B12965, Drywell Insulation Study
| ML20151H697 | |
| Person / Time | |
|---|---|
| Site: | Millstone  |
| Issue date: | 07/31/1988 |
| From: | NORTHEAST NUCLEAR ENERGY CO. |
| To: | |
| Shared Package | |
| ML20151H694 | List: |
| References | |
| B12965, NUDOCS 8808020008 | |
| Download: ML20151H697 (36) | |
Text
-
c e *s Docket No. 50-245 B12965 Attachment 1 Millstone Nuclear Power Station, Unit No.1 Drywell Insulation Study July 1988 8808020008 880721 !
i e ,.
I. I?TIRODUCTION h e referenced study (Reference 1) was initiated as a result of the Millstone Unit One ISAP Topic regarding the acceptability of fibrous insulation that had been installed in the drywell as a rep)acement for the originally installed reflective metallic insulation. his study evaluated insulation debris with respect to degradation of long-term ECCS recirculation capability. Specifically, as a result of a LOCA, drywell piping insulation can be dislodged from the pip--
ing and broken down into a wide range of debris forms. Once trans-ported to the torus, this debris could potentially plug the ECCS suction strainers, thus causing an excessive pressure drop across the cuction strainers and a corresponding degradation of ECCS pump t flow and NPSH margin. % e conclusions of this study indicate that ECCS NPSH margin and operability could potentially be compromised by the fibrous insulation in the drywell.
In arriving at this conclusion, conservative assumptions were applied as dictated by the regulatory guidelines on this topic. he conservatisms contained in the NRC's guidelines were developed on a generic basis for BNRs. In order to develop a plant-specific basis for relaxing some of the extremely conservative generic assumptions, extensive analytical and empirical research is regatred. Extensive corporate, financial, anr1 manpower resources would be required to ,
perform these analyses in order to substantiate realistic modeling c- .
of the insulation debris distribution and transport at Millstone Unit One. Engineering judgm2nt can be applied to demonstrate that the overall debris impact is not as severe as concluded in th'e attached study; however, for reportability determinations, the conservative approach has been taken since an alternate firm analytical basis has not been developed.
herefore, rather than commit to performing these extensive dynamic analyses, it is recommended that permanent modifications to increase the ECCS suction strainer area be pursued. Because there is a po-tential impact on ECCS operability due to the existing fibrous in-sulation and ECCS hydraulic design configuration, it is recommended that tb" condition be reported to the imC per 10CFR50 requirements.
Imples. stion of the recommended modifications to the ECCS suction st . caners is not feasible until the 1989 refueling outage; there-fore, justification for continued operation (JCO) for the remainder of the current fuel cycle must be developed. 10SC0 Generation Mech-anical Engineering feels that a very strong JC0 exists for Millstone Unit One. In the sections that follow, the technical basis for the JCO is presented.
II. JUSTIFICATIOi FOR COffritUED OPERATI0ti (JCO)
In order to alleviate the potential impact of fibrous insulation, replacement of the existing strainers with strainers of a larger ,
area is required. W e schedule for ultimate resolution of the l
l i 1
', , e potential insulation debris impact on ECCS wil.1. be determined under the ISAP program. @is JCO applies for the remainder of the current fuel cycle. %e JCO is based upon two major considerations. . h e
. first consideration is to cri.dit the measures taken to ensure the integrity of the high energy systems inside the drywell. nese measures, in conjunction with inherent design features, ensure that the probability of a LOCA of sufficient size to create insulation debris is extremely remote. h e second consideration of this JCO involves applying reasonable engineering judgment to develop a realistic assessment of the impact of the insulation debris on the ECCS recirculation capability. @is assessment demonstrates that the ECCS performance is acceptable, provided that the toCA thermal limit (MAPLHGR) is reduced to account for slightly degraded core spray pump flow and the EOPs relating to ECCS operation are revised to account for insulation debris clogging.
i
. .. j i
Evaluation of Drywell Piping Integrity As detailed in the attached study, the worst case quantifiable high energy line break in terms of fibrous debris generation and trans-port potential directly to the torus is a recirculation system double ended break below elevation 13'4" in the drywell, other high energy systems were identified; however, it was demonstrated that l the debris generation and transport potential for these. systems was bounded by the recirculation system pipe break. %ese systems in-clude the RWCU, Isolation Condenser, Feedwater, Main Steam, Core Spray, LPCI and Shutdown Cooling Systems. Because of the uncertain nature of insulation. debris generation and transport and the fact that a bounding case had been performed for the recirculation sys-tem, the impact of these systems on ECCS operability was not per-
'ormed. For the purycus of this JCo, these systems must be con-sidered because their indeterminate intermediate impact on ECCS operability may not be acceptable, i.e., a rupture of one of these systems would result in less insulation debris than a recirculation I system break; however, the amount of insulation debris could po-tentially be enough to conpromise ECCS operability.
