ML20235T962

Regulatory Analysis for the Resolution of Generic Issue 99, Loss of RHR Capability in Pwrs

ML20235T962
Person / Time
Issue date:	02/28/1989
From:	Spano A NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:
References
REF-GTECI-099, REF-GTECI-NI, TASK-099, TASK-99, TASK-OR NUREG-1340, NUDOCS 8903080541
Download: ML20235T962 (78)

Text

_

' NU REG-1340 Regulatory Analysis for the Resolution of Generic Issue 99: 1 Loss of RHR Capability in PWRs

)

U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research A.H. Spano I

I S;S RBRM "' " *on 1340 R L_--__________-____________

NOTICE Availability of Reference Materials Cited in NRC Publications Most documents cited in NRC publications will be available from one of the following sources:

1. The NRC Public Document Room,1717 H Street, N.W.

Washington, DC 20555

2. .The Superintendent of Documents, U.S. Government Printing Office, Post Office Box 37082, Washington, DC 20013 7082

3. The National Technical information Service, Springfield, VA 22161 Although the listing that follows represents the majority of documents cited in NRC publications, ,

it is not intended to be exhaustive.

Referenced documents available for inspection and copying for a fee from the NRC Public Docu-ment Room include NRC correspondence and internal NRC memoranda; NRC Office of inspection and Enforcement bulletins, circulars, information notices, inspection and investigation notices; Licensee Event Reports; vendor reports and correspondence; Commission papers; and applicant and licensee documents and correspondence.

The following documents in tne NUREG series are available for purchase from the GPO Sales Program: formal NRC staff and contractor reports, NRC-sponsored conference proceedings, and NRC booklets and brochures. Also available are Regulatory Guides, NRC regulations in the Code of Federal Regulations, and Nuclear Regulatory Commission Issuances.

Documents available from the National Technical Information Service include NUREG series reports and technical reports prepared by other federal agencies and reports prepared by the Atomic Energy Commission, forerunner agency to the Nuclear Regulatcry Comrnission.

Documents available from public and special technical libraries include all open literature items, such as becks, journal and periodical articles, and transactions. Federal Register notices, federal and state legislation, and congressional reports can usually be obtained from these libraries.

Documents such as theses, dissertations, foreign reports and translations, and non NRC conference proceedings are available for purchase from the organization spcnsoring the publication cited.

Single copies of NRC draft reports are available free, to the extent of supply, upon written request to the Division of Information Support Services, Distribution Section, U.S. Nuclear Regulatory Commission, Washington, DC 20555.

Copies of industry codes and standards used in a substantive manner in the NRC regulatory process are maintained at the NRC Library, 7920 Norfoll' Avenue, Bethesda, Maryland, and are available there for reference usi, by the public. Codes and standards are usually copyrighted and may be purchaseo from the originating organization or, if they are America National Standards, from the American National Standards Institute,1430 Broadway, New York, NY t0018.

___ ___________m

I I.

NUREG-1340 Regulatory Analysis for the Resolution of Generic Issue 99:

Loss of RHR Capability in PWRs.

Manuscript Completed: November 1988 Date Published: February 1989

' A.H. Spano Division of Safety issue Resolution Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washingtoni DC 20555 l

A

, ABSTRACT

(

Generic Issue 99 is concerned with the loss of. residual heat removal (RHR) )

capability in pressurized water reactors during cold-plant outage operations.

The issue focuses on two risk-significant common-cause failure modes of the RHR system: (1) air binding of the RHR pumps during reduced-inventory opera-

'tions and (2)' spurious closure of the RHR suction valves due to misapplication of the autoclosure interlocks.

Resolution of 'this issue involves consideration of the adequacy of plant capabilities for (1) preventing losses of RHR, (2) responding promptl and effectively'tosuchchallengesinordertopreventcoredamage,and(y)3 ensuring timely containment protection against the release of radioactivity to the environment in the unlikely event of core damage due to loss of shutdown cooling. This entails examination of (1) relevant operational and accident i response procedures, (2) the instrumentation available to the operator for  !

accident diagnosis and mitigation, and (3) the administrative controls avail-able for ensuring control' room cognizance of ongoing maintenance activities that could poter.tially affect the stability of the reactor coolant system.

This regulatory analysis provides quantitative assessments of the costs and benefits associated with several alternatives considered for the resolution

.of Generic Issue 99.

4 iii

v.n CONTENTS gPa e ABSTRACT................................................................ ' i '1 1

. LIST OFy.TABLES ......................................................... vi-

1. STATEMENT OF PROBLEM .............................................. 1~

2. '0BJECTIVE ....................... .................................

. 5-

3. . ALTERNATIVE RESOLUTIONS ........................................... 6

4. TECHNICAL FINDINGS ................................................ 9 4.1 R i s k ' E s t i ma t e s . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.

4.2 ; Summa ry of P roposed ' Requi rements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11'

-5. CONSEQUENCES ...................................................... 13 5.1 Alternative 1: No Action .................................... 13 5.2 Alternative 2: Containment Closure Without CDF-Reduction Fixes ....................................................... 14

'5.2.'l 'Value .................................................

5.2.2- Impact.

. 14

................................................ 16 5.3 Alternative 3: CDF'-Reduction Fixes .......................... 19' 5.3.1 Va1ue.....................................,............ 19 5.3.2 Impact'................................................ 20 5.4 Alternative 4: Containment Closure With CDF-Reduction Fixes ......................................................... 23 5.4.1 Value ................................................. 23 5.4.2 Impact ................................................ 23 5.5 RelationshipEt-o Generic Issue 94, " Additional Low-Temperature Overpressure Protection For Light-Water Reactors" ............ 24

6. DECISION RATIONALE ................................................ 25

7. IMPLEMENTATION .................................................... 27

-8. REFERENCES ....................... ................................ 29 Appendix A Shutdown Cooling Reliabili ty Improvements . . . . . . . . . . . . . . . . . . A-1 Appendix B C o s t A n a ly s i s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B-1 Appendix C . Generic Letter 88-17 ....................................... C-1 v

r. .

3,.t

li LIST.0F

TABLES Table'1..JCore' damage l frequency for cold-shutdown operations ............. 10 Table .2i Containment protection . sensitivity study results. . . . . . . . . . . . . . . . . 10.' .

cTable 3..' Proposed improvements for. reducing CDF due to loss of DHR .......:11 9

Table 4._. 0ffsite ris k reduction benefits for Alternative .3 . . . . . . . . . . . . . . .

Table 5. . Cost / benefit results for Alternative'3:.............~............. 22 LTable 6.. 0ffsite risk reduction benefits.for Alternative 4 ............... 23 Table.7.. Cost / benefit'results for Alternative.4 ..........>................ 23 Table 8. Containment closure cost / benefit results for.the generic site ...-24 i

u I

i vi p-L. .. ..

1. STATEMENT-0F PROBLEtt Generic Issue 99 addresses the problem of loss of residual heat removal (RHR) capability.in pressurized water reactors.(PWRs) during cold-plant outage operations. The1 issue focuses on two risk-significant modes of RHR system failure shown by experience to. occur frequently:

1.

Air ingestion level failure in the reactor of thesystera coolant RHR. RCS) pumpduring (s, which occurs mid-loop when the water

• mainten-  !

ance operations is . lowered excessively allowing air to be drawn into the P.PR pump suction lines and

2. Loss of RHR pump suction due to.the spurious closure of the RHR suction / isolation valves as a result of the-failure or misapplication of the pressure-cctuated autoclosure interlocks (ACIs) controlling )

v'alve position.

Both failure modes.'are of particular significance in that they entail the

. common-cause failure of both trains of RHR cooling.

It is ncted that Generic Issue-99 was originally directed solely to'the problem of the ACI-related spurious closures of the suction valves. In con-sideration of the results of the Deceleber 1985 AE00 report on decay heat rerroval (Ref.1), the scope of this generic issue was broadened to include the

'less frequent but potentially higher risk failure mode of inadequate water j level during mid-loop operations.

The autoclosure feature of the RHR suction valve interlocks provides for autoraatic closure of the valves whenever the RCS pressure increases above a

' level approximating the RHR system design pressure and is intended to ensure that the double barrier (two closed valves in each drop line) is not com-promised by operator errors .during plant startup. During power operation, the open permissive interlock (OPI) feature of the suction valve interlocks as well

'as lockout of electric power to the valves provides assurance that the closed.

valves will remain closed until, following reactor shutdown and restoration of electric power to the valves, the RCS pressure has decreased below the design pressure of the RHR system to the OPl setpoint permitting the operator to open the suction valves for RHR operation. On the other hano, during RHR operation, assurance is required that the suction valves will continue to remain open, and the presence of the AC1 feature provides the potential for inadvertent closure of the valves and consequent loss of cooling. Closure of the suction valves also isolates the RHR system pressure relief valves, which, in approximately a do:en plants, are the primary means for protecting the RCS f rom overpressure transients during water-solid plant operations. Closure of the valves also isolates the low-pressure letdown line, which, during water-solio plant opera-tions, upsets the balance between the charging and 1ctdown flows and may result The tern " raid-loop" rcfers to moir.tenance operations during plant shutdown i involving lowering the reactor vessel water level to a point below the top p cf the hot-leg piping.

[

I

'in a challenge of the systen provided for low-temperature overpressure pro-tection of the RCS (LTOPS). These negative aspects of the autoclosure inter-lock have prompted the NRC staff to reconsider the overall need for the suction valve interlocks. Removal of the ACI feature would be based on an evaluation of the potential negative safety impact. in regard to an interf acing-systems LOCA relative to the positive impact on safety.provided by the reduced potential for RHR system failures'and for low-temperature overpressurization transients. .

While the reported number of RHR system failures due to inadequate. water level is roughly two-thirds of that due to ACI-related valve. closures, the former mode of RHR failure poses an appreciably greater risk to the plant than does the latter for several reasons: (1) in comparison with the full RCS  ;

inventory condition generally found for the valve closure event, the reduced l RCS inventory condition of the mid-loop event may significantly reduce the  ;

allowable recovery . time for pr'evention of core uncovery from a matter of hours '

to perhaps less than an hour for unfavorable conditions, the importance of this reduced recovery time being enhanced by the f act that the containment may be open during such operations; (2) the decay heat rate for a mid-loop event is likely to be appreciably higher than it is for a valve closure event in that

! mid-loop operations in many plants are generally started early in the outage (e.g., about 2 days after reactor shutdown) whereas the expectation of a possible valve closure event tends to be spread out over the entire duration of the outage with the' related decay heat rate at a lower average . level; and (3) based on plant experience, the mean time to recover RHR is'tound to be greater 1or the mid-loop event than~for the valve closure event by a factor of four (about 60 min vs.15 min) partly because of the difficulty of restoring the air-bound RHR pumps.

Staff concerns regarding Generic Issue 99 include:

1. Probabilistic risk assessment (PRA) studies (Refs. 2, 3) of the risk at the Zion plant for power operations and for shutdown operations indicate that the-core damage frequency (CDF) due to loss of DHR able with that due coolingduringplantshutdownoperationsiscompg/RYforshutdown-to power operations: a mean CDF value of 1.8x10 operations compared with a mean value of 6.7x10-5/RY for power opera-tions.

2. The RHR system is normally called on to operate reliably for long periods of time amounting to an appreciable portion of the reactor year (i.e., a total of about 100 days / year for the average U.S. PWR).

As the plants age, the anticipated need for more and more frequent maintenance activities on loop components (in particular, certain classes of steam generators) implies increased risk due to possible loss of RHR. A considerable part of the time on RHR operation in-volves RCS conditions that preclude the use of the steam generators as a ready backup mode of decay heat removal requiring increased reliance on the reliability of the RHR system. As indicated above, this need for high RHR system reliability is significantly enhanced '

for mid-loop operations in comparison with other shutdown operations performed with the RCS filled.

2-

3. Over the 11-year period from 1976 through 1986, there have been a total of 177 reported losses of RHR, including 37 occurrences initiated during mid-loop operations.

4. Based on the numerous losses of RHR experienced at plants, the level of RHR reliability in PWRs cannot be considered good. In particular, the frequency of common-mode losses of RHR due to valve closure or inadequate water level has not diminished with time but has tended to remain at about 0.2 per reactor year since 1976.

5 Bulletin 80-12 (NUREG/CR-4005) issued in June 1980 to improve the availability of RHR systems in PWRs did not result in a reduction of the frequency of common-mode losses of RHR. In the period from issuance of the bulletin through 1986, there were a total of about 70 reported instances of common-mode RHRS failure, of which 25 involved partially drained RCS conditions. In the recent approximately one-year period from March 1986 through April 1987, four plants (San Onofre-2, Waterford-3, Sequoyah-1, and Diablo Canyon-2) experienced significant mid-loop losses of RHR that lasted from about 50 min to 220 min and had rises in RCS coolant temperature that approached or reached bulk boiling conditions (Refs. 4-7).

The sequence of events during various mid-loop losses of RHR have differed significantly among incidents in that the events were generally not anticipated or well understood at the time with respect to cause, effect on RCS conditions, feasibility of recovery procedure, or time avai16ble for recovery prior to the occurrence of certain consequences (e.g., bulk boiling or core uncovery). The results of the staff's review of the plant responses to Generic Letter 87-12 (Ref. 8), which was issued after the Diablo Canyon-2 event of April 10, 1987, to obtain information for assessing the safety of operations when the RCS is partially drained, have supported the staff's perception that the potential risks involved in such operations have not generally been adequately understood by plant operators. The possibility of various mid-loop operating configurations involving conditions not previously analyzed was one of the key findings of the staff's assessment of the Diablo Canyon-2 event (Ref. 9).

An example of an unanalyzed potentially high-risk mid-loop operational configuration is one involving intact hot legs and vented cold legs (e.g.,

openings required for the replacement of reactor coolant pump seals). In this case, boiling in the reactor vessel could lead to press.:rization of the hot-leg side of the loop and the consequent discharge of coolant from the cold-leg openings, with core uncovery an imminent possibility. This and other mid-loop operating conditions require analysis on a plant-specific basis. The need for plant-specific analysis of this and other mid-loop configurations was indicated in Generic Letter 87-12. Subsequently, the Westinghouse PWR Owners Group initiated a program *,o develop the analytical basis for understanding the risk of operations when the RCS is partially drained.

Previous NRC and industry reports and analyses (Refs. 3,10-11) of numerous instances of loss-of-RHR system failure, together with recommendations for related hardware and procedural improvements, have clarified the basis for

I proposing licensee actions required for resolution of Generic Issue 99. As indicated by the plant responses to Generic Letter 87-12, while certain of the recommended improvements have been or are being adopted by some plants, there is a general need for plants to examine how they conduct shutdown operations, in particular, mid-loop operations, and to implement needed improvements for the increased safety of such operations.

The need for correctivt action to enhance RHR system reliability was re-cognized in May 1980 following the Davis Besse 2.5-hour loss-of-RHR event that involved failure of the operating train while the other train was out for main-tenance. IE Bulletin 80-12 and the accompanying Generic Letter of June 11, 1980, requested operating plants to implement controls and technical specifica-tion changes that would ensure availability of redundant or diverse means for decoy heat removal as required by GDC-34 This was exemplified by the require-ment that both RHR trains be operable if the steam generator system was not available. In connection with the possibility of common-mode failures of both trains, the bulletin simply requested plants to ensure that all " reasonable" means be taken to preclude such failures.

The general requirement for written procedures on RCS/RHR operations is covered by Standard Technical Specification Section 6.8, " Procedures and Programs," and Regulatory Guide 1.33, " Quality Assurance Program Requirements (Operations)," and by Criterion V of Appendix B to 10 CFR Part 50, which requires the procedures to be appropriate to possible abnormal conditions. However, in the Waterford-3 loss-of-RHR event of July 14, 1986 (which lasted for about 2.3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> and resulted in boiling in the core), initiation of the event due to inocturate water level indication occurred during RCS draining operations that involved two methods of draining, one of which was not covered by written pro-cedures and led to a previously unanalyzed condition (Ref. 5).

With regard to areas where the existing requirements on RHP operations appear to be inadequate, two examples relating to mid-loop operations are indicated by the experience obtained on loss of shutdown cooling: (1) although the RHR system may be vulnerable to a single common-cause failure, the current technical s operating)priorspecifications specify to starting mid-loop only thatand operations two (2)trains ofno there is RHR be operable (one current requirement for closure of the containment during mid-loop operations, although experience indicates there is an appreciable likelihood of an extended loss of cooling water under reduced inventory conoitions.

2. 0BJECTIVE The general objective of the proposed requirements is to make the risk from accidental losses of RHR during cold shutdown operations a small con-tributor_ to the overall risk associated with operation of a PWR. On the CDF risk level, the target for resolution of Generic Issue 99 is that the con-tribution from loss of shutdown cooling be a small part (i.e., a few percent) of an overall CDF goal of 1x10~4/RY due to all causes. This target for loss of cold shutdown decay heat removal is compatible with the. interim guideline adopted by the staff for developing recommendations'for the resolution of USI A-45,i.e.,thatthequantifiablecontribug/RY.

removal capability 6should not exceed 1x10- ion of CDF The due to lossCDF loss-of-cooling of decay targetheat of a few times 10 /RY (for the plant at cold shutdown and the. containment open) may also be considered compatible with the proposed general performance guidelines given in the Commission's Safety Goal Policy, i.e., that the pro-babilityofalargg/RY.

greater than 1x10 release from A more an operating direct nuclear comparison power of the CDF plant target should with the be no policy guideline requires a definition of "large release" in the policy state-ment.

