ENCY

Vortee Suppressere - gg 3 ,,

                     .-M             -

Other Dete Hydrousee Yee _

                                                       **#                  ~

CORRECT NPSMR DEFICIENCY I '

Reeeeeme Restreuteteen
Medfy intercepter Ne
Repense ProMem insutottendel db t

vee AMS N. _ y . a i P 1 P PUMP PERPORMANCE Leeste Pump and

NPSMR DETERMINE NPSH PARAMETERS h h
Cowitation and Porticudete
Minimum Weter Loved
Se,e te Res. Leense M me P i- -
Suses and PipenS Leeses
Air aneset6en ENeste
Pusne enest CondN6ene.

P,T,a,$,U, see, %

Centeenment Cenenten.

DerimtT10#es

                                                                  < p                                                           NPSM . Not Poestive Suetion Head lePSHA . NPSM Aveelebie NPSMR MPSM Reevered le                                                            p . yeeg praetten i 7, by Vetumel
                 '-                               '-                                N-
                          ,8;,,,             =              AN P-t>

lep > e 7 Temperature U Vessesty in Pipe i i Note: Celeutete

                               @        $ .1.0+0.h P                                   -
                                                                                        ~

NPSHA NPSHR ehoved essount for line leases between evens and the pumpe. I P i.

                                                ;                          No                      j"^                        vee senseessiery3 ase n
                                                                                                        . ,,PS R.

Fihe 553 Combined technical Considerations for sump performance NUREG-0897, Revision 1 5-19

(1) Air Ingestion Potential (a) If submergence > 10 feet, intake velocity < 4 ft/sec, and Froude number < 0.25, a = 0 (see Table 5.1). (b) If a (see Table 5.2) > 2%, vortex suppressors should be considered _ to reduce a to 0 (see Table 5.6). (2) Debris Blockage Potential (a) If recirculation flow velocities are low (< 0.15 ft/sec), transport of any debris is highly unlikely (see Table 5.9 for a scoping assessment). (b) When considerable quantities of fibrous (i.e., fiberglass) insulation are employed, the significance of potential blockage can be quickly scoped by assuming material within the 7 L/D cone envelope (see Figure 3.25) is totally destroyed and that debris volume is transported to the debris screen. Because fibrous debris , blockage head losses (see Section 3.3.5) are a power function such as bc Equation (1) aHB=aUg which can be rewritten as AH B = a (Q/A)D(V/A)c Equation (2) where AB = head loss across blocked screen - Q = recirculation flow rate A = effective (wetted) screen area V = volume of fibrous debris transported to debris screen and distributed uniformly thereon NUREG-0897, Revision 1 5-20

p i I l Criteria for "Zero" Potential for Screen Blockage Criteria Criteria criteria ') 2 3 1 i Vb f 0 0 >0 Vrm 0 >0 any value Vee any value any value any value i Vhg 0 0 0 Uf any value 10. 2 f t/sec 1 0.15 ft/sec Hw 1 Hs 1 Hs 1 Hs i Vf b = volume of fibrous insulation employed Vrm = volume of reflective metallic insulation employed

Vcc = volume of closed cell insulation with a specific j gravity less than 1.0 (for Hw 1 Hs) this l insulation will float on water surface above the i sump.

! vhg = volume of hygroscopic insulation employed l Hw = water level at sump screen I Hs = sump screen height i Ug = flow velocity at the screen based upon the smaller of (1) the screen area that is shielded from prompt transport of insulation and below the minimum water level or (2) the smallest immediate, total approach-flow-area to the screens / grates below the minimum l water level. I 1 i Table 5.9 First round assessment of screen blockage potential i

NUREG-0897, Revision 1 5-21 I

Therefore, a quick assessment of the head loss across blocked screen area can be made and compared with the NPSHR. Figures 5.4 and 5.5 provide plots of transported debris volumes versus blockage head loss for high density and low density fiberglass debris and are based on experimentally derived head loss data for specific , materials (see Section 3.3.6). Material density dependence is illustrated by these figures, and necessitates obtaining similar correlation for other materials used. Thus, if a prior assumption is made that total transport occurs and the blocked screen calculated head loss is within NPSH margins, the most conservative calculation has been made. If unacceptable screen blockage losses are calculated, more extensive evaluations, such as outlined in Figure 5.6, will be necessary. (c) Reflective metallic insulation debris and associated blockage , effects should be evaluated on a plant-specific basis utilizing the debris considerations and findings discussed in Sections 3.3.4 and 3.3.5. (d) Combinations of insulations are more difficult to assess (see Section 3.3.7) and require estimating combined blockage effects. NUREG-0897, Revision 1 5-22

                                                                                                                                                                             . = :. :;-- .:. . .

High Density Fiberglas where: AH = 1653(OlA)*(V/A)* l Screen Area 3 100 - 200 ft2 p**# ~

                                                                                          /                                                                         ,
                                                                                  / ****
                                                                              /                                                                        e ===***
                                                                                                                                                                      ,100 ft 2 3*              10        -

E - 3 o f *** o p .*** 8 - **** s* .-=====~ 1 - 50 ft2 o

                                          =                     -                             -                                                               -

C = *""" >

9 =

e """"

u. -

1 /p t: ***# o 1 --

                                                                              / ***#                                                                                                            ..

g H -

                                                                -                              - - - Q = 0.000 g pm Q = 8,000 gpm j
                                                                -                              - - -- === Q = 10,000 g p m
                                                                -             I                I            I                                      I              I 0.1 1               2             3                                      4              5 Blocked Screen Head Loss, 2ft H0 3

Figure 5.4 Debris volume versus debris screen area, recirculation flow rate and blocked screen head loss, for high density fiberglass NUREG-0897, Revision 1 5 -23

Low Density Fiberglas where: H = 68.3(0/A)1 "(V/A)' 07 Debris Screen Area 2 1000 -

                                                                                                                         .**                                    r 200 ft
                                                                                                                                                                           ~
                                                                          #                                                                 ,,==='#*
                                                  -                                                      ** an**
                                                                      /*.***an                                                                ,,. -===~" '
                                                               /                                                          ,,                                     100 ft 2
                                   "x (100
                                                                           ,,,                                                                  pd
                                      .5!        -
                                                                                                                                 ,,no ****
                                      $          -                                                              .wa

c -

a S _ f .n**',,,,s*.*** .,<=== ' E - / j m -

                                                               /                                                 ,,  / .=                                        50 ft 2

a l10

                                                                                                                        #f n          -
                                                                                                        ,,** p
                                                                  /
                                                              /                                              --            Q = 6,000 gpm Q = 8,000 gpm
                                                                                               - -- - O = 10,000 gpm j_               l                               l                         I                  l                i 1                              2                          3                 4                5 Blocked Screen Head Loss, 2ft HO 2

Figure 5.5 Debris volume versus debris screen area, recirculation , flow rate and blocked screen head loss, for los density fiberglass NUREG-0897, Revision 1 5-24

(t)lSMAuCCATen.SAMO0.,E, STAT.ON.i (,) 7 (Si 4 ~ N (.) IP-E we P (Pwll IP-E - ACT m>I 147 -===T (al l (.i + l OtTtmuiset CONTAmutNT l VOLUWE INTtmCEPTED SY AT DETt9MsNE WOL OF FISROUS I'I M BeSULATIOes maae0Ves AS DETEmumt AT WOLutet DEf tswat AT WOLuut Stoutest SMAEODED Pt040US DESA4. SE0utNT OUT TO CONE Ante FROM CONE ARIS DISTANCE OF Vyw. Yp,. (SUSSCmsPTS DISTANCE OF F L/De T L/D* TO 0.8 sol. KtFER TO F0mMAft0N tee CMA MeS M S). (e) " 88 INSULATpDM EpeCAPSULATED ,ggyg , 04 MoosENCAPSULATED FISAOUST

                                                               ,g,                                                                                      DETEmutost AREAS OF AS FAgatCATED
                                                                                                                                                      eNSULATC88 DISLODDED SY A7.

I (18) A,

ARE A 0F Pem008..

DETEntmset YOLUME FmACTIDest A.* AAE A 0F mtFLECTfvE hetTALLIC DETimessesE VOLUut OF FOROUS OF SMatDDED FEROUS DESSIS WeSULATIO*f Ret #0Vic SY AT. PmotePTLY TmANSPORTED TO SUMP, voLuut,

        #Pw.*PL *A (1t)

CALCULATE WAIltsued FLOW YtLOCITY P0m Flow PATHS wrTMese CDetTAsteMENT Dustmo MCamCULATION WCDE. DOES MAzeh4Ute Flow VELOCITY EDUAL Ost i EXCEED DESmit inANSpom? YELOCITY - REOUIREEJ (tSS WOLUME OF GMREDOED Pen 0US.; - D 'I DE9meS' AT SUMP IS (,g) ASSUteE DES 8HS SEC0taES alt 08stD Yp ,eY,+Y,=V p NO VtmTpCALLY ON SUter SCMEN 70 AS-FASAsCATED DESM 68 0088 wieGeff 0F AS-F ASA8C Af tD te0T us0AATE TO sue ** WA seteWee DesEsosaces. M, (14) (31)

'I

WOLunet OF SHREDOED Fe#0US AREA NOT SLOCKED SY AS-DESAIS AT SunsP W FAtasCATED speSULATCM is 68 A,tet, Ap Ost (A

A.3/M, LESS ep,Vp,+ sp,Vpge s,Y3 eV A. A,. A- A m 0R AMA,* A.) TNAss TME SUuP Pimeutf ts.. PT -

1' (19) I' NO CALCULATED TM8CNNESS OF SMMODEO Pe#0US 0f 044 AT (22) SuwP. t

vp A , CALCULATE HEAD LOSS THAOUGH UNSLDCNED SUher AREA FOR AREA NOT SLOCnED SY AS*

Of tast TMcCNNESS. 4 FASalC ATED sesSULAfices e A w, (23) h DESRIS ANALYS13 MPUT TO SUup 0898086 (Sit F800mg S.4) Vpw = VOLutet CF SMM00tO Fe#000 peSULA7000 REMOVto SY P1Pt wM8P. (FTI) v p, = VOLUtst OF SumtD0to Fe#008 foetutAT?Ow Mhs0VfD SY MPE ItsPACT. (FT8 ) v p = VOLutet 0F SHetCD80 Fen 0US peSUL ATON Muovto SY AT hePteeOthetNT.(FT3 ) opw* FmACTICN OF v0Luut OF SMREDOED 1864ULA?sfN CAUSED SY PIPE wMap PROWPTLY TRANSP0ATED TO Suur. opg

FmACTsCas OF WOLuut OF Swnf 00E0 eseSULAflom CAUSED SY Pipt mePACT pmotePTLY TRANSpo#TED TO SuneP.

ej = PRACTIOes OF W0tUus OF SenE00tO SeSULAfsose CAUSED SY AT mspiseetutNT PRowPTLY TmaseSPORTED TO SUWP. L/D = RAfl0 0F AT LaseCTH TO PPE DIAtsETER. Y = TOTAL VOLutet OF SHMDOED OtSas4 TRANSP0mTED TO SUese SCMtst(PT8) A9 = A#tA 0F AS-FAS40CAft0 Fe#0US pe8ULAT60ef DeLOOGED SY A7. (FT3 ) Ag = AmtA 0F AS-FASAsCAf t0 REFLtCTIVE ptT ALLIC SeSULATION Dl8LODQ4 0 SY AT (FTI ) A = AAEA 0F SUtop SCatSIL (FTI ) , Ao p. SFFSCTfv5 UssSLOCatD Sune SCAEEN AntA (AmEA AVAILASLR FOR Flow)(FT2) Mg = leAEasute Liest Am Denotesse0es OF AS-FASmeCATED se'eULAff0es (PT) P = Pt$shotTEm 0F SFFECTIVE SUhsP SCREEN (FT) 9

CALCULATED TMsCassESS OF GMutDOED DeSanS 6es.T ON SunsP Scattes (ses)

Figure 5.6 Flow chart for the determination, of insulation debris effects _ NUREG-0897e Revision 1 5-25

6 REFERENCES U.S. Nuclear Regulatory Commission Documents Information supplied during a public comment period, 1983 (available in the _ NRC Public Document Room, 1717 H Street, NW, Washington, DC 20555). Information supplied by insulation companies during a public comment period, 1983 (available in the NRC Public Document Room, 1717 H Street, NW, Washington, DC 20555). NUREG/CR-2403, " Survey of Insulation used in Nuclear Power Plants and the Potential for Debris Generation," R. Reyer, et al., Burns and Roe, Inc., October 1981. NUREG/CR-2403, Supplement 1, " Survey of Insulation Used in Nuclear Power-Plants and the Potential for Debri.. Generation," R. Kolbe, and E. Gahan, Burns and Roe, Inc., May 1982. _ NUREG/CR-2758, "A Parametric Study of Containment Emergency Sumo Performance," G. G. Weigand, et al., July 1982 (also Sandia National Laboratory, SAND-82-0624 and Alden Research Laboratory ARL-46-82). NUREG/CR-2759, "Results of Vertical Outlet Sump Tests," Alden Research Laboratory /Sandia National Laboratory, joint report, September 1982 (also Alden Research Laboratory, ARL-4/-92 and Sandia National Laboratory, SAND-82-1286). NUREG/CR-2760, " Assessment of Scale Effects on Vortexing, Swirl, and Inlet losses in Large Scale Sump Models," M. Padmanabhan, and G. E. Hecker, Alden Research Laboratory, June 1982.

NUREG-0897, Revision 1 6-1

NUREG/CR-2761, "Results of Vortex Suppressor Tests, Single Outlet Sump Tests, and Miscellaneous Sensitivity Tests," M. Padmanabhan, Alden Research Laboratory, September 1982. t NUREG/CR-2772, " Hydraulic Performance of Pump Suction Inlet for Emergency , Core Cooling Systems in Boiling Water Reactors," M. Padmanabhan, Alden i Research Laboratory, June 1982. NUREG/CR-2791, " Methodology for Evaluation of Insulation Debris Effects," J. J. Wysocki, et al., Burns and Roe, Inc., September 1982. NUREG/CR-2792, "An Assessment of Residual Heat Removal and Con'tainment Spray System Pump Performance Under Air and Debris Ingesting Conditions," P. Kamath, T. Tantillo, and W. Swift, Creare, Inc., September 1982.

NUREG/CR-2913. "Two Phase Jet Loads," G. G. Weigand, S. L. Thompson, D. Tomasko, Sandia National Laboratory, January 1983 (also Sandia National l

Laboratory, SAND-82-1935). , NUREG/CR-2982, Rev. 1, " Buoyancy, Transport, and Head Loss of Fibrous Reactor 2 Insulation," 0. N. Brocard, Alden Research Laboratory, July 1983 (also i Sandia National Laboratory, SAND-82-7205). I NUREG/CR-3170, "The Susceptibility of Fibrous Insulation Pillows to Debris Formation Under Exposure to Energetic Jet Flows," W. W. Durgin, and J. Noreika, Alden Research Laboratory, January 1983 (also Sandia National l Laboratory, SAND-83-7008). NUREG/CR-3394, "Probabilistic Assessment of Recirculation Sump Blockage Due to Loss of Coolant Accident," J. J. Wysocki, Burns and Roe Inc. , July 1983

(also Sandia National Laboratory, SAND-83-7116).

l l i _. j NUREG-0897, Revision 1 6-2 L

NUREG/CR-3616, " Transport and Screen Blockage Characteristics of Reflective Metallic Insulation Materials," 0. N. Brocard, December 1983 (also Alden Research Laboratory, ARL-124-83 and Sandia National Laboratory, SAND-83-7471). Other Documents Brocard, D. N. , " Transport and Head Loss Tests of Owens Corning Nukon" ~ Fiberglass Insulation," Alden Research Laboratory, ARL-110-83/M489F, Ho1 den, MA, September 1983. Ourgin, W. W., and J. F. Noreika, "The Susceptibility of Nukon* Insulation Pill:as to Debris formation Under Exposure to Energetic Jet Flows," Alden Research Laboratory, ARL-111-83/M489F, Holden, MA, September 1983. Ourgin, W. W., M. Padmanabhan, and C. R. Janik, "The Experimental Facility for Containment Sump Reliability Studies," NRC Generic Task A-43, Alden Research Laboratory, ALO-132, Holden, MA,1980. Florjancic, D. , " Influence of Gas and Air Admission on the Behavior of - Single- and Multi-Stage Pumps," Sulzer Research Number 1970, Sulzer - Brothers, Ltd. , of Winterthur, Switzerland,1970. Glasstone, S., " Effects of Nuclear Weapons," Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. ,1981. Hydraulic Institute, Hydraulic Institute Standards for Centrifugal, Rotary and Reciprocatina Pumps, 13th edition, Cleveland, OH, 1975. Imatran Voima Oy, "Model Tests of the Loviisa Emergency Cooling System," Report No. 275, August 1976, Civil Engineering Department, Construction Laboratory, Finland.

NUREG-0897, Revision 1 6-3

Imatran Voima Oy, "Model Tests of Containment Sumps of the Emergency Core Cooling System," Report No. 291, April 1980, Civil Engineering Department, Construction Laboratory, Finland. Merry, H., " Effects of Two-Phase Liquid / Gas Flow on the Performance of , Centrifugal Pumps," Institution of Mechanical Engineers Conference on Pumps and Compressors, Paper No. C130/76, 1976, 1, Birdcage Walk, London, S.W. 1H 9JJ U.K. Moody, F. J., " Fluid Reaction and Impingement Loads, Specialty Conference on Structural Design of Nuclear Plant Facilities," Vol 1, Chicago, IL, December 1973, "American Society of Civil Engineers, 345 East 47th Stre'et, New York, N.Y. 10017. Murakami, M. and K. Minemura, " Flow of Air Bubbles in Centrifugal Impellers and its Effect on Pump Performance," presented at sixth Australian Hydraulics and Fluid Mechanics National Conference, Publication No. 77/12, December 1977, Institution of Engineers, Adelaide, Australia. Niyogi. K. K. and R. Lunt, " Corrosion of Aluminum and Zinc in Containment Following a LOCA and Potential for Precipitation of Corrosion Products in ! the Sump," United Engineers and Constructors, Inc., Philadelphia, PA, September 1981. Uzuner, M. S., " Stability Analysis of Floating and Submerged Ice Floes," Proc. ASCE, 103, HY7, Journal Hyd. Div., pp. 713-722, July 1977. i i I - i i NUREG-0897, Revision 1 6-4

.i 4 a APPENDIX A i SUHMARY OF PUBLIC COMENTS RECEIVED AND ACTIONS T5 KEN (REF. USI A-43) i i i i

r i

]

i i .: i j .

i ~~ ~ j NUREG-0897, Revision 1 - i i.

1 APPENDIX A !

SUMMARY

OF PUBLIC COMENTS RECEIVED AND ACTIONS TAKEN 1 INTRODUCTION , The technical findings related to Unresolved Safety Issue (USI) A-43 were published for comment in May 1983. Notice of the publication was placed in the Federal Reafster on May 9, 1983. The official comment period lasted for 60 days and ended on July 11, 1983. However, comments were received into September 1983, with followup comments received into November 1983. A listing of those who responded during the period and afterwards is shown in Table 1. Copies of the comment letters are on file in the NRC Public Document Room, 1717 H Street, NW, Washington, DC. A public meeting was held on June 1 and 2, 1983, at Bethesda, Maryland, to offer additional opportunity for public comments; however, attendance was very small. Followup discussions were held with.respondees to clarify issues , raised at this meeting and in the written comments. ,

 , An overview of the comments received is provided in Sectfon 2 below.

Section 3 contains summaries of significant comments and the actions planned to resolve them. 2 OVERVIEW OF COMENTS RECEIVED The major written comments received addressed seven specific subject areas. The comment categories and commentors are listed in Table 2 below. The commentors are identified in Table 2 as follows: Alden Research Laboratory (ARL); Atomic Industrial Forum (AIF); BWR Owners Group (BWR); Comeonwealth Edison (CEd); Consumers Power Co. (CPC); Creare Research and Development (CRO); Diamond Power Co. (DPC); General Electric (GE); Gibbs and Hill, Inc. (GH); Northeast Utilities (NE); and Owens-Corning Fiberglass, , Inc. (OCF). MUREG-0097, Nevision 1 A-1

Table 1 Persons who commented on the technical findings related to USI A-43* Alden Research Laboratory (ARL), M. Padmanabhan, letter to A. Serkiz (NRC),

 " Comments on NUREG-0897 and 0869," June 13, 1983.

ARL, M. Padmanabhan, letter to A. Serkiz (NRC), " Revision to Table A-3 in NUREG-0869," June 22, 1983. Atomic Industrial Forum, R. Szalay, letter to the Secretary of the

Commission, "NRC's Proposed Resolution of Unresolved Safety Issue A-43, Containment Emergency Sump Performance, Contained in NUREG-0869," July 22, 1983. - Atomic Industrial Forum, J. Cook, letter to R. Purple (NRC) and enclosure

 " Examples of Staff Review Going Beyond Approved Regulatory Criteria,"

June 4, 1984. BWR Owners Group, T. J. Dente, letter to T. P. Speis (NRC), "BWR Owners' Group Comments on Proposed Revision to Regulatory Guide 1.82, Rev. 1," October 18, 1983. BWR Owners Group, D. R. Helwig, letter to V. Stello (NRC), BWR Owners' Group comments on Regulatory Guide 1.82, Revision 1, July 16, 1984. Commonwealth Edison, D. L. Farrar, letter to the Secretary of the Commission, "NUREG-0897, Containment Emergency Sump Performance; Standard Review Plan Section 6.2.2, Rev. 4, Containment Emergency Heat Removal Systems; and NUREG-0869, USI A-43 Resolution Positions (48FR2089; May 9, 1983)," July 13, . 1983. Consumers Power, D. M. Budzik, letter to the Secretary of the Commission,

 " Comments Concerning Regulatory Guide 1.82, Proposed Revision 1 (File 0485.1, 0911.1.5, Serial: 23206)," July 15, 1983.

Creare, W. L. Swift, letter to P. Strom (SNL), " Comments on Figure 3-6 of NUREG-0897 and Table A-9 of NUREG-0869," June 13, 1983. Diamond Power Company, R. 6. Ziegler and B. D. Ziels, letter to K. Kniel (NRC),

 " Containment Emergency Sump Performance, USI A-43," July 11,1983.

Diamond Power Specialty Company, B. D. Ziels, letter to A. Serkiz (NRC),

 " HOR Test Result Summary, MIRROR Insulation Performance Ouring LOCA Conditions," December 6, 1984.

General Electric (GE), J. F. Quirk, letter to K. Kniel (NRC), " Comments on Emergency Sump Documents," July 11, 1983.

Including comments on NUREG-0869, NUREG-0897, proposed Revision 1 to Regulatory Guide 1.82, and proposed Revision 4 to Section 6.2.2 of the Standard Review Plan (SRP, NUREG-0800).

NUREG-0897. Revision 1 A-2

Table 1 (Continued) GE, J. F. Quirk, letter to T. P. Speis (NRC), " Comments on Proposed Regulatory Guide 1.82, Rev. 1," October 17, 1983. Gibbs and Hill, Inc., M. A. Vivirito, letter to the Secretary of the Commission, " Comments on Proposed Revision No. 1 to RG 1.82," July 11, 1983. Northeast Utilities, W. G. Counsil, letter to K. Kniel (NRC), "Haddam Neck, . Millstone Nuclear Power Station, Unit Nos. 1, 2, and 3 Comments on NUREG-0897, SRP Section 6.2.2 and NUREG-0869," September 2, 1983. Owens Corning Fiberglass (OCF), G. H. Hart, letter to A. Serkiz (NRC),

     " Comments on NUREG-0897 and NUREG-0869," June 23, 1983.

OCF, G. H. Hart, letter to A. Serkiz (NRC), " Updated Comments on NUREG-0897 and NUREG-0869," July 14, 1983. OCF, G. P. Pinsky, letter to K. Kniel (NRC), " Comments on NUREG-0879 and -0896," July 14, 1983. OCF, G. H. Hart, transmittal to A. Serkiz (NRC), " HOR 81owdown Tests with NUKON Insulation Blankets," February 18, 1985. Power Component Systems, Inc., D. A. Leach, letter to A. Serkiz (NRC),

     " Nuclear Grade 81anket Insulation," November 8, 1984.