W e purpose of this section of the Jco is to demonstrate that the potential for a double ended break for any of the above systems is exceedingly remote for the remainder of the current fuel cycle. i I
% *S h
WSCO component Engineering has completed an evaluation of the large diameter piping (i.e., greater than 6" diameter) inside the dry-well), to evaluate the potential for a double ended pipe rupture prior to the 1989 refueling outage, per reference 2. nis evalua-tion included the high energy systems detailed above. It consisted of reviewing the system's susceptibility to water hamer, intergran-ular stress corrosion cracking (IGSCC) and flow assisted corrosion, as well as reviewing cracking and repair histories and recent in-service inspection results.
Where appropriate, WSCO Reactor Plant Systems has also included inforration relating to the elevation and location of the high energy systems relative to insulation debris transport potential.
he results of this evaluation are provided below:
Main Steam and Feedvater Systens - A leak-before-break (LBB) eval-uation performed for the main steam and feedwater systems showed that the LBB criteria as specified in WREG 1061, Volume III, are satisfied for the remainder of the plant life. h is ensures that if an undetected crack were to propagate through the pipe wall, the leakage through the crack would be detected by plant personnel and the plant shut down well in advance of the crack propagating to a critical size resulting in a double ended pipe rupture.
A review of the water / steam hamer history indicted that no ,
reportable water hamer events in the main steam and feedwater
6-systems have been experienced at Millstone Unit No. 1. his is j consistent with a relative absence of water hamer events in BWRs for these systems. Additionally, the installation of a high ,eactor r vessel water level feed pu g trip prevents the water leval from rising above the steam line nozzle, thus, minimizing water hamer potential for these systems. Water hamer is not a probable event for the main steam and feedwater piping at Millstone Unit No. 1.
A review of the corrosion history for these systems was also per- l formed. We typical corrosion mechanism for carbon steel piping is !
flow-assisted corrosion which can lead to rapid piping failure. ,
his type of corrosion, however, has been more of a problem in pressurized water reactors (PWRs) than in boiler water reactors (BWks). Additionally, these systems have been inspected for flow ,
assisted corrosion with no significant wall thinning detected.
%us, the feedwater and main steam piping is not considered rusceptible to flow-assisted corrosion.
1
- Core spray A and n - A review of the core spray system indicated I that the piping inside the drywell was replaced during the 1980 refueling outage with IGSCC co:rosion resistant material (low carbon stainless steel). Pipe inspections in accordance with NUREG 0313 ,
guidelines have been performed during subsequent outages, with no j rejectable indications found in this portion of the system.
Recirculation System - Inservice inspection of the recirculation l
{
~ m- ~ < n.__-g + - - , _.n_--,- - . _ -,---,-_,--.,--.-_n _ - , . . , - - . __ v_- a. , . e, - - , n - , , - < .,- - ,-~ m e.
o . . .
system, in accordance with NUREG 0313, has been performed since 1978. In addition, 100% of the welds were inspected during the 1987 refueling outage.
All indications detected during the above inspections were evaluated in accordance with the guidelines of ASME section XI, IWB-3600. ne indications which were found not to meet the requirements of section XI were repaired by applying a full structural weld overlay on the ,
affected weld. One hundred percent of the welds in the recircula- !
tion system were also treated with induction heating stress improve-j ment (IHSI) during the 1984 refueling outage. %is process, which
! reverses the pipe through wall stress distribution, results in compressive stresses at the inside surface of the pipe eliminating i the crack driving force.
l In addition to the above results, it was also found that even if a l double ended pipe rupture were to occur in the recirculation system,
! the blowdown area would be minimized since the broken pipe would be contained by the whip restraints and therefore the dynamic effects (pipe whip and jet ispingement) would also be minimized. such a reduction in blowdown area minimizes insulation debris generation potential.
Isolation condenser - he isolation condenser supply line inside the
, drywell has been replaced with corrosion resistant material (Iow .
- _ . - ,--. ,,.,.,. _ ,,--,---- --.- - . n,-n, , , , , .- ..,,,,,_.nr,,_.a-----,,---'
l
1 ing methods increasing the compressive stresses on the pipe inside surface at these locations. In addition, approximately 20% of the return line inside the drywell has been replaced with corrosion
. resistant material. All welds, which have not been replaced with corrosion resistant material, have been inspected during the 1985 and 1987 refueling outages, with no rejectable indications detected.
System modifications have been completed to minimize the possibility of water hamer events. 'Ihe modifications involved installing a high reactor vessel water level feed pump trip which prevents the water level from rising over the steam line nozzle. Since the modifications were completed in 1979, no reportable water hamer events have occurred in this system. Additionally, from a systems perspective, Reactor Plant Systems has concluded that because of the relatively small pipe size and the minimum 33' elevation of this systea, the insulation debris generation and transport potential is minimized due to the intervening walkway gratings and piping.