To meet this target, the proposed recommendations for a defense-in-depth resolution of Generic Issue 99 involve improvements in key plant procedural and instrument capabilities aimed at preventing losses of RHR and, beyond that, ensuring that, if a loss of RHR does occur, the plant can respond promptly and adequately to prevent core damage. In addition, the recommendations would include assurance that containment protection would be available to prevent releases to the environment should an extended loss of cooling result in core damage.

t

i I

3. ALTERNATIVE RESOLUTIONS  !

1 i

t Four alternatives were considered as a basis for technical resolution of 1 Generic Issue 99:

1. Take no action.

' Require assurance of the plant's capability to provide effective 2.

closure of the containment prior to onset of an anticipated core uncovery condition arising from an extended loss of decay heat cooling during mid-loop operations. In the context of this alternative, evaluation of the cost / benefit aspects of such a con-tainment closure. requirement (referred to herein as-Fix C) is based on the assumption that improvements to reduce core damage frequency ,

have not been implemented. If such improvements are implemented, l the cost / benefit assessment of Fix C would be reduced. As detailed in Appendix.A, the containment closure requirement calls for all containment penetrations to the environment that cannot be closed prior to the estimated time of core uncovery be closed prior.to entering mid-loop conditions.

3. Require implementation of proposed CDF-reduction improvements (see Fig. 1), including:

a. Improvements'in plant capability to prevent mid-loop losses of RHR as well as respond effectively to such challenges (referred to herein as Fix A). These requirements are listed in Table 3 and detailed in Appendix 1.

b. In addition to the proposed actions under a, consider re-moval of the autoclosure interlock (referred to herein as Fix B) in order to eliminate inadvertent ACI-related losses of RHR.

4. In contrast to Alternative 2, Alternative 4 evaluates the cost / benefit involved in implementing Fix C given that Fix A or the combination of Fixes A and B has been implemented.

In the no-action approach of Alternative 1, an information notice could be issued.to all PWR licensees and CP Holders describing the Generic Issue 99 safety concerns and the related technicol findings on risk. This notice would be intended for information only and would involve no backfit requirements.

However, such an approach is not deemed appropriate in the light of the exten-sive experience to date on this problem, which shows that the overall rate of common-cause failures of PWR RHR systems has not tended to diminish over the past- decade in spite of NRC and industry information notices, analyses, express-ions of concern, and specific recommendations for needed improvements in RHR ,

operational procedures and instrumentation. In addition, as discussed in l Section4,theestimatedcontrigutiontocoredamagefrequencyduetolossof shutdown cooling is about 4x10~ /P,Y, implying a significant level of risk in the context of mid-loop operations being conducted with an open containment. The staff therefore concludes that the no-action approach cannot provide a satis-factory resolution of Generic Issue 99.

4

.I h .

4

With regard to Alternative 2, since the estimated core damage frequency l L due to loss'of shutdown cooling is primarily (about 80%) associated with losses of RHR during.mid-loop operations,-the question of allowing the containment to .4 be open during such operations becomes1important in ensuring that the risk to the public be. acceptably low. Accordingly, Alternative 2 considers the -

cost / benefit aspects of the proposed containment closure requirement separately (

o from actions taken to reduce core damage frequency. ~

Alternative.3 examines the cost / benefit aspects'of. actions to' reduce the l core damage frequency.without recourse to actions to' close-the containment. {

Actions to reduce the'high.CDF due-to loss of shutdown cooling are considered r to'be particularly important for the resolution of this issue. .

Alternative 4 evaluates the cost / benefit aspects of implementing the con-

'tainment closure requirement (Fix C) in conjunction with either Fix A or the.

combination of Fixes A and B to take account of measures to reduce.the likeli-hood of severe accidents due to. loss of decay heat cooling.

l

Improved Reliability of RHR Operation I

l I

-l Ensure Adequate Accident Prevent Ensure.

Response Capability and ACI-Related Containment Prevent Mid-Loop Losses Losses of Protection of RHR RHR- Capability (Fix A) (Fix B) (Fix C)

I l- I REQUIRE: REQUIRE: REQUIRE:

1. Prompt Operator Availa- Appropriate Assurance of bility of Plant Status Elimination of Timely Closure Information Autoclosure 'of Containment

' RCS/RHRS Indications / Feature of Penetrations for

. Alarms ._ Suction Valve Mid-Loop Reliable Water-Level- Interlocks Operations Measurement Capability

• Operator Aids

2. Alternative Procedures for. Recovery of Shutdown Cooling

3. Administrative Controls

4. Technical Specification Change-

5. Training on Loss of Shutdown Cooling 1

1 Figure:1: Schematic diagram of requirements considered for resolution of Generic Issue 99 i

8-

With regard to Alternative 2, since the estimated core damage frequency due to loss of shutdown cooling is primarily (about 80%) associated with losses of RHR during mid-loop operations, the question of allowing the containment to be open during such operations becomes important in ensuring that the risk to the public be acceptably low. Accordingly, Alternative 2 considers the cost / benefit aspects of the proposed containment closure requirement separately from actions taken to reduce core damage frequency.

Alternative 3 examines the cost / benefit aspects of actions to reduce the core danage frequency without recourse to actions to close the containment.

Actions to reduce the high CDF due to loss of shutdown cooling are considered to be particularly important for the resolution of this issue.

Alternative 4 evaluates the cost / benefit aspects of implementing the con-tainment closure requirement (Fix C) in conjunction with either Fix A or the combination of Fixes A and B to take account of measures to reduce the likeli-hood of severe accidents due to loss of decay heat cooling.

~7-

+ a.

i-

j. .' i Improved Reliability of RHR Operation l l l

Ensure-Adequate Accident Prevent;. Ensure Response Capability and ACI-Related Containment Prevent Mid-Loop Losses. Losses of Protection

.of'RHR. RHR Capability;

~

(Fix A)- .(Fix'B)_ _ _ . (Fix C) 1 I .I REQUIRE: REQUIRE: ~ REQUIRE:

1. Prompt Operator Availa- Appropriate . Assurance of.

bility of Plant' Status

~

Elimination of Timely Closure

-Information _ Autoclosure 'of Containment RCS/RHRS Indications / Feature of Penetrations.for Alarms Suction Valve Mid-Loop

• Reliable Water-Level- Interlocks Operations.

' Measurement' Capability

' Operator Aids

2. . Alternative Procedures for Recovery of Shutdown i Cooling- j

(- 3. Administrative Controls

4. Technical Specification Change

5. Training on Loss of Shutdown Cooling Figure 1: Schematic diagram of requirements considered for resolution of

. Generic Issue 99

3 4 TECHNICAL FIllDINGS 4.I' Risk Estimatg Probabilistic methods were used for a generic study of the risk-reduction l

benefits of requirements that would both prevent losses of RHR during cold shutdown. operations 'and improve plant capability -to respond to such losses. The

results of this study by the. Brookhaven National Laboratory (BNL) are reported in Reference 13.

In the BNL study, estimates vere obtained of the CDF-reduction worths of iroprovements i_n three areas: The first relates to improving plant capability l to diagnose and respond effectively to losses of RHR; the.second relates i specifically to preventing losses of RHR during-mid-loup operations by improving l operator capability to control RCS water level within a narrow range using an itppropriately reliable ano accurate systera for measuring RCS water level; the third relates to preventing losses of RHR due to ACl-related spurious closure of the RHR suction valves. For purposes of this regulatory analysis, it is con-E vehient to consider the combination of the first two of these in:provement areas as a single fix, Fix A, and the third fix, involving removal of the autoclosure l

interlock, as Fix B. Fixes A and B are incicated in Fig. 1.

Core damage frequency results obtained with and without Fixes A ar.d B in place are shown'in Table 1. The high risk significance of mid-loop operations is' indicated by the fact that 80% of the CDF from loss of cooling arises from l mid-loop operations.

l As part of the BNL study, a sensitivity study was used to assess the risk reduction provided by a containment closure requirement (Fix C) that would ensure availability of containment protection in the event of core damage.

Estiraates of the risk reduction provided by the containment being' closed or containment sprays being available are indicated by the bounding calculational results in Table 2, which were obtained by BNL using the liACCS consequences code (Ref. 14). Risk results for' intermediate containment conditions are given in Reference 13 (lable 4.6).

The offsite dose results it. Table 2 are given in terms of a high-population (High Pop) density site (represented by the Zion-site meteorological conditions and the 1980 50-mile-radius population distribution, equivalent to a uriform pcpulation density of abcut 890 persons /sq. mile), and a low-pepulation (Low Pop) density site (represented by a uniform density of 1C0 persons /sq. mile, combined with Zion-site meteorological conditions).

From the results in Tables 1 and 2, it is seen that, for the high-popula-tion site, the assessed base case risk due only to mid-loop losses of core cooling is 80% of the tabulated cose of 101t person-reras/RY shown in Table 2, or about 810 persen-rems /RY. For the low-pcpulaticr. site, the corresponding 50-taile population cose is about 150 person-rems /RY. These estirchtes of risk are much reduced if the containment (equipment hatch and other penetrations) is 9-

closed during mid-loop operations. Accordingly,'the most important immediate fix for reducing the offsite risk is that the containment'be closed during ,

mid loop operations or that there be assurance that the containment can be  ;

closed on a timely basis in the event of an extended loss of cooling. If the l containment:is effectively closed-(i.e., intact), release is by basemat failure, andlthe offsite dose can be expected to be negligibly small. However, the j results also indicate that, with Fix A in place to reduce the core damage .

frequency even if- the containment is open and sprays are not available, the risk _l level could'be significantly reduced to.about 0.80 x 96 = 77 person-rems /RY for the high-population site and, correspondingly, to about 15 person-rems /RY for the low-population site.

Table 1 Core damage frequency for cold-shutdown operations

Fix B Initiator Base Case  % Fix A (With Fix A in Place)

Loss of Cooling 4.3(-5)/RY* 50 4.1(-6)/RY 3.4(-6)/RY LOCA 4.3(-6)/RY 8 2.6(-6)/RY 2.6(-6)/RY LOSP 5.1(-6)/RY 12 5.1(-6)/RY 5.1(-6)/RY Total: 5.2(-5)/RY 165 1.2(-5)/RY. 1.1(-5)/RY About 80% of the.CDF due to loss of cooling is due to mid-loop losses, 8%

due to inadvertent suction valve closures, with the residual of 12% due to hardware failures and maintenance unavailabilities.

Table 2 Containment protection sensitivity study results Containment Conditions Dose For High Pop / Low Pop Sites *

(person-rems /RY)

Equipment ** Other** Containment Base Fix B Hatch Penetrations Spray Case Fix A (w Fix A)

Open Open No 1014/187 96/18 79/15 Open. Open Yes 535/99 51/9.4 42/8

" Closed" " Closed" No 13/2 1/0.2 1/0.2

" Closed"~ " Closed" Yes 6/1 0.6/0.1 0.5/0.1 Intact Intact No 0.2/- 0.02/- 0.01/-

i Risk due to loss of shutdown cooling only (excludes LOCA and LOSP).

Open, " closed," and intact refer, respectively, to probabilities of 1.0, 0.01, and 0 for the containment to be open for the release of fission products to the environment; thus, for the intact condition (i.e.,

probability = 0), the release category is purely PWR-7, i.e. , basemat ,

failure. {

4.2 Summary of Proposed Requirements Table 3 summarizes the proposed improvements in Fixes A and B for reduc-ing the CDF due to loss of decay heat cooling. These address various hardware l l

and procedural shortcomings that the NRC and inoustry have identified as I i significant contributors to common-mode RHR system failures. These requirements are discussed in Appendix A. It is recognized that their applicability may be i expected to vary somewhat from plant to plant depending on the design / operational specifics of the plant and on the steps that a number of utilities have taken or are taking to implement needed improvements. The adequacy of a specific action taken by a plant would depend primarily on whether the intent of the related requirement has been met.

As indicated in Fig. 1, the approach for resolving Generic Issue 99 in-volves examination of the cost / benefit aspects of Fixes A and B, which are aimed at reducing the core damage frequency due to loss of cooling, and of Fix C (containment closure), which is aimed at minimizing public risk after core damage has occurred. Estimates of the risk reduction benefits provided by these fixes are presented in Tcbles 1 and 2.

Table 3 Froposed improvements for reducing CDF due to loss of DHR Appendix A Regulatory Requirement Type Ref. Section Bases

1. Treno analysis of RHR flow Hardware 1.1(a) a, b with low-flow alarm

2. Instrumentation / procedures f or Hardware / 1.1(b) b, c l diverse independent water Procedural level measureraents and operational control

3. RCS water temperature Procedural 1.1(c) rneasurement 4 Protection against loss of RHR Hardware 1.1(c) b,c,d suctien valve position indica-tion on removal of power from valve

5. Operator aid on st6tus of Procedural 1.1(e) c RCS/RHR-related plant equipment

6. Opernor aid on tirres to bulk Procedural 1.1(e) c boilirg/ core uncovery

7. Upgraded loss-of-RHR response Procedural 1.2(a,b) c i procedures and controls (e.g.,

RHR pump recovery, alternative DHR recovery methods) l

Table 3 Proposed improvements for reducing CDF due to loss of DHR (Cont'd)

Appendix 1

-Requirement Type Ref. Section Reg. Bases

8. Direct communication between Procedural 1.2(c) b, c control room and containment during mid-loop operations

9. Decision criteria for Procedural 1.2(d) f designation of extended loss-of-RHR events

10. Decision criteria and Procedural 1.2(e) c, g procedures for closure of containment for mid-loop loss-of-RHR event

11. Control room authorization Procedural 1.3 c and coordination on work involving major plant evolutions

12. Upgraded communication between Procedural 1.3 c and within shifts on equipment status and work programs

13. Confirmation of correct Procedural 1.3 c alignment of critical valve positions

14. Operability of redundant means Technical 1.4 h fcr coolant injection for DHR Specification recovery

15. Removal of ACI feature Hardware / 2 d, e Procedural a 10 CFR Part 50, App. A, GDC-13 b RG 1.97 c 10 CFR Part 50, App. B d SRP 5.4.7, Rev. 2, July 1981, BTP ICSB-3 e SP.P 5.4.7, Rev. 2, July 1981, BTP RSB 5-1 f 10 CFR 550.72 and 10 CFR Part 50, App. E g 10 CFR Part 50, App. A, GDC-60 h 10 CFR Part 50, App. A, GDC-33

5. CONSEQUENCES This section examines the value-impact aspects of the proposed alternatives based on the guidance given in NUREG/CR-3568, "A Handbook for Value-Impact Assessment," and in NUREG/BR-0058, " Regulatory Analysis Guidelines of the U.S.

Nuclear Regulatory Commission."

In the assessment of values and impacts, " values" are defined as the improvements obtained in the protection of the public health and property. A measure of the benefit to public health is provided by the 50-mile population dose in person-rems as calculated by the MAACS consequences code. The code also calculates the offsite property damage costs involved in land interdiction and decontamination actions that may be required.

" Impacts" provide u measure of other consequences, mainly economic, that result f rom the implementation of the proposed requirements. As defined, I these impacts may be considered positive (e.g., the costs to the plant in 1988 U.S. dollars due to installing, operating, and maintaining the plant modifica-tions needed to comply with the proposed requirements, including the different-ial cost of any replacement power needed because of additional plant downtime) or negative (e.g., the savings to the plant in averted accident costs associated with plant repair or replacement, cleanup costs, power replacement). The net impact represents the difference between the positive costs and the present worth of the averted onsite accident costs summed over the remaining life of l the plant. The parameters affecting the estimates of values and impacts are recognized to vary from site to site and from plant to plant. In the following analysis, the point estimates of values and impacts are taken to represent best-estimate mean values for the entire population of U.S. PWRs.

l 5.1 Alternative 1: No Action No costs are usually attributed to a "No Action" alternative because the future costs of accidents are conventionally counted as benefits or averted costs in the assessment of the alternative &ctions. However, a severe core damage accident is estimated to result in about 4 to 5 billion dollars in onsite costs. If the accident also resulted in a large offsite release of radioactivity, the costs of relocating people, restricting food and water, and cleanup could cost a few billion dollars more. Reports from the U.S.S.R.

Indicate that the costs associated with the accicent at Chernobyl may amount to about $15 billion. In addition, eccidents less severe than a core demage accident might require shutdown and possibly cleanup. At $E00,000 to $500,000 l per day for replacement power alone, such events would also result in significant costs. Thus the convention of accounting for these 6verted costs in the resolution of other alternatives should not obscure the possible costs I cssociated with the "No Action" alternative.

i l

5.2 Alternative 2: Containment Closure Requirement Without CDF-Reduction i TIies 5.2.1 Value For the estimate of benefits associated only with the containment closure requirement, the staff considered the potential reduction in offsite consequences obtained by requiring that contairmient protection be available to prevent re-leases of radionuclides to the environment during mid-loop operations. At present;, there is no requirement for maintaining a closed containment during nid-loop operations similar to the requirement that applies when refueling takes place.