Table 2 Categories addressed in major written comments Comment Category ARL AIF BWR CEO CPC CR0 DPC GE GH NE OCF (1) Survey of insulation used is X X not current or complete. (2) Cost estimates are low. X X (3) Estimates of sump blockage X X X X probabilities are high. (4) Value-impact analysis questioned. X X X X (5) BWRs should be exempt; A-43 is a X X X PWR issue. (6) Insulation material definitions and X X descriptions need revision for clarity and completeness. (7) Technical comments on and X X X X X

X X X clarifications of subject matter in NUREG-0897 and NUREG-0869. NUREG-0897, Revision 1 A-3

By category, the actions taken in response to these comments are as follows: Categories 1 and 6: Tables have been added to NUREG-0897, Revision 1 and NUREG-0869, Revision 1 to include the additional plant insulation information provided during the public comment period. The text of the NUREGs has been , i revised to reflect recommended insulation definitions and the need to evaluate the specific insulation employed. Categories 2 and 4: The cost estimates provided by different industry groups have varied over a wide range. With the exception of Diamond Power Company, respondees claimed that the cost estimates in value/ impact analysis were too low. The revised value/ impact analysis reflects an averaged value derived from costs provided. Category 3: A detailed sump blockage probability analysis has been performed and is reported in NUREG/CR-3394. The results were used in the revised value/ impact analysis. These results show a sump blockage probability range for pressurized water reactors (PWRs) of 10 8 to 5 x 10 5 Rx yr and a strong , dependence on plant design. I Category 5: NUREG-0869 and Regulatory Guide 1.82 have been revised to , specifically identify areas of concern for boiling water reactors (BWRs) and , for PWRs. Category 7: Technical corrections and clarifications have been made in the appropriate sections of NUREG-0897 and NUREG-0869. i The NRC staff greatly appreciates the review and comments provided by the respondees. The time and effort they have taken to review USI A-43 has resulted in an improved report that will reflect current findings and a l balanced position with respect to this safety issue. NUREG-0897, Revision 1 A-4

  .   .                                                                                       i l

3 COMMENTS RECEIVED AND PROPOSED ACTION (OR RESPONSE) ACTIONS The NRC staff has given complete and careful consideration to all comments received on USI A-43. Summaries of significant comments and the actions taken by the NRC staff in response are provided in Table 3. Comments are , presented in alphabetical order, based on the name of the commenting institution. i } l 4 i i 1 ) l

I i s . NUREG-0897, Revision 1 A-5

i se Table 3 Comments received on USI A-43 and NRC staff response si i 2 ! SE Comment NRC Staff Response l 23 af Alden Research Laboratory

-~

ARL noted typographical errors and proposed These corrections and clarifications have been SL l 9 technical clarification to several tables incorporated into NUREG-0897 and NUREG-0869. l l Atomic Industrial Forum l The cost impact of $550,000/ plant used in Costs impacts were re-evaluated based on cost estimate value/ impact analysis is low by at least information received from AIF and other respondees a factor of 2. Economic considerations related to reduced The essence of a value/ impact analysis is that it 2 probability of plant damage should be excluded attempts to identify, organize, relate, and make as from the cost-benefit balancing. Decisions " visible" all the significant elements of value expected should be based primarily on the value/ impact to be derived from a proposed regulatory action as well ratio. as all significant elements of impact. The net values are compared with the net impacts. Thus if a proposed safety improvement is accompanied by an adverse side effect, the impairment is subtracted from the improvement to arrive at a net safety value for consideration in the

            '                                                             value/ impact assessment.

1 Similarly, when the immediate and prospective cost impacts I are summed, they should include all elements of economic l impact on licensees, such as costs to design, plan,

 ,                                                                        install, test, operate, maintain, etc.          Plant downtime or decreased plant availability is included when applicable.

The summed impacts, however, should be net impacts, for . comparison with net values. Thus, any reductions in l operating costs, improvements in plant availability, or reductions in the probability of plant damage are properly a factor in determining net adverse economic impact.

  ;                                                                       Future economic costs and savings are appropriately j                                                                      discounted.                                                         .

1 s

I

z Table 3 (Continued) .

E 8 S Comment NRC Staff Response l F Qualitative differences among impact elements are 1 respected, and distinctive elements of impact (of which i averted plant-damage probability, as a favorable rather 8 than adverse impact, is a prominent example) are separately _, identified, for appropriate consideration in regulatory decision making. The ratio of avoided public dose to the gross cost of implementation is ordinarily a major decision factor. However, it is not by itself always a good guide to a sound regulatory decision. The issues involved are often too l complex for a decision on this criterion alone. Other l 3 factors that enter, often in important ways, may include 4 any economic benefits that reduce a net adverse economic impact, the safety importance of the issue, and values and l impacts that cannot or cannot readily be quantified; for example, jeopardy to a defense layer in the defense-in-depth concept or expected reductions in plant availability that can be foreseen but not precisely estimated. 8 A sound regulatory decision rests on adequate consideration l . of all significant factors. An overly simple approach can l mislead if it simplifies away complexities that are the essence of the issue at hand. The assumption that sump failure will occur in A detailed sump blockage probability analysis has been 50% cases of the large LOCAs should be performed and is reported in NUREG/CR-3394. The results l justified. show a wide range of sump blockage failure probabilities . (i.e., 3E-6 to SE-5/ reactor year) and a high dependency on , plant design and operational requirements. These results I are reflected in a revised value impact analysis utilizing l a range of sump failure probabilities. ! The use of PWR release categories from The containment failure pro ibilities and release i WASH-1400 is too conservative. Containment categories used in the reg .satory analysis for USI A-43 failure probabilities used in WASH-1400 were based on information presented in WASH-1400, and ( e

l se Table 3 (Continued) E 8 oS Comment NRC Staff Response SE are inadequate to describe the nuclear also on other considerations. The comments presented by i industry's present knowledge in this field. an AIF subcommittee regarding the validity of continued Il Releases due to " vessel steam explosion" use of WASH-1400 assumptions, etc. are being evaluated 8 are unrealistic and should not be considered. through other activities such as: reevaluation of source terms, SASA studies, etc. USI A-43 regulatory analyses were based on the following considerations and for the reasons noted: (1) WASH-1400 assumptions were utilized to provide a common baseline calculations for reference plants and were used to estimate increases in releases due to a postulated loss of recirculation flow capacity. Until 2 revised failure mechanisms and new source terms are da determined, this approach provides a consistent set of calculations. (2) Although using a small containment failure probability associated with steam explosion would be more appropriate, release category PWR-1 (which includes steam explosion) was not a dominant

        '                                                               contributor to release. Release categories PWR-2, -4, and -6 were the dominant pathways contributing to increases releases due to a failed sump for the plants         ,

analyzed. l I (3) Basing release effects on the assumption of simultaneous failure of core cooling and loss of containment sprays is conservative. If containment . were not lost (as would be the situation for PWRs that have dry containments with safety grade fan cooler systems), the LOCA energy could be disspated without containment overpressurization and failure. Thus releases associated with PWR-2 and -4 categories could be discounted and PWR-6 releases only used. Such . considerations have been incorporated into this revised regulatory analysis.

se Table 3 (Continued) . E 2 8 Comment NRC Staff Response af (4) Other factors--such as containment structural design

i margins that argue against gross containment TL failures (as postulated in WASH-1400), realignment 8 to alternate water sources, controlled venting for

_. BWRs, etc.--have also been considered this revised regulatory analysis. The use of the CRAC Code and a "no-evacuation," The 50-mile radius reflects a substantial part (though 50-mile-radius model to develop public doses not all) of the total population dose, and is thus a is unrealistic. reasonable index of the radiological effect on the public. Standardization of calculations to that radius is helpful in comparing risks associated with different issues and l* average such risks for use with the 1000 person-rem /$M criteria. Evacuation of people is not considered because calculations suggest that, although it may sometimes be important for people directly affected, the effect of evacuation on the total population dose is likely to be small. 8 NRC should utilize information developed more Possible changes in the source terms are being considered

    . recently (i.e. , NUREG-0772) to reassess and       by the special task force established by the Commission reduce the source terms, rather than continue      to review the source-term issue. Changes would be to use the PWR-2 and PWR-3 release categories      premature before this group completes its evaluation from WASH-1400.                                    and the new values are accepted by all partics involved.

NRC'should utilize the " leak before break" Elastic plastic fracture mechanics analysis techniques to concept in evaluating the safety significance analyze pipe break potential has been used in USI A-2, with .

      ,of A-43.                                           a limited number of PWRs being analyzed. For USI A-2, the submittal of such analyses for specific break locations (on a plant-specific basis) will require obtaining an exemption from the requirements of GDC 4. Submittal of such analyses to address the USI A-43 debris blockage issues would be reviewed by staff on a plant-specific basis, should a licensee or applicant elect to utilize this approach.

l .

ge Table 3 (Continued) E - 93 8! Comment NRC Staff Response S af BWR Dwners Group 2. Af ter quick review of the proposed revision to The requirement for long-tera decay heat removal is 8 the regulatory guide, the BWR Owners Group and applicable to light-water reactors, both BWRs and PWRs. _., GE maintain that USI A-43 is not a generic issue for BWRs. The revisions to RG 1.82, which now proposes All types of insulation should be evaluated for the specific criteria for BWRs, should apply potential of debris generation, transport, and suction only to light-water reactors that have any strainer blockage. The wide variation in plant designs potential for harmful debris generation (i.e., and insulation employed does not support a generic light water reactors that extensively use statement. 3 fibrous insulation). c) These comments and any future comments by RG 1.82, Revision 1 (along with NUREG-0897, the BWR Owners Group should not substitute NUREG-0869 and SRP 6.2.2, Revision 4) was issued for the normal notice and comment procedure "for comment" in May 1983. Only 14 responses were that allows potentially affected licensees received as of September 1983. Some of these comments to respond to proposed regulatory guide (in particular GE's July 11, 1983 letter) cited a need changes. to specifically address BWR related concerns in the RG. 8 This was done and copies were sent to GE and the BWR Owners Group. Given the previous extensive distribution of "for comment" reports and regulatory positions and the rather small number of responses, the staff does not plan to reissue RG 1.82, Ravision 1 for comment. The NRC staff will incorporate additional valid technical points received from the BWR Owners Group and GE. . The most recent input from the BWR Owners Group (July 16, 1984) does not provide new significant findings;

                         .                                        rather this input re-expresses concerns previously voiced and stresses possible misinterpretations of wording in RG 1.82, Revision 1.                            .

1

C e Table 3 (Continued) . A 'P o! Comment NRC Staff Response & Commonwealth Edison 5. !L The Commission has not sufficiently justified the A-43 resolution does not mandate retrofits; rather, 9 need to impose retrofit requirements on either applicants are requested to assess long-term _. cperating or near-term operating license units. recirculation capability utilizing RG 1.82, Revision 1 and to then determine what corrective actions may be needed. The use of an information bulletin to the majority of the plants does not constitute imposition of a retrofit. Cost estimates for surveys, design reviews, and The A-43 value/ impact evaluation has been revised retrofitting are questionable. based on detailed sump blockage probability studies 3, (NUREG/CR-3394) and cost estimates received from

  ;                                                            industry responses.

The proposed RG 1.82 is overly conservative. The NRC staff acknowledges that conservatisms exist However, given the need for assurance that the in RG 1.82, Revision 1. However, such conservatisms recirculation sump remains a reliable source are prompted by the limited amount of available information of cooling water, the commentor agrees that an regarding insulation destruction due to high pressure evaluation of sump designs, potential for debris, jets and attendant debris generation, and the wide 8 air ingestion, and adequate net positive suction variability of plant designs and types of insulation head (NPSH) is fully justified. used. The commentor questions the assumption that 50% A detailed sump failure probability analysis was of LOCAs lead to sump loss, the value/ impact ratio performed and is reported in NUREG/CR-3394. The given uncertainties in estimated costs, tte basis " averaged" sump failure probability was 2E-5/ reactor-for' assuming 23 years remaining plant life, etc. year with a range of 3E-6 to SE-5/ reactor year. Consumers Power Regarding the proposed Revision 1 to RG 1.82, the Appendix A of proposed RG 1.82, Revision 1 was commentor stated (1) that Appendix A should be always intended to provide additional information clearly delineated as being an information and and/or guidance, not design requirements. Appendix A guidance source, not as presenting design require- has been clearly labeled as such. rents, and (2) that consistency is needed with respect to NPSH terminology.

me Table 3 (Continued) E 2 8B Comment NRC Staff Response 0 E' Regarding the value/ impact analysis, the commentor That 50% of LOCAs lead to sump blockage has been 5L questioned the assumption that 50% of the loss-of- reevaluated (see NUREG/CR-3394), and the results Il coolant accidents (LOCAs) lead to sump blockage and of that detailed study have been used in revising the 8 cites a sump failure frequency of 2 x 10 4 per A-43 release estimates. _. demand from another probabilistic risk analysis. The commentor questioned the direct application of The calculation of avoided accidents costs, core melt frequency reduction for computing avoided loss-of plant costs, etc., are consistent with current accident cost. The commentor disagrees with taking NRC staff evaluation practices. Recalculation of the credit for loss of plant cost. Rather, the parameters previously used will be carried out with commentor states that loss of plant costs should be the revised blockage frequencies. deducted from avoided accident costs.

 '.       Creare ro The beta factor used to predict a pump's                                     Efforts were made to obtain the original data required NPSH in an air / water mixture is based                             tapes and calculate the data's scatter; however, this on data whose scatter was not reported. The                                  information was not readily available. The suggested NUREG should note this and caution the applicant                            cautionary note has been added to NUREG-0897.

and reviewer to carefully consider the adequacy 8 of the NPSH margin if it is marginal. The use of an arbitrary minimum allowable NPSH NUREG-D897 and RG 1.82, Revision 1 no longer recommend . margin, either as a fixed value (i.e., I foot) a minimum allowable NPSH margin. Instead, they ! or as a percentage value (i.e., 0.5 x margin with note that whatever NPSH margin is available (after no screen blockage), is not justifiable. It should accounting for hydraulic and screen blockage r be 'ecognized r that what constitutes a safe NPSH effects) should be evaluated with respect to each margin is a plant-specific judgment. plant's long-term recirculation requirements. . Diamond Power Company NUREG-0897 resolves a significant safety problem in The NRC staff concurs. a thorough and equitable manner. I

Table 3 (Continued)

E in

 ?               Comment                                                   NRC Staff Response 8

0 The commentor provides recommendations regarding The proposed classifications have been combined with o' the classification of various insulating materials, other similar proposals to revise and clarify the i particularly on the need to distinguish between insulation classification and descriptions used in

 *I. totally encapsulated insulation and jacketed            WREG-0897.

8 insulation. The commentor provides listings of the types of The information has been added to WREG-0897 and " insulations purchased since 1980 and the types WREG-0869, along with Ata received from other of insulations used in recent retrofittings. manufacturers. The commentor states that the costs in the This cost information has been reflected in the value/ impact analysis are in agreement with its revised value/ impact analysis (WREG-0869), along costs and provides the following figures: with other industry cost figuras. < Cost of MIRROR" reflective metallic C insulation = $40/ft2 for material alone. Installation cost, excluding material

                 = $25/ hour.

8 Productivity = 1.24. hours /f t2 of insulation. 9 t t t I -

e Tabla 3 (Continued) E 8 83 Comment NRC Staff Response ?? Reflective metallic insulation is not the Information supplied by Owens-Corning Fiberglass Co. Il predominant type of insulation used in newer and the Diamond Power Co. regarding types of insulation Il plants. Recently insulated plants mainly used in existing and future reactors has been added to E use fiberglass insulation." NUREG-0897 and NUREG-0869. These reports have been

 .                                                             revised to reflect tnis new information. The trend appears to be toward a greater use of fibrous insulations.

A report on HDR test results on MIRROR" This report has been included as Appendix E in insulation performance during LOCA conditions NUREG-0897, Rev. 1. The results of this report was submitted to provide additional infornation do not support a hypothesis which postulates for the existing data base used in resolution free and undamaged inner foils being available of USI A-43.** to transport at low velocities and to cause 2 blockage. However, the limited data base I. precludes developing a detailed debris generation model. f

Letter of July 11, 1983.

       ** Letter of December 6, 1984.

i

g Table 3 (Continued) - 5 m h Comment NRC Staff Response 1 h General Electric Company 7 SRP 6.2.2 and RG 1.82, Revision 1 make no RG 1.82, Revision 1 and SRP 6.2.2 have been modified

 "       distinction between BWRs and PWRs; regulatory           to identify PWR- and BWR-related concerns. Regulatory
 "       criteria should differentiate between various           Guide 1.82, Revision 1, has been renamed " Water Sources plant designs.*                                         for Long-Tern Recirculation Cooiing Following a Loss-of-Coolant Accident."

Reference should be made to technical findings Based on the responses received, the A-43 technical that imply that A-43 concerns do not pose a findings will be revised to reflect (1) that there is serious problem for BWRs.* a more extensive use of fibrous insulations (i.e., NUKON") than previously identified and (2) that BWRs are reinsulating with NUKON". NUREG-0897 will reflect 2 current findings and identify both PWR- and BWR-h related concerns. Th6 value impact analysis utilizes a PWR for GE's point on utilizing a PWR probabili_stic risk the risk assessment and PWR-oriented industry assessment for drawing conclusions for a BWR is impacts and, as such, is not directly applicable acknowledged. Similar assessments have been made to BWRs.* for BWRs and those results have been utilized in i preparing this revised regulatory analysis.

     ' General Electric has reviewed the proposed                The requirement for long-term decay heat removal is revisions and has concluded that the design             applicable to both BWRs and PWRs.      RG 1.82,                               ,

requirements proposed in RG 1.82, Revision 3 Revision 1. Appendix A contains a series of tables (or ; are excessively prescriptive and not generically guidelines) that have been derived from extensive ' app)icable to the BWR.** tests and analytical studies. This information is provided for ease of referral and can, or need not, be used--at the user's option. RG 1.82, Revision 1 is

general, and not prescriptive. The applicant has the

                   ,                                             responsibility for design submittal and justification                         .

of the safety aspects thereof. !

Letter of July 11, 1983.

         ** Letter of October 17, 1983.                                                                                                        l
                                                                                                          .-s                . - . - -

l Table 3 (Continued) E a Comment NRC Staff Response

o

      @ The proposed RG should be revised so that no                  The technical findings in 1983 (versus earlier findings) 7 further requirements are imposed on designs that              are considerably different, particularly with 8

have already included precautions that preclude respect to insulation employed currently and the air ingestion into, or blocking of, suction lines transport characteristics of insulation debris. The

      " used for long-term decay heat removal.*                       air ingestion potential has been experimentally quantified and found to be small. However the 50%

blockage criterion in the current RG 1.82 permitted applicants to essentially bypass the debris blockage question. For those plants where design precautions can be clearly demonstrated, further actions (retrofits) are not necessary. 2 In addition, the proposed RG should be further The licensee and/or applicant always has the option i* revised to provide for alternative means of to propose alternate means to deal with a particular ensuring that long-term heat removal is not lost design or safety problem. as a result of suction blocking or air ingestion." In the SER for GESSAR, the NRC indicated that At the time the SER for GESSAR II was written, A-43 USI A-43 posed no problem for the Mark III concerns relative to BWRs were still under evaluation. containment configuration.* The staff's SER cited several elements of the GESSAR 8 II design that tended to reduce the probability for blockage of the RHR suction inlets due to LOCA-generated debris. The staff concluded that plants referencing the GESSAR II design could proceed, pending resolution of USI A-43, without endangering the health and safety of the public while the staff completed its evaluation of GESSAR.

          .                                                           The unique aspects of each Mark III plant design should be evaluated during plant-specific reviews of A-43 concerns.

Letter of October 17, 1983.

l

_ - - - . - _ . . _= _ - .-- - . _ . - _ - -- .__ - i .

          !!                                                     Table 3 (Continued)                                                      -

55 ,

          'P h
          %J Comment                                                        NRC Staff Response I          ![ The tests performed by Alden Research Laboratory               The comment is partially correct, because BWR RHR               !

(see NUREG/CR-3616) may even be very conservative suction inlets are located at some elevated distance for BWRs since it appears the tests utilized sump above the wetwell or suppression pool floor. However, 3 screens directly on the sump floor.* the insulation debris transport characteristics (see NUREG/CR-2982, Rev. 1) showed that low velocities (i.e., 0.2 - 0.3 ft/sec) can transport fragmented debris and are applicable to both BWRs and PWRs.

The proposed regulatory guide should be revised RG 1.82, Revision 1 states: "This regulatory guide i

to include criteria that will allow alternative has been developed from an extensive experimental and measures for precluding loss of long-term analytical data base. The applicant is free to decay heat removal due to air ingestion or select alternate calculation methods which are based

          ?"       blockage.*                                               on substantiating experiments and/or limiting analytical t3                                                                considerations." Thus, the applicant is free to

select alternate methods or measures for precluding loss of long-term decay heat removal.

i Earlier surveys on the use of insulation in light As stated above, current findings do not support the ' 1 water reactors have concluded that most BWRs utilize earlier surveys or conclusions. NUREG-0897 is being

               . metallic insulation, which minimizes the potential          revised to incorporate findings from public comments           t I for formation and subsequent transport of debris            received (particularly with respect to insulations

i to the sump screens.* currently used and the change to fibrous insulation 4

from previously used reflective metallic insulations). . Recent tests on the transport of thin stainless steel foils show that this material can be transported at low velocities.(i.e., 0.2 to 0.3 ft/sec).

a

                  ,*Letter of October 17, 1983.

a f , i 4 i , a f -- _ - -

t hm 12 Table 3 (Continued) S g Comment NRC Staff Response 1 Gibbs and Hill, Inc. g{ , Section B does not discuss the fact that sump- Appendix A (page 19) has wording very similar to the configurations that differ significantly from commentor's suggested wording, the criteria of Appendix A may be equally acceptable. Gibbs and Hill recommends adding the following concluding paragraph to Section 8:

                                                                              "If the sump design differs significantly from the 9uidelines presented in Appendix A, similar data from full-scale or reduced-scale tests, or in-plant tests can be used to verify adequate sump hydraulic 3,                                                            performance."
                 .L 05                                                            Tables A-1 and A-3 are inconsistent and Table A-2          The inconsistencies have been corrected.

has inconsistencies in water level noted. - Northeast Utilities Tests show that gratings are as effective as Gratings were very effective in reducing air j solid cover plate in suppressing vortices. ingestion to essentially zero.

                                                      .                          The procedure in Appendix B is too prescriptive.         Appendix B in NUREG-0897 presents the staff's The NRC should allow licensees to define and             technical findings for A-43. Appendix B was included develop their own evaluation methods.                   to illustrate major considerations. RG 1.82, Revision 1 is the regulatory document.

I 4 e 9

                                                                                                                                                                                                   .I i                                                        ,  .

r - E Table 3 (Continued) - iE

'?

h Comment NRC Staff Response - h Credit should be given for top screen area if For those plant designs and calculated plant 7 it is deep enough to reduce the potential for conditions where this point could be unconditionally { clogging (RG 1.82, Revision 1, Section C, Item 7). substantiated, credit would be given. The licensee should be free to determine methods Regulatory Guide 1.82, Revision 1 has been revised of inspection and access requirements (RG 1.82, accordingly. Revision 1, Section C, Item 14). RG 1.82, Revision 1 will be used to evaluate sumps The need for backfitting will be based on plant-in operating plants. This may require backfitting specific analyses that will reveal the need for at substantial costs. and the extent of backfitting that might be required. The cost of backfit should be weighed against core

 >'                                                                 melt costs.