Reactor Water Cleanup - Approximately fifty percent of the reactor water cleanup supply piping inside the drywell has been replaced with corrosion resistant material, while approximately ten percent of the return line has been replaced. One hundred percent of the welds in this system were inspected in accordance with the require-ments of NUREG-0313, during the 1985 and 1987 refueling outages.
'these inspections revealed no rejectable indications. Additionally, /
from a systems perspective, Reactor Plant Systems has concluded that l
h
-_-,,.m.. ._ -. __,, . _
y ,
-9 because of the relatively small pipe size of 8" and the minimum 28' elevation of this system, the insulation debris generation and transport potential is minimized due to the intervening walkway
- gratings and piping.
Low Pressure Coolant Injection - One Hundred percent of the low pressure coolant injection piping inside the drywell has been in-spected during the 1985 end 1987 refueling outages. No rejectable indications have been detected since the 1980 refueling outage.
Shutdown Cooling System 'Ihe portion of the shutdown cooling system within the reactor coolant pressure boundary (non-isolatable portion of the system) has been replaced with IGSCC corrosion resistant as-terial using IGSCC corrosion resistant welding techniques during the 1980 refueling outage. Inspections performed in accordance with NUREG-0313 requirements during subsequent refueling outages have detected no rejectable indications.
'Ihe above review indicates that a significant portion of the IGSCC susceptible piping (stainless steel) has been replaced with corrosion resistant material. In addition, the welds which have not been replaced were inspected during the 1985 and 1987 refueling outages. A leak-before-break evaluation performed for the carbon steel systems (main steam and feedwater) demonstrated that the leak-before-break margins of NUREG 1061, volume III, are ensured for the remainder of plant life. A ,
review of water hansner susceptibility for the feedwater and isoiation
condenser system concluded that the probability of a water hammer event in these systems is low. Based on the rigorous weld inspection program and the inherent design features for these high energy systems, it can be
. concluded that the probability of a double ended pipe rupture inside the drywell prior to the 1989 refueling outage is low. Continued operation on the basis of low probability of a high energy pipe break inside the drywell is justified.
/
~~
Evaluation of Conserv6tisms Ee referenced study (Reference 1) postulated that the design basis break wit.h respect to insulation debris generation and transport potential is a doubh r.nded break of the 28" diameter recirculation system piping in the lower level of the drywell, i.e., below elevation 13'4". It was conser-vatively estimated in this study that this break results in a total insulation debris generation of 17.0 ft.' of which approximately 5.1 ft.'
are fine fibers which are transported to the torus during the blowdown phase. h e balance of the debris is in larger pieces which are resistant to transport to the torus. In the unlikely event that the larger pieces are transported to the torus, their negative buoyancy characteristics ensure that plugging of the suctiori strainers will not occur.
We conclusions of the study were bounded by conservatively assuming that all (i.e., 5.1 ft.') of the fine fibrous insulation debris that is transported into the torus eventually plugged the suction strainers during ECCS recirculation. We purpose of this section of the JC0 is to apply realistic engineering judgments to the conservative assunptions and demonstrate that for the interim period of operation, there is reasonable assurance that Millstone !! nit one can be operated safely.
All forms of fibrous insulation debris, including <ine fibrous debris, demonstrate negative buoyancy characteristics in stagnant water ,
conditions. Note that the sink rates for fine fibrous debris have been h
l empirically determined for fibrous insulation with similar specific gravity. In turbulent water conditions, however, the buoyancy characteristics of fibrous debris is not readily determined. Fine. fibers may tend to remain suspended in the presence of turbulence. For the purposes of the referenced stu6y, the magnitude and duration of the turbulence was indeterminate; therefore, it was conservatively assumed that all of the 5.1 ft.' of fine fibrous insulation debris remained suspended and eventually migrated to each strainer. 'this bounding assunption was made in lieu of a rigorous analysis which would be needed to justify a finite duration. Based on engineering judgment, it can be estimated that the turbulence in the torus created by r.oCA blowdown jet ;
i forces is of a relatively finite duration. It is a reasonable estimate that within five minutes of the high energy pipe break, the turbulence in the torus would subside to the point where the negative buoyancy forces due to the heavier than water density factor for the fibers was dominant .
over the buoyancy forces due to the blowdown turbulence. At this point, all suspended fibers would sink and would no longer pose any threat of I further migrating and plugging the ECCS suction strainers. 'the estimate ,
of a five minute duration for the turbulence is based on qualitative and subjective engineering judgment. 'the duration of condensate oscilla-tiorv' chugging loads for a large break toCA (DaA) in Mark I containment is less than 65 seconds per reference 4. It should be noted that the post toCA pool swell phenomena is a short-term effect and will end before the condencate oscillation has subsided. After the condensate oscillation ends, the turbulence in the torus will begin to dampen out. Although no < \
extensive analysis or espirical research has been performsd on this l.