The base case risk e timate involved the combination of (1) the CDF estimate to losses thatof about occur during4.3x10~g/RY due(i sensitivity mid-loop operations,* to loss study of cooline} of which a estimates of the release categor medium (= 0.1), and low (= 0.01)yprobabilities probabilitiesfor obtained by assuming the equipment hatchhigh and (= 1),

other containment penetrations being cpen and the cretainment sprays not being avail-able during a mid-loop loss-of-cooling accident involving the release of fission products, and (3) a calculation of the 50-mile population cose (in person-rems) and of fsite interdiction / decontamination costs using the f4ACCS consequences code for the high- and low-population sites. The population dose is a measure of the total low-level dose risk to a population given certain c.ssumptions about such tactors as relocation, interdiction, and decontamination but yields no information about environmental contamination. Relocation, interdiction, and decontamination can reduce the populatior dose; this is modeled in the f4ACCS code by limiting the calculated lifetime dose to any individual to 25 rems. Without interdiction and decontamination, the calculated offsite dose levels coulc be higher by en order of magnitude (Ref. 15).

The full set of results of the BliL sensitivity study of offsite risk due to loss of shutdown cooling is given in Table 4.6 of Reference 13; the results for five cases are given in Table 2 of this report. For the upper-bourd case of an oper, containment and no sprays, the release catecory is taken te be PWP.-2; for the lower-bound case of an irtact containment witt sprays available, the release is through basemat failure, and a PWR-7 release is taken to be applicable. Accordingly, with the conditional core damage frequency due to loss cf cooling reduced in proportion to the 80% contribution from mid-loop operations, the related offsite consequences per reactor-year are:

For the evafuTtion of the rist roouction ber:efit of the containment closure requircrent (Alterr.atives 2 and 4), consideration is given only to the mid-loop contribution 10 CCF since the shutdown risk associated with the open ccr.tainment principally arises during reduced-inventory operations.

High Pop / Low Pop Offsite Containment Consequences

• State 50-Mile Dose Property Damage **

(No Sprays) (person-rem /RY) ($/RY) 5 Open 810/150 (2.8/1.1) x 10 Intact (Basemat failure) 0.2/- 1/0.4 5 Difference = 810/150 (2.8/1.1) x 10

• For mid-loop operations, base case CDF = 80% x 4.3 x 10-5/RY.

• • Sum of interdiction and decontamination costs.

Uncertainties in the estimates of offsite risks include those associated with the source term and human error contributions to CDF. In the BNL study, use was made of the WASH-1400 source terms modified to take into account the

, later release times and lower plume energies expected for accidents during shutdown. BNL reviewed various arguments regarding the applicability of the PWR-2 release category for shutdown accidents and determined that, unless a mechanistic calculation is performed with the new computer codes available for the accident conditions of interest, it is difficult to conclude that PWR-2 is overly conservative, i.e., tending to yield too high an estimate of the release.

A conservative factor not discussed in Reference 13 relates to the tacit assumption in the BNL calculations of no fission product retention in the open RCS and open containment systems. No analytical studies are available to describe the release behavior of radionuclides during a PWR loss-of-cooling accident at cold shutdown with the RCS and containment systems open. For this regulatory analysis, we have examined the effect of a retention factor of 50%

in both the RCS and containment systems, which implies a factor of four reduc-tion in the source term estimate used in the MACCS calculation. Because the interdiction / decontamination criteria embodied in the consequence calculations entail population relocation if the predicted long-term dose exceeds 25 rem in 30 years, the factor-of-four reduction in source term is actually reflected as a reduction of only a factor of 1.6 in the 50-mile population-dose estimate.

While the offsite property damages would be reduced by a factor of four in pro-portion to the source term, the effect of this reduction in economic costs on the total offsite costs (dose plus property damage) is found to be small since the property damage contribution is roughly an order of magnitude less than the population dose monetized at $1000/ person-rem.

On the other hand, other aspects of the BNL analysis with regard to the CDF estimates suggest the assessment of risk may be too low. The CDF results are bdsed on the scenario of a static boiloff of coolant from the open RCS for mid-loop loss-of-coolant accidents initiated at different times in the range from about 3 to 24 days after reactor shutdown. This choice of accident initiation times reflects Zion outage experience (Ref. 3). However, as indicated by the utility responses to Generic Letter 87-12 (Ref. 8), it is not unlikely that the mid-loop condition may be achieved as soon as two days after

I reactor shutdown at a time when the decay power is significantly higher than it would be a week or a month later. For higher decay power levels, the times to core uncovery are shorter and can accordingly be expected to entail higher human error probabilities with regard to recovery of the DHR function. Also, 1 other recently identified mid-loop accident scenarios not considered in the 1 BNL analysis appear to involve relatively short times to core uncovery and correspondingly high failure probabilities in recovering DHR (e.g., the scenario of hot-leg pressurization resulting in discharge of coolant from a cold-leg opening in Ref. 9). The possibility of such scenarios has been noted in Reference 13, but their contribution to CDF was not included for evaluation in the original study performed by BNL.

In considering these two areas of uncertainties, we judge that the effect of fission product retention in lowering the assessed risk is probably more than offset by the underestimate in human error contribution to core damage frequency. For piasent purposes, however, it is assumed that the two effects balance each other and that the assessment of offsite consequences (page 15) does provide a reasonable basis for evaluating the risk reduction benefit provided by the containment closure requirement.

For Alternative 2, the total value of the containment closure requirement in averted health effects is given.by the sum of all contributions from all affected reactors integrated over the remaining plant lifetimes. Thus, for a projected total of 78 licensed PWR units and an average remaining lifetime of 31 years, the induttry-wide total averted population dose provided by the con-tainment closure requirement is determined to be:

High Pop Site: 810 p-r/RY x 31 Y x 78 R = 2.0x10 6person-rems Low Pop Site: 150 p-r/RY x 31 Y x 78 R = 3.6x10 person-rems 5.2.2 Impact The principal cost consideration here is the effect of the containment closure requirement on the scheduling of shutdown maintenance activities and its possible effect on increased plant downtime.* An attempt to evaluate this impact generally is hampered by plant differences in the priorities and related scheduling given to specific activities requiring the large penetrations in the containment to be open (e.g. , running of cables, movement of large equipment).

In this connection, we take note of the following considerations:

1. lhe containment requirement, which calls for penetrations to the environment to be capable of being closed prior to the estimated time of core uncovery for postulated loss-of-cooling accidents, is seen to have a potential impact on plant operations primarily in regard to the large equipment hatch. Other smaller openings (e.g.,

personnel airlock door) are generally capable of being closed Other costs (revision of procedures, training, etc.) represent one-time costs and are clearly negligible compared with the recurrent costs from possible extra downtime; they have therefore not been included in the evaluation of impact.

within a matter of minutes. In addition, maintenance work on con-tainment penetration valves (e.g., large purge valves) should not be significantly affected by the containment closure requirement since such valves are series-lined double valves; thus if work is performed on one valve while the other is closed, the closure requirement could be satisfied. Also, with regard to steam generator secondary-side penetrations, the controlled closure of the main steam isolation valves and steam line relief valves could provide for timely closure of the containment if the secondary-side steam generator manways were open.

2. Although a small number of plants (e.g., ANO, McGuire, Catawba) in their responses to Reference 8 have indicated the capability to close the large equipment hatch in about one-half hour or less, it will be assumed here that PWR plants generically require at least two hours to close the equipment hatch. Thus, in the context of accident scenarios involving times to core uncovery of about an hour or less, the requirement for containment closure during mid-loop operations would appear to primarily affect the use of the equipment hatch.

Accordingly, the related impact on the outage schedule can be expect-ed to depend on (1) the amount of time spent at mid-loop (which, for plants using nozzle dams, may be expected to be appreciably less than if nozzle dams are not used) and (2) the extent to which such mid-loop operations requiring an open equipment hatch can be scheduled for the later stages of a refueling outage, at a time when the decay power level is substantially reduced relative to that a few days after reactor shutdown, and the time to core uncovery is correspondingly increased.

3. The formulation of the containment closure requirement in terms of time to core uncovery permits a certain flexibility in scheduling operations to minimize the impact on downtime. For Diablo Canyon (which does utilize nozzle dams), the indicated experience is that the plant has been able to schedule around their self-imposed con-tainment closure requirement without appreciable impact on downtime.

For other plants, based on assessments provided by representatives of the three vendor owner groups, estimates of the impact on down-time for a refueling cutage appear to range from a fraction of a day to about a week, a rough consensus indicating that the average extra downtime might be about 3 days. Based on this estimate of the down-time impact for a refueling outage and an assumed average frequency of one refueling outage per 18 months, the impact equivalent per year is about 2 days for refueling outages. If we add to this an assumed additional impact of I day / year due to forced maintenance outages, an overall impact of about 3 days /RY is obtained.

Various considerations suggest that this estimated average downtime impact for the industry as a whole may be too high. For example, B&W plants, in contrast to Westinghouse and CE plants, do not have to drain the RCS below the top of the hot legs to carry out various loop maintenance activities, which significantly reduces the probability of a loss of RHR due to air ingestion.

Further, some plants reportedly do not open the equipment hatch or may defuel

,j the reactor vessel 1 prior to opening the hatch. In other instances,.some plants y' alreaoy'have procedures in place requiring containment closure during reduced' l ' inventory operations. Furthermore, operating experience:(Ref. 16) shows that plant annual availability due to refueling and maintenance _ operations can vary widely= from plant to plant - from about 45 days- to well over 100 days depend-ing on how well the outage is planned (Refs. 17 18). Thus there is the general

- expectation that. the impact of the proposed containment closure requirement on-downtime can be appreciably reduced through' judicious' scheduling.of the mid-loop operations.

This last. aspect, while particularly important, is difficult to quantify

, in that it implies a performance evaluation of the ~ efficiency of outage planning on a plant-specific basis. Nevertheless, it is reasonable to expect that plants that niay currently experience an extra downtime of 3 days /RY would,.

over the 31-year period of time that constitutes'the average' remaining PWR lifetime, be able to. reduce the downtime considerably. In the. following analysis, an assumed 3-day /RY downtime will be used to obtain an initial estimate of the containment closure cost / benefit ratio with and without implementation of the CDF-reduction fixes. The cost / benefit results will be directly affected by.the assumed impact on downtime.

' Using an industry-average downtime ' cost of about $400,000 per day in 1988 dollarsbasedontheresultsgiven.inReference19,theestimated3-dgy/RY.

impact inposed by the containment requirement amounts to.about 1.2x10 dollars per reactor-year or,7for a total of 78 plants, a total industry-wide annual cost of about'9.4x10 $/yr.

To obtain the present value of this impact in terms of 1988 dollars, use

[ is made of the following discrete discounting annuity formula (Ref. 20):

I IntegratedCostFactor=h=II+" -

where (1 + r)

PV =;present value C = recurring annual cost (= 9.4x107 $/ year) r = assumed real discount rate (5% or 10%)

t = normalized average remaining lifetime (i.e., 31 years) i Thus, for a 31-year average remaining PWR lifetime, the total discounted impact in. posed by the containment closure requirement is estimated to be:

$1.5 x 109 assuning a real. discount rate of 5%

$0.9 x 109 assuming a real discount rate of 10%

I 1

Based on these estiniates of the containment closure impact, the cost / benefit ratios obtained for the proposed requirement (assuaing the absence of ar.y CDF,-reductionfixes)are:

Alternative 2 Cost / Benefit Site ($/ person-rem) r = 5% r = 10%

High pop W~ 450 Low Pop 4100 2500 5.3 Alterr.ative 3: CDF-Reduction Fixes 5.3.1 Value The results in Table 1 indicate that Fixes A und B taken together reduce the CDF due t loss of cooling by a factor of about 13. The residual CDF of about 3.4x10"g/RY reflects hardware failures drid niaint0 nance unaVailabilities in the RhR ar.d backup coolir g systei.is and in related support systems.

For Alternative 3, the benefits ir. of rsite dose reduction due to Fixes A er.d B derived from Tabler 1 and 2 are given in Table 4.

Table 4 Of fsite risk rcduction benefits f or Alternative :*

Increrrental Incremental Dose Reduction Fixes CDF Reduction High Pop Site Low Pop Site Fix A 3.9 x 10-5/RY 9.2 x 102p-r/RY 1.7 y. 102 p-r/RY fix.B** 7.E x 10 /RY 17 p-r/RY 3 p-r/RY

• For mid-loop ano non-mic-loop losses of RHR with the assumption of open certainment.

• • The increrrental benefits due to Fix B given th6t Fix A is ir place.

Thus Fix A provides for an arrta.1 averted 50-mile population dose of about EEL person-rems /RY for the high-population site, whic.h if integrated over 31 yeors and a prcjecteg totd of 78 units, yields an industry-wide total benefit of about 2.2x10 perspr-rems averted. The corresponding value for the low-population site is a.1x10' person-rems.

For Fix D, gisen that Iiy / is in place, the adcitional averted cose at thehigh-populationsjteof about 17 person-rems /FY yielos an industry-wide totol of ebout 4.ly]O person-ren.s averted. The corresponding inc'Ustry-wide berifit for Fix C for the low perulatiert site is about 7.3x10" person-rems averted.

Summarizing, the industry-wide benefits in averted population dose for Fixes A and B are-Alternative 3 Benefits (person-rems)

Fix High Pop Site Low Pop Site A 2.2x10j 4.1x10j B 4.1 x 10 7.3 x 10 5.3.2 Impact Evaluation of the impact for Alternative 3 involves consideration of (1) the costs of implementing Fix A and Fix B and (2) the savings in averted onsite costs derived from a reduction in CDF.

The estimated costs associated with implementation of the requirements listed in Table 3 were determined on the basis of a cost analysis (Appendix 2) prepared for the staff by Science and Engineering Associates, Inc., and on the basis of some directly available utility information on various costs involved in the removal of the ACI feature. The per-plant costs were estimated in five categories: hardware (materialplusinstallationlabor), procedure writing / rewriting, training, technical specification changes, and NRC review.

The first two cost categories were estimated for each of the proposed require-ments. The latter three categories involved " global" estimates, i.e., (1) a training course for key personnel (including the shift supervisor roomoperators,andmaintenancesupervisorforeachofsixshifts}fivecontrol to review the risk implications of a loss of shutdown cooling and the related problems of prevention and mitigation of such losses, (2) review of related technical specifications, and (3) NRC review of the proposed modifications.

The cost estimates for writing / rewriting procedures, training, and revis-ing the training manual were based on the findings compiled in Reference 21, with the point cost estimates in dollars escalated to 1988 according to the guidelines presented in Abstract 6.4 of Reference 21. The cost estimates for each of the five categories are:

1. Hardware (material / labor /QA-QC/ overhead / personnel exposure) Costs:

These costs for Fix A are er,timated to total approximately $470,000 per plant if it is assumed that all of the proposed hardware improve-ments listed in Table 3 are required to be implemented in the plant The water level instrumentation costs included within this estimated total cost cover the costs for a permanent standpipe /TV system as well as narrow- and wide-range differential pressure units. For Fix B, the estimated costs for removal of the autoclosure interlock feat-ure include the costs for a plant-specific analysis, cable disconnect-ing, interlock logic reprogramming 3 andreviewgthesecostsare estimated to be in the range $1x10 to $1.5x10 . For Westinghouse plants, where the Owners Group has provided a generic safety analysis of the ACI removal action (Ref. 22), the plant-specific analyses may be expected to be relatively simple to complete, and, the costs fcr Fix B would accordingly tend to be at the lower end of the above estimated range.

2. Writing / Rewriting' Procedures: Each of the proposed requirements in Table:3 were. reviewed to determine whether the required procedural

revision.was simple (approximately one page; with- a point cost .

estimate of_1915 in 1988 dollars) or complex-(about 10'pages, withta-1988 point cost estimate of $3650 per change). The total rewrite costs,.assumingLimplementation of all of the proposed requirements at

.a plant, amount to about $36,000 for Fix A and about an additional

$1000 for Fix B..

3. Training:. The training costs include those~ associated with the in-structor and training facility, the trainee's time (salary +. fringe),

and training manual revisions. In 1988 dollars,'these various costs, covering training.for six shifts of plant personnel,-are estimated to amount to about $75,000, 4 Technical Specification Change: A complicated technical. specification' change 'is estimated to require 16 ' staff-weeks of utility- technical, legal, management, and committee input (Ref. 21). With staff hours ccsted at $50 per hour (1988), this amounts to $32,000 p'er plant.

5. NRC Cost: -Based on information from Abstract 5.1 of Reference 21, NRC costs for a technical specification change review is estimated to total $28,000 (1988 dollars).

Accordingly, the per-plant cost estimates involved in implementing Fixes A and B.are:

Fixes A' Fix.B hardware. $ 470k $ 100-150k-(+ Radiation Exposure)

Procedures 36k 1k Training- 74k -

Technical Specification 34k -

DRC TS Review 28k -

6.3x105 $/ plant (1.0- 5)x10$/5 plant Based on an average of I staff-week of HRC review effort per plant at $50 per hour.