Appendix A to RG 1.82, Revision 1 requires obtaining Appendix A states: "If the sump design deviates performance data if sump design deviates significantly from the boundaries noted, similar significantly from the guidelines provided. performance data should be obtained for verification For operating plants, this may result in costly of adequate sump hydraulic performance."

sump testing. 8 NRC estimates for man-rem costs associated The value impact analysis has been revised based on

      ,   with insulation replacement are grossly                   cost data received during "for comment" period.

underestimated. The value impact analysis addresses only PWRs. The value impact analysis revision clearly addresses If the NRC has concluded that this issue only BWR and PWR concerns. applies to PWRs, the document should reflect this. . The commentor concurs with the comments The AIF comments are addressed separately; see above. submitted separately on this document by the AIF. ' I ,

E Table 3 (Continued) E 'P Comment NRC Staff Response

=

Owens-Corning { Detailed comments addressed the wide variation Detailed comments received on insulation types; ~ of insulations employed, descriptions, suggested descriptions,'etc. have been used to revise terminology, etc.* NUREG-0897. Comments recommended including transport and head Data from NUKON" tests have been referenced and loss data for NUKON" fiberglass tests.* major findings summarized in the revised NUREG-0897. The commentor questioned Table B-1, Criterion 2, Transport tests on reflective metallic toils that reflective metallic insulation foil debris revealed that they can be transported at low would not be transported at velocities less than velocities (0.2 - 0.5 ft/sec). p 2.0 ft/sec.* The commentor questioned the concept that if there Inputs received have been used in revising NUREG-0869. is all reflective metallic insulation there is no problem.* The commentor recommended changes to various Inputs received have been used in revising NUREG-0869. 4 tables as discussed at the June 1 and 2, 1983, I public meeting.* The commentor suggested word changes that would Inputs received have been used in revising NUREG-0869. minimize singling out fibrous type insulations as the screen blockage concern without considering blockages due to reflective metallic insulation ' mat'eri al s. *

       ,The commentor addressed the recommended revision         Inputs received have been used in revising NUREG-0869.

to reflect current status of insulations employed in nuclear power plants.*

Letter of June 23, 1983. ,

I ,

EE Table 3 (Continued)

IE

   'P o

0$ Comment NRC Staff Response

o y The potential for screen blockage by reflective A set of experiments to determine transport velocities DT metallic debris has not been adequately addressed. (similar to those performed on fibrous insulations) 27 In particular, the water velocities required has been completed by Alden Research Laboratory.

to transport debris and hold it against the sump The results are summarized in NUREG-0897 and used in screen have not been studied.* RG 1.82. The assumption that all fibrous blankets and The 7 L/D criterion is based on experimental studies pillows within 7 L/D of a break are destroyed is of representative samples of fibrous pillows exposed overly conservative. Different designs of pillows to high pressure water jets. These small water jet have varying resistances to destruction by water studies showed that increasing pressure (40-60 psia)

jets.* results in destruction of pillow covers and release of core material. Furthermore, blowdown experiments i' in the German HDR facility showed that fiberglass D' insulations (even when jacketed) were destroyed within 6 to 12 feet of the break, and distributed throughout containment as very fine particles. Unless conclusive experimental evidence is obtained that accurately replicates the variety of conditions that may exist in a LOCA, it is prudent to retain the conservative 7 L/D

      .                                                                criterion. The 7 L/D envelope is a significant reduction 8

from the previously proposed 0.5 psia stagnation pressure

      ,                                                                destruction criterion in NUREG/CR-2791 (September               ,

1982) and (in general) limits the zone of maximum destruction to the primary system piping and lower portions of the steam generators. f

Letter of July 14, 1983.

i .

E . Table 3 (Continued) iR '? o @ Comment NRC Staff Response ? k The commentor stated that estimated costs for DCF cost data are utilized in revisions to the 7 insulation installation and replacement are value/ impact analysis. too low. OCF cost estimates were* { Cost of NUKON" = $90/ft2 for material (as fabricated) Cost of reflective metallic = $100/ftz for material (as fabricated) Installation cost = $112/ft2 for labor and related support k The commentor provided recommendations for Descriptive classifications provided for insulation classification of various insulating materials, types have been combined with similar classifications stressing differences between NUKON" (an OCF obtained from Diamond Power Company for inclusion in product) and other fiberglass and mineral NUREG-0897, Revision 1 and NUREG-0869, Revision 1. wool materials. The commentor also noted the differences between NUKON" and high density

  ? fiberglass.*

t i The commentor identified 14 nuclear power OCF plant information has been utilized, along with plants that have been reinsulated with NUKON", information from Diamond Power Company, to develop are in the process of installing NUKON", or a current picture of insulation utilization in may install NUKON".* nuclear power plants. The major finding is that the number of plants using or planning to use fibrous insulation is larger than previously estimated. For example, the Diamond Power list reveals that 25 of 130 .

    ,                                                          operating and projected plants are utilizing fibrous insulation on primary system components.
                                                                                                                        ~

Letter of July 14, 1983.

I ,

I i EE Table 3 (Continued) - 5 l

o 88 Comment NRC Staff Response

  ![     The commentor recommended inspection surveys of             The recommendation for physical plant surveys (or g7     plants to identify actual insulations employed and          inspection to identify types and quantities of 27      recommended the modification of a draft generic            insulations employed) is a good one. However, the 3       letter to include this requirement.*                       generic letter for plant specific evaluation based on actual types and quantities of insulation employed will address this concern.

A report on "HDR Blowdown Tests with NUKON This report has been included as Appendix F in Insulation Blankets" was submitted to NL' REG-0897, Rev. 1. The tests demonstrated that support the capability of NUKON" insulation jacketed and unjacketed NUKON* blankets within 7 L/D to withstand the impact of a high pressure will be nearly totally destroyed. NUKON" blankets steam-water blast.** enclosed in standard NUKON" stainless steel jackets

3, withstood the blast better; not enough of these

  /3                                                                 tests were performed to allow conclusions to be drawn i

for similar insulations. Power Component Systems, Inc. j A report on " Buoyancy, Transport and Head Loss The formula provided for fibrous debris blocakge

Characteristics of Nuclear Grade Insulation head loss is included in Section 5 of NUREG-0897, J 8 Blankets," was submitted.to describe relative Rev. 1.

       ,  efforts in the area of fibrous insulations.***       -

i 1

Letter of July 14, 1983.

           **    Letter of February 18, 1985                                                       >
           *** Letter of November 8, 1984 h                                                 ,     s

APPENDIX B . PLANT SUMP DESIGNS AND CONTAINMENT LAYOUTS NUREG-0897, Revision 1 -

1 4 tk s ,, d

AW /

a e

                                                                                                                   'e            -

s# 's lI

                                 - %                    s'                                s     '     *   \   ,           .              -
                                                                        /l
                                                                   ,, # e ,
                                                                                     ,s' s '                   >             ..
                                                            <o           \s ,

s~

                                                $                       s'                                       $%

f . . l t -wz.

                                                                                                            ~
           .. = =.a   .'          p/"                       s,,.
                                            ~/'            _:                     cm 0
                     /

E.J

                                                ,           ON           e                      ,.

Bl p' ; a; -- R ( f / 4 2 W 4, r/ 1 .ym e - tv

                                         *,     '                       f         ki             I I < ~~

mieu unu I w

                                                                                   !fI anu _ _ m.
                              \                                    -

A '

                                                         -}
                                                                                                                                                 ~
                        .. .               ~

m,m . Figure B-1. ECCS Sump and Containtaent Building Layout,

Crystal River Unit 3 .

NUREG-0897, Revision 1 B-1

*N

l

( -

I I C)

                                                                 +
                                                               *0 4

8, f' +8

                                                                                              *k.                                                                                                                                     l
                                                           , s.                          t#
                                                                                                       '                  >8 i b b
                                                        ~%            ,,                      e          % . g j ,*'        '     i
                                                                                                                                                                         ~                  /u                               .

I I

                                                                                                            \'8 sisas esasnam -7                     :g          EEB\

r

b

                                                                                      .e                f                           4 k          ,

l 's ' ' ' . . _k

                                                                     .j                 .                                                                 _

geneetuCT s.

W r [...' smagten vtseth W T .d . f ,, ,_ u=r _aye . rae ,

                                                                                                                               , g.7,-.   -

I' .;~.f...)

                                                       ' : :   -   .       V :n UI              ,.        i .;        ,

I a:

                                                                           's IsrI  I

4 ~ ~ - ~

e 0 I h=t=' g

                                                                                                   ' '.(:,.
                                                                                                         .. f.4-l./..

J,.}

                                                    ==                  g  g, , d '.                                                                                       i
                                                                                                                                                                       ,i j 6==*

s, . ..=====a l

                                                                                       ,       g 5

r. . '

l ., '. s h

                                                                                                                                        ',1                    f d.
                                                                              ~

s w '% , i

                                                                                                %'                      j sous Arnws shst 'AthmFs fWr ,                                      ,-

tasuanties ashris to l reseh see. 1 Figure B-2. ECCS Sump and Containment Building Layout, oconee Unit 3 l l NUREG-0897, Reyision 1 B-2 _

o e

                                                                                                                                !* 'a.

f*d f 4 s,. 1 p *% l P o, f _

                                                                                                                                                                                                          ~
                                               ..,                                            %                                        /-                                   ji-               -

I y' N .-

               *                                                                           ,, i             : # , .. ,,                  teactor contaissent uau, M                                                           ss..= o.a.r.sor (e ... .
                                                                                       -                   .. . . .g . 1
                                                                              . A, p>                          i-                         4

! } d ! ~ y [.!. {

                                                                                ?
                                                                                             /

R?.? aW ' L n x m WA :- E=ad < f

                                                             ,                     h,                                                                   .

c p.n . . _., . n u : .r ,c . steam Generator e.. l Bote: grrer Mit 1 sange. shown. Unit 3 iseks' Arrows a 51. H 3'6" l

                                             ,gn ,. for insu2asten estria to pesan sump l

ECCS Sump and Containment Building Lafout, { Figure B-3. l Midland Unit 2 _

                                                                                                                                                                                                   ~'
                                                                                                                                                                       ~

NUREG-0897, Revision 1 B -3 _ _ _

s g sc*" 40' t i h

                                                                             ,       a                                so,                                                                                                                                            -'

I %

g,tj*,h # g .

                                                                                                       .wy
                                                            ;e*-                                                                                                                                                                  _,y
                                                                              #$                      Y                                     ;-

N.=

                                           ' 4'
                                                       /                                                        I                                                                                           **.

o. . > . , - .

                                                           . J'                                                   w                                                                                                          '~~           -
                                                                                                                                                                                                                                            .).

i.:::;.- g .o k ,

==.

n, .- ,,

                                                                                                                         ,                  ..                           .. .                                                     e.. j m

O

                                                                                                                         . xt-    -

l.da Ilrit-"' 2:i::: .

                                                                                                      *?'                                                                                                                     jd l

l ,

                                                                .s =     -

t. g;;g,,, y ow s_M al,'

                            ~
                          ,        .                                                                                             ,o e -
                                                                                         "                        e                         r,.

s ;, .

t. 9 /

                           /
                        .=.= =

=

u m / an.: arr . 3.sm r. r= s .a u e. net. 3.

                  'M             .                                                     .

31. -a' e =

l Figure B-4. ECCS Sump and Containment Building Layout, Maine Yankee , NUREG-0897, Revision I g_4 _ l , ws.-%+w w.w--,--.w-- .,.-w., w-,-,.-yc,----. -. _ -,--m - - . - , - - . - - - - - . - - - - - - - . _ _ _ _ _ _ _ _ . _ - _ - - _ _ _ - - - _ _ _ - - - _

i , . f N '0

4'd 0

9 b e N I

                                                                                 ~I                                                              -

a.

                                                                / /*                                                                                                                              v                                       .
                                                                                            \                                                                                                                        W
                                                                  %                                                                                .s -                              %
                                                                                                                                                                                                                               ~

s

                                                                                          %ek                   .

N. .

                                                                                                                                               %4kg.                                          -
                                                                                          /                        1>                                     -

f 7 "* 1

                                                                                                                                                   -               mi 3

fr YD [/3 J

                                                                         //                                                                  =G=.                              NN         i

/2 o Tl

l l ::::""

                                                                                                                                                             ~
                                           =:::.-

j U -

                            * $stei Arrows sker patmuers for

, tasuistsas estria to l sonet say. Figure B-5 . ECCS Sump and Containment Building Layout, ~ Arkansas Unit 2 _ l '

NUREG-0897, Reyiston 1 B-5 _._ _ _ j _n. , _ , , . . _ _ _ . _ , _ , _ _ _ _ _ . , _ _ _ , , _ , ,

a . f

f M I

                                                                              =Ac=                                -

Patssung ~ VESSEL - STEEL ComfarssesENT VESSEL

                                                 ,,,,..                                   I                              M, SRCAE                                                                   .EEAK LOCATNpe.                                                               LOCATION AAEA                                                                    AREA DAYWELL "7                                     .';

{ *.;; *

                                                                     . .. /                             g ,. *                                       -          .m& Ties 6 AT Q.';;.-
                                                                           ,                              ....;:.' } 9
 !                                    MPtseS                     * ** *                                    ! *. .'                                                   SR ATem0 Af sataa LetATaoss             * *!..

l

                                                                                                          / *.   ,g        .PIPIN aCas LOC    G ATeos,                   EL. son'-(

AREA . p . ., AatA g- - osv.ct.s..recoe

     ,x,
       , . ,a,rrttem a.

w m e

                                                 -          m    .
                                                                    .                                      ...    . C              CElll          C         -

,......v.. . ..... . . . , . . .

so coura -

                                                                                        .                                                                   ,.-     -.aowucousn t
   ,, a                                                          . ...

W I

                                                                                                                                                                    /.4T:=     tevn c t .4*c-o'd

.l w- a w- w= ?l. mW;;;- v "s' e

                                                                                                           ..,                                                          . . T..tk
                                                                                                                                                        -/

l un ,

                                                                 =.I"IE'I' 3.'                                     S.l:
                                                                                                          -                .,em.

c. ' ^

rx: l

                                                                 ..                                       c:7..
                                                                                                          -r.                                          ; .. -
                                                                 . . :a. . .

num'- r,

.... . 1. h;

                                                                                            . ... .... . . ; .. .q . . . , . .
                                                    ..    .:.g          ........      .. ..

mm ., n...r <... ... n .. . ,,,,,,,,,,,, .

              .m.n ...     ...

l Figure B-6. Primary Containment Vessel, WPPSS Unit 2 l l 1 1 l NUREG-0897, Revision 1 B-6 - m

h APPENDIX C INSULATION DAMAGE EXPERIENCED IN THE HDR PROGRAM . e NUREG-0897, Revision 1

 . o APPENDIX C
                                                                                       ~

INSULATION DAMAGE EXPERIENCED IN THE HDR PROGRAM HDR Program and Facility Description

The Heissdampfreaktor(HDR) safety program (PHDR) represents a major research effort in the Federal Republic of Germany addressing the safety of nuclear power plants. Funded by the Federal Ministry for Research and Technology (Bundesministerium fur Forschung and Technologie, or BMFT) and directed by the Kernforschungszentrum Karlsruhe (KfK), HDR experiments at'a decommissioned nuclear power plant cover a broad range of topics relevant to nuclear safety. The program was conceived with two basic objectives (1) to improve understanding of reactor system behavior under upset conditions and define margins of safety

(2) to evaluate and improve design and testing techniques fod nuclear systems and components HDR research is concentrated in the following five areas: (1) reactor pressure vessel blowdown from typical LWR operating conditions (2) response of structures and components to such extreme external loads as - i earthquakes and aircraft impact (3) structural response and fracture behavior of pressure vessels and piping under both thermal and mechanical loading

Source: Scholl and Holman, 1983.

NUREG-0897, Revision 1 C-1

(4) nondestructive examination of materials ^ (5) measurement of containment leak rates, both under normal operating conditions and following simulated accidents The HDR (Heissdampfreaktor or superheated steam reactor) achieved initi.nl criticality in October 1969 as a prototype 100 MWt boiling water reactor (BWR). Although the facility was originally intended to demonstrate tha commercial feasibility of direct nuclear superheat, numerous operating problems forced its final shutdown after less than 2000 hours of operation. Rather than restart the HDR as a nuclear facility, the BMFT decided in late 1973 to refit the HDR for light water reactor (LWR) safety research. The reactor inte-nals were removed and the facility decontaminated; new equipment was installed specifically for test purposes. The first blowdown tests at the recommissioned HDR test facility took place in 1977. . The HDR is a real nuclear power plant. That is to say, although it was originally designed nearly 20 years ago, the HDR still offers the following test capabilities relevant to more modern commercial plants: (1) Actual reactor systems and components can be tested up to 1:1 scale. (2) HDR systems and components are generically similar in construction and materials to those in use today. (3) The HDR containment provides a representative basis for investigating

pressurization and flow effects in multicompartmented structures following a loss-of-coolant accident. (4) The HDR can be placed under thermal-hydraulic conditions that subject systems and components to pressure, temperature, and n. ass flow loads typical of postulated accident scenarios. NUREG-0897, Revision 1 C-2

 . s The initial thermal-hydraulic state required for HDR blowdown tests (typically 110 bar, 310 C) reflects nominal PWR operating conditions and is produced by a specially designed test loop. The experimental test loop (Versuchskreislauf, or VKL) includes a 4-MW electric boiler for heating circulating water, a cooler with 8 MW of heat rejection capacity, and an appropriate volume and                      .

pressure control system. Warm water is fed in at the top of the reactor pressure vessel (RPV) and cold water at the bottom, and water at a mix . temperature is withdrawn through a feedwater inlet nozzle. The system is designed to produce either pressure vessel temperature gradients typical of normal PWR operating conditions or uniform (standby) temperatures. Initial tests on the VKL proved it capable of maintaining pressures stable within 1

bar, and temperatures stable within 3 C. Damage Incurred During Blowdown Tests

Blowdown tests conducted in the HDR facility showed there were high dynamic loads in the vicinity of the immediate break area. Inspections following these blowdown tests revealed: spalled concrete (attributed to thermal shock), blown open and damaged hatchways (in some ccmpartments doors were ,

torn from their frames), bent metal railings, damaged protective (or painted) coatings, peeled and heavily damaged thermal insulation on the piping and vessels, and scattered insulation debris throughout the containment building. The damage to, and the scattering of, glass wool insulation was particularly severe. I *See Holman, Mueller-Dietsche, and Miller,1983. NUREG-0897, Revision 1 C-3

I Figure C-1 shows the HDR containment and break compartment. The large number of compartments at various elevations should be noted and utilized when making use of findings for application to U.S. nuclear plants, which are generally much more open, without many intervening compartments. Figures C-2 to C-5 are photographs illustrating damage that recurred. Figure C-7 shows , a typical pressure and temperature plot for containment following a blowdown. Insulation Damage Experienced During Blowdown Tests NUREG/CP-0033 reports insulation damage as described below.

(1) Insulation (Vessel and Piping) Standard glass wool insulation with sheet metal covering was torn away within a radius of 3 to 5 meters and distributed throughout containment. A significant improvement was achieved through replacing the glass wool insulation with foam glass insulation on the pipes. 1 Insulation on larger vessels in the pressure wave path could be protected by steel bands as long as the pressure loading was from outside to the inside. However, at times the wave pressure loading penotrated beneath the surface and lifted off the protective sheathing. (2) Insulation (RPV) The RPV, with its nozzle openings and complex flow patterns, is an exception because the pressure wave propagates to a certain extent from inside to outside. Several types of insulation were tested here with the following results:

Glass wool with sheet metal sheathing was peeled off and destroyed.

 ' Foam glass was destroyed by larger inner overpressure because of its-brittleness.

NUREG-0897, Revision 1 C-4

                                                     -+50                                                  3605:0*

i i , i' .f.o l l G 80 - +30 %

7- - 4ffl 4 --

270 A 4 7 -

                     '1                 +              -.20 q  .   .      g

_. ..u,_ 1 1 . 10 === ,

180' 0 Cross Section A-A Cit # , 10nr 1.--~2 0'm f

HDR containment. i j installation shaft stairs crane acess hatch l

            ,.:                                                               , elevator break opening Figure C-1 HOR containment and break compartment
                                                                                                                                                     ~

NUREG-0897, Revision 1 C-5 . _

                                                                                                                                                 ~
                                 - - - - - - - - - -                .,-_ ,,_,        -------n,-_,,,---v-           ---   ,----e   ,oww - - - - -

s 7" 's

                                                                                                                                                                                                                                                            ;                               s
                                                                                                                                                                                                                                                                                                      /

s - f x. ,,e,

                                                                                                                                                                                                             's
                                                                                                                                                                                                                   ,2g                                                                         /

s ,, s j

                                                                                                                                                                                                                                             '/s            j                 7                                                                                                      I A

8erststutzen l

                                                        --                        ~

J / .-- ~ s i I N N - x_ - s.nete

                                                                                                                                                                                                            '                                                                                                                                                                        i l

1

1 l l 5*- .

                                                                                                                                                                      ..                 ,,       .-s,                                                                                            - -

ry' ' t , < : g. . <, ; , . , -

                                                                                                                                                                           ~,(
      ..,: L . ,# ; ,.                      I.               . .           Q                                       .
                                                                                                                                                     .- . ,; "                                                                                                                  q . gg, .' ..; -.-' "
                                                                                                                                                                                                                                                                               , j,
     ;.                                ,' ii.              -
                                                                                                                                                    " ,[  (                   h,hh-                                                                                         ,                                                                                       , ,.'
                                                                                                                                                                       ..-+g,
                      ... . ..,C-                         . . + . .

l . . ., I .

                                                                                                                                            ....-z...,

4 I4. - u- -

                                                                              'y                                                                                                                                                                                                                                                                                     >y
                 .s. +.y                                  4                                 .,                                                                         s                                                                                                                                                                                                     .
     ;" y m y : . ._ .r . ,; .,e'.

F

             .y .,
                                                        ... . , ,                                             -         r,.

s- ,, . ;

                                                                                                                                                                                                                                                   .*.    .'.,?..
                                                                                                                                                                                                                                                                                                 ~. .                 .
                                                                                                                                                                                                                                                                                                                                  . -='., ,7 ;
                                                                                                --#'s                       ,
                                                                                                                                                                                                                                                                          ;,+.,,7
                                                           -y.                                                                                                                                                                                                                                                                     , -
   . 9 ;, ,.                                                              , .                                                                                      "

... . .- :e :

                                                                                                                                                                                                                                                              ..~*y.                                   ':
                                                                                                           ...:. - - ,%'~-e-
                                                                         - ,l                                                                          , . ;;;

f ... 3

                                                             .-.y

_ /. ;; _ ,

                                                                                                                                                                                                   .-. . ;: '                                               ,..                                                                                                         .:h
                                                                                                                          ~*Tr. W.;
                                                                                                                                                                                                            ,                                                                           s y

sr : ,.,4 7 =y'* .9.. .- - ==ar

            'p' . . N . . .                   .s                      '.: f     s.f           .

s a e Af;, .

f g v 'S F r. , . (. ,fg -

                                                                                                                                                                                                                                                            ,", '-

pt , i '

                                                                                                                                                                          ..'y f 'f ;::,.,,;;.                                  . .
 ,                                     &                           .. .T-                                                                                                                                                                                 . c.:' y '                                -
                                                                                                                                                                                                                                                                                                                        . . . . ,                                    a~
   .l f,r if                                                                                                *sl..f.y. ' : 9*               .
                                                                                                                                                                           ,- ' ' l,                                     ;                           .

f .fgyy ['. i , Y . I 7-- -. ( l ,i ], , . . f.. : e, J ; h..Qtj '- ![ . ? 1. .' 3: r . , ,. p, ;i . , , . .

                -(,'. v.' }(%.'                                                '      ',.,                   %                 -
                                                                                                               * - '".                                             .y- j "                                                                          p?ly,f,'.#)..

3, ;. . 4

                                                         ,. .; 4
                                                                                              .,y %;.-                                                                                      '.#,      ,,,,.,                 ..cy                   . .,_ p , .,                                         ;                  ,,J 4' ga. (y . ' - r ,._ w . ,. Q ., . g,p lsi' i .g';g'.                                                                                                                             - ? y .g.                                           , ,.

m ., s

                                                                                                                                                                      .                                                                        , s . ,, ,>                                        -

_[. 3. ,

                                               ' . . ~. ne y-                                                                                                    y: .. %. ,,j c                            .-
                                                                                                                                                                                                                . 3 s q                                                                                                                                                                                                                                       W,5.> r .
,;                                                                                                            e.: t wx:,         .- -        ,                                                                                                                              t . .~ , .-
                       .M
  *), '. :-                                                                                                                                                                                                                                                                                          /

.. t',- .c, = . , ' ' - , > " -9 '

[,3 c- . . ;. g . 7

m. ;,.. .; .9,, a. p,,4 < > . s., ;.. i m_.g .

