ML20058C408
| ML20058C408 | |
| Person / Time | |
|---|---|
| Site: | Catawba  |
| Issue date: | 11/22/1993 |
| From: | Rehn D DUKE POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| TAC-M87867, TAC-M87868, NUDOCS 9312020447 | |
| Download: ML20058C408 (34) | |
Text
.-
l l
l a
l Dukel%wer Company D. L Ron i
l Catawba Nuclear Genero! ion Depanment liceItesident 1
4893 ConcordRoad (803)S313205 Offnce i
York.SC29:45 (803)S313426 Fax DUKEPOWER t
i
.J November 22,1993
)
U. S. Nuclear Regulatory Commission ATTN: Document Control Desk i
Washington, D. C.
20555 l
l
Subject:
Catawba Nuclear Station l
Docket Nos. 50-413 and 50-414 i
Response to Second Request For Additional Information CLA Water Volume And ECCS Subsystem Surveillance Requirements TAC Nos. M87867 AND M87868 On November 19,1993, Duke Power received a second request for additional information from the NRC staff related to a proposed Technical Specification change submitted by _
Catawba Nuclear Station on October 5,1993. This amendment application proposed changes to the cold leg accumulator water volumes,. charging and safety injection pump l
heads and flow rates, and residual heat removal pump flow rate..The information received j
from the NRC staff requested responses to four questions. Enclosed is Catawba's response to the questions contained within this second request for additional information.-
-l l
Pursuant to 10 CFR 50.91 (b)(1), the appropriate South Carolina official is' being provided 4
a copy of this letter, j
i 1
Very truly yours,
/Y Y
D. L. Rehn l
RKS/
l i
Enclosure 9322020447 931122 DR ADOCK 05000413 PDR i
[
awe ww rw.,
s.
l U. S. Nuclear Regulatory Commission November 22,1993 Page 2 xc: Mr. S. D. Ebneter Regional Administrator, Region II U. S. Nuclear Regulatory Commission 101 Marietta Street, NW, Suite 2900 Atlanta, GA 30323 Mr. Heyward Shealy, Chief Bureau of Radiological Health South Carolina Department of Health &
Environmental Control 2.600 Bull Street Columbia, SC 29201 American Nuclear Insurers c/o Dottie Sherman, ANI Library The Exchange, Suite 245 270 Farmington Avenue Farmington, CT 06032 M & M Nuclear Consultants 1166 Avenue of the Americas New York, NY 10036-2774 INPO Records Center Suite 1500 1100 Circle 75 Parkway Atlanta, Georgia 30339 Mr. R. J. Freudenberger
.NRC Resident Inspector Catawba Nuclear Station Mr. R. E. Martin Office of Nuclear Reactor Regulation U. S. Nuclear Regulatory Commission One White Flint North, Mail Stop 14H25 Washington, D.C. 20555
I i
U. S. Nuclear Regulatory Commission Page 3 D. L. Rehn, being duly sworn, states that he is Site Vice-President, Catawba Nuclear Station; that he is authorized on the part of said company to sign and file with the Nuclear i
Regulatory Commission this revision to the Catawba Nuclear Station Technical Specifications, Appendix A to License Nos. NPF-35 and NPF-52; and that all statements and matters set forth therein are true and correct to the best of his knowledge.
m -
D. L. Rehn Subscribed and sworn to before me this 22 day of A/0V,1993.
Synav7/0 s
[otary Publip/
My Commission Expires:
// w.2 1, 2 cr o
s.
DUKE POWER RESPONSE TO SECOND REQUEST FOR ADDITIONAL INFORMATION -
APPLICATION FOR AMENDMENTS REGARDING COLD LEG ACCUnfULATOR WATER VOLUME AND ECCS SURVEILLANCE REQUIREMENTS l
4 NOTE:
Questions 13 and 14 were mistakenly omitted from the DPC response dated November 15,1993 (response to initial request for additional information).
Q13.
Duk.e Power indicates that the RHR flow rate will be increased from 3648 gpm to 3900 gpm. Generic Letter 88-17 recommends that RHR flow be reduced for mid-loop operation to avoid vortexing. -Will this proposed increase in flowrate have an impact on RHR operation in'mid-loop operation?
A.