Based on these per-plunt costs, an upper bound estimate of the associated industry-wide cost is obtained if we assume that each of the 78 plants incurs all of the above costs' required to implement Fix A and Fix B 7 On this

' bcsis, we have6 a: total industry-wide implementation cost of $4.8x10 for Fix A and$(8-12)x10 for Fix B.

l.

L 1

_ _ _ _ _ _ _ _-__-__ _ __ - - _ __ a

To estimate the savings associated with the averted onsite costs of a reactor accident, consideration is given to the following cost elements:

1. Replacement energy costs,

2. Plant capital: costs,

3. Plant decontamination / cleanup costs,

4. Plant repair costs,

5. Early decommissioning costs,

6. Worker health impacts,

7. Electric utility business costs,

8.  !!uclear power industry costs,

9. Onsite litigation costs.

Theaggregateofsuchonsitecostsofanucleagpowerreactoraccident has been estimated by the staff to be about $4.6x10 in 1988 dollars, based on the analysis presented in Reference 19. Accordingly, ta ing account of estimated reductions in core camage freauency of 3.9x10~g/RY and 7.2x10~fhe /RY providedbyFixesAandB,respectively(Table 4),ghecorrespongingaverted onsite costs per plant-year amount to about $1.8x10 and $3.3x10 . Thus, for Fix A, ghe industry-wide present value of averted onsite costs amgunts to l'i.2x10 if the assumed real discount rate is 5% or about $1.3x10 if the assuurd real discount rate is 10g. For Fix B6 the corresponding present-value avertu ensite costs are $4.0x10 and $2.4x10 ,

Summarizing, the industry-wide impacts ussociated with Fixes A and B are:

Averted Onsite Costs Net Impact

_Fix Implementation Costs r=5% r=10% r=5% r=10%

0 8 8 7 Fix A $4.8x10 6

$2.2x10 6 $1.3x10 6 -$1.7x10 6 -$8x10 6 Fix P. $(8-12)x10 $4.0x10 $2.4x10 $(4-8)x10 $(6-10)x10 Combining these estimates of net impact with the above person-rem benefits estimated for Fixes A and B results in the cost / benefit ratios given in Table 5.

Table 5 Cost / benefit results for Alternative 3 Cost / benefit ($/ person-rem)

Fix Site r=5% r=10%

A High Pop 20* 20* ,

Low Pop 120* 120*

B High Pop 100-200** 140-240** j Low Pcp 550-1100** 820-1400** l I

Not including averted onsite ccsts, which, if included, would result in a net savings.

• • Includes averted onsite costs.

1 l I l

i j

5.4 Alternative 4: Containment Closure With CDF-Reduction Fixes

' This alternative (.onsiders the cost / benefit aspects of the containment closure requirement, given that Fixes A and B are in place to provide a reduced core damage f requency.

5.4.1 ,Value Based on the data given in Tables 1 and 2, Table 6 gives the savings in averted offsite consequences if the containment is closed during mid-loop operations and Fixes A and B'are in place:

Table 6 Offsite risk reduction benefits for Alternative 4 Fixes Industry-Wide Averted Offsite in Annual Averted Offsite Dose Dose

• Place High Pop Site Low Pop Site High Pop Site Low Pop Site 1 5 f A 77 p-r/RY' 14 p-r/RY 1.9x10 p-r 3.5x10fp-r A&B 63 p-r/RY 12 p-r/RY 1.5x10 p-r 2.9x10 p-r 4 Assuming 78 plants x 31 years.

t 5.4.2 Impact The overuli industry impact of the containment closure requirement in potential extra downtime isg the sarce with Fixes A and B in place as it is without9them, i.e., $1.5x10 for an assumed real discount rate of 5% or l $0.9x10 for an assumed rate of 10%.

Based on these results and those in Section 5.2, the cost / benefit ratios for the containment closure requirement with and without the implementation of fixes A and B ore summarized in Table 7.

Table 7 Cost / benefit results for Alternative 4 Value/ Impact ($/ person-rem)

F_ixe s Site r=5% r=10%

None Hir) Pop 750 450 Low Pcp 4100 2500 A High Pop 7900 4700 Low Pop 43000 26000 A&D High Pop 10000 6000 Low Pop 52000 32000

__________._._________m________

The results in Table 7 show the variation in cost / benefit ratio between the hich- and low-population sites to be a factor of five. In comparison, the cost / benefit ratio is much less sensitive to the real interest rate parameter, varying by a factor of only two-thirds for the 5 and 10% interest rate values assumed. In the interest of simplification, these results suggest (1) eliminating interest rate as a parameter by adopting an intermediate rate of, say, 7.5% and (2) in lieu of the two bounding site oer sity cases chosen, giving consideration to a site that is more representative of a U.S. " generic" site, i.e., one based on an assumed mean density of 340 persons /sq. mile as projected f or the year 2000 (Page T52 Reference 23).

Thus, fcr an interest rate of 7.%, the present-value industry-wide 1mpact due to the containnient closure requirement is estimated to be $1.2x109 ,

assuming the requirement implies an extra downtime of 3 days /RY. For assumed downtimes of less than 3 days, the inpact would be proportionally lower.

Accordingly, MACCS code calculations of the reduction in population dose provided by the containment closure requirement for the generic site yield the cost / benefit estimates given in Table 8.

Table 8 Containment closure cost / benefit results for the generic site

• Value \alue/ Impact ($/ person-rem)

F i,xg herson-remsaverted) 3-Day D:;wntiue 1-Day Downtime None 8.0x10 5 1,500 500 A

7.7x10] 15,000 5,000 A and B 6.3x10 19,000 6,300

• 340 persons / sq. mile; interest rete = 7.%

These results indicate that, for an extra downtime impact in the range 1 to 3 days /RY, the cost / benefit ratio without the CDF-reduction fixes in place is of the order of 1000 $/ person-ren whereas, if the fixes have been in.plement-ed, the cost /ber.efit ratic 1s an order of magnitude greater.

5.5 Relationship to Generic Issue 94, " Additional Low-lenperature Overpressure F rotectier. For Light Pater Reactors" ~

The regulatory analysis perforrned for Generic issue 94 (Ref. 24) shows that the frequency of low-temperature overpressure events can be expected to be reduced if the ACl-reidted spurious closures of the RHR suction / isolation valves were to be eliminated through removal of the autoclosure interlocks. The anal) sis for Generic Issue 96 further shows that ren. oval of the ACis woulo 2x10;/RYfor68PWRunits.provp e for a best-estimate reduction in cure damage frequency of ub If this additional benefit is combined with the ccr.tainment closure berefits obtained with Fiyes A and B in piece, the total benefits for Fix C can be shown to increase by about 13%, with

• correspondingly small upward shif t in the estwated cost /tenefit ratio.

6. DECISION RATIONALE The rationale for deciding on an appropriate regulatory resolution of Generic Issue 99 involves two main considerations. First, based on the frequent losses of shutdown cooling experienced in PWRs and the related high contribution to core damage frequency, there is a clear need to improve plant capability to prevent or otherwise respond effectively to loss-of-cooling challenges during cold shutdown operations. Se coi/' in the light of mid-loop operations now being conducted with the containment open, there is a need to ensure that the probability of a loss-of-cooling accident involving the release of fission products is acceptably low or that the containment will either be closed or be capable of timely closure whenever mid-loop operations are under-taken. ,

1 Insights on the relative cost effectiveness of alternatives relating to '

these considerations have been provided by the application of probabilistic risk assessment methods to obtain estimates of the related risk-reduction values and by engineering cost estimates to determine the related impacts. The cost / benefit assessments of actions proposed under Alternatives 3 and 4 have .

been listed in Tables 5 and 7 for two sites: a high-population site (Zion) '

characterized by a 50-mile-radius population density of about 890 persons /sq. mile and a low-population site characterized by a population density of 100 persons /sq. mile. For Alternative 4, containment closure, the cost / benefit results for a U.S. generic site are given in Table 8.

With regard to Alternative 3, the cost / benefit results indicate clear support for the staff proposal to require implementation of Fix A that could reduce by a f actor of ten the estimated core damage frequency for shutdown loss-of-cooling accidents. This conclusion holds for both high-population and low-population sites. For Fix B, however, the corresponding cost / benefit results indicate less definite support for a regulatory requirement to remove the autoclosure interlocks. With regard to effecting a reduction in core damage frequency, the results in Table 1 show that, whereas implementation of Fix A by itself would reduce the core damage frequency by 3.9x10 /RY, implementation of Fix B given that ix A is carried out would provide for a further reduction in CDFofonly7x10~y/RY,i.e.,about2%oftheCDF-reductionpotentialprovided by Fix A.

With regard to the containment closure requirement, if the CDF-reduction fixes A or A & B have been appropriately implemented, the cost / benefit results in Table 8 for the generic site are seen to be approximately 15,000 to 20,000

$/ person-rem for an assumed 3-day /RY extra downtime or proportionally less if the downtime is assumed to be less then 3 days /RY.

Other considerations bearing on the need for containment closure or closure capability during mid-loop operations are discussed in the following paragraphs.

On the one hond, in support of the proposed requirement, there is the basic concern by NRC and the industry that all reasonable steps be taken to ensure that accidents involving the release of fission products or the

)

. appreciable threat thereof will not occur during plant operations, especially when the containment is-open. . Coupled with this is the expectation that the estimated 3-day downtime impact imposed by the. containment closure requirement can be~ expected to decrease in future refueling outages as.the plants learn to schedule around the requirement, which would substantially reduce the estimated  :

~

industry-wide impact and thereby lower the assessed- cost / benefit ratio for the.

requirement.

On the other hand, some' increase in the cost / benefit ratio is to be expected in taking account of certain-tacit conservatism implicit in the risk assessment calculations, e.g., (1) the assumption- that core uncovery is tantamount.to core damage whereas, depending on the decay power level and other aspects of the accident sequence, the time between such core uncovery and actual onset of release of fission products from the core may amount'to an appreciable part of an hour or more and (2) the incomplete modeling of.the spectrum of recovery actions that may be available to the operator on an emergency basis for which quantification is generally difficult unless the plant has' specific operator action emergency guidelines in place for shutdown operations.

In summary, based on the staff's evaluation of the foregoing arguments, the regulatory actions recommended by the staff. call for:

1. Requiring implementation of the CDF-reduction Fix A;

2. Recommending, but not requiring, appropriate remov.al of the autoclosure interlock feature from the RHR suction / isolation valves; and

3. Requiring implementation of the containment closure requirement, at least pending satisf actory implementation of Fix A.

].4 ti.'*"-

^

yi 4 ..

t h 17.. IMPI.EMENTATION!

~

Thes issuance by the Office of. Nuclear Reactor Regulation of Generic! Letter:

.88-17, dated October 17,1988, frequenting PWR-licensees and 'applicarits' to;

implement plant ~ improvements. pertinent to the concerns of Generic Issue 99,s

.: has.provided for resolution of this-issue.;

^

1

(

t.

u-(.1-c . .

6. REFERENCES

1. " Decay Heat Removal Problems at U.S. Pressurized Water Reactors," AE0D Case Study Report, AE0D C-503, H. Ornstein, December 1985.

2. " Zion Probabilistic Safety Study," Commonwealth Edison Company, September 1981.

3. " Zion Nuclear Plant Residual Heat Removal PRA," Electric Power Research Institute (EPRI), W. B. Reuland and G. Vine, NSAC-84, July 1985.

4. San Onofre-2 Licensing Event Report, " Loss of Shutdown Cocling Flow," LER 86-017, Event Date: March 26, 1986.

5. Waterford-3 Licensing Event Report, " Extended Loss of Shutdown Cooling Due to Steam Binding of Shutdown Cooling Pumps," LER 86-015, Event Date: July 14, 1986.

6. Sequoyah-1 Licensina Event Report, " Loss of Shutdown Decay Heat Renoval Resulting From False Indication of RCS Level in Sightglass," LER 87-012, Event Date: January 28, 1987.

7. Diablo Canyon-2 Licensing Event Report, " Interruption of RHR Flow During RCS Midloop Operation," LER-87-005, Event Date: April 10, 1987.

8. NRC Generic Letter 87-12, " Loss of Residual Heat Removal (RHR) While the Reactor Coolant System (RCS) is Fartially Filled," July 9, 1987.

9. " Loss of Residual Heat Removal System, Diablo Canyon Unit 2, April 10, 1987," NUREG-1269, NRC Augmented Inspection Team, June 1987.

10. IE Information Notice 86-101, " Loss of Decay Heat Renoval Due to Loss of Fluid Levels in Reactor Coolant System," December 12, 1986.

11. NRC Information Notice 87-23, " Loss of DHR During Low Reactor Cooling Level Operations," May 27, 1987.

12. " Residual Heat Removal Experience Review and Safety Analysis - Pressurized Water Reactors," Electric Power Research Institute, G. Vine and W. Lcyman, NSAC-52, January 1983.

) 13. 1-L Chu et al. , " Improved Reliability of Residual Heat Removal Capability in PWRs os Related to Resolution of Generic issue 99," Erookhaven National

> Laboratory, NUREG/CR-5015, April 1986.

14 D. J. Alpert et al., "HELCOR Accident Consequence Calculation Code System,"

Sandia National Laboratories, NUREG/CR-469), to be published.*

• Available in the NRC Public Document Room, 2120 L Street NW., Washington DC.

15. NRC Memorendum from A. R. Marchese, through Karl Kniel, to T. P. Speis,

" Specific Person-Rem Calculations With and Without Offsite Protective Actions," October 22, 1985.

16. USNRC, Office of Information Resources Management, " Licensed Operating l Redctors," NUREG-0020, Vol. 11, No. 1, January 1987. l

17. " Nuclear Unit Operating Experierce.: 1985-1986 Update," EPRI, EPRI hP-5544, December 1987.

18. "The Influence of fuel-Cycle Duration on Nuclear Unit Performance," EPRI, EPRI HP-5042, February 1987.

19. J. C. Van Kuiken et el., " Replacement Energy Costs for Nuclear Electricity-Generating Units in the Unitec States," Vol. 2 1987-1991, fiUREG/CR-4012. January 1987.

20. R. P. Burke et al., " Economic Risks of liuclear Power Reector Accidents,"

NUREG/CR-3673, April 1984 1 21. Science and Engineering Associates, Inc., " Generic Ccst Estimates,"

NUREG/CR-4627, June 1986.

22. L: CAP-11736, Letter from R. A. Newton, Westinghouse Owners Group, to Wayre Hodges, NRC, April 22, 1988.

23. USNRC, " Demographic Statistics Pertaining to Nuclear Power Reactor Sites,"

f!UREG-0348, November 1979.

24. B. F. Gore et al., "Value-Impact Analysis for Generic Issue 94, "Ac'ditional Low Temperature Overpressurization for Light Water Reactors,"

NUREG/CR-5186, November 1988

APPENDIX A SHUTDOWN COOLING RELIABILITY IMPROVEMENTS

1. FIX A: ACCIDENT RESPONSE AND PREVENTION OF MID-LOOP LOSSES OF RHR The basis for ensuring adequate operator capability for prevention of mid-loop losses of RHR and for accident diagnosis and mitigation is seen to involve three interrelated elements: (1) the ready availability to the operator of accurate, reliable plant status information affecting RHRS operability, RCS inventory, and feasibility of recovery procedures; (2) the availability of well-established DHR recovery procedures, including alternative backup pro-cedures; and (3) the availability of appropriate administrative controls cover-ing shutdown operations irrportant to safety. Included is the need for related training of key control room / maintenance personnel with regard to the risk implications of a loss-of-shutdown-cooling event and the related problems of prevention and mitigation.

1.1 Plant Status Information RHR System Variables Instrumentation should be provided to ensure a relit.ble, adequately sensitive, direct measurement of RHR flow, together with one or more other sensitive indicators of RHR flow (e.g., motor current, pump suction pressure). These indications should be visible to the operator and trended to provide an audible and visual alarm of a low-RHR-flow condition.

(The Safety Parameter Display System (SPDS) available at most plants or use of the plant process computer may provide a practical means for obtaining this trending capability.)

RCS Water Level Measurement Requirements Reliable, accurate roethods for measuring the RCS water level during mid-loop operations are essential to assist the operator in the difficult task of maintaining the water level within a narrow range defir.ed at the upper end by the elevation constraints imposed by the various hot- or cold-leg maintenance activities being performed and at the lower end by the need to avoid entrainment of air into the RHR suction line.

In this connection, the results of studies (Refs. 1-0) indicate that an information factor determining the likelihood of air ingestion into the RHR system for given flow ccnaitions is the depth of the RHR suction line opening below the surface of the water in the hot leg. This depth is significantly less for plant designs where the drop line exits the hot leg piping at a 30 or 45 angle to the vertical (as in inany Westinghouse plants) than it is for plants where the drop line is oriented vertically downward. The implication of this finding is that plants with a 30' or 45 drop line orientation, in comparison with those with a vertical drop line, can be expected to require greater vigilance on the part of the operator to avoid air entrainment for operations with the loop less than full.

t

[

The measurement of water level during mid-loop operations constitutes an activity important to safety requiring well established procedures in regard to installation, maintenance, calibration, testing, and operation of the level-measuring instrumentation. The licensee should:

1. Review and upgrade as required the procedures calling for strict I limits on water level during mid-loop operations, frequent ,

comparisons of level indications, calibration checks of levels.in I reference legs, precautionary actions if RCS draining or water addition operations are not appropriately reflected in water-level indication changes, and appropriate control of valve positions that i

can affect level indication.