                                                                                                                                                                                                                                                                              - :; , g

_ , . p. ,. yn,u

                                              ;; .-                                                                                                  .e
                                                                                                                                                                                                                                                                        . i^ + ' , '.

7

                                                                                                                                                                                                                                                                                                                                     --. [
                                                                                                                                  ~

r .f-. 'hn ~ m c n_ 0 Y _,l * ._.

                                                                                                                                                                          , .- , ...j.- . . . ..                                                                             ; , . c; y

N 7 ,:-;xx . ,.s _,. . c . . A5 . .. .4 . t..3, .

                                                                                                                ...y .
                                                                                                                                  , ., , .           w                                 ,

s.. . .. , ,a Figure C-2 HDR blowdown compa tment and photo of local Jamage near the break nozzle NUREG-0897, Revision 1 C-6

       -M, . d. . . im s                            .
                                                    .o.                 2..

M. Y

                                                                                                                                                                                                       < a ,,6*@4O-                     -            x. , 4. . ' . ' . N .1 o, w                                                                                                                                                                                                                    .

e. . . . .s. .

                                                                                                                                                                                                                                                                                        . ?.. n Y' ".L '.%

4, h. \ . ' .-/ I[. - . I' l",gfM.44 , ,,3aispipMuesseuW6demeest".**.* $..s7 % ,T 4

                                                                       .p.
                                                                     ,.g. , ,, ; _                         7 ...a c;n,, n , v .c 4
                                                                                                                                                ~ q.
                                                                                                                                                                   ..' ,, p , 4
                                                                                                                                                         % 4, . , .. ..
                                                                                                                                                                                                                                    .s    ,..,.
                                                                                                                                                                                                                                                              ....t. > . .                          ..
                                                                                                                                                                                                         . .z,                                                                      'p
                                           %i                                                                                                                                                                           . . . <f ..                                                           - . .
                                                                                                                                                                                  .,4
                                                                                                                                                                                                                              .1'                     ,y
                                                                                                                                                                                                                                                         -        'S% a '                                                    _.
     -T .. ' _..) , ,..          , . , _

y (.

                                                                                     ' . ' . ' .;;p .
                                                                                                                                                                                                                       ). ~g ' . , : ? Ye' ?Q4 s
                                                                                                                                                                                                                                             . . %. ~ . 'W '* W
                                                                                                                                                                                                                          ' Q*,F'.l.

Tg.V . > o . w; . . q . 3 . f , ; ,. . _ .

                                                                                                                                                                                                                                                                                                -A ,.                -g
                                                                              .4*..,.                         A.w ,.. . . t
                                                      ,4, ,. . .                               /                                                                                                                                                                                                    - -
                                                                                                            #                                                                                                                                      s.
                                                                                                                                            .            i                    -

i 7, _ -z _* ' ' '? , r ?Q, :' !..Q .; - ..,~

                                                                                                                                                                                                                                                                   - _"~ d*'s
    ..-                                                                                                 ,+                                                                           .

w(, v...n .. a;4 o *.- . . , 3 4

                                                                                                             .,          . .: . :' w . . :- yk ., -- . . . , ~ ,
                                                                                                                                               .., .                                                            ,s.-

_n. o .

                   ...'.                                                      ,g,. %, - .- .> .
                                                              . .                                                                                                                                  e,      .v, r '.s.                                                     ', s                                                                                           y
  ',,.,.:. ..;a,'                                                                                                                                                                                                                                  . . :.y                                  .

l :, r, 'f-t , .

                                                                                  ;.~   . , . ,' ..;. r v.
                                                                                                                                                                      ...     . r    *        '   .e , c -                       - .                   c 4

z.. _

4. . . .

                                                                               - w_=                   .                                                                                                                  _ .                      .
                           ..,-_-.=....,,2 7-                                                                                                      .                                     --
                                                                                                                                                                                     . . . - .-. .-                                                               . c .,-                                         - .;
                                                                                                                                                              -t.,

g. g.. p, . ..q

    .1                                          q. . , , .U ,,
                                                                                    ., q 4 ., .,k., .... .p ,.. ,.;
                  ,                                       . .                                                                                                                                 . . , . y. . ..         .g g                    '.

n ., .

                                                                                                                                                                                                                                           ., .7. -
  .,          p.              s ..                       - . -                   .               .
                           . ..                           . ;, -

p. ,.r. .. :- ..*

x . . ,

,4 .. - ,.

4 ., -

                                                                                                                                                                                                                                                                                           '.                         ',. y.

u

                                                                                                                                         .. . , . '                     ..- .-      N".....,>                                ..                                                                        '.

w...

                                                                                                                                                                                                                                                 . . . I.
                                                                                                                                         ,' .pd. .:, ; ,, ,A                    - .,                            .,
  *'l
                                                                                                                                                                                          ') . . ..                                        >                                   b.
                                                                                                                                                                                                                                                                                                                        .T-y,;, . n. .                          . *4 . . .. . ' A. . yy-                                                   -

t'., .

                                                                                                                                                                                                                                                                                         .-             ,'8'
                ,..o                                             .i,n                       ,
                                                                                                                    ,                                            5                                                                                                        _
                                                                                                                                                                                   -     t*
                                                                                                                                                                                                                                                        .                     ; i ..
    ,.'/y.>,

e

                                                                                                                                                              ,3            g
                                                                                                                                                                                        ,*e...
                                                                                                                                                                                                                                                                                           .,,              ,4,.     +t'       -

e. r .. . . . . . . . . ...

                                                                                                                                                                                                                                                                                                                    ...i-
                                                                                                                                                                                                                                                                              .3...,.g i

i l l i 1 1 1 1 1 l Figure C-3 HDR post-blowdown damage to concrete structures 1 l within blowdown compartment l l l NUREG-0897, Revision 1 C-7

t i . i l l ( l l l p 4- g. ' .

                                                                                             '.%                    .. ,.:: .       ;"                        .!     w.' ..(,       " E+. '.4*:"2',,Q j;%f..                                                                                                       l t

g.-?-..". *:g .+.

                                                                                                                                           ,. . . ,'*y r.x-]                       . 7a q"q:                 & ;;

g.L.., -: .

-6

, . .. c.

                                 - g  ; ,                                       .: ._, - . .              *- ~
                                                                                                                                                                         ..w          . ,7

_..,..s

                                                                                                                                                                                                                                         <v l                           .,4;,,; r. 4 v
                                                                                                                                                                 . .. e .. . .-

.. ;~,.. -

                                                                                                                                                                                                                                 ~ ^ .. . ' % : J
                                                                                                                                                                                                                                         .._g, v v_: .. e ..
                                                                                                                                           ,_.      .. ..;_-               g_,                                                  .i 4.% , :-                                                                                                                                                                                       y
                                                  .               .:          F,.                    s m %.           s,            s,           . . . -                               .
                                                                                                                                                                                                           . e

_,. _ l ,, +

2- .. . , ., . t. - .. ?, .'**. 4 1e. .

                                                                                                                                                                                                                                           .                    'n
                                         , y,.                                                    e d                           '.- ;, y ~. s :

it - ._, .- . . '. .. .. % t, - .

w. ;

, ~ . ,t ..

                                            ' I , .%

G.; ,:!: :.

                                                                                                                                . . hh&. - '*' ' (,                                         ., , .                                                                                                               l l
                                   ? .hi ' . . ' YVk,
                                " ,g '                                                            i . ...?        .

yS. '. i> e ! i .

                                                                        .*?                                                                       '
                                                                                                                                                                                                                 .                                                                                               l Gr. ..; } ,;f; , <? ' ',.                                                                                                                 ll                                 .
                                                                                                                                                                                                                                ., jl*. , G ::

s

                                             * *k',                      n.;                           .
                                                                                                                                                                                                                             , .         v. ,: . ,,.
                                                                                                  ,                                                                                            ,..              .                                                  ..                                            l
                                                       ..: ,q                                                                           :                                             a.                        .

y- .;n= t

                                                                                                                 . . . C.;

l ,,.

                                                      ,s   9 6.r  .
                                                                                                                                                                               ' s.s         s
                                                                                                                                                                                    .. . . . ' ,                                         >, . . . c%.

i j

                                                          . -s
                                                          .,                                                                                                                         .f.. .. ,                                      t         - ,.
                                                                                                                                                                                                                                             .s...
                                                        .* %. ,r,v .;r J.                                                't.; ., s ,,

4:\.. .c

                                   ~                                                                                                                                                                                                                       '

l f s,

                                                    , yy.

g 4

                                                                                                                                                                                . r
                                                                                                                                                                                     .E.               <
                                                                                                                                                                                                        .*-           . , 1' . ,;,_.;
                                                                                                                                                                                                                                      .                n
                                                                                                                                                                                                                                                                   ,J.w;t 'y . ; . .y -

i u _,- p....

                                                                                                                                                                                                                                           .-         +

f- ,4...','..'.x .- , , . , , - r- ri

s ,,. . g , 1 . . -, ' . ' .,g. l ,

l % .b 4 A L.- N ' . [ .',' N. . , ,.{- ,.y . . , #. ,. . .. . . . .

                                                                                                                                                                                      'h '
                                                                                                                                                                                       .. ,                   . ..,f. .- - .             , . N ' I.<

e l

                                              ; pk l, ' . ,' t '.l '. .

f, _ ,;, -

                                                                                                                                                                                                               *      . ,\_ [-                                %              -

) ; . ~ .v .. .%, . , '.'. '... . y-.r.... ,

. .v__ _ . 9;- : * . ~. .' . . . ,

j f .. ____,,/.,,5,

t. <-

                                                                                                                                                                -4..            ..,              .
                                   %.4./g.k.., ' .j.;- ,c ,
                                                                                                                                                                                                                                   . . o -
                                                                                                                                              - ,A '?. * -
                                                                                                                                                                                                                      , .f . . . ;j ,' ,j;.. '

' . s,.. semi , . + #.T ' ' f g,. ,' ,.

                                                                                                                                                                                                 $d "' . . . ~ ; T, $
                                 .. . c . N . .

yi, ., . , ,.. x.

c. - ,,.

                                                                                                                  ;A 2
                                                                                                                                                                                           ~,,.3,.,
                                                                                                                                                                                                                                                                   .+                                             -

r. ' - ' . ; ~)'J ' a . , J v" .',."W, .

V.3. . .x . .y ,: " b > 6.* ^ v., ' [.;.0 k[i y'cS.Ik. l .. J- #,sdt: .

                                                                                                                               ' ' 7F .M  7# " :' \ .I ."Y'- Y E - N 7'I l

l Figure C-4 HDR post-blowdown damage to railing structures in compartment near blowdown chamber l i l t l

                                                                                                                                                                                                                 -                                                                                               I NUREG-0897, Revision 1                                                                                                         C-8

i i i l l l l ! i l i i [I s

                                                                                                                                                                                                                                 . .., 7                                             1
                              i, N:pl,i'7-
                                                                                                                                                                                                                                                               .)i, ,3 e,

l

                                 ;.;...                                                                                                                                                                                    4 'l dl' ;\                           4 l

l r L . r. t ,

                                    *f,
                                    ,1                                                                                                                                                                                          Mif.                     . . . . .._ _ . . . 4, I                                      ** i                                                                                                                                                                      w. ,                                                                 l I :f.                                                                                                                                                                     iff,'jG                                                              l
                                                                                                                                                                                                                 - %;-y                                                              l

,1 .s

                                                                                                                                                                                                                                                         ==                          e

)

i ,;,

( ~ 1 .:. .i wJ -

. ..s ,

l _ U.d A& I Y %

                                                                                                                                                                 *tv. ,    *
                                                                                                          .a-                                                             ' ~ 4.;. -g ?3                           ,.

r. -- . f ~-

                                                                                              .. :; 7,- T g ..";q, W
                                                                                                                                                                                                           - , 42 ;

_ ,.yx..LA' e 'h,e g

                                                                                                  .,.                                                                                                 <    :--                                   .2 M'*                                                                                                      g( ~ -(

3: CE .s. :

                                                                                                                                                                           , .l L. W                                      '-  .

Figure C-5 HDR post-blowdown damage to compartment doors due to pressure wave. Debris shown is spalled concrete from stairwell located near blowdown compartment.

L NUREG-0897, Revision 1 C-9 l . - - . . - - _ . - - , - . . _ _ _ _ _ _ _ _ _ - _ _ . _ _ - _ _ _ . . _ . _ _ _ . . - ..-l

l l .

[

                                     'i                                 '

l v;;

                                                                     .,                                                                                            l
                                                                                                                   ~

l i 4 l o, 5 ?

                                                                    *t                                                           ,
                                                                                                                                                                   ?
                                                                 ~
                                                                                                                                 ~

3 s .. , c; hY., !

                                                                                     ,'d 1

i l Figure C-6 HDR post-blowdown damage to insulated piping within the blowdown compartment

                                                                                                                                                                 ~i
                                                                                                           ~

NUREG-0897, Revision 1 C-10

l

l O

                                                                                          'l
                                                                                               . h$     .~ l s ;

a m

   '?
   =

i [ T CA P barixesi k 160 - -2.80

y angeR of Peak Pressures

e PRSB 120 - 7 2.20-a ng

Tdome C ' T

                                                       -                                                         break comp.

80 - r 1.60

4 -

       '                                                                                                                                      ^

l - 1.00-40

l. Maximum Blowdown Duration (<1 to 60s)

! "P : : : : : > 1 0 1000 2000 3000 4000 5000 Time Af ter Break ( Sec.) - Figure C 7 Typical pressure and temperature in containment after blowdows j . 1 l t .

Foam glass insulation sheathed in stainless steel proved more resistant to pressure waves and jet impingement loads because its connecting joints yield to inner overpressure and suppress it. Insulation mats with glass wool inserts and pure textile or wire-weave- , strengthened covers resisted pressure waves and jet forces equally well. Figures C-8 and C-9 illustrate the insulation damage incurred. Two letters from the HDR staff that provide further information regarding insulation are included in their entirety in this appendix. Two other documents that are pertinent to this subject are " Investigations of the Transpcrt Behavior of Particles During a Blowdown Test at HDR," GKSS Report 83/E/9, and " Considerations Related to Accident Induced Debris Distribution in a Pressurized Water Reactor Containment," GKSS Report 83/E/8, December 1982. Both documents were written by M. Kreubig and translated by G. Holman of Lawrence Livermore National Laboratory. References Holman, G. , W. Mueller-Dietsche, and K. Muller, " Behavior of Components Under Blowdown and Simulated Seismic Loading," paper presented at the ASME Pressure Vessel and Piping Conference, Orlando, Florida, July 2,1982. Muller, K. , G. Holman, and G. Katzenmier, " Behavior of Containment Structures During Blowdown and Static Pressure Tests at the HDR Plant," in Proceedings of the Workshop on Containment Integrity, Vol II of II, U.S. Nuclear Regulatory Commission Report NUREG/CP-0033, October 1982 (see also Sandia National Laboratory, SAND-82-1659). Scholl, K. H. and G. S. Holman, "Research at Full Scale: the HDR Program" in Nuclear Engineering International, January 1983. NUREG-0897, P.evision 1 C-12

.i .l . i I l t

                               ,l $,1-[A'}*'                */*             )f d          ..
                                                                                               '                                           f
                                                                                                                                           ~
                                                                                                                                                         ~

M

                                                                                                                                                     . .t                         ,

i [ ;9 ' '* ;1 4o - i e, , y- 6 Y[ - & l ..

                                                                                                                                                                                )
  < - ~                     . ._ fg,;,                                                                           ,

g ., I

a. ,. l .

g ; i

                                                                                            ~~
                                                                                                            ,, f i                                                                                                         {                            f g,                          q
                                                                                                                                               ~
                         \                                                  .J                                 _

i

                                                                                                                                           +)                                                                   ,

c. r ..,y ,

f? R

                                                                                                            ?                                               '1         '

4

                                             ..                     ~,.            .
                                                                                                             'fr
                                                                                                               ,\h.

f-t K, I * : )>~ ' ^ >- ejf;lt,' - l l i Figure C-8 Damage to jacketed fiberglass insulation located on the i HDR blowdown test. Source: letter from G. Holman to l A. Serkiz, NRC, " Photographs of HDR blowdown damage," l i April 18, 1983. NUREG-0897, Revision 1 C-13 _ 1

i i,

                                                                                                                                                                             't                           i l

1 I i 1

                                          ;l
                                                                                                                                  .. . 'qv,~ ~ > -

4 .l 4L . . . . .:, .. Q.N

' ..?'.._ N

                                         .:                                                       xs.j;', ..'.wQ..: .gp: y's."*;c1-- : _ '.sg                              ,

i r;. .r: ' ' . < :., n. . .*

                                                                                                      ;~g.v. .. s. . .e.. .~ r 4Pl, .'s ...

_. ' % ; '- . l i i i l l i Figure C-9 Foam glass insulation damage following a blowdown in the HDR. Foam glass insulation withstood blowdown tests better than fiberglass. Source: letter from G. Holman to A. Serkiz, NRC, " Photographs of HDR Blowdown Damage," April 18, 1933. I 1 - l NUREG-0897, Revision 1 C-14 1

Kcrnf2rschungszontrum Karlsruho - Geseaschaft mit beschrerster Haftung Kemforscrungszentrum Karieruno Garew Postfach 3640 D-7500 Kartsnsw t Mr. A.W. Serkiz Projekt HDR-Sicherheitsprogramm Task-Manager te,ter c.u ing w vu er-o,etsche n Generic Issues Branch Division of Safety ~ Technology U.S. NRC oaium: Aug7 02, 1983 - bo i Mail Stop NL-5650 so.re.te,. Kl. Muller , i reieton: o724n 82 4343 ) Washington, D.C. 20555 m,e u,et%

                                                                                                                                                  ~
              ~U.S.A.

Dear Mr. Serkiz:

I will send copies of our papers concerning equipment qualifi-cation next. week to G.S. Holman (LLNL) for translation. These papers conclude

                -     behavior of components during blowdown and in post blowdwon atmosphere distribution of isolation materials
                -      distribution of debries during blowdown in direction to the
    -                  sump-area
                 -     behavior of containment structures during blowdown.
                  -    proposal of using HDR as a equipment qualification testbed.

NUREG_0897, Revision 1 C-15 Kernforschungszentrum War.srune. WssenschatthCn TeCfnsche Emntungen und verwanung 15to Eggenstevi-Leocaussnelen. n Tel 107241) 821. Tenen. 7826444 Orantwort- Reaktor Karterune. Stadttmro u Postanschrift 0-7500 Karvu e 1 weoerstrade 5 Postfacn 3640 vorsdiender des Aufschtsrats Staatssekretar Hans Heger Haunsched. Vorstant Prof Dr. Rudo68 Haroe. Vorsdaender. Or Heemut Wagner. Steev vorsdaender; Prof Dr Horst 8cewn. Dr Hans Henrung Hennes Prof Or Westgang Kiose N N M ROP 1JKO)G C3 200 20t Commer

Mr. Wind of HDR-project will join the lith WRS-Meeting end October 1983. Perhaps you can contact him together with G. Holman. He will answer additional questions and if needed from your side, he can illustrate component behavior and - damage by slides. In this case please contact me during September 1983 by phone or telex. _ With best regards Kernforschungszentrum Karlsruhe GmbH - Project HDR Safety Program I O I t [ / .' (. fu NUREG-0897, Revsion 1 C-16 M e

Kcrnfarschungaz^ntrum Project HDR Safety Program Kl. M(iller Karlsruhe Aug. 1, 1983 - bo NOTA: INSULATION DAMAGE IN THE HDR BLONDOWN EXPERIMENTS -)

1. Glas Fibre Insulation HDR was equipped with this typ of insulation. at the begin of the experiments. In the break compartment all glas fibre insu-lation was destroyed at 2 m around the break nozzle and dis-tributed through the whole reactor in very fine particles on the walls and floor. The iron wrappers were thrown away from-vessels within 4 m around the break nozzle, the glas fibre -

being untouched. With enforced shieldings (steel bandages) around the vessels nothing happened.

2. Glas Foam Insulation Glas foam insulation around pipes up to 200 mm # withstood the blowdown impact even in a distance of about 2 m around the nozzle, except a small area where the mass flow touched the pipe. At these placed the insulation was cut out. The insulation of the pressure vessel .was destroyed at great areas around the break nozzle caused by the first pressure wave cracking the material (short break nozzle RDB-E experi-ment).

NUREG-0897, Revision 1 C-17 6 @ Kernforschungslentrurn Kartaruhe' WeserieCmaftsch-feCnr. sche Evectifungen und Verwanung- 1514 Eagenstevi-Lecooksinefen fel 1072471821.bei Durctiward 82 , fees: 1826484. Orarwwort- peantor Katierune. Stadecuro u Postanschnft 0-7500 Karteruhe 1. Weeerstrade 5. Posefech 3640

o . The glas foam then was cracked into great pieces not leaving the break compartment and a great amount of fine particels following the blowdown pathes up to the sump inlet.

3. Glas Foam with Stainless Steel Shielding .

This material withstood all impacts and retained intact even - installed about 1 m around the break nozzle. .

4. Insolating Matrazes They consist of an special cloth outside eventually reinfor-ced by steel wires filled with glass fibre or. stone wool. This material withstood all impacts even good as material point 3.

Nevertheless there were some corrosion effects on the cloth caused by demineralized water at high temperatures. More detailed information you will get on request from Mr. Wind of K. Mueller. ( H NUREG-0897, Revision 1 C-18

*M

e .

Kcrnfcrschungsz:ntrum Karlsruhe ' Geseeschaft tml beectwenkter Hettung Kerntorschungsgentrum Karter@e OmeH Poettaen 3640 0 7500 Marierune 1 Projekt HDR-Sicherheitsprogramm Mr. A.W. Serkiz L*'er: D* -'as W Muaer O'etsene Task-Manager Generic Issues Branch Division of Safety Technology . . U.S. NRC ' Mail Stop NL 5650 o.eunt Sept. 12, 1983 ~ 8"*" K. Muller Washington D.C. 20555 reston or24n 82 4343

   . U.S.A.                                                                    ,                  in,g u,,,%                            ,

Dear Mr. Serkiz,

I read the reports you send to me with great interest. There are some additional remarks concerning Nureg/CR-2982 coming from our experiments. We found out; the jet forces are main cause for debris generation and distribution; pipe whip etc. are neg11 gable. Jet forces act only in a diameter of 2 - 5 m around the nozzle,de-pending on break diameter and break geometry. We did these experiments with pure steam and pure water jet with nozzle diameters of 200 - 450 mm 9. - First the pressure wave mainly destroys covers around fibre-glass - and mineral wool and brilt le insulation materials as glas foam. Than the impact of the fluid peels off the unprotected " wool la-yer" or cuts out the foam glas around pipes. The jet and the following turbulences transport even heavy weight fragments to the next compartments. Here heavy parts are normaly fixed by drag force and only light wight particles will be trans-ported further especially into the dome. NUREG-0897, Revision 1 C-19 - - Mornicrechungssentrum Marierune. Wiseenecnettech.feceresche Envechtungen und verweeung 75t4 Eggenstein-Lecocidenaren, fel 407247)821.fones F826444 Orane ort Reektor Marterune' $racteuro u Postanschrift- D 7500 war'arune 1. Weeerstrede S Postfach 3840 Vortsliender des Aufeschtsrate $f astesekretar Hans Hager Haunecodd. Vorteent Prof Dr Rusost Haros.Vorestaeneer Dr Heemue Wm 5 fee, voretaeneer. Prot Dr Horst Schm, Dr Hane Hennen9 Hennies. P of Dr Wolfymose__

All grids and components within the building act as a screen for fixing these light wight particles, so in containments with a com-plicate interior most of the generated debris are fixed before reaching the sump area.