{
,w- . ~ - - - - - - - ---n -e , ,- _ - -,---en ,n ,- , -,----,-n n ~. -n -- --, --, - , , ,,, - -- - $ -
I topic, it is reasonable to assume that the torus turbulence effects would dagen within five minutes to the point where the fine fibrous insulation debris will sink. .
Applying these realistic assuqtions with respect to torus turbulence and debris suspension, a more realistic pressure drop across the suction strainers can then be estimated. Specifically, given a turbulence duration of five minutes, the strainer plugging can be calculated as detailed below. We Attachment 1 table sumarizes the methodology utilized to assess insulation debris i gact in the study and the JCO.
o Amount of Debris Generated he referenced study estimates 30% of the total insulation debris takes the form of fine fibers, i.e., 5.1 ft.8 of 17. ft.' total.
his percentage distribution of debris can be attributed to the nature of the jet i gingement forces resulting from the initiating pipe break. Because of the expanding jet flow field there are three natural divisions of the field. For the region that extends from the break location to a distance of 3 IA, the jet stagnation pressure is of a magnitude to cause total destruction of the insulation. For the tvd region between 3 and 7, the energy of the jet is such less concentrated and the anticipated damage will range from high levels of destruction at 3 IA to minor damage (blankets .
dislodged and torn but little or no shreds) at 71A. For the third C
o .
l l
region which extends beyond 7 L/D the stagnation pressure is reduced to the point where the insulation would be dislodged in the as-fabricated form. Division of the jet impingement influence into these regions has been empirically verified for PWR primary coolant system breaks as detailed in Reference 5. BWR jet expansion fields, due to lower pressure, decay more rapidly than PWRs, thus, the debris estimates are conservative. Reference 6 also indicates that there is a tange of debris damage as a result of the ux:A jet interaction with fibrous insulation. his reference endorses a debris assessment methodology that estimates 30% of the total debris forms fine fibers.
For the worst case break noted in Reference 1 (recirculation pipe below elevation 13'4"), it was noted previously that the actual recirculation pipe deflection would be limited due to whip re-straints. Limited deflection of the piping shadows the break area and reduces the debris generation potential. Sus, the actual insulation damage for a worst case recirculation pipe break will be less than estimated by the methodology endorsed by references 5 and
- 6. Because of the complex methodology to estimate debris generation on a three region basis, it was determined that the most straight-forward method was to apply the 30% criteria endorsed by reference
- 6. % e 30% fine fibrous debris assunption is conservative.
%e C
l o Magnitude of Strainer Plugging
- h e torus is physically divided by ring girders into 16 bays, Due to the turbulence in the torus during the blowdown phase, insulation would be uniformly distributed throughout the torus. It is assumed that all of the debris contained within a bay containing a strainer
, (3 total) is immediately deposited on the strainers. We remaining insulation debris in the torus can only plug the strainers while the torus is in turbulence with the insulation fibir suspended, i.e., 5 minutes.
)
It is assumed that this debris would be uniformly 6?stributed over the 3 strainers. Wis is a reasonable assumption because as a strainer becomes clogged, the velocity of flow near the unblocked strainers will increase, causing debris to migrate to the unblocked high velocity region. We same argument would be valid if a particular strainer experiences a high flow rate due to its location on the ring header with respect to the suction of the operating pump (s). As that strainer gets blocked, more flow would be obtained from the other strainers, thus inducing a velocity gradient.
h erefore, it is reasonable to assume that the debris is uniformly and equally distributed on three suction strainers.
I With these assumptions, the amount of debris which is deposited on ,
the suction strainers can be reduced from 5.1 ft.' to approximately
1.65 ft.3 %e calculation method to make this determination is detailed below.
. Amount of insulation debris deposited on the suction strainers inmediately:
3/16 X 5.1 ft.' .95 ft.'
he amount of the remaining insulation debris deposited during the recirculation phase is calculated using the following approach.
Total torus volume turnover time under maximum ECCS recircula-tion flow is calculated to be 29.1 minutes. Because the in-sulation debris is uniformly distributed throughout the entire torus volume, the amount of insulation debris that can clog the strainers is directly proportional to the percentage of time the insulation remains suspended. Werefore, the total amount of debris that accumulater on the strainers As limited to the period of turbulence in the torus.