2. Ensure the capability.for redundant, diverse, and independent measure-ments of water level within the reactor vessel or hot legs with con-trol room indications and an audible low-level alarm to protect against air ingestion into the RHR system and a high-level alarm to protect personnel against spillage from openings in the RCS.

3. Provide adeouate resolution in level-measurement capability to ensure reliable operational control of the water level during mid-loop operations. (Adequate resolution may be expected to depend on the RHR design and operational characteristics specific to the plant, in particular, for a given-flow of RHR, on the depth of the RHR suction line opening below the surface of the water level in the hot leg.)

4. For direct-reading water-level measurement systems, provide for the repiccement of temporary tygon tube arrangements by a permanently mounted visuhl standpipe level system with a calibrated level scale and a TV camera system to provide continuous readout' of the standpipe level in the control room. The remote-reading capatility ensures that water-level information will be available to the operator even if the containment must be evacuated.

5. Provide for design protectior against instrumentation errors caused by changes in RCS pressure (e.g., the Katerford-3 event of July 14, 1986, in whicn insufficient nitrocen pressure resulted in collapse of the tygon tube and consequent inaccurate indication of level).

6. Provide for appropriately established correlations of RHR flow rates and minimum wa.ter levels for operator assistance in the control of water level during mid-loop operations.

7. Provide means (e.g., SPDS) for trend analysis of reactor vessel water-level changes, RCS water additions, and RCS and RHR system leakages.

A-2

RCS Water Temperature Measurement. Since hot- and cold-leg RTD temperat-ure measurements are not meaningful following a loss of RHR flow, provision  ;

should be made for a minimum of two measurements of reactor vessel water temp- I erature (with. control room indications) as may be obtained through use of the existing core exit thermocouple (CETs) with appropriate administrative con-trols to ensure that the CETs are connected prior to initiation of mid-loop operations.

Valve Position Indication Protection. Protection is required against the RHR suction / isolation valve position indication lights being disabled by a single power source failure.

Accident _ Diagnostic Aids. The following should be provided to assist the l operator in diagnosing potential accidents:

1. Status of RCS/RHRS-Related Equipment. Means'(e.g., status board or SPDS) should be provided for ensuring that the operator has prompt access to information relating to the status of all plant equipment .

and related maintenance activities on components that could affect '

RHR system operability, RCS inventory control, and the operability of systems required for the implementation of alternative decay heat removal schemes.

2. Time to Bulk Boiling / Core Uncovery. Curves of estimated time to bulk boiling and core uncovery for all anticipated mid-loop operating configurations and decay power conditions should be t provided to assist the operator in accident diagnosis.

l 1

1.2 Accident Response Procedures For Recovery of Shutdown Cooling j i

- For a loss of RHR initiated under unf avorable conditions (e.g., reduced I RCS inventory, high deca be significantly short (yless heatthan rates), theI hour),

about estimated time so that to core uncovery successful recovery may  !

will be highly dependent on the operator taking prompt correct action to {

mitigate accident conditions. This implies the availability of established 1 abnormal operating procedures and related training capable of addressing the  :

variety of conditions anticipated. Accordingly, licensees should: {

l

1. Review and upgrade procedures and training for responding to a j loss-of-RHR alarm during mid-loop operations (e.g., procedures for 1 the prompt recovery of air / steam-bound RHR pumps, inclusion of an  ;

operational precaution not to start the second RHR pump until the i

) conditions leading to the air binding of the first pump have been {

corrected). j

) 2. Establish procedures and related controls for several alternative feasible DHR methods, including consideration of feed and and bleed {

operations, identification and availability of alternative water i sources / flow paths / injection modes, and the effectiveness of steaming i in the reactor vessel combined with reflux condensation in the steam I generators.

i f I

)

l I

A-3 1

i

)

3. Provide for direct communication between the control room and per-sonnel in the containment working on equipment important to safety of operations under partially drained RCS conditions (i.e., when the RCS level is more'than 1 ft below vessel flange level).

4. Establish decision criteria for operator designation of an extended loss of decay heat cooling during mid-loop operations as an Unusual Event if the DHR function is not restored within 15 minutes and an Altrt if the temperature of the water in the reactor vessel is deter-mined to have reached 200 F.

5. Establish precedures and operator decision criteria to ensure the closure of containment penetrations prior to the estimatea time to core uncovery under extended loss-of-DHP, conditions during mid-loop operations.

1.3 Administrative Controls During plant shutdowns, the many different maintenance ano surveillance activities going on in the plant can provide for a working environment con-ducive to human errors. Consequently, a well-planned outage program incorporat-ing well-established procedures is required to minimize the time spent under reduced inventory conditions and, in general, to reduce the likelihood of personnel errors leading to losses of RHR. Beyond this, plant experience on RHR losses indicates that adequately tight administrative controls are essential to ensure that the operational command in the centrol room knows and has control of all plant activities aff ecting plant safety. For example, in the Diablo Canyon-E event of April 10, 1987, tighter administrative controls could have precluded the 75-min delay in the operator knowing that the two steam generator manways were intact and that coolant could be safely added to halt the thermal transient.

Accordingly, administrative controls should be reviewed and upgraded as required to ensure that:

1. Primary responsibility for all activities in the plant resides in a single inoividual (e.g., the shi1t supervisor) who authorizes all work involving major plant evolutions (in particular, those that could affect RHP avaiMbility, RCS integrity ana inventory, and the feasibility of alternative DHR recovery procedures) and with whom coordination is required prior to the actual initiation of such work.

2. There is adequate communication between shifts and within each shif t through updated reviews of equipment stotus and work programs held at least once each shift, especially as related to equipment and testing affecting RCS inventory.

3. Adequacy of the controls regarding installation, testing, calibrat-ion, and operation of instrumentation required for measurement of RCS/RHR variables important to safety for mid-loop operations as related to the requirements set forth under Section 1.1.

A-4

1

4. Critical valve. positions'are in the correct alignment.to preclude inadvertent 10sscs of coolant (LOCAs) during RHR operations, in particular, under reduced-inventory operating conditions.

1.4 Technical. Specification Change Prepare or revise:the technical specification relating to plant operations' .

With the RCS partially drained (i.e. , STS-3.4.1.4.2) to require:

.1. Operability of two RHR loops with at least one RHR loop in operation.

2. At least two pumps (e.g., charging, safety injection) of edequate capacity to be operable or capable of being made operable prior to a rise of 20 F in reactor vessel water temperature following a loss of decay heat removcl tapabilit,y.

3. Operabih ty of at least one flow path from the RWST or, accumulators to ensure coolant injecticn on loss of til AC as perraitted by RCS pressure conditions.

. 4. Modify. the single-a:>terisk footnote in STS-3.4.1.4.2-to state that one.RHR loop may be inoperable for up to 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> for surveillance i testir.g provided the other RHR loop is operable and in operation ar.d 'I at'least two pumps are operable or capable of being made or ..?ble'as I provided for under 2 above.

)

5. Modify the double-asterisk footnote in STS-3.4.1.4.2 to state that one RHR pump rray be 'deenergized for up to I hour provided (1) no operations.are permitted that wculd cause dilution of- the RCS boron concentration ano (2) core outlet temperature is maintained at least 75"F below saturation tercperature.

2. FIX D: PREVENTIOM 0F ACI-REl.ATED LOSSES OF RHR Staff guidance on removal of the ACI featcres to. reduce related spurious

-losses of.RUR was set forth in a Jaruary 1985 RSB memorandum (Ref. 9). This celled f or review of plant proposals for removal of the' ACI f eature on a Case-by-Case ba!is, each proposal to derrenstrate that the change constitutes a net irrprovement it, safety, including consideration of the following aspects:

1. The means available to minirr,ize Event V concerns, The alarms to alert the operator of an irrproperly positioned RHR i

2.

motor-operated valve (M0V).

3. /,dequacy of the RHR relief valve capacity.

I A-5

4. Means other than the ACI to ensure that both MOVs are closed (e.g.,

a single switch actuating both valves).

5. Assurance that the function of the open permissive circuitry is not affected by the proposed change.

6. Assurance that M0V position indication will remain available in the control room regardless of the proposed change.

7. An assessment of the effect of the proposed change on RHR reliability as well as on LTOP concerns.

Requests from the Kewaunee and Diablo Canyon-1, 2 plants for the removal of the ACI feature in accordance with the above guidance have been reviewed and concurred in by the staff (Refs.10,11). For Diablo Canyon, the staff reviewed the supporting analysis submitted by the utility and concurred that the removal of the ACI as proposed by the utility did not constitute an unreviewed safety question and did not involve a change in technical specifications. The staff has also approved the requests from the Callaway and Wolf Creek plants (Refs.

12, 13) for a different approach to the prevention of ACI-related common-cause failures of the RHR system. Both plants have double drop lines posedmodificationinvolvedtheremovalofpowerfromthe(open}andthepro- inboard valve in one drop line and from.the (open) outboard valve of the second drop line while retaining the autoclosure feature on the remaining two valves. By this means, a double ACI-related failure is required for failure of both trains of the RHR system; in addition, electric power is always available to one of the valves in each drop line to provide for isolation of the RHR system from the RCS in the event of an RHR-related LOCA. This approach is clearly not applic-able to plants with a single drop line. In general, complete removal of the ACI is of advantage in that it effectively designs out the problem of inadvert-ent valve closures due to ACI-related human errors or hardware failures.

The Westinghouse Owners Group has submitted a generic reliability study (Ref. 14) evaluating a proposed action to remove the ACI feature in four classes of Westinghouse plants. The related modifications would be along the lines of the Diablo Canyon request and entail (1) installation of the safety-grade alarm system to alert the operator if any of the isolation / suction valves is open above a given pressure setpoint, (2) protection against loss of power to the valve position indication lights if power to the valve motors is removed, (3) related procedural changes, and (4) adequate capacity of the RHR safety relief valves with regard to overpressure transients. If the study is approved, it is intended that it serve as an analytical basis for ACI removal requests from individual Westinghouse plants.

3. FIX C: CONTAINMENT CLOSURE CAPABILITY FOR MID-LOOP OPERATIONS Current plant operating procedures and technical specifications do not generally require the closure of containment penetrations to the outside environment during reduced inventory operations. Since plant estimates of the time to close the large equipment hatch are indicated (in the plant responses to Generic Letter 87-12) to be typically upwards of two hours for most PWRs, it is unlikely that timely containment protection against offsite releases can A-6

generally be ensurea, at least for certain mid-loop accident scenarios where the estimated times to core uncovery may be one hour or less.

Accordingly, the proposed staff requirement calls for revisirig the mid-loop operating procedures to require that all containment penetrations that cannot be ensured of being effectively closed (e.g., by at least four bolts for the equipment hatch) within the estimated time to core uncovery (starting from the time the operator decides to close the penetrations) be closed prior to entering operations with the RCS drained to the mid-plane of the loop.

A-7

APPENDIX A REFERENCES

1. ." Containment Emergency Sump Performance, . Technical findings A..W,'Serki:$1A-43,"NUREG0897, Related to U Revision 1, November 1984,.

2.- NRC Memorandum from A.LH. Spano to K. Kniel, " Issue of Air ingestion and-Vortex Formation During Mid-Loop Operations," June 8, 1987,

,1

3. P. S. Kamath et al., "An Assessment'of Residual Heat Removal and Containment-Spray Pump Performance Under Air and Debris Ingesting-Conditions," NUREG/CR-2792, September _1982,

4. V. E. Schrock et al., "Small; Break Critical Discharge . The Roles of Vapor and Liquid Entrainment in a Stratified Two-Phase Region Upstream of the Break," NUREG/CR-4761, December 1986,

5. H. Zuber, " Problems in.Modeling' of Small Break LOCA," NUREG-0724, October 1980.

6. Y. R. Reddy and.J. A. Pickford, " Vortices at Intakes in Conventional

. Sumps," Water Power, March 1972.

7. B. T. Lublin and G. S. Springer, "The formation of a Di Liquid Draining-from a Tank,"'J. Fluid Mech., Vol. pp. 29,3G5 p on1967, 390, a Surface of a

8. J. L. Gordaon, " Vortices at Intakes," Water Power, pp. 137 138, April 1970.

9. .NRC Memorandum from B..W. Sheron to RSB Members, "Autoclosure Interlocks for PWR Residual Heat Removal (RHR) Systems," January 28, 1985,

10. Letter from S. A. Varga, NRC, to D.'C. Hintz, Wisconsin Public Service Corporation,' January 16, 1985.

11. Letter from H. Rood, NRC, to J. D. Shiffer, Pacific Gas and Electric

" Safety Evaluation of the Removal of the Autoclosure Interlock Function,"

February 17, 1988.

12. Letter from B. J. Youngblood, NRC, to D. Schnell, Union Electric Company, July 30, 1985.

13. Letter from B. J. Youngblood, NRC, to G. L. Koester, Kansas Gas and Electic Company, August 16, 1985.

14 WCAP-11736, Letter from R. A. Newton, Westinghouse Owners Group, to Wayne Hodges, NRC, April 22, 1988.

A-8

APPENDIX B Cost Analysis of Proposed. Requirements to Improve Residual Heat Removal Reliability-l NRC Contract No. NRC 04 87-086 l Task Order No. 005 '

1.0 BACKGROUND

There is a long history of problems associated with the operations of Residual Heat Removal (RHR) systems on Pressurized Water Reactors (PWRs) when the Reactor Coolant System (RCS) is partially drained for maintenance. Some of these RIIR problems have resulted in temporary loss of decay heat removal. The Reactor and Plant Safety issues Branch (RPSIB) of the Office of Nuclear Regulatory Research of the Nuclear Regulatory Commission (NRC) identified the RHR problems under the general title of Generic Issue 99: Loss of Residual Heat Removal System When Plant is Shut-Down. The RPSIB had also identified, in a preliminary manner, several potential improvements to the PWRs RHR systems while operating with the Reactor Coolant System (RCS) partially drained for maintenance. These suggested modifications include both hardwr.re modifications as well as operational / procedural and training improvements.

2.0 PURPOSE The Regulation Development Branch (RDB), also of the Office of Nuclear Regulatory Research, has been approached by the RPSIB to provide a cost analysis of 21  ;

proposed modifications to improve RHR reliability. The RDB, in tum, requested Science and Engineering Associates (SEA) Inc. to provide a rough estimate of the costs to ,

implement each of the RPSIB suggested modifications with particular sensitivity to paperwork related costs (e.g., analysis, procedural changes, training, etc.). This report documents the efforts conducted under Task Order No. 005 of the NRC contract No.

NRC-04-87-086.

Due to the quick tum around requirements, the cost analyses performed under this task were not comprehensive and constferable " engineering judgement " was used to produce the cost estimates. The cost estimate for each task should be treated as a rough ,

estimate only. The relative costs of each task, however, can be treated with a higher degree

{

B-1

of accuracy, since common underlying assumptions were used to estimate the costs for each area.

l 3,0 TECIINICAL APPROACil i The RPSIB originally identified twenty one (21) areas of potential improvements to the RHR, transmitted to SEA on a draft statement of work on April 21 1988:

1. Trend recorder indications of RHR flow with low-Oow nlarm.

2, Upgrade procedures for water level measurement / control during mid-loop operations.'

3. Instrumentation for diverse, independent water level measurements, control room indications and alarms.

4. Operator aid correlations of RHR flow rates and minimum water level.

5. Instrumentation for treral analysis of RCS inventory.

6. Procedures to ensure connection of redundant Core Exit Thermocouple (CETs) during mid-loop operations.

7. Protection against loss of RHR suction valve position indication on removal of power from the valve.

8 Operator aid on status of RCS / RHR - related plant equipment.

9. Operator aid on times to bulk boiling / core uncover y.

' 10. Upgrade procedures for response to loss of RHR related alarms (e.g.,

e RHR pump recovery procedures).

I 1. Establish procedures and controls for alternative, feasible, decay heat removal (DHR) recovery methods.

I 2. Direct communications between control room and containment during mid-loop operations.

13. Establish decision criteria for designation of extended loss-of-RHR events.

14. Establish decision criteria and procedures for closure / evacuation of containment.

• Mid Loop operations refers to those operations conducted on the RCS of PWRs while the reactor coolant level has been lowered to the mid-point of the hot-leg nozzles.

B-2

l

15. Control room authorization and coordination on work involving major plant evolutions.

16. Upgrade communications between and within shifts on equipment statu:

and work programs.

17. Upgrade procedures / controls on instrumentation for mid-loop measurements of RCS/RHR variables.

18. Confirmation of correct alignment of critical valve positions.

19. Technical specification on containment closure capability for mid-loop operations.

20. Operability.of redundant means for coolant injection for DHR recovery.

21. Implement removal of RHR Auto Closure Interlock (ACI) feature.

The approach taken by SEA was to perform the following activities on each of the above areas of potential RHR reliability improvements:

A. Review available pertinent background material provided by the NRC as well as other sources ofinformation.

B. Define elements of each of the suggested RHR improvements which could have a cost impact.

C. Prepare an initial rough cost estimate for each of the suggested RHR improvements.

A review of the issees concerning some of the loss of RHR events, Generic Letter 87-12 " Loss of RHR while the RCS is Partially Filled" as well as responses from Pacific Gas and Electric (PGandE) Diablo Canyon Plants and San Onofre Nuclear Generating Station (SONGS) was performed by the SEA team of engineers to become familiar with the background as well as to understand the basis and technicalities of the proposed RHR modifications. Additionally, comments and inputs received from m.mbers of the RDB were also used to further clarify requirements of the proposed modifications.