So only a break location with direct access to the sump area will block the screens in the way described in your papers. . In the post blowdown phase when the emergency cooling system is fed by the sump water there are only some " main water ways" left leading from the nozzle to the sump. These " main water ways" will not cause pump failure. , From my opinicn you will get more debris collected and settled wit-hin the core barrel and other core internals than reactivated by the back flow of the water to the sump. Even if activating the ' Containment spraysystem you will get more problems with the blockage of the injection nozzles of a wa te r spray system by the debris than blocking the pump or sump' inlet. Yours sincerely, Kernforschungszentrum Karlsruhe GmbH ProjektHDR-Sicferheit rogramm e / L/ i I 5: G. Holman, LLNL F. Wind, PHDR NUREG-0897, Revision 1 C-20

                                          *M  '

3 .

APPENDIX D DETERMINATION OF RECIRCULATION VELOCITIES

                                                           ~~ ~

NUREG-0897, Revision 1 -

       ..                 .      .      --           -          .- --                           _     _   . - - _                                       .-    .~.

APPENDIX D DETERMINATION OF RECIRCULATION VELOCITIES 1.0 General During the recirculation mode of operating the ECCS, water on the reactor floor will drain to the sump, the source of water for pumps which provide long-term cooling of the reactor. This flow of water on the reactor floor may be at sufficient velocity such that insulation debris is transported with the flow, resulting in blockage of the sump screens and a pressure drop across the screens. Of major concern is the impact of this potential pressure drop on the pump flow and on the available pump NPSH ccmpared to the required NPSH.

                         - Various types of insulation materials have been tested to determine what flow velocities will initiate movement and transport of this debris.                                                        Of equal l                            importance is the determination of what flow velocities will exist in a given plant during the recirculation mode, as it is the relative magnitude of the actual recirculation velocities to the experimentally determined transport velocities which determines the probability of insulation debris blocking the sump screens.

i l Due to the arrangement of plant walls, structures, and equipment, there will be only certain flow paths available from each postulated break location to the sump (s) . Some plant layouts will result in a few obvious flow paths in , other plants, the flow paths may be numerous and not so easily defined. Those paths having the shortest length and offering the least resistance (lesses) - will produce the greatest velocities (i.e., have the most water surface slope). For a given velocity, the flow path with the largest cross-sectional area will carry the largest discharge. Local velocities will be considerably different from average velocities due to local flow contractions. Losses may be produced by surface friction, drag due to the flow past appurtenant struc-tures, equipment, or pipes, expansion losses downstream from constricted l !, NUREG-0897, Revision 1 D-1 .

    ,-    --.-----.-,,---,..-,n.                    . + . . - .       _    -...,----.,_-n.                        - .----.- -- -.- _ _ _ - , - - -

openings, bends of the ficw path, and any other phenomena causing turbulent energy dissipation. This appendix will review various means for determining the recirculation velocities, such that an assessment of debris transport can be made. If a - preliminary analysis using simplified methods indicates recirculation veloc-ities are within a factor of about two (2) compared to the experi~mentallf derived transport velocity for the insulation type (s) under study, then more l refined analyses are warranted. For example, if recirculation velocities are up to about 50% less than the predetermined debris transport velocities, f transport may still actually occur since many approxima* tons are inherent in the preliminary analyses. On the other hand, if the recirculation velocities are up to about twice the transport velocities, transport may be less severe than indicated for similar reasons. To be conservative, it should be assumed i that all flow is returned by the safety injection system since this maximizes recirculation velocities on the containment floor. 2.0 Review of Network Resistance Method A preliminary method of estimating recirculation velocities is to define a i system of possible ficw paths with varying resistance. This flow / resistance network is simplified by finding equivalent resistances to series and parallel paths, until one equivalent flow path remains. Since the total flow is known and the equivalent resistance may be estimated from coefficients of friction j and losses available in handbooks, the total head drop from the break to the sump may be calculated. As all parallel flow paths are subject to this same total drop, the individual flows in all other paths may then be determined. Knowledge of flows per path allows local velocities to be determined from the known local cross-sectional areas. This preliminary analytic method is [ presented in NUREG/CR-2791, and is summarized below (using conventional t hydraulic terms). t i N'JREG-0897, Revision 1 0-2 1

As an illustration, assume the simplified situation shown in Figure D-1 (taken from NUREG/CR-2791) . Flow from the break may reach the sump in all combina-tion of the paths illustrated, and this combination may be reduced to the . flow / resistance diagram shown, where resistances R1 through R8 correspond to the similarly numbered flow paths. The resistance may be determined from the . following set of equations, starting with the well known Darcy-Weisbach resistance formula (4, 8) . g y2 (D-1) L 4R 2g H where h = the drop in water level (piezometric head) (ft) f = friction factor (dimensionless) L = flow path length (ft) Rg = hydraulic radius = flow area / Wetted perimeter (ft) V = average cross-section flow velocity (ft/sec) g = acceleration due to gravity (ft/sec ) values of f, which depend on the relative roughness of the flow path and the flow Reynolds number, are available from standard text and handbooks, such as (4, 8). Since V = Q/A (D-2) where Q = flow rate (ft /sec) A = cross sectional flow area (ft ) , letting C = 1/2g K= 4R g

~ ~

NUREG-0897, Revision 1 D-3

then

h = L 2 Q A By setting , R = KC/A we obtain the usual system loss equation h = RQ (~ l indicating greater resistance (higher values for R) for paths having greater i j friction, longer lengths, and smaller cross-section areas, and vice versa. 1 i Equivalent resistances may be found for combined flow paths by use of the ! above equations and continuity, noting that the loss for each parallel flow l path equals the total loss. The result is that resistances in series add, and resistances in parallel follow a reciprocal law. i Therefore, the network in Figure D-1 may be simplified to 6 , h where RA = R1 + R8 R2 RB = R3 + R6 m RC = R4 + R5 : U RB v Parallel resistances such as R2 and RB may be combined by finding an equiva-lent resistance i NUREG-0897, Revision 1 D-4 1 e

o .

                                                                                       '2 1

RD =

                                                                        /R2, 1         1 RBJ                                                                                                                                                 -l
                                                                                                                                                                                                                                      ~

such that the network is now simplified to U . v@

                                                                             @v where RE = RD + R7 which in turn is reduced to one equivalent resistance RF by application of the reciprocal law for parallel resistances. Therefore h = (RF)Q

and h may now be calculated, because RF is estimated from the individual branch resistances, and the total flow of the ECCS is known. Given the calculated h, which is the same for all parallel branches, flow in each L

branch may be calculated using the individual resistances for that branch. For example, L 0 1 *9 8

R1 + R8 ,

and the velocity at any section along flow path 1-8 may be determined by dividing the above determined flow rate by the cross-sectional area at the section of interest. It is important to consider local flow contractions to less than actual structural openings. A typical flow contraction can be as low as about 0.65 of the actual available opening, depending on the geometry involved. -

NUREG-0897, Revision 1 D-5 .

a . The above summarized method appearing in NUREG/CR-2791, although sound in principle, includes many approximations. A basic problem is that values for f are available only for straight, prismatic channels, and that average values of f and R are use r e en e pa . s may be percome by using H much shorter flow paths, each having the proper value of f and R ' ~ H makes the calculation more laborious. It should also be recognized that most of the flow resistance is due to drag of various objects in the flow ' path, to bends, and due to flow expansions from contracted areas. Drag losses may be expressed as (4) D" D where C = a d mensionless drag coefncient. D A similar expression is used for losses due to bends h a" B where C will vary with the bend radius. B Values for DC and BC are ava a e f r a va en s apes in staMad ten and handbooks (4, 8). Head losses due to flow expansions are given by (1, 4 & 8) 2 V y2 h E

                    =[l-       ]        =C  g[

where a = contracted flow area (ft ) A = downstream cross-sectional area (f t ) V, = contracted area velocity (ft/sec) NUREG-0897, Revision 1 D-6

l l l . l I The contracted velocity may be related by continuity to the average flow l velocity of the branch, and C expressed in terms of V instead of V . The , [ E a r l total head loss for a given flow path may thus be calculated from e L y2 ,, . h = [f 4, +CD+CB+CI E 5=R7 (D-5) l where l R = ( +C

D B E 2 H 2A 9 The above illustrated calculations will be improved by the addition of these terms, but numerous flow paths must be defined such that the available values l of R H' D' B, and CE really apply to that section, as average or effective l values of these coefficients for varying path characteristics cannot be 1 i

determined, i i l l Despite the possible refinements to this method, not all flow / resistance networks can be simplified to one equivalent resistance. Consider, for , example, the following simple case. l i R1 R2 , BREAK SUMP p R4 R5 L This problem may be overcome by using a different type of analyses, as illus-trated below for a more complex flow network postulated for a given plant. l ~ NUREG-0897, Revision 1 D-7 l _ - - _ - - - - - _ _ . . _ . - - . _ . _ _ . _ _ _ , , , _ _ _ _ _ _ _ . _ ~ , _ _ - - . - _ _ _ , , . . . _. - .

a . 3.0 Complex Network Analysis In the example illustrated in Figure D-2 there are 28 flow paths and 18 . junctions, A to R. For each flow path Eq. (D-5) is applicable. For example, for the flow path 5, - H -H

95 C B where H = piezometric head at the junction identified by the sub-script (ft)

        .          Q = flowrate along the flow path identified 'by the subscript (ft /sec)

R = an overall resistance factor as defined in equation (D-6) S for the flow path identified by the subscript (sec /ft ) Similar to Eq. (D-7), 28 equations corresponding to the 28 flow paths are . available. Also, for each junction the continuity equation can be applied. For example, in Figure D-2, for junction J, assuming inflow from flow paths 16 and 21 Q g7 =Q16

021 (D-8)

Combining Eq. (D-8) with head loss relationships similar to Eq. (D-7) gives

H

                   #                   H* - H            H   -H

( - H^ )

                                  =  (     ,   #)   +(     I
                                                              ,  #)

R R g7 16 21 For each of the junctions, one could write an equation similar to Eq. (D-9). Hence, if flow directions are first assumed, 18 junction equations are ob-tained to form a system of nonlinear equations with the 18 unknown piezometric NUREG-0897, Revision 1 D-8

heads at junctions. One of the most widely used method for solving such a system numerically is the Newton-Ralphson method (5) which iteratively solves the system of equations. Computer programs using the Newton-Ralphson method , are readily available in many books on pipe network analysis (5, 10) or on numerical analysis of nonlinear equations (3). To use the Newton-Ralphson - method, one assumes the flow directions and provides an initial estimate of the piezometric heads conforming to the assumed flow directions. Since the method is iterative, the acceptable error in final solution should also be indicated. The method usually converges very fast, although convergences may not be obtained if initial values are unreasonable and too far from actual values. The flow directions, if wrong, will be autcmatically corrected by the calculation procedure to conform to the values of the piezometric heads obtained after each iteration. For the example considered, the piezometric head at the sump and the total flow into the sump Q w uld be known. T Referring to Figure D-2, H, the piezometric head at junction A is known. If the piezometric heads at each of the junctions B to R are determined, one could calculate the flows using the flow path equations similar to Equation (D-7). There are 17 unknowns, namely H B "R, and the required 17 equations can be obtained by writing the conti-

 . nuity equations at each of the junctions A to Q.                       For example at junction A, H                           H              !                      !

H ( - H* )1/2

                                            +   (

0

                                                          - H^ )
                                                            ,          +(       U
                                                                                   - H^ )

R 17 2 6

                                          !                          /                          - 0)

H 8 H -H a

                      +(          -. H^ )     +   (           , A)        ,
                                                                                  ,g R                           Rg 7

The Newton-Ralphson method can be used to solve the 17 equations similar to (D-10) for the 17 unknowns H B "R. e me s iterathe and soWes a linear matrix as explained below: NUREG-0897, Revision 1 D-9

s 4 Let the 17 non-linear equations be, F =0 , F =0 2 F 7

                                      =0
           'A linear matrix is written as,
                      #                                                                                         9    p       %          #       1 0F           OF                                                0F 1                              1             ~~

1 Z F OH B OH C

                                                                                      ~~'OH R F
)                                   2                              2                                     2             Z         =        F 2

(D-11)

                                                                                 ~~~~

C OH ' OH '

                                                                                            ' OH B                            C                                       R i                       _                _____________                                                                 _                    _

t t 4 0F 0F g7 0F 17 ~~~~ 17 :: F OH OH OH B C R , , , , , ) Using the initial guesses of H B "R, he v ues of 0F3/6HB ' I 1! "C # * " " j F, F etc. are calculated first and the linear patrix (D-11) is solved to 2 obtain, Z B R'

                                                                      *#       "*#      "*                 "

W***** B R*

the values of F , F etc might be non-zero, since the initisi guesses are not 2 actual values. The corrected values of H g to H are used for the next iteration, and the R calculations are repeated until Z " ** ###* * * ' "

B R ***

l margins. Af ter several iterations, the final corrected values .2f H B "R i will be considered as the actual values, and these are then useri in the flow path equations similar to (D-7) to obtain the flows in each flow path. l , D-10 NUREG-0897, Revision 1

l Alternatively, a network analysis based on corrections to the flows in each loop could be performed. The flow system given in Figure D-2 could be trans-formed to an eight loop network as given in Figure D-3 by replacing parallel ' ~ ' pipes with equivalent pipes. In this case, initial guesses of flows along each flow path should be made such that the continuity equation is satisfied . 4 at each junction. Referring to Figure D-3, there are 8 loops. For each loop t the algebraic sum of the head losses around the loop would be zero. The i positive direction of flow must be defined, such as clockwise around each l loop. For example, referring to Figure D-3, for loop 6 with assumed flow directions let the initial guesses of flows be Q13, Q ,4 Q11, and Q12 * "9 *

flow paths 13, 4, 11, and 12 respectively. Since the algebraic sum of the head losses around the loop would be zero, we get

] F ~ 6" 13 (013

096 4 94~ 096
001 J

i

                    - R.          (Q g
                                          - 606

095 +R12 ( 12 # 006 ~ 00 7

                    =0                                                                             (D-12) 1 I

where AQ g is the correction to flows in the loop i required to convert initial estimates to actual values of flows. When a flow path is common to more than one loop, corrections from each of the loops have to be included to get the actual flow for that flow path. Writing similar equations for each loop, we get r =0 1 r2"0 i r8"O I

~ ~

NUREG-0897. Revision 1 D-11 ~

e e As a first iteration, the unknowns SQ to AQ 8 are solved by Newton-Ralphson method for the assumed initial guess of the flows around each loop. Then the flows are corrected with the obtained values of AQ to AQ 8 "" * "'* - iteration is carried out. The procedure is repeated until AQ g to AQ ecome 8 acceptably small. .

 . This method is quicker in that a lesser number of unknowns (equal to n' umber of flow loops) is involved. However, it is difficult to give initial estimates of flows satiafying continuity equation at each junction.

Instead of the Newton-Ralphson method, other iterative methods can be used, such as Hardy-Cross or linear methods, to solve the nonlinear system of equations (5). l Irrespective of the method of analysis for large networks, the time consuming part is providing the initial data of R values for each flow path and the initial estimates of piezometric heads or flows. It must also be realized that many break locations must be considered', with each location requiring re-evaluation (perhaps redefinition) of the flow network. Therefore two other methods to predict flow patterns and local velocities are addressed below. 4.0 Two-Dimensional Analyses Rather than pre-defining flow paths, another approach is to use a two-dimensional numerical model which, by its nature, secounts for the shape and size of the various flow paths and obstructions in the containment build-ing. The flow to the sump being basically horizontal, the complete three-dimensional flow equations are integrated vertically over the water depth (depth averaged) and solved numerically using one of several techniques. The two basic classes of techniques are the finite differences and finite elements methods. In the former, a grid is defined covering the flow field, and the derivatives appearing in the differential equations are approximated based on the values of the variables at the nodes of ths grid. The most

NUREG-0897, Revision 1 D-12

I common type of finite differences grid is rectangular, with possibility of a variable resolution, but other grids are possible, particularly circular grids for problems with obvious circular characteristics. , In finite elements methods, the variations of the variables of interest are - approximated continuously over elements through pre-defined " basis functions" (or interpolating functions) and nodal values. The most common ' type of element is triangular with nodes at the vertices, but there is no limitation on the shape of the elements that can be used (rectangular and curvilinear are common), and the number of nodes per element depends on the choice of basis function. One of the advantages of the finite elements method over the finite difference method is that the flow domain can be approximated more closely and that variations of resolution are more convenient with finite elements. As an example, a grid of triangular finite elements is shown in Figure D-4 for the previously discussed application. Finite elements solutions, however, tend to require larger computation times than finite differences solutions. There are many other differences between available two-dimensional models. These other differences concern the details of the numerical technique used, such as the way in which the nonlinear terms are treated or handling of the advective terms (which tend to create numerical instabilities), or the way tire integration is performed. Another important difference between available models is the way in which turbulence and the corresponding Reynolds stresses are simulated. A common approach is to use an eddy viscosity concept but flow separation is then difficult to reproduce, and the values of the eddy viscosi-ty has a large effect on numerical stability, making the selection of this parameter all the more critical. The so-called e-c method of turbulence simulation has recently been shown to be very powerful, at the expense of an increased number of differential equations to be solved. For auch two-dimensional analyses, the break flow is simulated by a flow source term (s) at one or more nodes at the break location. The sump may be simulated either by sink terms for nodes around the sump, or specifying values NUREG-0897, Revision 1 D-13

of normal velocity ccmponents at these locations. Various assumptions regard-ing the distribution of velocity or flow around the sump may be made. Losses due to friction and distributed drag from small pipes or structures are , estimated and appropriate values of f selected. Losses due to flow eddies and large-scale turbulence may be simulated depending on the grid detail and on . the analytic model. For practical grid sizes such as on Figure D-4, proper modeling of flow separation is doubtful. Initial values must be pr'escribed for velocities and water depths at all nodes, and zero velocities and a horizontal water surface are convenient initial conditions. At solid bound-aries, zero normal (perpendicular) velocities must exist, although the tangen-tial velocity component may be either zero or unprescribed. Several two-dimensional models which are applicable to this problem are available, including those by Wang & Connor (9), Leendertse (7), Benque et al (2) and Launder and Spaulding (6). Application of any of these models to the calculation of recirculating flow patterns in containment buildings should, , however, be subject to careful evaluation as a number of features exist in the proposed application for which the analytic models have not been fully tested. A notable feature to be checked is the flow separation that can be. expected behind obstructions. Results of two-dimensional models are flow velocities and water surface elevation at the node points versus time. For this application, transient effects would probably be negligible, but the computation time would remain large because of tne fine grids required to account for the geometrical details of the domain. In spite of the relatively. dense grid shown in Figure D-4, it is not possible to closely follow the actual bounding geometry in regions of small clearances and local contractions. None of the analytic techniques described above includes consideration of the initial break flow momentum, nor do they closely simulate the complex geometry of the containment and appurtensnt equipment, as either one- or two-dimensional approximations are made. Also, losses must be independently NUREG-0897, Revision 1 p.14

estimated. If complex flow patterns have significant effects on the problem under consideration, it is accepted practice that a physical (hydraulic) model study may be necessary. 5.0 Hydraulic Model Studies - Depending on how close any analytically predicted recirculation velocities are to the experimentally determined debris transport velocities and the need to further refine the evaluation of potential debris tr insport to the sump, it may be advantageous to use a physical (hydraulic) model. Such a model would include all geometric features of the containment floor . area which could affect flow patterns. A portion of the type of model which wculd be suitable is illustrated in Figure D-5. Although a full-scale simulation of the reactor floor and sump geometry may be considered, it is more efficient to use a reduced scale model (and there is no technical reason to the contrary). NUREG/CR-2760 reports on studies specifically designed to evaluate potential scale effects on sump hydraulics. These studies show no scale effects as long as medel flow Reynolds numbers exceed certain limiting values, such as typi-cally achieved at geometric scale ratios of about 1:4. The advantages of using a hydraulic model are . (1) There is no need to make assumptions regarding loss and contraction coefficients as these are implicitly included. (2) Flow paths are reproduced to their actual geometry rather than simulated by one- or two-dimensional techniques, allowing accurate spatial defini-tion of velocity variations. (3) The break flow momentum can be scaled, and numeroos break locations can be evaluated without model reconstruction. NUREG-0897, Revision 1 D-15

4 (4) Basic debris transport phenomena, such as relative volumes moved and downstream settling in lower velocity areas, can be demonstrated using

;                                                          simulated (scaled) debris.

5.1 Similitude Requirements . l The main similitude requirement is based upon scaling the two dominan't forces , in free surface flow, gravity and inertia. These primary forces are embodied in the Froude number, F, i l' r- 1 . VgH , (where V, g, and H are as previously defined) and equality of Froude number between model and prototype leads to proper scaling of flow patterns from the break to the sump. The selected geometric scale ratio must be large enough, however, such that viscous forces involved with friction and drag are properly scaled. This will be true if the model Reynolds number is large enough auch that loss coefficients are equal to those of the prototype. Alternately, adjustments in the size of components causing losses may be made to compensate for the lower model Reynolds number. The use of standard laboratory velocity l meters may also influence the choice of the model scale ratio. i l It should be noted that the actual reactor pressures and water temperature do l not have to be scaled in the hydraulic model. The gas pressure over the water I is constant in space and will have no effect on flow patterns. Water tempera-ture affects the water viscosity and surface tention, but neither parameter E influences flow patterns for sufficiently large geometric scale ratios and ! model Reynolds number. t Simulation of the insulation debris transport, if desired, is more complex. l Since it may not be possible to directly scale all relevant parameters as is j the case for other analogous hydraulic models simulating material transport, i NUREG-0897, Revision 1 0-16 1

s

                                                                                                                 . 6 test results are more qualitative than quantitative.                                                                                   One approach is to find w

a model material which is transported at the model velocity scaling the known actual transport velocity for that insulation material. Alternately, the actual insulation material ruy be used (at scaled size and volume) if the model flow and velocity is increased to actual (prototype) values, while - maintaining the water depth. 'For a scale ratio of 1:4, this involves doubling the model flow and velocities from that given by normal Froude scal'ing. It should be demonstrated that such flow increases do not change the flow pat-turns as determined from running the model at Froude scaled flows. t s

.(

1

s I % i

                                                                                                               )                              ',

i S

x NUREG-0897, Revision 1 0-17 .

REFERENCES

1. Baumeister, Theodore, editor, Marks' Standard Handbook for Mechanical Engineers, Chapter 3, Eighth Editions McGraw-Hill Book Company, Inc.,

1978. .

2. Benque, J.P., J.A. Cunge, J. Feuillet, A. Hauguel, and F.M. Holly, "New Method for Tidal Current Computation," in Journal of the Waterways, Port, Coastal and Ocean Division, ASCE, Vol. 108, No. WW3, August 1982.

3. Conte, S.D., and C. DeBoor, Elementary Numerical Analysis, McGraw Hill, 1965.

4. Daily, J.W., and D.R.F. Harleman, Fluid Dynamics, Addison-Wesley Pub-lishing Company, Inc., 1966.

5. Jeppson, R.W. , Analysis of Flow in Pipe Networks, Ann Arbor Science,1979.

P

6. Launder, B.E., and D.B. Spaulding, Lectures in Mathematical Models of Turbulence, Academic Press, New York, 1972.

7. Leendertse, J.J., " Aspect of a Computational Model for Long Period Water Wave Propagation," Rand Corporation, Memorandum BM-5294-PR, May 1967.

8. Streeter, V.L., Handbook of Fluid Dynamics, Section 3, First Editions '

McGraw-Hill Book Company, Inc., 1961.

9. Wang, J.D., "Real Time Flow in Unstratified Shallow Water," in Journal of the Waterways, Port, Coastal and Ocean Division, ASCE, Vol. 104, No. WWl, February 1978.

10. Watters, G.Z., Modern Analysis and Control of Unsteady Flow in Pipelines, Ann Arbor Science, 1979.

NUREG-0897, Revision 1 0-18 ,

CO?!TAINMENT

                                                         ,.3,,.;;j'.:$g* . $ :.

s

                                                       . :.p . ,                   *{.*;y,g
                                                . *-                            N4
                                                                                              ~

ey . SUMP

                                           ? .,.;.N'!.e
                                                                                                        ,9 .
                                                             .g J ,< f

J *I L* t,g*

                                                                                    * . y,       A         a PATH 8                -

f,f $.