5.1 ft.' X (16/16 - 3/16) X 5 min./29.1 min. = .71 ft.'
h erefore, the total fine fibrous debris accumulation on the strainers is 1.65 ft.' (0.95 ft.' + 0.71 ft.8) /
u
o Pressure Drop Across Debris Accumulation
. h e resultant pressure drop across the suction strainers with the 1.6E ft.' of insulation debris equally and uniformly deposited over the surface of the strainers is determined using a method endorsed by Reference 7, Specifically, Reference 7 provides a relationship where pressure drop across the debris accumulation is a function of the fluid flowrate, area normal to the flow, fluid viscosity, depth L
of debris, fiber-specific surface, specific volume of the fibers, and the density of the insulation. In this relationship, fluid flow, and fluid viscosity are variable. All other parameters are constant. A calculation was performed per Reference 8 to quantify the resultant pressure drop for the expected range of torus temper-ature and total ECCS flow rate. Results of this calculation are plotted in Attachment 2.
o ECCS Pump Operability calculations were performed per References 9 and 10 to determine if sufficient ECCS punp NPSH nargin is available or can be maintained under all credible torus temptrature and pressure conditions. he pressure drop attributed to the insulation debris as determined y above was factored into these calculations. A positive UPSH margin must be maintained to prevent the ECCS punps from cavitating.
similarly, these calculations verify that while maintaining adequate
NPSH margin, the ECCS pumps are able to fulfill their core cooling and torus cooling design functions.
n ese calculations evaluated two phases of ECCS operation. We first phase evaluated is the ten minute period of operation follow-ing the IccA, when no operator action to throttle or secure ECCS -
pumps is credited. Reducing the ECCS flow effectively reduces the NPSHR and increases the NPSHA. For this ten minute period of opera-tion, the calculations minimized the margin in NPSHA versus NPSHR by assuming that all LPCI and CS are running and discharging flow to the RPV, which is depressurized to the torus pressure. Torus water temperature was also maximized during the blowdown to conservatively 2.-
reduce the NPSHA (Reference 11). %e results demonstrate that for the maximum unthrottled ECCS flowrates, the NPSHA exceeds the NPSHR for calculated torus temperature and pressure conditions, hus, during this ten minute period, the ECCS pumps will not cavitate due to the additional pressure drop across the ECCS suction strainers created by the insulation debris.
We effect of insulation debris plugging the strainers on ECCS delivery is small. At purip runout conditions, reduction in flowrate for CS is approximately 70 gpn per pump and for LPCI, it is approxi-mately 110 gpn ps 9 (reference 9,10). GE has performed an analysis to detr eine the effect of a reduction in core spray flow i on the design basis large break locA. mis analysis conservatively ,
i asstaned the following:
=
- a. All debris inmediately collects on the strainers so that the full pressure drop penalty sust be taken.
- b. Core spray flow rate was degraded by 100 gpa per pump. Note that degradation of LPCI flow is not isportant since the assumed single failure for the design basis large break LOCA is failure of the LPCI injection valve; thus precluding LPCI flow into the vessel.
Results of this analysis (Reference 12) show that there is an p increase in Peak Clad Temperature (PCT) of <20'r. h is increase requires lowering the MAPLHGR limit (T.S. 3.11.A) by 1% to compensate for this 20'r increase. However, to be conservative, a 2% MAPLHGR limit decrease will be applied.
We design basis small break LOCA is not adversely inpacted for the following reasons:
- 1. A smaller break results in auch less debris. hus, pung flow degradation would be such less.
- 2. W e PCT for the limiting small biceak IOCA is 50'r less than for the large break. Even adding the full 20'r penalty from the large break case results in Pers that are bounded by the design basis large break LOCA. ,
he second phase of ECCS operation that was evaluated involved the period beyond ten minutes. For this phase, the torus temperature / pressure conditions are more limiting from an NPSH perspective, however, it'is assumed that the operator throttles the pu g flowrate according to the NPSH limits provided in the EOPs. We revised NPSH limits which account for the effect of insulation debris will be incorporated in the EOPs.
%e analysis shows that reducing flowrate is a viable option. Also, the EOPs will be modified to restrict the use of torus and drywell sprays only to control the containment pressure below its design limit. Se restriction on the use of sprays is necessary to ensure that enough 2 containment backpressure will be maintained to provide the necessary NPSHA. his will have the added benefit of resolving a potential concern currently being evaluated associated with the torus spray interlock setpoint. With these changes in the EOPs, NPSH margin can be maintained after ten minutes by reducing ECCS flowrate without compromising the ECCS core cooling or torus cooling design function.
As detailed above, when realistic assunptions are applied to the insula-tion debris impact model, acceptable ECCS pung NPSH is demonstrated for the ten minute period following the ux:A. Beyond that point, it has been demonstrated that operator actions will maintain acceptable ECCS pump NPSH conditions, thus, ensure ECCS operability. Bese operator Actions are dictated by the plant Bors. Se Bors will be revised to ensure that proper NPSM is available to prevent pump cavitation.
4
III. SLMMARY 1
A. Conclusions
- 'the purpose cf this document was to demonstrate that continued operation over the ren.ainder of the 1988/89 cycle is justified given the existing 4 configuration of the fibrous insulation and the ECCS hydraulic design.