Seven categories of costs were determined to provide sufficient elements for estimating the costs of the proposed modifications on a per plant basis. These categories were: material / equipment, installation labor, engineering and quality assurance / control, procedure rewrites, training, technical specification changes, and NRC review. An estimate of worker radiation exposure costs was also determined. The exposure was dependent on the nature of the modification, the location of the work within the plant, shielding measures taken during the retrofit phase, and several other factors.

B-3

A four loop,1100 MWe Westinghouse PWR design was used for estimating the costs of hardware related modifications (material and installation labor elements).

Training, technical specification changes, and NRC review costs were estimated as global costs, i.e., these three costs elements were determined to be independent of any particular set modifications and would be accrued if any modification to the RHR system were performed.

Radiation exposure cost estimates were derived based on guidelines presented in Abstract 4.1 of Generic Cost Estimates (NUREG/CR-4627). In this cost analysis, radiation exposure costs were incurred by only two activities: RHR Modifications #2 and 1 #5. ALARA considerations such as the installation of temporary shielding and system decontamination were not included inthe estimate. The collective radiation exposure was estimated by taking the product of the labor hours necessary to perform the task and the work area dose-rate associated with that particular task. In order to balance these exposures against economic costs, a $1000/ man-rem conversion factor was used.

Procedure writing, reviews, and rewrite cost estimates were determined for each of the proposed modifications based on whether the requirements were considered simple or complex. A simple operating procedure change entails a straightforward modification such as changing a valve setting or the value of a sensor setting. Revisions and modifications to the written procedure are limited to an equivalent of one page for the simple category. A complex operating procedure change could require engineering analysis at.3 research. The point estimate cost for the simple category is $915.00, whereas the complex category estimate is $3,650.00, both in 1988 dollars. Further discussions of these estimates can be found in Section 4.2.

The engineering and quality assurance / control costs account for the cost of engineering and design, as well as quality assurance (QA) and quality control (QC) activities, associated with implementing the requirements. A study of the relationship of these costs with the total direct cost of material, equipment, and labor was conducted by SEA under a separate contract to the NRC. The study concluded that a reasonable approximation of the combined cost for engineering, design, QA, and QC can be obtained by using factors of 25% for changes to plants well along in construction (typically more than 75% complete) and operating plants, and 30% for new plants. The basis for these value.s and a more detailed breakdown of engineering and quality control costs by EEDB code of accounts is available in the document Encineerine and Ouality Assurance Cost Associated With Nuclear Plant Modification (NUREG/CR-4921). There are cases where Engineering /QA/QC costs are greater than normally anticipated (i.e., minimal structure / system modifications but major engineering analysis effort) or are lower than B-4

anticipated (i.e., installation of off-the-shelf items requiring a minor amount of engineering). For this study, a 25% Engineering and QA/QC factor was applied to total direct cost, i.e., material / equipment and labor costs.

A preliminary cost estimate was performed for each of the original twenty-one proposed modifications. After a review by the NRC, the proposed modifications numbers 2,6,11,17, and 19 were either eliminated or combined with ene or another of the proposed modifications. The final list of proposed RHR modifications now has sixteen items. The cost estimate for each of the modifications, as well as the basis for these costs are discussed in the next section.

4.0 COST ANALYSIS 4.1 Basis of Individual Cost Estimates

. The basis and underlying assumptions used to obtain the cost estimates for each of proposed modifications to the RHR system are presented. The basis for the global cost estimates of training, technical specification changes, and NRC review costs are presented in Section 4.2.

4.1.1 Trend Analysis of RHR Flow With Low flow Alarm This proposed modification is to provide a display of RHR flow trends as well as generation of an alarm if the RHR flow falls below a predetermined flow rate.

Maximum use of currently installed equipment was presumed. Signals from the currently installed RHR flow sensors could be provided to the plant computer or the Safety Parameter Display System (SPDS) for RHR flow trending, detection oflow RHR flow conditions, and display of RHR flow and trends.

Costs for this modification are presented in the summary table. Modifications to existing pmcedures were estimated to be complex.

4.1.2. Instrumentation / Procedures for Diverse, Independent Water Level Measurements - Control Room Indications and Alarms This proposed modification entails the installation of two diverse RCS level measurement systems for use during mid-loop operations. The proposed systems are variants of currently used temporary Reactor Vessel Refueling Level Indicating System B-5

l (RVRLIS) instrumentation. One of the water level measurement system would be based on a dedicated wide and narrow range differential pressure sensor and a display in the control room with high and low level alarms. The second water level measurement system would be based on leve measurements by a permanent stand pipe monitored with a dedicated TV camera and a TV monitor in the control room.

SEA estimated the costs based on a straightforward conceptual designs. Both of the systems we aid use available fittings to communicate with the interior of the RCS - no new RCS pressure boundary penetrations were judged to be required. Pressure boundary isolation would be accomplished by using currently installed valves. Each of the systems would have six new manually operated valves: two (in series) at the RCS water level input, two (in series) at the RCS reference input, one at the low point of the tubing for draining, and one at the high point of the tubing for venting. All tubing, valves, and sensors are not required to be rated for high pressure since these systems would only be used when the RCS is depressurized. Approximately 50 ft. runs of tubing inside the containment building were estimated to be required for each of the systems. A total of 300 ft. ruc of cabling,40 ft. inside the containment building, was estimated to provide the signals to the control room area. Existing cable trays were presumed to be used. The differential pressum (Ap) sensor system signals would be tied into the plant computer for logging and trending. The stand pipe / TV system signals, to be mainly used as a back-up and confirmation of the Ap system readings, would not be electronically recorded. ,

Costs for this modification are presented in the summary table. Generation and modifications to curmnt procedures were estimated as complex.

4.1.3. Operator Aid Correlations of RIIR Flow Rates and Minimum Water Level.

This proposed modification entails the preparation of simple aids to the operator to correlate RHR flow rates with RCS water level indications. The operator would be provided either tables or nomographs correlating the two variables and depicting caution zones as well as recommended operational zones.

SEA estimated that this modification would not require hardware modifications.

The estimated individual cost entails procedural changes rated as complex.

4.1.4. Trend Analysis of RCS Inventory.

This proposed modification is to automatically provide a status of the RCS liquid  ;

inventory based on existing flow meters. Flow signals of liquid flow into and out of the B-6

RCS could be monitored by a plant coraputer and a current estimate of the RCS inventory maintained.

Costs for this modification are presented in the summary table. The estimated procedural changes were rated as simple.

4.1.5. Protection Against Loss of RHR Suction Valve Position Indication on Removal of Power From the Suction Valve This proposed modification is to provide a power source to the position indicators of the RHR suction valves independent of the power source which operates these valves. In this fashion, on loss of power to the RHR valves, the operator would still be able to determine the valve position.

SEA estimated that this modification would entail installing four new cables between the valve position indicators in the control room and the valves in containment.

Two motor operated valves,in series, were presumed to exist on each of two RHR suction lines. The cost estimate for this modification are presented in the summary table and the procedural modification was rated as simple.

4.1.6. Operator Aid on Status of RCS / RHR - Related Plant Equipment This proposed modification is to provide the operator with a check list of recommended line-ups and operations of RCS/RHR related plant equipment during operation of the RHR system. Specific attention would be given to mid-loop operational requirements.

SEA estimated that this modification would not require hardwarc modifications.

The check list could be used in conjunction with a currently installed plant computer as well as individual displays. The estimated individual cost entails procedural changes rated as simple.

4.1.7. Operator Aid on Times to Bulk Boiling / core Uncovery This proposed RHR enhancement would provide the operator with a simple set of curves to assist in determining the remaining times to bulk core boiling and / or core uncovery based on variables such as RCS inventory, temperatures, and estimated residual decay heat power levels.

B-7

SEA estimated that this modification would not require hardware modifications.

- The curves would be used with variables obtained form the currently installed plant computer or individual displays. The estimated individual cost entails procedural changes rated as complex.

4.1.8. Upgrade Loss of RHR Response Procedures and Controls (e.g.,

RHR Pump Recovery, Alternative DHR Recovery Methods)

This proposed RHR enhancement entails reviewing current procedures 'or responding to out-of-limit indications of RHR equipment, specifically during mid-loop operations. Attention should be given to situations such as air ingestion of RHR pump.

SEA estimated that this modification would not require hardware modifications.

The estimated individual cost entails review of current procedures and procedural changes which are rated as complex.

4.1.9 Direct Communications Between Control Room and Containment During Mid loop Operations Reviews of recent events involving deficient operation of the RHR during mid-loop operations revealed extraordinarily long times for the control room operators to obtain information on equipment status in the containment building. This proposed enhancement would be to review current mid-loop operational procedures and verify that adequate communication is established and maintained between the control room and personnel warking inside the containment building.

SEA estimated that this modification would not require hardware modifications.

The estimated individual cost entails procedural changes rated as simple.

4.1.10. Establish Decision Criteria for Designation of Extended Loss-of-RHR Events In light of current events associated with RHR operations, there are indications that there may be a need to improve the criteria associated with the designation of the extended loss of RHR category. The proposed enhancement would be to review the current procedures for loss of RHR and verify that clear criteria exist to designate the occurrence as an extended loss of RHR event.

B-8

SEA estimated that this modification would not require hardware modifications.

The estimated individual cost entails procedural changes rated as complex.

4.1.11 Establish Decision Criteria and Procedures for Closure of l

Containment for Mid-loop Loss of RHR Event l

This proposed enhancement is to verify that the containment building can be closed expeditiously in the event ofloss of RHR, specifically during mid-loop operations.

This enhancement essentially entails review of all operations which are conducted during RHR operations to assure that the containment building can be isolated within a reasonable time.

SEA estimated that this modification would not require hardware modifications.  ;

The estimated individual cost entails procedural changes rated as complex.

4.1.12. Control Room Authorization and Coordination on Work Involving Major Plant Evolutions. J This proposed enhancement is to review all operational procedures which could occur during RHR operations and insure that the control room operators are fully cognizant of their occurrence.

SEA estimated that this modification would not require hardware modifications.

The estimated individual cost entails procedural changes rated as simple.

4.1.13. Upgrade Communications Between and Within Shifts on Equipment Status and Work Programs This proposed enhancement is to review all operational procedures which would occur during RHR operations to insure that adequate information on the status of equipment and opemtions is transmitted between shifts.

SEA estimated that this modification would not require hardware modifications.

The estimated individual cost entails procedural changes rated as simple.

4.1.14. Confirmation of Correct Alignment of Critical Valve Positions.

This proposed enhancement is to review all RHR related procedures to verify that adequate checks of valve positions are performed prior to re-alignments or changes to .

RIIR operations.

i B-9

_ _ _ _ _ _ _ _ _ _ _ _ - - - - _ _ _ _ _ _ _ _ _ _ l

SEA estimated that this modification would not require hardware modifications.

The estimated individual cost entails pmcedural changes rated as simple.

4.1.15. Operability of Redundant Means for Coolant Injection for Decay Heat Removal Recovery This proposed enhancement would be to review RHR related procedures and ascenain methods and equipment required to pmvide redundant coolant injection far decpy heat removalin the event ofloss of the RHR system. The procedures for operatioq of the j redundant means of coolant injection for decay heat removal would also be reviewed to assum adequate testing and verification of their operation.

SEA estunated that this modification would not require hardware modifications.

The estimated individual cost entails procedural changes rated as complex.

4.1.16. Removal of RIIR Auto Closure Interlock (ACI) Feature This proposed enhancement would remove the Auto Closure Interipck (A CQ feature from the RHR system. Experience has indicated that the ACI feature pan causa inadvertent shutdown of the RHR due to spurious signals or under special loss-of-power corxiitions not related to RHR operations.

SEA estimates that an electrician could remove ACI feature related cables in 39 hours4.513889e-4 days <br />0.0108 hours <br />6.448413e-5 weeks <br />1.48395e-5 months <br />. Approximately 648 hours0.0075 days <br />0.18 hours <br />0.00107 weeks <br />2.46564e-4 months <br /> were estimated foi review and reprogramming interlock and RCS pressure related logic. An additional $70,000.00 was factored in this cost estimates to account for additional plant specific engineering analysis. Procedural modifications were rated as simple. The present worth of the annual operation and maintenance cost savings (associated with the removed equipment) over the remaining useful life of the reactor (25 years) was also calculated.

4.2 Basis of Global Cost Estimates This section addresses the basis and underlying assumptions used to obtain cost estimates of training, technical specification changes, and NRC review costs. SEA presumed that these costs would be accmed in the event that a plant institutes one or more of the proposed RHR recommendations.

The cost estimates for write /re write procedures, training, and revision of the training manual were based on information obtained from The Identification and B-10

Final Report (NRC contract number 33-84-407-006). Findings of this report are also compiled in Abstracts 2.2.2, 2.2.3, and 2.2.4 of NUREG/CR - 4627, Generic Cost Estimates.

l 4.2.1 Write / Rewrite Procedures Ooeratinc Procedure Change Simple - a straightforward, simple modification such as changing a valve setting or the value of a sensor setting. Requires modifications on a single page of a procedure. Direct labor, geographical variance plus fringe benefits (0.7 added to cost) are considered. The average point cost estimate is $915 (in 1988 dollars).

Complex - revisions to 50% (approximately 10 pages) of the operating procedure requiring considerable research and some innovative analysis. Direct labor, geographical variance plus fringe benefits (0.7 added to cost) are considered. The average point cost estimate is $3.650 (in 1988 dollars).

The following table (Table 2-1 in the SEA report) presents four costs estimates for re-writing procedures. They were supplied by experienced personnel at four nuclear power plants and represent differences in plant organization and management philosophy. A detailed breakdown of the personnel categories and labor-hours is found in the SEA repon mentioned above.

Routine Cha:.ge Cost (1986$) Complex Change Cost (1986$)

Direct labor + Labor + Direct labor + Labor +

Geoc. Variance Frince Geoc. Variance Frinze Plant #1 262- 333 445- 566 2583 - 3287 4391 - 5588 Plant #2 745- 949 1267 - 1613 2371 - 3017 4031 - 5129 Plant #3 629 - 801 1 % 9 - 1362 1786- 2658 3036 -4519 Plant #4 142 - 180 241 - 306 574 - 738 976- 1255 AVERAGE (rounded) 750 - 1000 3100- 4100 POINT ESTIMATE (rounded 1986 $) 900 3600 B-ll

i The point cost estimates were escalated to 1988 according to the guidelines presented in Abstract 6.4 of NUREG/CR - 4627, Generic Cost Estimates. Their new values are 3915 and $3,650 for routine and complex change costs, respectively. They will provide order of magnitude cost estimates.

l 4.2.2 Training Assume: Training costs include:

- costs associated with instructor & training facility, plus

- costs associated with trainee's time (salary + fringe)

Ia) 6 shifts- Operations 1 Shift Supervisor (SRO) 1 CR Operator (SRO) 2 CR Operators (RO) 2 Operators (unlicensed)

Ib) foreach shift-16 hrs. classroom training  ;

8 hrs. simulator training 8 hrs. on-the-job training i

The following labor rates (or costs) were used to calculate training costs (in 1986 dollars) and are based on infonnation obtained from The Identification and Estimation of the Cost ,

of Reauired Procedural Chances at Nuclear Power Plants, SEA Final Report (NRC contract number 33-84-407-006).

Trainees

- Shift supervisor's salary = $30/hr x 1.7 (fringe) = $51/hr

- SRO's salary = $25/hr x 1.7 (fringe) = $42.5/hr

- RO's salary = $22/hr x 1.7 (fringe) = $37.4/hr

- Operator (unlicensed) = $15/hr x 1.7 (fringe) = $25.5/hr Instructor & Traininc Facility Costs

- Simulator "in-house"; costs = $33/ student hour

- Classroom training; costs = $17/ student hour

- On-the-job training; costs = $7/ student hout B-12

The costs for training the shift supervisor are:

simulator training costs = ($51+ $33) x 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> = $672 classroom training costs = ($51+ $17) x 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> = $1,088 on-the-job training costs = ($51+ $7) x 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> = $464

! Total Shift Supervisor's Training Costs = $2,224 To reflect 1988 dollars the above rates and costs were escalated according to the guidelines presented in Abstract 6.4 of NUREG/CR - 4627, Generic Cost Estimates.

For each shift- 16 hrs. classroom training - costs = $5,200 8 hrs. simulator training - costs = $3,375 8 hrs. on-the-job training - costs = $2,115 For 6 shifts, costs = $64,140 (1988 $)

iia) 6 shifts- Maintenance 1 Shift Supervisor- 8 hrs. requalification on control procedures 8 hrs. classroom training - costs = $550 For 6 shifts, costs = $3,300 (1988 $)

REVISE TRAINING MANUAL Assumb: 100 pages @ $70/page Cost = $7,000 (1988 $)

Total Training Costs = $64,140+$3,300+$7,000 = $74,440 (1988 $).