                                        -i ' '                                             1,.
                                                                                                      '%' h* 3 :.
                                 . ff[

h ,o TA "3, 3 *g "

                                  ,?          ,.N                                                                      U:o,
                          .j 1              p                       e                                  w           o p'
                        ;fe.yf             ;/                                                            tb       D     . ;9: ..       -

f CRANE R'EACTO R CA,VITY [ h d id.P

                                        .)k
                                         .:. WALL                          \                                               -

y. \ .s' J I .J. *.;-

x.Q-

o g / ::i (.-

7.j,#

                                           '6                         BREAK g           1,                   LOCATlON         PAj'g p                             , .;.

C

                                .h. ' 'i. .                                      p
                                                                                                 . r. . ,            :
                                  .;:o w                     g         r-
                                                                                             . .s
                                                                                                       } ~eq.,.: '
                                       .;,                        - 7. py, 7 t*d / ., .g.3 '
                                                                                  -ge?.g                  .
                                                     ?%bll&.. s. %. _.;%.. :;.}l:'b ~
                                                            @                         8 R2 BREAK                                                                                                      SUMP R7 R3                      R6 FIGURE D-1 SIMPLIFIED FLOW PATTERN WITHIN CONTAINMENT                                                                              _
                                                                                             '~ ~

NUREG-0897, Revision 1 D-19 - w.- 'm, m

O o STEAM GE$ERATOR BREAK MN eF,- %** *. . t . . . .

                                                                                                           .                            REACTOR eA                                 .
                                                                                                              . 'a . . .                CONTAINMENT 7                                          .-f 3                              'v.                BUILOING t*Y                                                                  L                   ' '. g .

y :.r,,,v.:.w.4 ,

                                          ,1          IL                 JI                           ,'                                                          .

Y. '

                                                                                                   't                   0
                     .J '                                                                                 .
                                                                                                   .; n4.}'., ,N
                  . v,
                                 /                :,

A

                                                   ,                               i                                                         3 q,                                       n't 23                 3                                       .,                                                  4 4                                  .r                               'y m                            [l           \0              9
            ;        V                                                                                              h                             b
                                 l.

u Fer. .4 . wm ., 1 3.

ze n -

                         ~"
           !.          j'         p 9              t                                       'i                                        g
            $.       Esta b                      N                     k,g,,?.:i'Q"~@y.

l: ,, . lI N '

               '              45                 L','                   15 -

3, h

                                                                                                                                  ,      ,     ,j
                              /                  '

9' i j7 ya 26 L.t

                                                                  ':n .?' '              A.s .                     >:
                      \c                         #                           '              "

f[

                                           ,,                         19
                                                                                                         \m

p,

                           >:                                                                         vie n.,,Q<k<.><.v              2t -                                         \
                                                                                                         .y.>,
                                                     .s,
                                                       .                         l                            .
                                                                       . c. , , s ;u REACTOR PRESSURE STEAM GENERATOR                                                     VESSEL F                  E                   D                                         C                                  B 23                 22                             13                                     S 12                                   4                              6               8
                                                                            "                                                        7 Oc                                            ap 10                R                  3
                                                                           ^              '                                                  SUMP Ne                                            O BR AK                    Q                                A' 14 MC                                                                                    17 0                                                               16 19 26        g       27          g                             '21 G                    H                   I                                                                             J FIGURE D-2 COMPLEX FLOW NETWORK FOR CONTAINMENT                                                                                           _
                                                                                                                                                             ~

NUREG-0897, Revision 1 0-20 _

. = STEAM GEMERATOR BREAK wm

                                                    *Vo:-          ' ' *. n . e .. .
                                     .. A , ,                    i

s.

REACTOR CONTAINMENT 5

                                                                                  , ip              .                      *)

10 'l. '} ,{. g

i Oe i (' r_ y' n 2$ f.c~ I4 . ,5) 's..y ;$ k IQy' A Y- .f . h ,' .$ h

                 ,4       25           $y            I5-          $ $ -:-3,,f (" % Aj ,)l@!
                                                                                            ~

rg,4r (f. s t'- g

                                                                                                                     '7 J',
                ', g        2g         g,                .
                                                                           }                 g..              f ,jf c

4: g7 _ l]x . in ,3 (>irW4 s e.s h:@4 W: ,y

4. El w

_ e. 7 .

                                         *, 1,
                                                * ** 4 v l

.G ,3 4

                                                                                **p*'
                                                                                   .r

REACTO R PRESSUR E STEAM GENERATOR VESSEL

                                ;          O                    :               ;

12 @ 4 C '

                                                               ;               a               @                        s          EQUlVALENT
                          @                         @                                                                 ,             F LOW-PATH c         BREAK                      ';
                                                                                                                   ) SUMP l                         c i

l e i e e a FIGURE D-3 COMPLEX NETWORK WITH FLOW LOOPS IDENTIFIED , NUREG-0897, Revision 1 0-21 -

                                                                           =
                                                                               .l f// // /7'W A////// ////%
                 ///// // / // //// h                                                .
             / / // /C                               \ / /^
             /// / / / /                              V d/ /A                -

I l if'/ // / 'l /// > , h 7 h\

     // / / / / / /                                        ////s
    //   ////               /                   l/         / / // /      \

F// ' t // V //

                            //
                            //
                                       %'Q% l V             N     j' ?"

Aar f

(~

m

                 //
                 //
                            //
                            //                         n
                                                           'f'Nll l sua      fk
                 //         //                         s( ff           6g l l ll\                                   OfYAnt hi 'I- // ///\                                           w.i                   -

V/ /// / ////?u}

    'h/   ////

V/ // / /lk /4Q44d7 'y h)

h/// / " ,//) /// !/ . 1

         \/ ///// )                           //// //       //

l \/// /(~ N/ /// /////////}/ M/ / / / / / / / / / MI ////JF ~ FIGURE D-4 TWO DIMENSIONAL REPRESENTATION OF CONTAINMENT RECIRCULATION FLOW NUREG-0897, Revision 1 D-22 6-m e a

1 i l ~ - 1

                                                                             ~
                                                                                          ^                m i                                                                                                                                                                              ~

j ) -!

l l l NOTE: PHOTO COURTESY OF ALDEN RESEARCH LABORATORY, WORCESTER POLYTECHNIC INSTITUTE l l l l 1 l FIGURE D-5 HYDRAULIC MODEL OF CONTAINMENT FLOOR AREA 1 NUREG-0897, Revision 1 D-23

                                                                                                                                                     ",e -

j l

A 4 APPENDIX E MIRRORe INSULATION PERFORMANCE DURING LOC.) CONDITIONS PROVIDED BY DIAMOND POWER COMPANY A i

.~

i NUREG-0897, Revision 1 _

HDR TEST RESULT

SUMMARY

MIRRORe INSULATION PERFORMANCE DURING LOCA CONDITIONS DCN AE6609-111984-02 - Prepared ega bB : K. T. Gilbert Senior Engineer Technical Support Mirror Insulation Reviewed By: e.L.74 R. L. Patel . Senior Group Supervisor Technical Support Mirror Insulation Approved By: 4 $ B.D.Ziek Manager Mirror Engineering NUREG-0897, Revision 1 ---- - m- < - - , , . , , , e - y -.- . - g _ ,,-.,- -- y -rm -e+

e . Introduction Insulation reaction to LOCA jet forces ultimately relates to the - Emergency Core Cooling System's ability to perform its intended

function, since insulation debris has the potential to block sump screens and reduce the pump's, ability to recirculate the cooling medium. Based on postulated damage thresholds and modes, testing has been performed to determine the transport potential and resulting screen blockage patterns for components, of metallic reflective insulation (NUREG/CR-3616). The subject testing was designed to answer questions related to how reflective insulation reacts when exposed to LOCA magnitude jet forces: o How is metallic reflective insulation damaged by jet forces? o Will it be removed from the pipe or remain in place as installed? o If the insulation is torn apart by the jet forces, what sizes and shapes of debris will be generated? - l l 1 NUREG-0897, Revision 1 E-1

*M-

m b-

M The test results summarized in the following pages provide valu-able insight for understanding the fundamental questions of damage potential and mode, so that the most realistic assessment of _ screen blockage potential can be made. The test results sum-marized here are the only test results available that provide information on reflective insulation reaction to high pressure' jet forces. For more details of the test program, please refer to Diamond Power Specialty Co. report #DCN AE6609-111984-01, "HDR TEST RESULTS, MIRRORe INSULATION PERFORMANCE .DURING LOCA CONDITIONS." i l l l [ I NUREG-0897, Revision 1 E-2 O e

. O Test Facility Description A decomissioned nuclear reactor (100MW prototype BWR) has been refitted to allow full scale testing with conditions representa-

tive of comercial LWR operation. The "HDR" (paraphrased meaning "superheated steam reactor") is a real nuclear power plant which offers test capabilities relevant to modern comercial plants. The initial thermal hydraulic state for this specific HDR blowdown test was 110 bar, 318*C (saturated steam), which reflects nominal PWR operating conditions. The reactor vessel design parameters include 10m height, 2.96m I.D., 75ta3 volume, 110 bar . design pressure and 360*C design temperature. The design basis break was simulated by means of a set of rupture disks mounted in a 450m I.D. nozzle. Approximately fove feet from the nozzle end, a large deflection plate is located to guide the jet away from direct impact with the compartment walls. NUREG-0897, Revision 1 E-3 _ m

Four specimens of MIRRORe metallic reflective insulation were provided for simultaneous testing. Each of the specimens was designed and manufactured using standard practice and materials. , , All materials used were 304 stainless steel and all fasteners

(screws, buckles, pop rivets) were standard production components. The location of the test specimens relative to the test nozzle'and deflection plate are shown in Figure 1. NUREG-0897, Revision 1 E-4

                                           *+m w

                                                                                               + Y.

O Origin = Center of nozzle discharge vs O X Axis along axis of nozzle N / k !! N" l '~ ~vE-.. O Dimensions to approximate ; center of installed specimen 1 -i -i -i.@,1 ,

                                                                                ,                             'g I   -
                                                                       ,Yi
                                                                         ,               gj e -e.i
                                                                                                          /      * *

O Coordinates are (X,Y,Z) V ' '- W~ DISTANCE FROM +z +x SPECIMEN NOZZLE DISCHARGE **"za*".#-a. no NUMBER (APPROXIMATE)

          @                              2.5 FT
          @                               22 FT
          @                               11 FT
          @                               1O FT
                                                                                             -z
                                 +Y
                                                             .gia.-is.-m                         o.-e,-ie
                                                 .                          -e -a.-e ll"Df "
  -x                                                   +x       -x                                                             ,x
                             -.                                                             G
                  -e -a -e O o.-e.-io
 *O-i s.-i s.-i o                                                                           +1 View From X-z Plane or Plan View
                                 -Y Vlow From X-Y Plane or Elevation View                                                                              _

FIGURE 1;ISOMETilICLOCATIONDRAWING - NUREG-0897, Revision 1 E-5

O Test Results "Before" and "after" photographs support the following general ,

observations: o No large, flat pieces of sheet metal were, released - from insulation units. o Forces required to " tear apart" insulation units tear and deform thin gage liner material into many irregular shaped and/or small pieces. o Insulation installed farther than approximately 10 feet from the break location remained in its installed location and essentially undamaged. . o Metallic components / debris did not affect test (measurement) instrumentation or plant instrumentation. o No insulation debris was transported outside the test com-partment by either the blowdown jet forces or subsequent flow velocities. NUREG-0897, Revision 1 E-6 .

Representative photographs are included on the following pages. The arrows superimposed on the "before test" photographs represent I the best estimate of the steam jet directions relative to the

insulation specimens. P NUREG-0897, Revision 1 E-7 em i

   . - , _ , , . . . . . - . - , - . - , . - , , . , , , , , _ , - , , , , , , . - , .                             . . , _ - . _ , - _ - - . . ~ , . . - , , - - , . -

2 .. .. . ~ . . . .. - ,,. : .

                                                                                                         ~     .;       ; . ..<:.. ;;, :'?        ..                          .        :"i :                                ...m; .

Q '. . . . '. l . ' . I ' :l . '.~ , _*

                                                                                                           'hifb,.,.,'IN_l;t?5 .I                                                                                            .           .-     h, ?;
                  ;        g . ,:                                  . .                                                             . h. , . . :. . . .                        . .

.. n.:;, _ .,y 'S -

                      -, b m ..k.-~.                                                                                                                                             - - :-                                                                                          SPECIMEN 1,
                                            .; * -.'?                                                                                                                             "

4%j BEFORE TEST e ,

                                                                          '.h,                                                                                                  ,

3 i ,. _f .n arg, . J f .l . t4 ' MQe'. . i . . Figure 2 shows the

                         .5                    . 1.5 m ;. 5,
                                                                                       '                                                                                                                           "-                                                      insulation      (in       the g'4.!         ry. c c .. W                                                                        .

lower left corner)

              .?g J v % J ' < M j..rw 'g p. ; .

H f . ' :' E . %';..Pf y relative to the nozzle

                                                  - .                                                                                                                                                                                                           +
                                                                                ,                                    .         1..
                                                                             ~
                                                                                                                                                                                                                                                       -       t           and     the      deflector _              )

4 *.99 g..q . '% :,' . & *- r..' .' '. A r h .

                                                        #' .Y , 4 2                         p'                "IllWSPA' -*                      'f .                    yl           ,,

I * .

                                                                                                                                                                                                                            ' S:'            * 'l l q . .- -                                h.hW js;                             w ffd . 'C"j.. 7 - - g,- 4.# - g;,,,' .~.l . '~. /,~,.. .,':s-                                                                                                             .. .. 3, .N.&,'y,_'               y Ji.'fk .

h l

l _R . . . . l 4 ,

                                                                                                                                                                                            .*                                                      .;) -

s.:...,'.f..,. ,p -

l ,.

                                                  ~4                                                                                                     '..?>-                                                                                                                                                    .

l 4 -

                                                     ....*..,,ah
                                                                                                                                                                             ~. , .
                                                                                                                                                                                                                               .'.',.s"..    ;.i 7         .,M
           .,.N                                                                   .
                                                                                                                                                                                                                         '.                            I ' ;: -   '                                                 '
                                                                                                                        .g'l '*k " .;*. ., . . .. ,' ; -_; , z. '.Qy l
                                                   %.                .I y'
                                                                         ' .' '; _ ' -                                                                                   ,1;.                      :
              %gS                                                                        ' , :qf,.c         . ' . , . ,*,.%* . ' ... .._ .:'.

k ' t'

                                          , ' . -;    s,'-
                                                             *  . . , . , . . . . . _>.a
                                                                                                    **==

_/ " * . . ; hy . ..,

                                                                                                                                                        *(.. g i

9' $,8,.e os . . s ! . .i,; qi ;- R % s g.

                                                                              -                                                                                                       s
                                                               .?%r.    . ' -g-k.                                                                                                                                .%g:l]!k3j '

l . .,

                                                                                                                                                                                  ~,                                                                               :

_~ r. 97'

                                                                                            .                           .?...
                                                                                                                            ,                   ..                                                        w ., sp, . . . , .                                                                                        .

j '.[ /- .1.' , ' [, i. . - ; . ' .') ,' . W' ~ U : l[ .' e'.,q, ' , t . -, ;. _,.( -

                                                                                                                                                                   .a           o -
                                                                                                                                                                                                                                                       -*l 1

r

                                                                                         .s              - :
                                                                                         . f :. , v g - 4,-i,1e .*..'3,.;.,. . .. . .. . .                                                                  . .,
                                                                                                                                                                                                                                             $' . .f .. .. . , ' -
                                                                         ..                  .p        . , s3
                                                                                                                                             ',,e.           -
                                                                                                                                                                         .       .e.                             ;
                                                                                                                                                                                                                                /. -             .,9 - .:.1
                                                                                                                                                        .5 .[/ [ ' .'

-' \ , ' ' , , ' )* ..* _ .tf

                            ~
                                                                                                                                                        - '                               ,                                                  .? .h.' .' .)                                                          l

' P.,

                                                                        . 4..
                                                                                         .c. 2:                                                                                      ; $. .- sg i ' .g. ..                                                                                                         i
                                                                                    .         g           .     '.." l .. Lc . , i . ; .j
                                                                                                                    -                                                                                                                                     ~ - - -

g .'9

                                                                                                                                                                                                                       .- .f.),gy,'y j                                                                                                                        ,                                  .
                                                               .,;g     .- '&. M:..,... 7:.
                                                                                                                                                                                                                         }, . . g . ,.;. ~
                         ,.m                               -
                                                                  ,.             .                              , .r .            .r..             : . ..,.. , . . ,                                                                     . ..                                                                    _
                                                                                                                                                                                                                                                ~

FIGURE 4, SPECIMEN 1 - l

 .                                                                                                         BEFORE TEST I

NUREG-0897, Revision 1 E-8 -

  . - _ _ = _ _ _        -                   _ . - _              _ _                                                              - _ - - . _ _ _ _ _ _ _ _ _ _ _                                        _ _ . _ _

i. * 'b '

r" ld!fY.' ' u./ ..k SPECIMEN 1 , ,7;;{i / i+

q. ,

                                                                                                                    ~l                              - -

AFTER TEST 1 m;; .;., - 2 . > 2 --

. c' < .. Figures 5 and 6 show i !.8 ;f ~

3

                                                                                               -4              7. ;., , ,                        . r                              debris       from      test
                                        '% c-.                              ~
                                                                                                  '    N f'.)~ - v,f/       ,
                                                                                                                                 ~,-

specimen photos 1. indicate These that

                                                                                      . ' ' 2;-        ..

thin gauge liner _

                                                                      "' ' ' f*%'
                                                                  ,. 3 ,                            .e                  : .*%                                                      material debris is torn
                                                                  - '%              W                      -
                                                                                                                          ~

and crumpled by forces required to tear apart FIGURE 5, SPECIMEN 1 the insulation section. I.FTER TEST Note that no evidence of large, flat sheets

                                    '                                                                  -c                                                                          of      liner     material

! '. s , v ; ..+ , , . . . resulted from the test. 1 * '

                                                            . M ' '                                     .~,                                                               ~ Figure 7 demonstrates
                                            'je s

J ' the tendency of insula-

                                        -.,                                                                                                                           - .          tion sections to remain              i 1                           i                                 7                    , . ,              g                                                              ' p-intact,      even   under j                                .
                                                                                   ' g- 16     .OT                            -

severe destructive con-

.m

                                                          't -

Q .'~ ditions. No components from this section were l

                                                                ~,                                                                              --' .9                             set loose, even though A
                   ^
                                                                                                                                                                      ,            it is severely crushed and deformed.                       (

l

.1 - --

                                                                                                       , :p         ,.

effp,,e -

., n .. I FIGURE 6, SPECIMEN 1 AFTER TEST .

g- -

                                                                               /
                                                            /
                                                                  /
                                                                                                                                  ;A..-
                                                                                 ..         -\         N h/Q                                               g FIGURE 7, SPECIMEN 1 AFTER TEST NUREG-0897, Revision 1                                                   E-9                                                                                                                    ;

l y . v.' p f. l , I c.,) '

                                            '% _                                                                                              ff.                             9 j                                   g               SPECIMEN 2 t

y .

                                                                                                                                     ' gf                         9         g                                          ,

t -

                                                                                                                                                             *i                 Figure             8  shows    the     l s                                                                                                                        ,         insulation specimen in-                I stalled on the pipe                    ;

prior to the test. 1

                                            .u
                                                    '..p.Q-ki;p"fq*

A < . Details: - 4". . . - A.n "a.,c's< .. -e

s. p . y:W -

                                      .z    ; c,                                                                                                                                0.0. of insulation =24"                 '

Length of u' nit =15.75" FIGURE 8 SPECIMEN 2 Thickness of insul=3.0" i BEFORE TEST Liner material = .0025" Material = All 304 S.S. ! Distance from nozzle 4 m 22 ft. t

-e.

Figures 9 and 10 show tne test unit on the 42 h. pipe following the test W*- ,. , n' s (photos from opposite i.. s -i .. side of pipe). Note o , that the test unit has

                                                                                                                                                                         'i     been moved along the
                                                                                                                                                                    ",          pipe and has sustained a           small    amount     of i

deformation on one end

!                                                       ,                                    ,x                             4        j      .                                   disc due to the motion

g. relative to very rough

pipe surface. . FIGURE 9, SPECIMEN 2 AFTER TEST i

                                                                                                            &    T 1 -3
                                                                                                                                        ~

l

                                           , q g <. A { T

[ f p3 .Y.

                                                                                                                                                   r       -
                                                      *~

a:n k .- ' f , Ms t

                                                                              ..,        ..u j                                     EOlp*; hfb                        d, 1

J

.l                                                                                                                                                                                               .

l FIGURE 10, SPECDi:N2 -

 !                                                                                     AFTER TEST j                   NUREG-0897, Revision 1                                                      E-10

    --Am._-       _ . -a..A         ...h2-                ------a.A---                        . _ _ -            - * *              -u---is-AilAm.4                  .s_+            ---m..-- - - - _ - . - - _ _                                           . - - - - - - -                          - - - - - - - - -

j ;. - .

                                                             , g.                        _
                                                                                                                                                                .,...=,...;.-             .

f ' ,. f1k .

  ;\
                                                                 .; 1 9                                                Z 3 A . ;.v s 4,,: ,. <.                                                                                                                                                                           l

p. .

                                 &                                                                                      ?
                                                                                                                                                -e ll       ._ g T.

L

                                            .' g                                           -
                                                                                                       + MN~ ? ,-r-
                                                                                                                               - ~_
                                                                                                                                                                                  ~         : r'                                                                                  SPECIMEN 3 i                                                                                                                                                                                                                               -

l } '/ci/s/i!!!#5555 g: Figure 11 shows the i

. insulation installed.

XI- -/I!!!!f5fi~s if piii @G The insulation is fas-

!                                                                                                                                                     ff                   - 5/i5: j                                                E-                                      tened       together wi th -

l ' ff

                                                                                                                                                   'I                     -

l' E :-- buckles and screws and is supported on the edges on I-Beams.

 ;                                                                                                          FIGURE 11, SPECIMEN 3 BEFORE TEST i

i- *. "# g Details:

                                                                                       '5' 5E . .                                                                                                                                                                           Panel size: 11.7x46" i

7.i.T

                                                                                                                                                                           -.                                t~q  .

Thickness-of insul=4.0"

                                                                                                                                                                                                                                                                          Liner material =.0025"
                                                  "                                                                                                                                                 i"-              '

Material =All 304 S.S. l . . .. i

                                                                                                                    *                                      .  ,                                                                                                             Distance from l
              =

j.,, [ lh.g d :~ g; n zzle q = ft

!                                                                                              /                                                                                        -

4 Figures 12 and 13 show I'- - - the insulation sections j , after the test was performed.

                                                                                                                                                      -                                                                                                                                      No damage j                                                                                                                                                                                               9                                                                          was apparent from above the insulation.          The j

slight damage observed l

                                                                                                                                                                - '                                                                                                         from below the specimen

(Figure 13) must have l FIGURE 12, SPECIMEN 3 been caused by impact AFTER TEST from a foreign object,

                                                                                                                                                                                                                                  '...                                      since no damage was

(..d :. M; l .: . . yN

                                                                                                                                                                                                                    , s cf 'g.1                    ,,.
                                                                                                                                                                                                                                                          ,                 observed from above.

I W . ,.? W g' . ., , .;f,

                                    .y. . .. :c. , .
                                                                                              . .. # t                              ,
                                                                                                                                                                                                                                        ;;/.u 3 ; ? , . ' . L ' .M                                                                                                                                                                    . ' ' .' ' .;;
                                 ;a          ;. ;;- 4. .'.y
                                                                                           . ,, ,g ; ~         .
                                                                                                                                                                                                                  . y y$;y.

t., .y , . ed 6 @. ',1f . %*,

                                                                                                                          ~
                                                                                                                                                         ,*                                              [j,.[.f I f: t' i .% ~ ' '

. . .q. ' -,. . : . :

h. ..- . . D ': *- : w..:! '

                                                                                                                                                                                                                                           .. D.
                                                                                                                                                                                       ./3, + x.?,

m j , 1. ; . .n.

                                             ._                                ; . c z.;*
                                                                             * :: , n*4 ;, ,. . . if 'L y * '.

sL '.. . . . ,?