- It was clearly identified that continued operation is justified on the bases of two technical arguments. First, Millstone Unit one has taken extensive measures to ensure that the probability of a toCA is minimized by performing IsI and iglementing material changes for crack susceptible i
systems. Each high energy system in the drywell has been thoroughly reviewed with respect to ISI results and inherent design features and it has been concluded that the probability of a double ended pipe break is
! extremely remote. In fact, the NRC has i g licitly agreed with this conclusion per Reference 3 (1987 IGSCC Inspection and Repair Review) in which they conclude that Millstone Unit one can be operated safely for the current fuel cycle. Secondly, it has been demonstrated in this
! document that the actual amount of insulation debris that clogs the i strainers is less than conservatively estimated in the referenced study.
i With this reduction in insulation debris, it has been demonstrated that g with reconnended tooP changes, sufficient ECCS puq NPSH margin is avail-i able or can be maintained for all credible torus conditions and ECCS flow l alignments. Additionally, it has been demonstrated that with the j reconnended MAPtJtGR limit changes, the insulation debris clogging will j not prevent the ECCS pu gs from fulfilling core cooling or torus cooling ,
! requirements.
e Generation Engineering concludes that there is reasonable assurance that l
Millstone Unit one can be operated safely for the current fuel cycle un-
~
til the revised final strainer design is implemented.
i B. Reconnendations
- 1. Replace the existing suction strainers with ones of a larger surface aree during the next refueling outage.
- 2. Revise the EOPs for the remainder of cycle 12 as follows:
l l
- a. Change the core rpray and LPCI NPSH requirements to account for debris and the effect of all six ECCS pungs 1
running simultaneously.
- b. Restrict the use of drywell and torus sprays to only l
control the conteinment pressure below the Pressure Suppression Pressure Limit (rigure 5 of EOP-580).
- 3. Administrative 1y reduce the MAPLHGR limit bv 2% (.2 kw/ft) for the remainder of Cycle 12 (see Attachment 3 for the revised MAPLHGR curves).
L
1 .. . .
1 i
i I ! ,
l 4. Initiate a Technical specification change to incorporate the 2% g IMPLHGR reductica into the Tech. Specs. for the remainder of ;
cycle 12. -
l s
i a
i e
l t
a e
1 v
b 6
k f
i l
g-I h
i
== !
,w_ ,*vw-w--,--wwm_v- --,-e-w----e.--eve +--.- .%w, - rvw-+-----*% -*wrw-- m-v-t-v
ATmonof!S l i
- 1) Insulatirm Debris Ispect Assessment Sunmary )
i
- 2) Results of NUSCO Calc. 486-071-626GM, Predicted Pressure Drop Across ECCS suction strainers as various Pressures and haperatures L
- 3) Revised MAPLHGR curves 4
1
- i l
9 d
t ;
i l
h
, l t
i ,
1 i
i h
l I
,_ ,_ ~ - _ _ _ - .. _ , _ . . - . _ _ ._.__._.-.z,.,m.,. , __.- _ . _ _ ,. . ,-__. . . . , , _ . . . . . . , , , . . . - . . . , , - . . . . _
i References
- 1. Millstone Unit one Drywell Insulation Acceptability Study, performed by J. D. Brassord, dated 3/8/88
- 2. Memo, PSE-CE-88-147, N. F. Azevedo to J. D. Brassord, "Millstone Unit No. 1 Drywell Insulation Study", dated 3/4/88
- 3. Letter M. L. Boyle to E. J. Mroczka, "MP-1 1987 Refueling Outage IGSCC Inspection", dated 10/26/87
- 4. GE Report t NEDO-21888, Rev. 2, "Mark I Containment Program Definition Report"
- 5. NUREG-0897 "Containment Emergency Sump Performance"
- 6. NUREG/CR-2791 "Methodology for Evaluation of Insulation Debris Effects"
- 7. NUREG/CR-2982, Rev. 1, "Buoyancy, Transport, and Head Loss of I'ibrous Reactor Insulation"
- 8. NUSCO Calc. 486-071-626GM, "MP-1 ECCS Suction Strainer Pressure Drop calculation", performed by J. D. Brassord, dated 5/2/98
/
- 9. NUSCO Calc. 084-044-416GM, Rev. 1, "MP-1 Allowable Operating Ranges
)
. . . e for LPCI Pumps Following a 14CA", performed by R. T. Peruolo, dated 5/9/88
- 10. NUSCO Cale. 084-044-453GM, Rev. 1, "MP-1 Core Spray NPSH Limit curves at 0, 5, and 10 psig Torus Pressure", perforrned by P. M. p vukas, dated 5/9/88
- 11. NUSCO Calculation e Wl-517-85--RE, "Millstone Unit 1: Post !4CA Torus Pressure and Wrus Water Tenperature", by Nirmal Jain, , dated 5/4/88.