4.2.3 Technical Specification Change The estimates of licensee costs for technical specification change were obtained from Abstracts 2.2.1 of NUREG/CR - 4627, Generic Cost Estimates. The estimates do not include costs ofimplementing the revision. An additional cost allowance should be included if the technical specification change considered will require public hearings.

Complicated task assumes 16 staff-weeks of utility technical, legal, management, and committee input. Staff hours are costed at $53 per hour.

Cost = $34,000 (1988 $)

B-13

ooo ocoNeoooooooom comMcNooooecooaw J

4 ogo ooo aqqqqoqqqampoo.

mm-neomNNoNno p

o o T T_N,C g q q -1,q q q o o m ommsmoNNoNosooo*.

MNMNWemW mWO N a MNOWmmNmW WW-N N 9 g N m - N N,M, N eN- ,m-Nm p M -

< Nb 3 oooooMnNoo@@ comb e oooooMNooWWWW*oNmn -

-o-Me

-eneo M

o ocooowemooo-ow-o e cocoowmoco-oe-oane m e a m i e. o soi me ad ,ow--N-~a, ni n o. a h e- o d o m i N d o d n i n N e e,*.W e n. N

$! NNanN m- eme-owan

• eme-monn--N-nN *

v. . ~. M = ~. . n N n n . n = ~. w . ~. n =. . N. n n n n o. *. =. < o. < ~

< ~

<, wm*~ * .

a N. u. . .-neNnn--m- , eme-nNen--m-nNN ,

ao oUooooooooooo ooooooooooooooo Nm3 PooooNNNooNNNS ..Q ,, =mmmmm--mm-----n

= o 9 a N e o o.N N N N a n 3 .... p e ang y.. p

$ ,w -n - mamme---mm-- g -

NNNNNnnNunannnnggf o*

en NNNNNannNNnnn n&8 5 n88 5x n&B 88xB o 8 48x o asxch iH omooNwoeoomne uh o oeooNoeooeneaea SB O E

ao no y" a $ NnNewnneeeeNo 22 NnNwonewweNoNuo

• 0, We jg ne o mWessWWeedsde 2 E N mWdeWWnadefeWWo cowwmoeoNwnon

~C

. momemooNwnwnene NNN -- - G.E/ &

<g a -y# w NNN eN -- ~ g# e 58

• a + 8 - +8

-E - a- o z S

A o m. *

• o S. v.

o B. v. #$ "m f $. $ 5$. $H< m.

Bf Om

.U.

Q #

ed E #f e e w<

wm ao B I

E ooooooooooooo 3 ooooooooooooooo MMMMMMMMMMMMMMo g

$ e y

x. No MMMMMMMMMMMMN NNNNNNNNNNNNN n. NNNNNNNNNNNNNN-03 N- a.

o o o so

<0 O O O y

am 0 0 o o

• 4 4 2 p o

, h5 oo ao

@ ooooooooooooo

• • • • • • • • • • • • • 5 N ooooooooooooooo "He***********o -l 3 ou 5 ,we,,v,*------- P D< -- g ,,,,,,**,---- - G E c m S o w E w

-P w

- A

• oo ooomoonoooono ooomonoooonocoo 0000 005N009 009990 0 0

• M o - o 9 0. o 9 O $g M 09 .

em oWooNeomoooom o@ooNoeoomoeoWo N-e T N- - -- N- --

B -e 4

a

- t1

• z

. D b ^

5 mb z c< ooooooooooooooo w m ao ooooooooooooo

-e --ee --o-ee wee--ee wowe woOO E w oo e ENN ANA Y NA NNNNN NNN NN0 N N N' h y*

< 5 20W coNnmoooooooo comemnoooonoooo an oo oo ***NN*noooono o o N *.N - o o o o m oooo oo--comeomomooo

<e

-O ao oo-nNoommocoo mm Neo NN N o om Nm NN N s nN - a, - - mN -

1 0 -

nn n

• z 3D G- G-n -gu~ -so-GE8 %E GEE GN um--8s z

3 x zg s oc- o a x-rg3~

~sW.

a ggs a

sg -- ~3%.a3~c ca.

~

ca..ac --

> -- - ome m a,--- o.- - omeo,----o----

m n N N p

z D

0 a "I  ;

Ea.

~seu  :

2 es 8 g$ a. 8 s

2 es s

! Eli via s

g II ss 3 ,5j ess as s II ss E$ -

s E

e l 35% at Eoyo8 $.44 3 cc 2 E uj 42- Ed s 44 sE.g c e g Es rs&

z . a s _g z. E s.es c 2 2 matg u

=

.c w t m N. 2s7. 5=; s f f - E s. a. 2 .3=

=3o w. ige m=g3- N. awr-3t E na=w.ag==.eg gom a -

g E .oa- ggarss e, a c E i

!h,wmeh.Rhn8En,o) !AffE wwe0 w3883*E v3E

!Ik xxo x ffE xe.

3 ff .ek$

o m.o.j co-e.soz*f.ff.

z me noeco . f. <oOA =lkh -

l zwu eNeerausuccusE B-14

l l

L 0053520000800004 1 00 0 64008 A 0044795505950005 00 74553 I T 7 RS 055780220207000 523261 56 579 4 5 00 7522 1, EO 1. $ 1 29, 2 9 7 39551 1 8 T

A C 3 1

1 5 M 3 0000032006664 1 0753 1 T 1 0638 3 9228233 1 R

OT S 00000680001 061 040582405353268 069767 2 5

4, S

O C

1 6590 69 9, 8 8 0

5, 89829757,7 0045 1 6

5, BO 8881 80341 1 21 324836 6 22232 4336575 4,1 4,2 8 1 2 2 2 2,1 20,46404 45 7 6 9,7,2 0 1 2,1 6 A

L C 6561 32331 1 51 322 1

2

1. 3 4

1 2 1 436 4 1

D 000000000000000 O 00 0000 000002200222221 M 21 l RS l a  :

a  : 2222.l a : )  :

OE 555551 1 551 1 1 1 1 3  : ) 1 3t : )

222223322333333 to )tTssts T 3333 toTss 1 1 1 1

)tts BT S L

A 33 o )t ts  :

T

_ AA Tss T _

LP bu cc oo CO T O

b oo u cc O S b es

( t i

O'

_ Sl r T Sl aor C Sl ao s C

_ 060029800878060 iao r b L I ib

_ R ea A 60 r ea L 1 380 nb L 232653666828270 T 70 2664 tea O taf l

O taf l

A af l

A BS AR 979857398848770 505456024363978 222 mo

+

r. %

T p

3 76 71 mo

+

% T OT 21 34 6509 mo

+% T OT

_ LU 1 1 1 1 c5 a 3 r o5 1 1 1 r o5 LO b 9 2, b 9 4 b9 5 AH la( E 7 la( # la( #

T f d G d d O

T o ae N 3 1

f o ae D O fo ae D O

%h A 5 e r

R %h r M %h M r ST 2 v 5 e 2 v R 5 e 2 v R DC (

O E d (

O H

(

O H EA s D a s s EF 000000000000000 t s

I e 00 t s

R 0000 t R R .

777777777777770 o W h r 70 o 7777 s o

Ot V 22222222222:221 C e 21 C 2222 C BU v C O C C AO O O O LE A A

A Y Cg O Cg T 'g I

in n n V

I R r 8 i r

i r

TO e 6 e e s 000000000000000 e n

00 e 0000 e e CT 888888888888880 8, _ 80 n 8888 n t

a UC 444444441 1 1 1 1 1 1 ig n

ig ig m DO A 2 1 1 n 41 41 n t

i F E 4 E E s

E R t P s

C Co RO 000503000030000 00 3300 004701 00001 0000 O 00 0000 d

B R S 060020800808060 A 60 0080 e 2' 1 1 21 8 1 1 4 t

a A C le L U I O R R TH G e

r NI N a U E w .

dr a B TC H D A E F 000000000000000 00 0000 R E . 1 1 1 1 1 1 1 1 01 1 1 000 5 00 1 1 1 1 H 1 R Tl. V 222222221 2221 1 1 0 1 1 2222 Au McE 7 2

009653000030000 l 00 1 1 00 LT 002621 00001 0000 i a 00 - 4400 A S 001 1 60550505000 r e 00 0055 I O 05 25 22 2 7 22 R

E C 52 1 1 4 t

e ..

TTt 1 M

AN MU

)

C) )

T ) RC) 5 ) C I ) Ct utRC 5 C)R N ) CRouR t 4, )RCt Ct U CR(( o u iiR(hh( oi md i

dd( t cct ( d ttt 4 RuRuO(

( O (t t a at aa t oaao a ooo 4 t t t t ( oo EEf eef L eeLft el t t 1 loo t FFLL Y 1 1 05601 1 1 1 01 1 1 1 r 1 1 001 1 TI 5 4 6 o 64 T 2 b 1 0 a

N L 1

A U

O s

R pe s y lo oge e ss a r e

gan ac v ee lii l lp r r t

n N nRf la V

bb s in oe8o tccRC O a Ret r

e se mm ee D i

ie T

uu dd - -

diIn I

T ea r

t s ss s m r r s PI deW/ nG noAA s o sn r mte la ye up 2 e e n oon s,.no

1. n CC mi R

C W r,egns egel al aeh nn t

b al eAa t ym s ig 33 .ta 2t tnta au ub n oo nb lp D 4 n 5 . , .n S

eal uF nai5 rni alasSo O a. i d st FT/MaCmaamCuDr E r t r t r s

i C M 8 nE s221 1 .m

. r D dar 4 eeee e d ee 4 i t oie onSSsg S1# TnnT MASSES r

r ee lePPlbes/epr t Vcointu a m L A

ct Miur t p oseTT M , ,

eeee RT-H p 2 "2 n*2 d nmg o T Rcm Rlbt 9 bbl l 1 /

C b a//a88aCuoor / O Hle o H aaaa RD1 1 H1 2CRRC2ACCP T REC RCCCC 1

D C1 r.

C llll

600269 6 08 064 2 5 R

O T S 380 6. S 5 8, BO AC 1

4 026 1 2,4,0 2 6 2 9 1

L 1 1 0 5

_ 1 27 00A 21 /

RS Nl a  :

OE 1 3 t

s o )t)

BT 33 Ts t

s T

S AA b oo O LR u cc Sl r C iao r b L R 80A ea

/ A O 8 0.N taf l

T BS AR 88 34 mo O T

LU 6 r+%

o5 6 LO b 9 1 AH la( #

T d O foae D T O

%h r M

6T. 5 e 2 v DC (O R EA s H EF 00 A ts R R / o OtV. 722 7.N C Bu C AO Q LE A Y C T g I

V R u r

I e s TO 00 A e e CT / n ta UCA 8 5.N 1 1 ig n

m D F E n

s O E R t P s

o R 00A C /

d O S 0 80 0.N e B ta A R 6 L U 1

le O R TH I

e r N a U w .

dr a S TC H D A E F 000 R E . 000 H

R Tt.v 1 1 1 AU MQ E

l 000 ) )

LT 000 90 A S 000 04 903 4 0 0 I O 05 8 5 9 7 0 R C 0 1 2,5 9, $ S, 0 E 9 0 4 9 0 (

6 4 TT 82 272 3 1 A Nf 1 $1 1 5 ( 5 MU $ $$$ $

: r RLCD:

a e d)

OAOA-I E

y e d

T I B A E R AH r e n N L OR p u U T/E o AR V (

r tooo t t MGON L LLL E A Y 1 1 1 T 7 T 6# )n 7 T S O TI O 1 ica0 T N C Di9r A Ot c D D N U C M e :ry A O M c R le R i L O 5 G g A 2 o T Fh t L

O Sn Go%,

P T N g N m0 i1 O n I V s i s mes IT A rh r ,

P a r ly o c ga t

I R 6 npA e on r S2(te M e C 1 F

S n eg &

E D

soRm d c oe s/ r O V Dwee P

M een Rlb iigv H aen RCRE g[

ll1

Updated Summary of Cost Estimates for RHR Modifications Estimated Costs Engineering / Sub. Total RHR Modification Material Craft Labor QACC Overhead Procedure Mod. Cost

1. 1 0 2,956 739 2,808 3,650 10,153

1. 2 27.015 144,455 42,868 137,233 3,650 355,221

1. 3 3,650 3,650

1. 4 0 2.956 739 2.808 915 7,418

1. 5 1,138 15,972 4,278 15,173 915 37,476

1. 6 915 915

1. 7 3,650 3,650

1. 8 3,650 3,650

1. 9 915 915

1. 10 3,650 3,650

1. 11 3,650 3,650

1. 12 915 915

1. 13 915 915

1. 14 915 915

1. 15 3,650 3,650

1. 16 0 22,662 75,665 21,529 915 120,771 Totals $28,153 $189,001 $124,289 $179,551 $36,5 20 $ 557,514 0 & M SAVINGS FOR MOD #16: ($6,800)

Radiation Exposure: $ 72,600 Training: $74,440 Tech Spec: $34,000 NRC Costs: $ 27,600 l Grand Total (rounded): $759,400 l B-17

h  ; '

s APPENDIX C sua UNITED STATES .

3 - o- NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C,20555

.:. tj

%,  ! october 17, 1988 TOLALL HOLDERS OF OPERATING LICENSES OR CONSTRUCTION PERMITS FOR' PRESSURIZED

WATER REACTORS (PWRs)

SUBJECT:

LOSS OF DECAY HEAT REMOVAL (GENERIC LETTER NO. 88-17) p, 10CFR50.54(f) l Loss'of? decay heat removal (DHR) during nonpower operation and thel consequences ofEsuch a loss.have .been of increasing concern for years. Numerous industry and NRC. publications have. addressed the subject.. The Diablo Canyon event of April,10. 1987, and ensuing work by both the staff and. industry. organizations have provided additional insight. Yet the problems continue, as-illustrated by

(1) the inadequac'es demonstrated by many licensees lin their response-to

-Generic Letter (GLj 87-12; (2). the event at Waterford on May 12,-1988; (3) th'e event at Sequoyah on May 23, 1988; (4) the.DHR perturbations due to inadequate l

. level at San'Onofre.on. July 7, 1988; and (5) the apparent lack of a completeL industry understanding of the potential seriousness of.such events.

The report of the Diablo Canyon event, NUREG-1269, stated.that operating a

-plant with a reduced reactor coolant system (RCS) inventory was a particularly.

sensitive condition and identified m.,y generic weaknesses in DHR. GL 87-12, which requested information from all .WR licensees, provided additional in-

' sight, and NUREG-1269 was transmitf.ed with the generic letter to ensure that licensees had the' latest information. Despite this,~many of the responders tc GL 87-12' demonstrated that they did.not understand the identified problems.

Deficiencies exist in procedures, hardware, and training in.the areas of (1) prevention of accident initiation, (2) mitigation of accidents before they potentially progress to core damage, and (3) control of radioactive material if a core' damage accident should occur. Although deficiencies exist-in all'PWRs, certain-design features make initiation and the time available for mitigation

-in-the4 Westinghouse and Combustion Engineering designs of more concern than in the nuclear steam supply systems (NSSSs) designed by Babcock and Wilcox~- .

Nevertheless, we believe expeditious actions are necessary at all PWRs to rectify these deficiencies. These should be paralleled by' programed enhance-ments whi'ch supplement, add to, or replace the expeditious actions to accom-

-plish a more comprehensive improvement. Recommendations covering these items a're sumarized in the' attachment, and additional information and. guidance are provided in the three enclosures.

8810180350 l

C-1

2 Pursuant to 10 CFR 50.54(f), we request your response regarding your plans with respect to each of the recommendations as related to operation following placement of the NSSS on shutdown cooling, or following the attainment of NSSS conditions under wfiich shutdown cooling would normally be initiated. Your ,

response is to include the following: j (1) A description of the actions you have taken to implement each of the eight recommended expeditious actions identified in the attachment. Your reply shall be submitted to us within 60 days of receipt of this letter.

(2) A description of enhancements, specific plans, and a schedule for imple-mentation for each of the six programmed enhancement recommendations ,

identified in the attachment. Your reply shall be provided to us within 1 90 days of receipt of this letter. I Individual deviations from the recommendations will be considered on a case by case basis provided compensatory measures are provided which will achieve a comparable level of protection.

No further responses are required to GL 87-12 and licensees or construction permit holders need not provide any supplemental information in a response to  !

GL 87-12 to which they previously committed.

We will accept documents such as technical reports, action plans, and schedules prepared by industry groups when accompanied by commitments from participating licensees in lieu of individual documents from those licensees. Alternatively, ,

such industry group documents may be incorporated by referen.ce in licensee documentation. We encourage your participation in cooperative efforts to effectively resolve these issues.

Your written response shall be submitted under oath or affirmation under the provisions of Section 182a, Atomic Energy Act of 1954, as amended. Your written response is needed to determine whether actions to modify, suspend, or revoke your license are necessary. An analysis as required by 10 CFR 50.109 has been performed regarding this request.

The original copy of your written response shall be transmitted to the U. S.