The proposed increase in the RHR surveillance requirement will have no impact on RHR operation in mid-loop. Physically, there will be no changes to the RHR system involved with the proposed TS changes. Analytically, credit will be taken for more RHR injection flow. In order to take credit for more RHR injection flow, the surveillance seguirement must be increased to ensure actual RHR performance remains above analysis assumptions. The latest CNS RHR injected flow test data, which is corrected for uncertainties, indicates that the 3900 gpm proposed TS will be acceptable.
Q14.
Duke Power states that the LOCA reanalysis to determine the impact of the proposed TS change meets the criteria of 10 CFR 50.46, including the PCT.
being below 2200 degrees F. Were the changes from the previous analysis such that they are small enough not to be considered to be a significant change (greater than 50 degrees F)?
A.
As discussed in the response to question 8, the large break and small break PCTs as a result of the reanalyses are 1945 and 1264 deg F, respectively. The current Westinghouse large break and small break PCTs, as given in Section 15.6.5 of the CNS FSAR (Reference 4), are 1954 and 1440 deg F, respectively. For simplicity, the above PCTs do not include the effects discussed in Reference 5.
Therefore, the large break PCT does not involve a significant change, but the small break PCT does. However, Duke Power typically reports significant changes only for the limiting PCT (large break in this case). Up until the operation of CNS Unit 1 Cycle 8 (CIC8), the current large break and small break analyses presented in the CNS FSAR have been the licensing basis. However, due to changing plant conditions (primarily steam generator tube plugging), it will be necessary for the LOCA reanalyses to become the licensing basis for startup of CIC8. Duke Power plans to keep the current LOCA analyses presented in the Page 1 of 4 1
i
?*
l CNS FSAR the licensing basis until the proposed ECCS TS changes are approved.
l Q15.
Please provide additional information on the cause for safety injection and charging pump flowrate limitation (NPSH versus cavitation) as discussed in Reference I to the DPC letter of November 15,1993.
A.
A copy of DAP-91-0074/DCP-91-0074 (Reference 1) is included as Attachment 1.
Specifically, Appendix I of Reference 1 discusses cavitation and NPSH requirements for the SIPS and CCPs. Per Reference 1, Westinghouse and Dresser / Pacific recommend a NPSH of 30 feet in order to support runout limits of 560 and 675 gpm of the CCPs and SIPS, respectively. Table 6-87 of the CNS FSAR (Reference 4) lists the available NPSH values for the SIPS and CCPs for the most limiting conditions. The available NPSH for each is approximately 60 ft. which well exceeds the 30 ft. requi ement.
l Q16.
Please provide a more explicit discussion of the reason for the increase in the pump head values of TS 4.5.2.f.1 and TS 4.5.2.f.2.
A.
In the Technical Specification Amendment for Catawba Unit 1 Cycle 7 (Reference 6), the CCP and SIP minimum developed head values were decreased.
Specifically, the CCP minimum developed head requirement was reduced from 2380 to 2223 psid, and the SIP minimum developed head requirement was reduced from 1430 to 1341 psid. These changes were made to provide for more test margin in performing the CCP and SIP pump head curves. These changes were consistent with the LOCA minimum ECCS system requirements in effect at that time. In the CIC7 submittal, the CCP and SIP runout limits were also revised per the information provided by Westinghouse and Dresser / Pacific in Reference 1. The changes to the CCP and SIP runout limits did not affect the LOCA minimum ECCS system requirements in effect at that time since the CCP and SIP minimum injected flow requirements (i.e., flow balance requirements) were not revised. Since CIC7 TS changes, new LOCA analyses have been performed which have different minimum ECCS system requirements.
As discussed in the response to question 2, the CCP and SIP flow balance requirements were reduced for the LOCA reanalyses. These reductions were necessary to completely address the mnout concerns identified by Reference 1.
However, because the LOCA reanalyses included additional input assumption changes that were expected to be PCT penalties (see response to question 8),
Westinghouse (LOCA vendor) strongly suggested that the ECCS injected flow assumptions be maintained to at least 99% of the previous total integrated injected flow assumptions. Since the flow balance requirements for the SIPS and CCPs were being reduced, it was necessary to take away some head curve test margin Page 2 of 4
e 1
)
i in order to satisfy the 99% integrated flow constraint.
Coincidentally, the proposed SIP and CCP minimum deveJoped head requirements (1418 and 2349 psid, respectively) required to satisfy the 99% integrated flow constraint are nearly the same as the requirements that existed before the CIC7 TS changes.