W -

3Jy.;t . :..:^ % .v .;.; ; . . :- .q . ;-- : ,o -Y ... . . .:

                                            .-..49 q:q  . ..-

n g y - +e

                                                                                                                         ..-                                                                           i ( - - ,7 3 ,. -
                               % . '",                                                                       p.                                     -;,
                               ;j-                                   **
                                                                 . . . ,       L.,['-,q    4 ,,.                       ,l '                                                                       ,, .;l ~
                                                                                                                                                                                                                           ..,.; 7
                                                                                                                                                                                                 . . , ..,;.9;_
                                                    .                    . . . . , . - + ; * 'g,i., .; y.
                                                                                                                                                               < .4    .            . v,                          .

y - %,u. - ,. .. ;;,_. -

                                 <: -      - - .                                                                                                         .         .          .     + . %. . ., ~                                         .      .z - -

FIGURE 13, SPECIMEN 3 - AFTER TEST _ NUREG-0897, Revision 1 E-ll

SPECIMEN 4

' ~

o P lt >

                                                                                                                       '    Figure 14 shows the                       l
                                                                      [,Q                            .i         L         U-shaped box insulation
                                                ~r        __              --

J installed on a tee.

                                         %sr                                  0                                  -

Details g .

                                                                                                       --                   Length of unit =16.0" Thickness of insul=2.0"                  !

FIGURE 14, SPECIMEN 4 Diameter of circular i BEFORE TEST section = 12.0" f ner inaterial = All

                                 ;"c-1           -7                                  h Distance from
                     - g'    .
                                     '3 nozzle 9. m 10 ft, d                                                                                                  Figures 15 and 16 show                  ,

the test unit after the M test was performed. Damage was confined to I g.W - g, Q-F ,J, local areas around the 3 .. . disc (vertical g

                                                     'g-y                            ,

o end surfaces). Damage shown

                                                         , # 't .                                                           in the lower photograph
              .1                                                                .
              '-                                                    'I g             -

is believed due to

                                                              '                                                          .,                    om a foreign         i f(                 ' b-                               8'
                                                                          *g                         ,

AFTER TEST t

            .          s       )

y

        ' 8
              /

7 , l n ~

                                                 /. '

e6.y" N !

4 e

                               . Wk W .'                          .   .

l FIGURE 16, SPECIMEN 4 i AFTER TEST _ L-_- MLLREG-0_897, Revision 1 E-12

APPENDIX F HDR BLOWDOWN TESTS WITH NUK0N INSULATION BLANKETS an-PROVIDED BY 0 WENS-CORNING FIBERGLAS CORPORATION NUREG-0897, Revision 1 O

*6

n .

Test Report: HDR Blowdown Tests - With NUKON Insulation Blankets Gordon H. Hart Insulation Technology Laboratory Owens-Corning Fiberglas Corporation Research and Development Division - Granville, Ohio Revisea: March 27, 1985

                                                                                ~
                                                                    ~

NUREG-0697, Revision 1 .. _ e

__ =_ _ .. . . _ _ _ _ _ _ . _..__ ____ _ - _ _ _ _ . _ __ . _ - _ _ _ . HDR Blowdown Tests

!.                                                                                                                  Ow:;ns-Corning Fibsrglas March 27, 1985 SUM 4ARY:

4 This report sumarizes the results of the two tests conducted this past sumer at the HDR facility in West Germany. For Owens-Corning, the objective of these tests was to determine the capability of the NUK0N nuclear containment area insulatiort to withstand the impact of a high pressure steam-water blast and to determine the size distribution of the fibrous insulation debris resulting from the impacted blankets. The report sumarizes the test , procedure and the results; it contains, in addition, "before" and "after" photographs and weight tables of the various components. , In short, the tests demonstrated that unjacketed NUKON blankets, or NUKON blankets covered in a metal mesh, that are located within nine pipe diameters of the simulated pipe break, can be totally destroyed but may not be, i. depending on the orientation (i.e., over 90 percent of the wool insulation was reduced to fine fibers). However, NUKON blankets enclosed in the standard NUKON 22 gage stainless steel jackets vithstood the blast to such an extent that less than 50 percent of the metal jacketed wool insulation was reduced to fine fibers (for pipe insulation within seven pipe diameters from the simulated pipe break). These test results are unique to NUKON insulation i systems since they likely depend on type and thickness of stainless steel 2 jackets, strength of jacket latches, type of insulation wool, type of fabric covers, strength of fabric to fabric seams, strength of the Velcro joints, and strength of the Velcro to fabric seams. Further, it should be emphasized that i the success of the metal jacketed NUKON pipe insulation in resisting the blast 4 constitutes only two data points. These should not be used as' points of extrapolation to cover different materials or conditions. While it is ~ reasonable to assume that a flat NUKON blanket, covered with 22 gage stainless l steel jacketing, would also resist damage by a water-steam jet blast, no actual data for this configuration was obtained in these tests. ANALYSIS OF THESE RESU_LTS: A. Overview: l l Attachments 1 and 7 show the layouts of the nozzle, the impingement plate, and the support strut for both Tests No.1 and 2. The center of

, the impingement plate was positioned 1.5 m from the burst plate of the
nozzle. The impingement plate was a 2.6 m diameter disk with its center 4

positioned 2.0 m from the ceiling, or its upper edge 0.85 m from the ceiling (note that the plate and the ceiling are not perpendicular). l Set-up for Test No. 1 (conducted on June 15): } A single blanket of NUKON pipe insulation (measuring 870 m long, 50 m , thick, and of adequate stretch-out .o cover a 100 m by 120 m ; i rectangular bar that was used to simulate a pipe) was placed on one of i i three rectangular steel struts. See Attachment 2 for positioning of the pipe blanket relative to the 450 m inner diameter nozzle. This blanket i} was left unjacketed. The center of this blanket was located within a I 350m cone of the nozzle, representing 0.8 pipe diameters (0.8 D). j However, as it was likely for this blanket to be hit by water reflected

                                                                                                   '~ ~

} NUREG-0897, Revision 1 p.j - i

       , , - - - - -                     --          . . - _ _ -                       -.               - - - - . - - - - - - - - - , - - - - - ~ - - - -

HDR Blowdown Tests Ow:ns-Corning Fibsrglas March 27, 1985 from the impingement plate, the reflected distance was 1850m, or 4.10 (to the center of the blanket in both cases). Two flat blankets, each measuring 500 m by 750 m by 50 m (thick), were attached to the ceiling, directly above the axis of the jet nozzle; see Attachments IA and 18. This plate was oriented perpendicular to the axis of the nozzle. See Attachment 3 for photographs of the flat blankets before the test. Attachments 1A and 1B show the position of the nozzle, the insulated bar, the impingement plate, and other support elements. The impingement plate was positioned 1.5 m from the burst plate of the nozzle with the insulated strut extending between the two . and slightly below the center axis of the nozzle. The flat blankets were not located within a "90 degree cone" extending out from the center nozzle. Therefore, for the purposes of impact, their distance of the,he from t nozzle, was calculated to be 3320m or 7.4 0 pipe diameters, assuming that they would be impacted by water from the impingement plate. See the NRC report, " Methodology for Evaluation of Insulation Debris Effects", pp. 22-26, for an explanation of the 90-degree cone extending out from the nozzle. The blankets were attached to the ceiling with Velcro strips and pins with speed washers (with the pins imbedded into the concrete ceiling). Results for Test No. 1: Both types of unjacketed NUKON blankets were badly destroyed by the jet blast which originated from water-steam heated to 310 degrees C and 110 bar. The flat insulation was totally destroyed, with only pieces of cloth, which were caught on pipe supports, able to be retrieved. The pipe insulation was largely destroyed although 15 percent of the wool was left intact, enclosed in the fiberglass fabric. All pieces were located around the test room but none in original positions. See Attachments 4 and 5 for photographs of the resulting fabric material. These results are inconsistent since material located 7.4D from the nozzle was totally destroyed while other material, located 4.1D from the nozzle, was left 15 percent intact. Set-up for Test No. 2: (Conducted on July 4) For the second test, the impingement plate was angled about 30 degrees from the center line axis. With this orientation, a greater strut length was available for insulating. Therefore, two pieces of pipe insulation blankets, each of the same size as for Test No. 1, were able to be placed on the strut that was on the side impacted by water reflecting off the angled impingement plate. Each pipe blanket was covered with standard NUKON 22 gage stainless steel jacketing. These shall be referred to as Pipe Blankets A and B. See Attachments 7A, and 7B, and 8, lower photograph. The flat blankets were positioned exactly as for Test No.1. This time, however, they were covered with metal scrim jacketing. See Attachment F, upper photograph. The test conditions for Test No. 2 were about the same as for Test No. 1: 310 degrees C, 110 bar.

NUREG-0897, Revision 1 F-2 _

    ,     .                                                                         HDR Blowdown Tests
  ,                                                                                 Owens-Corning Fiberglas March 27, 1985 Results for Test No. 2:

The flat blankets, located 7.4 D from the nozzle, were totally destroyed by the blast, as in . Test No.1. The fabric .and the metal scrim were again strewn about and caught on components in the test chamber. Of the two pipe . blankets, the one closest to the nozzle (Pipe Blanket A) was just slightly damaged, retaining 93 percent of its original weight of wool (the end of the blanket was slightly torn up). Its center was located 125m, or 0.3D, from the nozzle itself, although it is likely _ that reflected water-steam had the greatest impact. For the reflected case, the distance was about 2830mm, or 6.30. Blanket A remained in its original position. The one closest to the impingement plate (Pipe - Blanket B) was partially damaged, retaining 25 percent of its original 4 weight of wool. Its metal jacketing, badly deformed, remained. on the bar as did the piece of blanket that contained the wool. This latter blanket had its center located 1350m , or 3.00 from the nozzle, although for reflected water-steam, its distance was 1930m, or 4.30. See Attachments 9 and 10 for photographs of this pipe insulation after the blast. Attachment 11 shows photographs of the metal jacketing for Pipe Blanket B. A more careful examination of this jacket leads to several

;                         conclusions. The majority of the jacket damage can be attributed to the water-steam pressure. The rivets for the latches all appear to be blown radially straight out by an internal pressure. Most cracks in the steel

,i had been initiated frcm an internal pressure pushing out. The fracture shown in Attachment 11 occurred along the initial bend of the rectangular jacket; apparently, internal water pressure ripped the metal 1{ Jacket where the added stress of the bend caused a weak spot. There was - ] some question as to whether or not the burst plate damaged the steel jacketing. Two cracks in the jacket showed very clean edges and

evidence of abrasion. It is quite possible that they were caused by the flying burst plate. Dents and cracks in Attachment 12 strengthen this conclusion.

B. Summary of the Tests: Table 1 gives a sumary of the weights of the blankets both before and after each test. The flat blankets are, of course, separated from the pipe blankets on this weight table. The Velcro, used to attach the pipe blankets and sewn onto the fabric, is treated as part of the fabric. i The weights of the NUKON base wool are separated since the wool, not the fabric, is considered to pose the greater sump blockage potential and hence its fate was of most interest in this testing. In Table 1, what i is of greatest significance is the difference between the results of Test No. 1 and Test No. 2. By metal jacketing the pipe insulation, the percentage of pipe wool reduced to debris was dropped from 85 percent to

41 percent. This is significant because 't demonstrates the i effectiveness of metal jacketing in protecting the blankets. It also

, demonstrates that a substantial portion of the wool insulation, that started within seven pipe diameters (70) of the break, was not reduced to fine fibers. This contradicts the assumption made by the NRC that all fibrous insulation located within 70 of a break would be reduced to fine fibers by a blast. NUREG-0897, Revision 1 F-3 - - _ e,n . , -.

                                                                           -                  .~.,.    -
                                                                                                           -....m

HDR Blowdown Tests Ow;ns-Corn,ing Fibsrglas March 27, 1985 On the other hand, the flat blankets, placed on the ceiling directly above the impingement plate and at 7.4D of the break, were totally reduced to fibrous debris. The addition of the metal mesh jacketing apparently had no effect whatever in protecting the blanket. On this basis, if calculations show there is a need to reduce the sump blockage potential, it is recommended that flat, or nearly flat, blankets placed on steam generators be covered by stainless steel jacketing, not by mesh. However, it would be advisable to obtain actual test data on flat, metal jacketed blankets subjected to a blast. C. Previous det Impingement Tests: The NRC assumption, that all blankets within 70 of a break will be reduced to loose fiber debris, is a rational one. It is based on " jet impingement" tests conducted in 1982 and 1983 at the Alden Research Laboratories (ARL). These tests demonstrated that, in the worst case, blankets made of fibrous insulation will be torn apart by dynamic water pressures of 20 psig or greater when located within a "90 degree cone". Using this pressure, the NRC backed out the "7D" assumption. On this basis, the assumption is rational. However, the ARL tests were performed using cold water emerging from a two-inch diameter nozzle. In an actual two-phase blast, such as would occur in a pressurized water reactor containment area accident, the water-steam jet would have less momentum at 20 psig than a cold water jet, hence it would have less destructive potential. Also, because it could not constitute a defined jet, it would likely have less destructive potential. However, the NRC was justified in using the ARL data because it was the only data available at that time. The two HDR tests, showing that metal jacketing can be used to protect fibrous insulation, really only constitutes a single data " point". That data point should not be extrapolated in other directions to predict the behavior of other types of wool, fabric, stitching, metal jacketing, latches, or insulation system designs. A variation in any of these variables could have had a profound effect on the results presented in ~ the two OC Itests conducted at HDR. D. Issue of Size Distribution of Fibrous Debris: One of the original objectives of this testing was to obtain a size distribution of the fibrous debris. This distribution, it was reasoned, could then be used with ARL water transport data to calculate the quantity of debris that could reach a sump screen in a specific plant sump analysis. However, such a size distribution could not be obtained. The wool that was torn from the blankets was not able to be found and, hence, was assumed to be entirely reduced to lcose fibers. All the woel retrieved was still enclosed in fiberglass fabric; hence, its size distribution was not an issue (i.e., enclosed in fabric, it would not transport to the sump screen as loose fibers). Therefore, the results of the test were binary: wool that remained enclosed in the fabric was not transportable, while wool that was torn from the fabric enclosure was reduced to loose fibers. NUREG-0897, Revision 1 F-4

HDR Blowdown TGsts . Ow:ns-Corning Fiberglas March 27, 1985 E. Conclusions and Recomendations: l From the HOR Blowdown Tests No. I and 2 on NUKON insulation blankets, several conclusions can be drawn:

1. Unjacleted blankets, and those jacketed in metal mesh, located about 7.4 pipe diameters from the jet nozzle, were reduced to shredded fabric and unretrievable loose insulation fibers. Most of the fabric cenerated by the tests was caught on protrusions in the area.

The wool not retrieved was assumed to be reduced to loose fibers. , On the other hand, unjacketed. pipe insulation, located within 0.8D of the nozzle, was only 85 percent destroyed.

2. Some of the 22 gage metal pipe jacketing in Test No. 2 was badly bent by the blast; however, it was not torn away from its position around the strut it had covered. The suggestion is that the reflected jet, rather than the primary jet, inflicted the greatest damage.

3. The use of metal jacketing over pipe blankets was effective in reducing the level of wool destruction from 85 percent (Test No.1) to 41 percent (Test No. 2), or 75 percent for Pipe Blanket B and 7 percent for Pipe Blanket A.

It is recomended that for sump analyses involving pipes insulated with metal jacketed NUKON blankets, Attachment 13 replace Figure 3.26 in the NRC report, NUREG-0897. The curves on these graphs have been redrawn by using data collected in these tests. It is also recomended that, if possible, further testing be conducted. This would include metal jacketed flat NUKON blankets and insulation samples placed at various other positions and orientations. Ideally, the impingement plate should be removed and insulation samples should be impacted by the primary jet, not only by reflected water. Finally, it is recomeded that these results not be extended to insulation materials fabricated with different gage metal jacketing, matal latches, compressibility of insulation, etc. Variations could have a profound effect on the results. Also, caution should be urged on extrapolating these results to so-called " encapsulated" insulation since that is not a clearly defined type of insulation and since its behavior could be significantly different. NUREG-0897, Revision 1 F-5 . r . _ _ - _ - - - - - - -

HDR Blowd:wn Tssts Ow;ns-Corning Fibstglas March 27, 1985 TEST #1 Original Weights (kg)* Piece of NUK0N Cloth Scrim Velcro Wool Total

1. Pipe 1.03 0.09 0.025 1.50 2.66

2. Flat A 0.54 0.05 ----

0.70 1.27

3. Flat B 0.54 0.05 ----

0.70 1.29 Comparative Weights from Test #1 (kg) ' Piece Cloth & Velcro Wool j Before* After % Lost Before' XTier % Lost

1. Pipe 1.05 0.83 21% 1.50 0.22 85%

2. Flat A 0.54 0.51 12," 0.70 0 100%

1

3. Flat B 0.54 0.44 0.70 0 100%

TEST #2 Original Weights (kg)*

,             NUKON Blanket            Cloth          Scrim           Velcro               Wool               Total

1. Pipe A 1.03 0.09 0.025 1.50 2.66

2. Pipe B 1.03 0.09 0.025 1.50 2.66

3. Flat A 0.54 0.05 ----

0.70 1.30

4. Flat B 0.54 0.05 ---- 0.70 1.29 Ccmparative Weights from Test #2 (kg)

NUKON Blanket Cloth & Velcro Wool Before* After % Lost Before* 77ter % Lost

1. Pipe A 1.05 1.03 1.50 1,39 7%

2. Pipe B ) 80% ) 41,*

i 1.05 0.64 1.50 0.38 75%

3. Flat A 0.54 0.70 0 100%

4. Flat B 0.54 ) 0.27 ----

0.70 0 100%

5. Unidentifiable ---- 0.49 ---- ---- ---- ----

Fabric Scraps l 6. All Blankets 3.18 2.43 24% ---- ---- ----

Based on average values of the weights of the materials for six different blankets constructed for the tests.

I

NUREG-0897, Revision 1 F-6

E

o m AYAcJeMagt- ta

      'P                                                                                           c ; t <_ ,     r. O g                                                                                             B.stor view a

g' ATT ACM MENT g

     <                enst +o i    w
    ^

A. Elll> VIE. VJ \ s i y i' " " ' " ' " ' FL AT >JUKo ej pdMFET' W MM 4150o My o

s g 'N y

                                                                                                      #       #"E                uc;.ut Aitt> :: rear
                                   '                                    *J3;5*
   -'                                                       g+                      ..
                                              ,e e,                                      -

w _ 1

c

                                                                                    -                                    ' 450 su i.o.
!                                                                                                                              WOzzLE w'TH 8                                     -

DUST PL AT E.

                                                                                  ,### g S m.
                 )              o~

j

   ?
   ~           /

4

                                    \    us g'                                  J-
                                                                                                        -=
                                                                                                  ,. % o cone s n ot ,ek o nt .'n A n L
             \

_t\ R - - MUKow P4PE

                                                                                                '             'MPIN 'f MEIIY I'LAIC
                      ' /             S-INsd ATiou GJ A IOO MM ,
                                                                                            /

C id l 12o Mu mg A j i

                           .IX SSCT IC'J A-A l

I l 1 I ' - 3 .

Attachment 2 l TR 35947 bg \'

                            ,.;.                 e. . .

Mx - if fp : . .

                        .        l-d i

l K u 4 , e 8- - hh,vi

                                                                    ?                                                    ,t                    _ :?

t'- t af

                                                                                                                                                 + .7        is.-

eW

                                                                                                     ~
                                                                                                                  . , s. t.                    x y
                                                                                      . . , ,   .i*P*t*                                      -

y

                                                                                                                               ".i . ,, _ ,r l , :,
                                                                                                                                                               . 73

e .I (

INSTALLED INSULATION BEFORE TEST 1 . l NUREG-0897, Revision 1 F-8

1 Attachment 3 TR 33947

                                                                           . y. -
                                                                           ,s          q-t
                                                                                  > 'x y . ., g                   , :. ,

t . E -

                                                                                                    ~
                                                                                                                            ,i r.h,
                                       &                                            k ,s                                        '

l

                                                            .          . , , . . .                                                                                                                            l
                                        ,-         .ef$ g ' , -      v   2'

W I'i_kL*

l i l l l e l_. :k o., F',.Q y- M r gf U1 - _ 9., w 7 l _,,.7 m t. ' ; r .- ~ .,A.,.J.1

                   .~~.

ll.QCyM g,. ;;g h y

                                              ' , O. y
                                             .13,9
                                               . .u -. m,               . , ..s_
                                                                                ,y  .,

3

                                                                                                   ;aj
                                                                                                 . , ,y
                          . ;g -

7;;;;*

                                           .. l t >.                   %* ; . . t.f a    ..,
                                                                                              ..+<

y p ,;.c t . , ; /. . s 3. n , ; <.. .- ,c u 4f-};;Cs. , L T- , ,> ','rc.*

                                                       ,.       .,.,....w l                                               !,'.y W t%v.t  ,        ;

i 1 1 1 l l INSTALLATION OF FLAT BLANKETS BEFORE TEST 1 ,

~ ~

NUREG-0897, Revision 1 F-9 _ 1 1

r . l Attachment 4 TR 35947 Te l l

                                                                                                                                                                #f-_

i t i._ u. n :t , i

                                                                                                                                               .4 : -
        ~ , "we < ~,,., ,, ,. ., , ,.
                                                     .a.c.      ,.                                            ,
                                                                                                                   *(:~ ty :                                  C 4114                              .

i

                                                                                                                   . l$'&
                                                                                                                      ^

I l 7.. * 'M Mghm 3 r e -.seMeduxqu nu q.2 7 r.an l

                                                                                  - c     n
                                                                                                                                      . n c z.p q y w b .!
                                                                                              -                              ,, [N...k...$i   .

[& i ' .' . icnfqp 3.s4.-f,:, w~.$ Ei, :.. k,

                                                                                                                    !*,ly:4 jh?"'"

l

                                                                     ,- ? ,               %,9   H                      @,%;.%*Xn' .                                                        l
                                                                                ,          pt                     . ,s > ;Da                                                              i l                                                             lpll              .;                   ,.           ., g i                                                                   1 jl-               i                          ;       ei
                      , ;?
                            ;;.]$ h' p,
                                                                            '1                   ,       ;-
                         ; L .. . y:.

4 3.[j ff , j

,..nSy;J. /;ct.' &,

                       ; .cwyn                             -
                                                                   '\                           ,.

4 s

V'-gt .g;;p . . . -F 4 4 lpg ,.

                                                                                                       -       t
                              -sp , ,,. %1[.                            '

, ' " If,$56 t 3, . . u .aw.

                                .gre;! anA                                s                                                                                                               l l

l RESULTS OF TEST 1 __ NUREG-0897, Revision 1 F-10 .

Attachment 5 l TR 35947 l M *C - - . - .- hp:W;g?b ;n - . s.%. .w, , =: .-x,~ .

                                                                                                ..                   x.                                 ,

, ; ;. y ~ >..:.: # - - 77 ,,4 Lv.4 3 s r.i e 4 s . ". ..- 4 i hW, !% < :~1': ,: -

                                               ~ - ~                                   , - ,, #Yo. ..c           ; f- :,..
                      ,      r, ta; ..<                                                 dVy' . w' gg. ::                                                                 ~9
                       ..                             ,[ . ...
                                                                                   .(. . . . ,.~
                                                                          .N
                                                                                      .p s

t, -

                       .,             - ;.- - ':                       A. q:

s l_ \ an ; ,- . . . . ~ -, s i ava,. s:-?.% g.::.;tt%l . l} -). T. :.

                                                                                                                                          ~
                         . ,+ ':~ ' ;, : J. . e -a; '-lff },:)

j .;'i :

                                                                    - >({ V.
                                                   ,                  I            ,
                                                                 /. ,

4

                                     - .               ,    ;;, 4,r,-                                             e
                           ~.a~,.&,.             .
                                                      <~             <

1 l i RESULTS OF TEST 1 _

~ ~

NUREG-0897, Revision 1 F-ll _-

Attadhment 8 l TR 35947 m M. "~ :~ ,, l <..