Nuclear Regulatory Commission, Document Control Desk, Washington, D.C. 20555 for reproduction and distribution.

This request is covered by Office of Management and Budget Clearance Number 3150-0011 which expires December 31, 1989. The estirnated average burden hours is 200 person-hours per licensee response, including assessment of the new requirements, searching data sources, gathering and analyzing the data, and preparing the required reports. Coments on the accuracy of this estimate and suggestions to reduce the burden may be directed to the Office of Management and Budget, Room 3208, New Executive Office Building, Washington, D.C. 20503, and to the U. S. Nuclear Regulatory Commission, Records and Reports Management Branch Office of Administration and Resour os Management, Washington, D.C.

20555.

C-2

3

.if'you have technical questions regarding this matter please contact Wayne:

Hodges at 301-492-0895.- Other questions may be directed to the NRR Project-Manager assigned to. this issue Charles M. Tramell,(301-492-3121) or to the

-Project Manager, assigned to your plant.

j@.

s'f $ Al'd '

Denn ActingsAssociate M. Crutchtferf Direc tor for Projects Office.-of-Nuclear Reactor Regulation

?

Attachment:

Recommended Actions

' Enclosu'resi P 1. Overview and Background Information Pertinent l'

.to Generic' Letter 88 2. Guidance for Meeting Generic Letter 88-17

3. Abbreviations and Definitiorts A

w l

s

]

C-3

LIST OF RECENTLY ISSUED GENERIC LETTERS Generic Date of Letter No. S&bject _ Issuance Issued To 88-16 REMOVAL OF CYCLE-SPECIFIC 10/04/88 ALL POWER REACTOR PARAMETER LIMITS FROM LICENSEES AND TECHNICAL SPECIFICATIONS APPLICANTS 88-15 ELECTRIC POWER SYSTEMS - 09/12/88 ALL POWER REACTOR-INADEQUATE CONTROL OVER LICENSEES AND DESIGN' PROCESSES APPLICANTS

'88-14 INSTRUMENT AIR SUPPLY 08/08/88 ALL HOLDERS OF SYSTEM PROBLEMS AFFECTING OPERATING LICENSES SAFETY-RELATED EQUIPMENT OR CONSTRUCTION PERMITS FOR' NUCLEAR POWER REACTORS

, 88-13 OPERATOR LICENSING 08/08/88 ALL POWER REACTOR L ' EXAMINATIONS LICENSEES AND APPLICANTis FOR AN OPERATING LICENSE.

88-12 REMOVAL OF FIRE PROTECTION 08/02/88 ALL POWER REACTOR REQUIREMENTS FROM TECHNICAL LICENSEES AND '

SPECIFICATIONS APPLICANTS 88-11 NRC POSITION ON RADIATION _ 07/12/88 ALL LICENSEES OF EMBRITTLEMENT OF REACTOR OPL9ATING REACTORS VESSEL MATERIALS AND ITS AND HOLDERS OF IMPACT ON PLANT OPERATIONS CONSTRUCTION PERMITS  ;

88-10 PURCHASE OF GSA APPROVED 07/01/88 ALL POWER REACTOR SECURITY CONTAINERS LICENSEES AND HOLDERS OF PART 95 APPROVALS 88-09 PILOT TESTING 0F FUNDAMENTALS 05/17/88 ALL LICENSEES OF ALL EXAMINATION B0ILING WATER REACTORS AND APPLICANTS FOR A t

BOILING WATER REACTOR OPERATOR'S LICENSE UNDER 10 CFR PART 55 88-08 MAIL SENT OR DELIVERED TO 05/03/88 ALL LICENSEES FOR POWER THE OFFICE OF NUCLEAR REACTOR AND NON-POWER REACTORS REGULATION AND HOLDERS OF CONSTRUCTION PERMITS FOR HUCLEAR POWER REACTORS i C-4

ATTACHMENT TO GENERIC LETTER RECOMMENDED ACTIONS Expeditious actions and programed enhancements are recommended concerning operation of the NSSS during shutdown. cooling or during conditions where-such cooling would normally be provided. The recommendations apply whenever there is irradiated fuel in the reactor vessel (RV). These recommendations are summarized below and discussed further in enclosure 2:

Expeditious actions:

The following expeditious actions should be implemented prior to operat-ing in a reduced inventory condition *:

(1) Discuss the Diablo Canyon event,,related events, lessons learned, and. implications with appropriate plant personnel. Providestraining shortly before entering a reduced inventory condition.

l (2) Implement procedures and administration controls that reasonably assure that containment closure ** will be achieved prior to the time at which a core uncovery could result from a loss of DHR coupled with an inability to initiate alternate cooling or addition of water i

to the RCS inventory. Containment closure procedures should include l' consideration of potential steam and radioactive material relea r from the RCS should closure activities extend into the time boiling takes place within the RCS. These procedures and administrative controls should be active and in use:

(a) prior to entering a reduced RCS inventory condition for NSSSs supplied by Combustion Engineering or Westinghouse, and (b) prior to entering an RCS condition.'herein the water level is lower than four inches below the top of the flow area'of the hot legs at the junction of the hot legs to the RV for NSSSs supplied by Babcock and Wilcox, and should apply whenever operating in those conditions. If such procedures and administrative controls are not operational, then either do not enter the applicable condition or maintain a closed containment.

• A reduced inventory condition exists whenever RV water level is lower than three feet below the RV flange.

• • Containment closure is defined as a containment condition where at least one integral barrier to the release of radioactive material is provided.

Further discussion and qualifications which the integral barrier must meet are provided in enclosure ? and in the definitions provided in enclosure 3.

C-5

-(3) P'rovide at least two' independent, continuous temperature indications that are representative of the. core exit conditions whenever the RCS~

is in a mid-loop condition

• and the reactor vesse1~ head is located on top'of t'he reactor vessel. Temperature indications should be.

periodically checked and recorded.by ~an' operator or automatically '

and continuously; monitored and. alarmed. Temperature monitoring should be performed.either:

.(a) ' by an: operator in the control room (CR), or

-(b) from a location outside of the containment building with provision for providing 'inanediate-temperature values; to an operator in the CR-if.significant changes occur. Obserirations-

-should be recorded at an interval no greater than 15 minutes during normal conditions.**

'(4) Provide at.least two independent, continuous RCS water level indica .

tionsJwhenever the RCS is in a reduced inventory condition. Water 1 level indications should be periodically checked and recorded by an 1 operator or; automatically and continuously monitored and alarmed.

~

Water level-monitoring should be capable of being. performed either:  :

1 (a) by an operator in the CR, 'or-T(b) from a location other than the CR with provision for providing imediate water level values to an operator in the CR if significant changes occur. Observations should be recorded at an interval no greater than.15 minutes during normal condi-tions.** 1

'(5).Implemeutproceduresandadministrativecontrolsthatgenerally- {

avoid operations that deliberately or knowingly lead to'perturba-tions to the RCS and/or to systems that are necessary to maintain  ;

the RCS in a stable and controlled condition while the RCS is in.a l reduced inventory condition.

1 If operations that could perturb the RCS or systems supporting the l RCS must be conducted while in a reduced inventory. condition, then 6 additional measures should be taken to assure that the RCS will remain in a stable and controlled condition. Such additional measures include both prevention of.a loss of DHR and enhanced monitoring requirements to ensure timely response to a loss of DHR  !

should such a loss occur. l I

• -~A mid-loop condition exists whenever RCS water level is below the top of the flow area of the hot legs at the' junction with the RV.  ;

• • Guidance should be developed and provided to operators that covers 1 evacuation of the monitoring post. The guidance should properly balance reactor and personnel safety.

' i I

l-1 C-6 t: _ _ _ - _ .-__-__

I

3 (6) Provide at least two available* or operable means of adding inventory to the RCS that are in addition to pumps that are a part of the normal DHR systems. These should include at least one high pressure injection pump. The water addition rate capable of being provided by each of the means should be at least sufficient to keep the core covered. Procedures for use of these systems during loss of DHR

, events should be provided. The path of water addition must be specified to assure the flow does not bypass the reactor vessel before exiting any opening in the RCS.

(7) (applicable to Westinghouse and Combustion Engineering nuclear steam l

supply system (NSSS) designs) Implement procedures and administra-tive controls that reasonably assure that all hot legs are not blocked simultaneously by nozzle dams unless a vent path is provided that is large enough to prevent pressurization of the upper plenum of the RV. See references 1 and 2.

(8) (applicable to NSSSs with loop stop valves) Implement procedures  ;

and administrative controls that reasonably assure that all hot legs are not blocked simultaneously by closed stop valves unless a vent path is provided that is large enough to prevent pressurization of the RV upper plenum or unless the RCS configuration prevents RV water loss if RV pressurization should occur. Closing cold legs by nozzle dams does not meet this condition.

Programmed enhancements:

Programmed enhancements should be developed in parallel with the expedi-tious actions and they r.ay replace, supplement, or add to the expeditious -

actions. For example, programmed enhancements may be used to change expeditions actions as a result of better understanding or improved procedures. This may lessen the initial impact of expeditious actions such as the speed with which containment closure must be achieved and may include consideration of such factors as the decay heat rate. Additional guidance is provided in enclosure 2. For example the first paragraph of section 2.2.2 and the first paragraph of section 3.3.2 illustrate the flexibility we have in mind as long as safety is adequately addressed.

We intend that programmed enhancements be incorporated into plant opera-tions as they are developed when this results in significant safety improvement or enhancement of plant operations with no decrease in safety. Procedural and hardware modifications may be implemented without prior staff approval where the criteria of 10 CFR 50.59 are met, although it is our intent to review and/or audit such changes.

Programmed er.hancements should be implemented as soon as is practical, but no later than the following schedule:

• Available means ready for use quickly enough to meet the intended functional need.

C- 7

v +'

l. 4

.(1) . Programmed enhancements consisting.of hardware installation and/or modification,'and programmed enhancements that depend upon hardware installa. tion:and/or modification, should be implemented:

l l

() byithe end of. the first' refueling outage that is initiated 18

. months or later following receipt of this letter, g (b) by the'end:of the.second refueling. outage following receipt.of this. letter, whichever occurs first. If a shutdown for refueling has been initiated'as of the date of-receiptLof this letter, that'is to be counted as the first refueling outage.

(2) Programmed enhancements that.do not depend upon hardware changes should be implemented within 18 months of receipt of-this letter. .

I We recommend'you implement the following six programmed enhancements:

(1) Instrumentation Provide reliable indication of parameters that describe the state of the RCS and the performance of systems normally used to co'ol the RCS

'for both normal and accident conditions. At a minimum, provide the following in the CR:

(a) two independent RCS level indications (b) at.leastitwo independent temperature measurements representa-tive of the core exit whenever the RV head'is located on top of the RV (We suggest that temperature indications be provided at all-times.)

(c) the capability of continuously. monitoring DHR system perfor-mance whenever a DHR system is being used for cooling the RCS (d) visible and audible-indications of abnormal conditions in temperature, level, and DHR system performance (2) Procedures-Develop and implement procedures that cover reduced inventory operation and that provide an adequate basis for entry into a reduced inventory condition. These include:

(a) procedures that cover normal operation of the NSSS, the con-tainment, and supporting systems under conditions for which cooling would normally be provided by DHR systems.

l 1

C-8

5 (b) procedures that cover emergency, abnormal, off-normal, or the equivalent operation of the NSSS, the containment, and support-ing systems if an off-normal condition occurs while operating under conditions for which cooling would normally be provided l

by DHR systems.

(c) administrative controls that support and supplement the proce-dures in items (a), (b), and all other actions identified in this communication, as appropriate.

I (3) Equipment (a) Assure that adequate operating, operable, and/or available equipment of high reliability

• is provided for cooling the RCS and for avoiding a loss of RCS cooling.

(b) Maintain sufficient existing equipment in an operable or available status so as to mitigate loss of DHR or loss of RCS inventory should they occur. This should include at.least one high pressure injection pump and one other system. The water addition rate capable of being provided by each equipment item should be at least sufficient to keep the core covered.

(c) Provide adequate equipment for personnel communications that involve activities related to the RCS or systems necessary to maintain the RCS in a stable and controlled condition.

(4) Analyses Conduct analyses to supplement existir.g information and develop a basis for procedures, instrumentation installation and response, and equipment /NSSS interactions and resp'.nse. The analyses should encompass thermodynamic and physical (configuration) states to which the hardware can be subiected and should provide sufficient depth that the basis is developed. Emphasis should be placed upon obtain-ing a complete understanding of NSSS behavior under nonpower opera-tion.

(5) Technical Specifications Technical specifications (TSs) that restrict or limit the safety benefit of the actions identified in this letter should be identi-fied and appropriate changes should be submitted.

• Reliable equipment is equipment that can be reasonably expected to perform the intended function. See Enclosure 2 for additional information.

C-9

i-.

(6) RCS' perturbations:

Item (5) of the expeditious actions should be reexamined and opera-tions rdYined as necessary to. reasonably minimize the likelihood of loss of DHR.

Additional information and guidance are'given in enclosure 2.

1.

REFERENCES (1). C..E. Rossi,'"Possible Sudd q Loss of RCS. Inventory during Low Coolant-

' Level Operation,"-NRC Information. Notice 88-36, June 8, 1988.

t (2) R. A. Newton, " Westinghouse Owners' Group Early Notification of Mid-Loop Operation. Concerns," Letter from Chairman of Westinghouse Owners Group.to Westinghouse Owners ' Group Primary Representatives (IL,1A), 0G'-88-21, May-

=27, 1988.

j Y

)

C-10 1

11 _. . . . _ _ _ _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _

Gi=====is8 SON 1. REPORT NUMBER fi, _ by P,%fB:DPS, add Vol No., d enyl mac FOpM sit U 5. NUCLE A R R EGULATOR T m

3o 3 Pj' BIBUOGRAPHIC DATA SHEET si .NseavCT.oNs oN ma ne v =se .

NUREG-1340 3 LE AVI SL ANiL 31sV Lt AND $v9 fif Lt Regulatory Analysis for the Resolution of Generic Issue 99, " Loss of RHR Capability in PWRs" , o , , ,,,o,, co,,u , , o mon r 4 veAa g

e Aumoais' November 1988 5 DATE RtPORT ISSUED Alfred H. Spano l I'eb ruary 1989 v 6 PMOJECTIT A5E/WOME UNtf muMe&M 7 FEMS DRM#NG ORGAN 12 Af ION NAM 6 AND M A* LING ADDRESS Haca e tW Com/

Division of Safety Issue Resolution e nN oa GaA=i Nuwaia Office of Nuclear Regulatory Research U.S. Nuclear Regulatory Commission Washington, D.C. 20555 If a T YPE OF REPORT to SPON50HtNG ORGAN 62 Af SON Navt AND MAeLING ADontS5 Haciverle Ced.A Division of Safety Issue Resolution Technical Report Office of Nuclear Regulatory Research ,,,,,OoCo . oy . ,

U.S. Nuclear Regulatory Commission Washington, D.C. 20555 12 SUP*LE Mt,NT An v Not t s 13 ASSTR ACT (20p e. ores er ient Generic Issue 99 is concerned with the loss of residual heat removal (RHR) capability in pressurized water reactors during cold-plant outage operations. The issue focuses on two risk-significant common-cause failure modes of the RHR system:

(1) air binding of the RHR pumps during reduced-inventory operations and (2) spurious closure of the RHR suction valves due to misapplication of the autoclosure interlocks.

Resolution of this issue involves consideration of the adequacy of plant capabilities for (1) preventing losses of RHR, (2) responding promptly and effectively to such challenges in order to prevent core damage, and (3) ensuring timely containment protection against the release of radioactivity to the environment in the unlikely event of core damage due to loss of shutdown cooling. This entails examination of (1) relevant operational and accident response procedures, (2) the instrumentation available to the operator for accident diagnosis and mitigation, and (3) adminis-trative controls available for ensuring control room cognizance of ongoing maintenance activities that could potentially affect stability of the reactor coolant system.

This regulatory analysis provides quantitative assessments of the costs and benefits associated with several alternatives considered for resolution of the Generic Issue 99.

,. A A m A .v1,

.. oOCom, AN A rs,. . . ..oaos,oi sc .10.s STATEMENT Residual heat removal, decay heat removal, generic issue, autoclosure interlocks, mid-loop operations Unlimited 16 $f CUWIT T CL A$$tf lC ATION ITo;s paget e 4D(Ntd ag nsroPtN gNot o tg pM5 , g qq pgg (T u reoorti nnri nnni f i nd IP NUUgt de Q6 P AGk b ig PH.(t

z UNITED STATES E-NUCLEAR REGULATORY COMMISSION ""E8Effl0Esil?"

"S""' $

. WASHINGTON, D.C. 20555 PERMlf No G 67 y 1

o OFFICIAL BUSINESS PENALTY FOR PR!VATE USE, $300 120555139531 1 1 A N 11 S 19 H19'U1 US NRC-0ADM DIV FOIA & PUBLICATIONS SVCS TPS PDR-NUREG P-209

' WASHINGTON DC 20555 E

o C-O k

Z O

r">Q Ma O

O

?"

"i 9m sa*

cP da

s. a T

'k Ro 9

9 5

02 C

m b..

E E

R a

8

_ _ _ _ _ ___