It should be noted that the most recent CNS SIP and CCP head curve test data will easily satisfy the proposed minimum developed head requirements.
l 1
l I
l f
l l
l l
Page 3 of 4 l
REFERENCES 1.
DAP-91-074, DCP-91-074, D. L. Fuller (Westinghouse) to R. C. Futrell (Duke),
" Emergency Core Cooling System Pump Runout Limit Issues," October 3,1991.
2.
Kabadi, J. N., et al, "The 1981 Version of the Westinghouse ECCS Evaluation Model Using the BASH Code," WCAP-10266-P-A, Rev. 2, March 1987.
3.
N. Lee, et al, " Westinghouse Small Break ECCS Evaluation Model Using the NOTRUMP Code," WCAP-10054-P-A, August 1985.
4.
Final Safety Analysis Report, Catawba Nuclear Station, October 1,1992.
5.
Ixtter from M. S. Tuckman (Duke) to USNRC, "McGuire Nuclear Station Docket Numbers 50-369 and -370, Catawba Nuclear Station Docket Numbers 50-413 and -414, Repon Pursuant to 10 CFR 50.46, Changes to or Errors in an ECCS Evaluation Model,"
October 18, 1993.
6.
Ixtter from M. S. Tuckman (Duke) to USNRC, " Catawba Nuclear Station, Docket Nos.
50-413 and 50-414, Technical Specification Amendment, Unit 1 Cycle 7 Reload," April l
13, 1992.
i i
l Page 4 of 4
i i
ATTACIIMENT 1 1
i
12-zs-ins eT:sean man a c rexem emirmim To ess2s131 e.ee g
Q:.T.
-.,,9, Westinghouse Energy Systems g5 Electric Corporation P
Idctober3,1991 1, :.: /
-- DAP-91-074 DCP-91-074 l
Mr. R. C. Futrell, Manager 1
Nuclear Safety Assurance Duke Power Company l
P.O. Box 1007 Charlotte, North Carolina 28201-1007 l
Dear Mr. Futrell:
Duke Power Company McGuire Units 1 & 2. Catawba Units 1 & 2 l
Emeroency Core Coolina System Pomo Runout 1.imit Issues i
i.
The purpose of this letter is to inform Duke Power l Company that an issue has been identified regarding the amount of runout margin available' for the Emergency Core Cooling System (ECCS) pumps.
In addition, this!1etter gives a synopsis of previous ECCS performance issues that inay affect pump runout and should be assessed to assure that the current ECCS configuration does not result in runout flows that exceed the runout limit for the pumpt, The ECCS j
pump runout margin and the previous ECCS performance issues are cescribed below:
j ECCS Pumo Runout Harain j
f Discrepancies have been found in information regarding the' pump runout limits (i.e., margin available beyond the design runout flowrate) of ECCS pumps manufactured by Dresser / Pacific Pumps for three operating plants.
In each of these cases, the discrepancies were addressed on a i
plant-specific basis. As part of the resoluti'on process, the pump vendor and Hestinghouse developed guidelines for determining the amount of margin i
available, which may be more limiting than the' assumptions lused in past runout margin assessments.
The runout limit of the ECCS pumps depends on the pump manufacturer, i
model, impeller type, and impeller casting type. Westinghouse records i
indicate that the ECCS pumps supplied to Duke Power Company by Westinghouse were manufactured by Dresser / Pacific Pumps.. Although Westinghouse has information regarding model, impeller type, and impeller casting type for the pumps that were originally supplied, Westinghouse
~
1 i
m 11-23-1993 07336RM FROM DPC NUCLERR ENGINEERING
.TO 88313191:
P.07 j
i l
October 3, 1991 Page 2 l
does not have conclusive information on the cuIrent status of the pumps (i.e., the impe11ers or pumps may have been modified or rep, laced). Note, the available runout margin beyond the design runout flow was evaluated j
only for pumps manufactured by Dresser / Pacific' Pumps, since'no other pump manufacturer has identified any change to their pump runout margins.L j
Previous ECCS Performance Issues l
i In addition to the issue of available runout margin, there are several previously identified system performance issue' that may increase the-s potential to exceed the runout limit of. the pump. These issues are:
miniflow, reactor coolant pump seal injection,i Technical Sp'ecification j
Verification. ECCS (branch line and total system) resistance, suction boost during recirculation, and ECCS flow measurement inaccuracies. These j
issues have been described in previous genericicommunicatio's. All of
)
n these issues should be. assessed, regardless of, pump manufacturer, to l
assure that the Duke Power Company ECCS configuration does not result in pump operation beyond the runout limit. Operating a pump beyond its runout limit may challenge its operability, cause pump dama,ge, and i
possibly result in a loss of function.