- 12. General Electric Evaluation Sumary "ECCS Performance Sensitivity
'Ib Reduced Core Spray Flow For Millstone Unit One Nuclear Power Station", DRF A00-03243, dated 5/J 0/88.
GMB15:47 /
mev. 2
+
ATTAdWEENT 1
!aselation Debris tapact Assesseest Summary ,
Steer Justification of 3,etuaties seethed N r. Il Leo Deviation Free study ceaments
- 1. Seelga beels recits. Socire. Systee Same N/A Pipe break location based pipe break below elevattee on most direct short-term 13*4* transport path to torus duriag blowdena.
- 2. Anse6,t of fibreos ineele- 17.0 ft.' fibrees same s/A All insulation withis blow-ties debris generated inentation debris deva jet strees is conserva-total. tively assumed to be ree-dered embris due to jet impingeoemt. Realistically, the design bests double ended recire, pipe break will met create 17.0 ft. of de-bris because of the pipe dip restrelats which limit e to pipe deflection thus M e
causlag "shadowing" of the jet streme.
- 3. Amoest of f ame fibrosa .38 I 17.8 ft. . same s/A Justificaties for 30% fine debris created 5.1 ft.3 fibrous is based on en-perimental results detailed in referome:es 5 a 6. This estimate is ceaservatlwe.
- 4. Ameest of fine fibreos 5.1 ft. Same s/A g of the time fibreos
. debris tremeyerted to debris is conservatively the torno daring blaw- assumed to transport to deem torus during short-term blowdown phase. This assumption is conservative.
[: _ _
.e ste(y
- Justifiestiet. of .
Swelustf ee Stathed (Bef. Il y Deviatise Free Study Ceemmets S. Amoest of !*rger freg- Det critical to sees N/A Larger fragonets of debris meets of debris debris impact rapidly slaks which tod=scos mesemed to treesport essessenet transport poteetial. We to the torne during elsgging potential exists e
obert-term 'leudeus due to rapii sink rates a leeg-tere treceport. t elevated locaties (5 ft. free the bettes) of the secties stralmers even it debris tramsports to the torus.
- 6. .'ist *bution of fame Uniferely distri- Same N/A Unifera distributies due fibrw , debris in terug betod la the torus to turbulence caused by initial peel swell &
sabee, seat condensate oscillaties.
e
- 7. Durattee of terius Indefinite S stantes Finite turbulence based Turbulence in torus created turbulence en engineering judgment by IDCA bleedeus phase which e derived free the actual laducee poet swell and com-duraties of condeesste doesate oscillation. Assump-escA11 sties ed 65 seceeds, ties of indefinite turbe-See JCO for details. leece in the stw. is swer-ly conservative and uns made to bound analysis.
Study Justification of .
twelmettee seethed (per. il Deviatlee Free study comments a
Leo .
- 8. Fine fibrous debris laderinite fadetA/ i te for Cao 07. In stageant condittees, fine suspeestos poteetial. the taras buys fibrous eshibits a dettalte containing the sink rata. Turbulence main-section straleers. talas fine fibers suspeeded 5 ala. for reesir;- until the turbulence dampens lag fine fibrous to the point abere the slak debris. rate is predeolaaet ever the opmerd velocity compeeemt of the turbulence. In order to be conservative, the fine fibrous insulation debris da
- the bays costelaing section strainers is arbitrarily
- ssue_m to reenia la sus-pension.
- 9. Amount of fine fibrous 5.1 ft.I distri- 1.45 ft.' dis- The estimated tores ter- -Distributies of debris is dowris that clogs the but 4 uniformly tributed umi- bulence duraties of 5 malform (5.1 ft.33 to e
section s*rainers s equally ever fora'y a equally aim. results la a shorter -16 bays in torus-3 contain e ever the surface evet the surface period of time in ubich section strainers a:oa of the 3 area of the 3 suspended debris is preme -Insulatica la strainer bays secties strata- section strata- to transport to the assumed to immediately clog er:. era. secties strainers. See strainers 3/16 I 5.1 ft.'
comments for esiculaties -Torus volume turnover is mothed to reduce ameuat 29 minutes. Turbulence free 5.1 ft.3 to 1.65 ft. . duration is 5 mim.
-pensiador of insulaties embria susceptible to transport to stralmers during turbulsece (5/29) I (13/16) I 5.1 ft. .
h i _ _ _ _ _ __
e
. e Study Justificaties of _
Evaluaties mothed faef. 1) LCO Deviaties Free Study Comments
- 10. Desultaat pressure drop Pressure drop as a JCO pressure drop Difference in pressure drop Pressure drop across a bed screes sucties strainers functies of toep- et any gives flew caused by a reduced accumu- et fibers is linearly ese to accumulation of erature end flow- rete and tempers- of fibers as detailed in 99. propertional to the depth debris. rate developed. ture condition is of fibers. The pressure appremiestely 1/3 drops were calculated per of the pressure the methodelegy detailed drop calculated la WUREG/CR-2902, mee. 1 la the study. (aeference 7).