                                                                               .,7               . , Y','~7 a ._4        -

l p es *! " ' . -9 . l

                                                                                                         ,k
                                                                            -                           ;                                     -4           ,
                                                          %    &            h Tk;.                             .

g

                                                                                                                                                -m         1.
                                                                                                        <;w ,.. .. ,                                 .   .
                                                                                                           .::mL,::..;-

v. u-:_;> ; Y:,.;::i.r%:

1 1

                                                                                                    ?p': o.w.g. g. y;,,                                                                                   l Lw :.: u i
                                                                                                          %** W.'C.

V G

                                                                                   , , _                               t l

f ., .l

                                                                ! l,,..j;.                                                                  '

3 w;

                                               %                             t s    - l.'

37 . d'; IL x 0 - t l ' I

                                                                                           -                                                                                                             r 3

r

 !                                                                                                                                                                                                       i INSTALLED INSULAT'04 BEFORE TEST 2 NUREG-0897, Revision 1                                            F-12                                       -

4

Attach:nent 9 TR 35947

                                    . . _ ,             .r n                                                                                                 ---'-m
                                                                                                                                                                                                         .(-
                                                                                                                                                                                                          .. . d.$ .
                      ..     .c
                                                                                                                                          #                                         j    -p:& yp ^wy. qJ .'.',              ,

rd'y- 'a , ,- l

                                                                                                   "                                                            ,.s c.

vt q.., . .. . . ,. _ .A . . ,.

                                                                                                                                                     . -.g .                                                            .,

c. s .

                              < c.                .--                                                                                              ,                                   .                   . . ..

FA: p. 3 3....; . o . c.p .

                                                                                                                                                                                 ,S,:y

g . rt.

                                    , u..                                                        .                                 s                       .
                                                                                                                                                                                         . t ae q                       '                                     -
                                                                                          .                           .                                                         /.q-{:,/;s         .'

p.. '% < . . . s v

                                                                                                                                     +
                                                                                                                                                           .s                        -
                                                                                                                         %<r 1                              , ~, .' v e . &                       <
                                                                                            /                                                            .t
                                                                                                                                                                .L.~

g;;s,.<. y. .. .

                                                                                                                         .,!).Q;' 4 a f5
                                                                                                                         / . r,r - , ,- J ya t
                                                           ../:s                 .

y

                                                                           ..y                                                                                                     ;a
                                                            ;.,,;:+,:v

. u ; ~, m . p ,<. r.w: , i. .,;

a

                                                                                                                                                                , - n .;z.         .*
                                                                - y. ::};f w,.~.c,e,.                                    %:;:.<D,r-               y , %r
                                                                              < 4.; 7 p 7p l

l ff' '

                                                                                             . .                      ^ W }-                                .
                                                                                                                                                                                          -                i

m. . . . ; f. , . '

                                                                                                                                                           ~ .. ..                             u.,       w Q                                                                8 "

l'. * (

                                         . . t , %'s,[         ,

2-v ?' 3

                                              .ge '

g , .

                                                                                                                 , .e .           -3 r               ,e                                                         .A wm j

paps, ', .1g ' {- , , ,L, - */ ' ' y' g/em a "3 { 3.-

                                                        -                                                , . . -                                                                                                      . . v.
,                                                                                                          g;            i
                                                                                                                                                                                                                      -. A..e ' y: 5 9
                                                                                                         .q            -

v, b. ;, ,

                           .                                                                                                                                                                                                7 4 , ,., .

m .

                                                                                                                             .h                                                                                            *i .e . .. y,
                                                                                                                                                                                                                              ,_r er . .:

a, ;

                                                      ,                       ,                                                                  c :c;;> . s                                     ,
                                                                    ..s                                                                                n c
                                                                      .." ~ . . ..$ .).
                                                                                                                 .         \::                          3
                                                                                                                                                                                                --                                                                          1
                 ,            d' Y                                   ) % 'a,n = e .                                 ,? .

r

                                                                                                                                                   'g'--

ns ;,'~~ f & . 7_ l l RESULTS OF TEST 2 i

                                                                                                                                                                     .                                                                                                     I NUREG-0897, Revision 1                                                                     F-13                                                                                                                                                       -

l i

Attachment 10 TR 35947

d __

k c, . ; .; *

A . . _ .

                                                                                       ' s.

p

                                    ,       .                 t        ;           ..
                                                                                                                                                   ~
                                                                                                                            ~~

i .

                                                                                    ,                  .       a ..

6

                                                                                 ,' 4, t

f6 .' . v.

                         ;$              i .                          .

P o

                                        ..d                                                      ,
                            .%P4
                                        #gp                      r 7

j RES!JLT3 0F TEST 2 ^ l NUREG-0897, Revision 1 F-14 -

                                                                                                                                                                           ~
                                                                                                                                                                                         \

l i

       ,     ~

Attachment 11 TR 35947 l

l

!                                                                                                                                                                                               1 i

l l i . A ' 7 k i , . j . . sin g ;; W

                                                              < e. '            1 . . , , . _:.                    . .;.;.yar.,,.:,.g..vg;qp+6 g ,

JACKET 'IEAREST THE IMPINGEMETT PLATE i l + 4 y I ;, l 4 . l l 4 l l

FRACTURE CAUSED BY '4ATER PRESSURE I STUDY OF STAI'ILESS STEEL JACKET FAILURE NUREG-0897, Revision 1 F-15

Attachment 12 TR 35947

                                                                                            .            e m,,
                                                                                  ~

l .. 4 '. [Yfl A31-pyre ~ ' f,

f. _

k.t j.r,~ A

                                     . . Js u ~ '

4

                                                                                                                         .;c k SUSPECTED BURST PLATE DAMAGE M
                     .g y(

y 4

                       -l%,                                 f;.:' i:y                                               t
                        = ^

q:,g'. _ Ash _. \'# lll

                   , e ,: '. . $
          . , z. . .                          -. , . , * .. .

v'

y ' c' -Mq me,4;S,.2. , p,. .,,

s

                                                                                                        - -cw 3<     .,

r ' X

                                 ,, . ' b' i                                           e pib 41' _  -

t l S l INITIAL BURST PLATE DAMAGE PROPAGATED BY INTERNAL WATER PRESSURE STUDY OF STAINLESS STEEL JACKET ?AILURE NUREG-0897, Revision 1 F-16 ,

l

                                                                                                                                                                    . . .: 4 b               4 07/24/85 EXCLOSURE 4 DRA?T 3                                                                                               .

PROPOS3D RG 1.82, R3VISIOX 1 (Comparative Tert) 1 Ref: USI A-43 Resolution i ( l

l I e l l l

                                                                                                                                                                .Jr 1                                                                              DRAFT 3 2                                                                           Division 1 3                                                                        Task MS 203-4 July 24, 1985

4

Contact:

W. E. Campbell, Jr. (301) 443-7856, MS 212 NL 5 PROPOSED REVISION 1 TO REGULATORY GUIDE 1.82 6 WATER SOURCES 7 [SWMPS] FOR [EMER6ENEY-60RE-600 tin 6 8 ANB-EONTAINMENT-SPRAY-SYSTEMS] 9 LONG TERM RECIRCULATION 10 COOLING FOLLOWING A LO55 OF 11 COOLANT ACCIDENTI 12 A. INTRODUCTION 13 General Design Criteria 35, " Emergency Core Cooling," 36, " Inspection 14 of Emergency Core Cooling System," 37, " Testing of Emergency Core Cooling 15 System," 38, " Containment Heat Removal," 39, " Inspection of Containment Heat 16 Removal System," and 40, " Testing of Containment Heat Removal System," of 17 Appendix A, " General Design Criteria for Nuclear Power Plants," to 10 CFR 18 Part 50, " Domestic Licensing of Production and Utilization Facilities," require 19 that [a] systems be provided to [ remove-the-heat-reieased-to-the-containment] 20 perforn specific functions; e.g., emergency core cooling, containment heat 21 removal and containment atmosphere clean up following a postulated design 22 basis accident [(BBA3-and-that-this-system]; These systems must be designed 23 to permit appropriate periodic inspection and testing to ensure [its] their 24 integrity, [ capability,] and operability. General Design Criterion 1, " Quality 25 Standards and Records," of Appendix A to 10 CFR Part 50 requires that struc-26 tures, systems, and components important to safety be designed, fabricated, 27 erected, and tested to quality standards commensurate with the importance of 28 the safety function to be performed. This guide describes a method acceptable 29 to the NRC staff for implementing these requirements with respect to the 30 sumps / pools performing the functions of water source for the emergency core 31 cooling, containment heat removal, or containment atmosphere clean up [and 32 containment-spray-systems]. This guide applies to light-water-cooled reactors. 33 34 2 Comparative text based on "For Comment" version published May 1983. 35 Proposed deletions from it are shown in bracket with overstrike ([-]) and 36 additions to it are shown underscored ( ). Tables and figures are not 37 shown " comparative". " Active" revision was published June 1974 .

Page datas are: 05/06/85 4, 9, 20, 22-30; 05/10/85 21; 05/31/85 2, 3, 6, 16-18; 06/04/85 5, 7, 8, 10, 14, 19; 07/23/85 12, 13, 15, 31; 07/24/85 1, 11, 32.

t

1 Any guidance in this document related to information collection 2 activities has been cleared under OMB Clearance No. 3150-0011. 3 B. DISCUSSION 4 [Samps-or pomp-intakes-serve-the-emergency-core-eeeling-system-fEEES) 5 and-the-containment spray-system-(ESS)-by providing-for-the-collection-of 6 reacter-coolant-and-chemically-reactive-spray-seintion-and-aif ewing-its-recir-7 eniation-for-additional-cooiing-and-fission predact-removal. 8 Pi ac eme nt- o f- the-EEES- s amp s - at- the-i ewe s t-l ev ei-pr a c ti c a4 - ens o re s -maxi mum 9 utilization-of-available-recircalation-coolant --However;-there-may-be pieces 10 within-centainment-where-ecoisnt-cenid accamaiste-during-the-containment-spray 11 peried; providing-these-areas-with-drains er-fiow paths-to-the-samp-iocation 12 wiii-minimize-cociant-heidop --This guide-dees-not-address-the-des +gn-of 13 such-drain paths:--However;-since-debris generated-by-a-iess-of-cociant-seefdent 14 (t0EA3-can-migrate-to-the samp-via-these pathways;-these-drains-are-best 15 terminated-in-a-manner-that-wiii prevent-ceeris-frem-being-transported-te 16 and-accomniating-en-the-EEES-samp:--Appendix-A-addresses-cencerns-related-to 17 debris-transport-and-the-effects-of-attendant-samp-screen-bieckage. 18 Smaii-drainage-samps-that-are-used-to-celiect-and-meniter-normai-ieskage 19 fiew-for-ieakage-detection-systems-within-containment-are-separate-frem-the 20 EEES-samp-and-are-at-a-iewer-elevation-that-the-EEES-samp-to-minimize-inadver-21 tent-spiilever-inte-the-EEES-samp-doe-to-minor-leaks-er-spiiis-within-the 22 c ontai nme nt:-- The- fi ce r- adj ac e nt- to- the- E E ES- s amp-we ni d- no rmaiiy- s i e p e- down-23 ward;-away-from-the-EEES-samp-teward-the-drainage-celieetien-samps:--This 24 downward-siepe-away-from-the-EEES-samp-wiii-minimize-the-coilectien-of 25 dsbris-against-the-samp-screens]. 26 B.1 Pressurized Water Reactors 27 In pressurized water reactors (PWRs), the containment emergency sumps 28 provide for the collection of reactor coolant and chemically reactive spray 29 solutions following a los',-of-coolant accident (LOCA); thus the sumps serve 30 as water sources to effec t long term recirculation for the functions of resi-31 dual heat removal, emergoney core cooling and containment atmosphere cleanup. 32 These water sources, the related pump inlets and the piping between the sources 33 and inlets are important safety components. The sumps servicing the emergency 1.82/2 05/31/85

1 core cooling systems (ECCS) and the containment spray systems (CSS) are 2 hereinafter referred te in this guide as ECC sumps. Features and relation 3 ships of the ECC sumps pertinent to this guide are shown in Figure 1. 4 The primary areas of safety concern regarding ECC sumps and pumps inlets 5 are: (a) post LOCA hydraulic effects, particularly air ingestion, (b) block-6 age of debris interceptors resulting from LOCA destruction of insulation and 7 its transport, and (c) the combined effects of items (a) and (b) relative to 8 recirculation pumping operability (i.e. , impact on net positive suction head 9 (NPSH) available at the pump inlet). 10 Debris resulting from a LOCA has the potential to block ECC sump debris 11 interceptors (i.e., trash racks, debris,screensl and sump outlets resulting 12 [rescit] in a degradation of or 2 loss of [ net positive-section-head-(]NPSH[3] 13 margin. [The-t8EA generated] Such debris can be divided into the following 14 categories: (1) debris that is generated early in the LOCA period and is 15 transported by blowdown forces (i.e., jet forces from the break), (2) debris 16 that has a high density and will sink, but is still subject to fluid transport 17 if local recirculation flow velocities are high enough, (3) debris that has 18 an effective specific gravity near 1.0 and will float or sink slowly but 19 will nonetheless be transported by very low velocities, and (4) debris that 20 will float indefinitely by virtue of low density [or-composition] and will 21 be transported to and possibly thru the [ sump] debris screen. Thus, debris 22 gerieration [dee-to-the-t8EA], early transport due to blowdown loads, long-term 23 transport, and attendant [sereen] blockage [ effects] of debris interceptors 24 must be analyzed to determine head loss effects. Appendix A provides relevant 25 information [ guidelines] for such evaluations; References 1 through [5] 12 26 provide additional information relevant to the above concerns. 27 The design of sumps and their outlets includes consideration of the 28 avoidance of air ingestion and other undesirable hydraulic effects (e.g., 29 circulatory flow patterns, outlet designs leading to high head losses). 30 The location and size of the sump outlets within ECC sumps is important in 31 order to minimize air ingestion since ingestion is a function of submergence 32 leyel and velocity in the outlet piping. It has been experimentally deter-33 mined for PWRs that air ingestion can be minimized, or eliminated, if the 34 sump hydraulic design considerations provided in Appendix A are followed. 35 References 1, 3, 6, 7, and 8 provide additional technical information relevant 36 to sump ECC hydraulic performance and design guidelines. 1.82/3 05/31/85

e WATER FROM SPRAY LOCA

      . COVER PLATE POST LOCA                                       .                                     DEBRIS SCREEN '

WATER LEVEL' DEBRIS INTERCEPTORS TRASH RACK DRAINAGE SUMP s \ /

w c \ _. a .s y n N gOEBRIS E:N: k$l.hf~t ',', ]I.N?.$'!? - SUMP PIT VORTEA L'c*i ' ' ' ' . .} .. g 3 SUPPRESSOR {~~~ NNft;-:.:- c.- .

                                                                                         -z. ,-

oo (ol-- --- [

                                                                        - ::          k.

e cg . - l R 2 m . . ~.?z:

                                $.'                           I       ?.
                                                                      . . .$t -      tofi)
                        '.:$rig.

7 7293 ,.U~.. fo) .d.: ! .: ..;d

                                                                           . .y .{k. .k?-
                               . :. - ;. . tc.+                 -: c.. .::3 .y

J. in,. T. RECIRCULATION

                                         -Q to (i)

PUMP s i

                                                                                                               )
                 .                                                                     from (o)     3 I

(o) = SUMP OUTLET (i) = PUMP INLET !'""d " ^ l li)

AS DETERMINED DURING SAFETY ANALYSIS R ** CUBIC OR HORIZONTAL SUPPRESSOR MAY BE USED ds'08/77 g WITH EITHER SUMP OUTLET s

CD (P

                                                                                .                                                                     i i

1 Placement of the ECC sumps at the lowest level practical ensures 2 maximum utilization of available recirculation coolant. However, since there 3 may be places within containment where coolant could accumulate during the 4 containment spray period, these areas can be provided with drains or flow 5 paths to the sumps to prevent coolant holdup. This guide does not address 6 the design of such drains or paths. However, since debris can migrate 7 to the sump via these drains or paths they are best terminated in a manner 8 that will prevent debris from being transported to and accumulating on or 9 within the ECC sumps. 10 Containment drainage sumps are used to collect and monitor normal 11 leakage flow for leakage detection systems within containme.us. They are 12 separated from the ECC sumps and located at a lower elevation than the ECC 13 sumps to minimile inadverdant spillover into the ECC sumps due to minor 14 leaks or spills within containment. The floor adjacent to the ECC sumps 15 would normally slope downward, away from the ECC sumps toward the drainage 16 collection sumps. This downward slope, away from the ECC sumps will minimize 17 the transport and collection of debris against the debris interceptors. 18 High density debris may be swept along the floor by the flow toward the 19 trash rack. A debris curb, upstream of and in close proximity to the rack, 20 will decrease the amount of such debris reaching the rack. 21 It is necessary to protect sump outlets by debris interceptors [pamp 22 intakes-by-screens-and-trash-racks-(coarse-eater-screens] of sufficient 23 strength to withstand the vibratory motion of seismic events, to resist Jet 24 loads and impact loads that could be imposed by missiles that may be generated 25 by the initial LOCA and to withstand the differential pressure loads imposed 26 by the accumulation of debris [er-by-trash]. Considerations in material 27 selection for the debris interceptors includes; long periods of inactivity, 28 that is no submergence; and possibly pericds of operation, that is partial 29 or full submergence in a fluid that may contain chemically reactive materials. 30 Isolation of the ECC[5] sumps from high-energy pipe lines is an important 31 consideration in protection against missiles, and it is necessary to shield 32 the screens and [ trash] racks adequately from impacts of ruptured high-33 energy piping and associated jet loads from the break. When the screen and 34 [ trash] rack structures are oriented vertically Eieested-above-fieer-ievel]. 35 the adverse effects from debris collecting on them [sereen-structure] will 36 be [at-a-minimum---Separating-r] reduced. Redundant ECC[5] sumps [sereens] 1.82/5 06/04/85

l i l i 1 and sump outlets [ pump-section intakes 3 are separated to the extent 2 practical [wiii-help] to reduce the possibility that an event causing [a-3 partialiy-eiegged-screen-or-missfie-damage-to-one-screen-wenid] the inter-4 ceptors or outlets of one sump to either be damaged by missiles or partially 5 clogged could adversely affect other pump circuits. [in-addition; proper 6 design-of-section-intakes wiii-avoid-fiew-degradation-by-sir-ingestion; 7 swiri--er-vertex-formation:] 8 [The-iecation-of-the pomp-section-intakes-within-the-EEES-samp-is-import-9 tant-in-order-to-minimize-air-ingestion-that-is-a-f anetion-of-submergence 10 ievei-and-samp-eatiet-veiecity:--8ther-factors-to-consider-are-vertex-forma-11 ti o n-(whi c h-ean-i e nd- to- ai r-i ng e s ti o n3 - a nd- swi ri- e f f e c t s - a t- th e- s ecti o n-i ni e t: 12 it-has-been-experimentaliy-determined-that-air-ingestien-can-be-minimized-or 13 eliminated-if-the-hydraulic-design guidelines provided-in-Appendix-A-are 14 foliewed:--References-it-3;-67-7;-8;-and-9 provide-additionai-technicai-infor-15 mation-relevant-to-samp-hydraulic performance-and-design guidelines: 16 As-neted-above;-the-design-of pump-suetion-intakes-ineindes-censideration 17 for-avoiding-af r-ingestion-or-ether-andesirable-hydraulic-ef f ects-f erg:; 18 swiri t-section-iniet-design-effects 3 --However;-for-smaii-amounts-of-air 19 ingestien--the-recirculation pumps-can-stifi-be-considered-eperabie provided 20 sufficient-NPSH-margin-is-demonstrated:--Appendix-A provides guidance-for 21 e s ti ma ti ng- N P S H-margi n-i f- es ti mat e d-l e v e l s - o f- si r-i n g e s ti o n- a re-i ew-( 4 - e r ; 22 52%) --References-i-and-18 provide-additionai-technical-findings-relevant-tn 23 pump-operatien-and-NPSH-effects.] 24 It is expected that the water surface will be above the top of the debris 25 interceptor [sereen] structure after completicn of the safety injection. 26 However, the uncertainties about the extent of water coverage on the [ screen] 27 structure, the amount of floating debris that may accumulate, and the potential 28 for early clogging do not favor the use of a horizontal top [sereen] interceptor. 29 Therefore, [because-of-this-uncertainty;] in computation of available interceptor 30 surface area no credit [can] is to be taken [in-computing-the-evailable-surface 31 ares] for any [tep] horizontal [seeen] interceptor surface, and the top of 32 the interceptor [sereen] structure [wenid] is preferably [be] a solid [ deck] 33 cover plate to provide additional protection from LOCA generated loads and 34 designed to provide for the venting of any trapped air. 35 Debris which is small enough to pass through trash rack and which could 36 clog or block the debris screens or outlets is to be analyzed for head loss 1.82/6 05/31/85

1 effects. Screen and sump outlet blockage will be a function of the types and 2 quantities of insulation debris that can be transported to these components. 3 A vertical inner debris screen would impede the deposition or settling of 4 debris on screen surfaces and thus help to assure the greatest possible free-5 flow through the fine inner debris screen. Slowly settling debris that is 6 small enough to pass through the trash rack openings could block the debris 7 [ inner] screens if the coolant flow velocity is too great to permit the bulk 8 of the debris to sink to the floor level during transport. [A-vertfeally 9 meanted-inner-screen-wenid-minimize-settling-of-debris-en-the-screen-surface-10 and-if-sufficient-enbiecked-screen-area-is provided-to-keep] If the coolant 11 flow velocity [at] ahead of the screen is at or balow approximately [6] 5 12 cm/sec (0.2 ft/sec), debris with a specific gravity of 1.05 or more will 13 likely settle before reaching the screen surface and thus hcip to prevent 14 undue clogging of the screen. 15 The size of openings in the (fine] screens is dependent on the physical 16 restrictions [--including-spray-nezzles-] that may exist in the systems that 17 are supplied with coolant from the ECC[5] sump. The size of the mesh of 18 the fine debris screen is determined based on consideration of a number of 19 factors including the [The] size of openings in the containmant spray nozzles, 20 coolant channel openings in the core fuel assemblies, and pump design 21 characteristics, for example seals, bearings, and impeller running clearances 22 [wiii-need-to-be-considered-in-determining-the-size-of-the-fine-screen]. 23 As noted above, deoraded pumping can be caused by a number of factors, 24 including plant design and layout. In particular, debris blockage effects 25 or debris interceptor and 5, ump outlet configurations, and post LOCA hydraulic 26 conditions (e.g., air ingestion) must be considered in a combined manner. 27 Small amounts of air ingestion, 52% i.e., will not lead to severe pumping 28 degradation if the " required" NPSH from the pump manufacturer's curves is 29 increased based on the calculated air ingestion. Thus the combined results 30 of all post LOCA effects need to be used to estimate NPSH margin, as calcu-31 lated for the pump inlet. Appendix A provides information for estimating NPSH j 32 margins in PWR sump designs, where estimated levels of air ingestion are low l 33 (52%). References 1 and 8 provide additional technical findings relevant to 34 NPSH effects on pumps performing the functions of residual heat removal, emer-( 35 gency core cooling and containment atmosphere cleanup. When air ingestion is 36 5 2%, compensation f}}

NUREG-0869, Forwards CRGR Review of Proposed Resolution of Unresolved Safety Issue A-43, Containment Emergency Sump Performance. Draft NUREG-0869 & NUREG-0897 Re Unresolved Safety Issue A-43 Also Encl

Contents

Text

SUBJECT:

Enclosures:

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

Subject:

Reference:

Reference:

SUBJECT:

Enclosure:

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

Dear Mr. Serkiz:

Dear Mr. Serkiz,

SUMMARY

Contact:

Navigation menu

NUREG-0869, Forwards CRGR Review of Proposed Resolution of Unresolved Safety Issue A-43, Containment Emergency Sump Performance. Draft NUREG-0869 & NUREG-0897 Re Unresolved Safety Issue A-43 Also Encl

Text

SUBJECT:

Enclosures:

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

Subject:

Reference:

Reference:

SUBJECT:

Enclosure:

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

SUMMARY

Dear Mr. Serkiz:

Dear Mr. Serkiz,

SUMMARY

Contact:

Navigation menu

Search