- 11. prsW nargia Evaluaties trith the pressure The NPss margia W/A The JCO WPSW margia eval-drepe calculated evaluation (ref. mation divided into two la tie, the NF5WA 9 & 10) demo - phasest 0-10 mia. & beyond
- 1. clearly on- strates that ad- 10 min following a 14CA.
coeded by SF1WR. equate WF55 margia For the first phase, no 52 detailed is inherently meta- operator acties is credited analysis was toined for le alt. to throttle the ECCS pumpe necessary to period following e and improve the NF55 com- a support this IDCA. Beyond 10 min., dittees. Per this period, $'
conclusion. restricties en the with the 6esige basis torus use e* containment toep./ pressure conditions spray a operator a most limiting flew com-acties (throttling ditions assumed, ref. 9 & 10 ECCS pumps) are used demonstrate there is ad-to malatain WPSR equate WPSW margia rer the margia without com- period beyond 10 sia.,
promising core coel- throttling of the ECCS pumps tag en torus cooling can be credited to improve esquirements. WPSW conditione. Also, the use of torus and get sprays will be restafeted to control the costaiament pressure below its desige limit to ensure adeguate WPSWA. Beatricties en the use of spreys and the revised NPSW limits which account fer strainer plog-ging will be incorporated la EOPs.
'ti
,e '
5 9
3 X v
f 8
v
' T I
0*
k$
15*
- I.
l :t -!
[ -i N=
i 1
ll 11 l
Y }* I 8
(
t
}
I
}
C 4
I I I 2
\
1 0 <
a ,
l i ! I S s l l
h I
+ , , - , , , , - , ,.----,-,--m~ , , - -..% e--- - ,----.w-m.w r- w-w--------- -,,v.gr
A7 T A < HMEl3T 1 7 ./, ._/0 INSULATION PRESSURE DROP
. HL (FT) TEMP (F) 30 -
H i E :
! 100 ,
D .
25 -""" " + " - "--
"I - " - " " -
L +:" +! i O : : i - 120 S ! '
S -
- A125 A 20 - - " " " " " l " " " " " "" " t " ""- " " " - l " - " "" ">""" """" "" ; ".""140 C
R i :
. . /
f '
S 33 _.
l
. . ; .. . . . . . . . . . i . . . . . . .
[
.,.. . . .g/ ..
[
.i.... .
160 2,0...
l .
j i !
. S i T 2 203
.6 220 R .
A ; i, 1 10 -i- -
n i- - ' - -
N . .
E .
R .'
I S i 5-"""""r [.
r " ""- -
i "" -
0 O 5 10 15 20 25 30 35
- TOTAL ECCS FLOWRATE O(GPM*1000:)
elke L
[
d_ 4g y '
5 s .
n Ql V ,e .
g
, [ '
J
- w
/ / giii.
. , e 3
. r_ u gMa w
)
l4je 8m g
e pn2k1 e<
/.4 9 c k
R .e
/ h.ea %{
b
,i, s Si.:s til.
. wer 5 /, E =
- gFalx{s]e A W0t:we3 3 J.5" Q h) b
- 55 'e if i 43't r- gd
$1-
{<t
- %d t .-
gig Ruu g ,/ ,.; 12 713 , 'a
- c. 3 a -
g ,; .:
=}f,
. o. e
~
rk A _ it I'! E 1
- $ -t .j !3*f.,3 g
.=
r . ..
hb5g h3:.h $
, 4
'd ic -
A 4 }"
32
- a_' _
E g : *-
g4 . d
. l' e' s ?
1 g
- -e {
i X$ 'l l -
n - . . .
2 O oss mumns aan mm mvu 2m2w Ani- ~
'2 Afd. .
{ J.W 3W h7 V AAp mf" '
6e t
t= ,b L- r -
1
/
e O
s
~.
3) l s
- l. .
A, !
M $ -
I W
/__
g I sg
. l
< E ,
I o
, ' .5 EE g **
i* E.
) .I.N
. 2 4'
.gw -
(. J t . .
x 4 h 3 l i
i T it Els1 ;
! I L h -
$gI~ l
~
! vt I N [f.
o.
Igl i
- -2
% is
- 1 I< lI
-[ .
l I
R I IN ;
,\ r . :
4 > g g.
- U
- d i <-t i I_
- a
.* 2 : 2 a C 4
r 1
(4/a9 Tl2 tavu wuvunn tv>w vv3=il nwvid un3Av wwiryw I 0
2