?
At present, there are too many variables for Westinghouse to as'sess the.
cumulative effect of this new ECCS pump runout mar' gin issue alo'ng with the
{
previous ECCS performance-issues on a generic basi;s. Therefore, it is
}
recommended that Duke Power Company assess these 1,$ sues for Duke Power Company. To assist you, attached is a discussion' f the variods factors that i
o effect ECCS pump runout flow determination.
j
. l 4
i If there are any questions, please contact the undersigned.
Very truly yours,
' g.
. L. Fuller, Manag'er.
Carolina Area Attachment 0145K l
1
11-Z3-1F33 07337An FROM DPC NUCLEAR ENGINEERING TO br.,m W A F.D8 EMERGENCY CORE COOLING PUMP RUN0tJT LIMIT ISSUES
SUMMARY
Discrepancies have been found in information regarding pump runout limits for Emergency Core Cooling System (ECCS) pumps at three operating plants. The ECCS pumps for these plants were manufactured by Dresser / Pacific Pumps.
In each of' these cases, the discrepancies were addressed on a plant-specific basis.
However, as part of the resolution, Dresser / Pacific Pumps and Westinghouse developed guidelines for determining pump runout limits, which:may be more limiting than the assumptions used in past runoutjassessments.-
BACKGROUND The ECCS provides emergency core cooling water to the Reactor Coolant System (RCS) following a loss-of-Coolant Accident: (LOCA). The; ECCS operates in two distinct modes:
injection and recireviation.. During the injection mode, emergency cooling water is supplied by the accumulators, the Charging / Safety Injection (CH/SI) pumps, the High Head Safety Injection (HHSI); pumps, and the Residual Heat Removal System (RHRS)/ Low Head Safety Injection pumps, depending on the ECCS design and RCS pressure. The ECCS pumps normally take suction from the Refueling Water Storage Tank (RWST) during this injection mode. During the recirculation mode, the RHRS pumps take suction from the containment sump.
Depending on the ECCS design, the CH/SI and HHSI pumps may receive suction flow from the RHRS pumps or recirculation pumps during the recirculation modo.
Prior to initial plant startup and plant startup,following ECCS modification, utilities are required to perform tests to assure adequate system performance.
i One of these tests assures adequate Safety Injection (SI) performance, which j
might include total pump flow, branch line balance, or verification of system flowrate distribution. Generally, Standard Technical Specification Surveillance Requirement 4.5.2.h defines the flow distribution test acceptance criteria. This specification provides a requirement for the minimum total flow through all SI branch lines, excluding the highest flow line. The highest flow line is assumed to have ruptured and will spill its flow into the containment or spill its flow against reactor coolant backpressure, depending upon the postulated size and location. of the pipe break..In addition, the specification provides a requirement for the maximum flowrate to preclude CH/SI and HHSI pump runout.
ISSUE DESCRIPTION During the review of the ECCS operating parameters and proposed revisions to Technical Specification surveillance requirement 4.5.2.h for three operating plants, the CH/SI and HHSI pump runout margins were evaluated.
It was Page I of 5 TAM 111er/ET-NSL*MFSL-91-ZZs/Z:09/19/91
}
u -zs-isss e7:3BRM FROM DPC NUCLEAR ENGINEERING TD um
,w
{
i i
i 1
EMERGENCY CORE C00 LING PUMP RUN0UT LIMIT ISSUES l
t i
discovered that an improper assumption regarding the configuration of the pumps
.j l
were made and, as a result, the available margin was less.than previously-5 believed. Specifically, the runout margin was assessed based on criteria for l
an investment cast impeller when the pumps, in fact,- had sandcast impellers.
l e
}
The available margin for sandcast impellers is not.as high as the margin for the investment cast impellers. 'After further investigation into the issue, l
i Dresser / Pacific Pumps and Westinghouse have defined new guidelines for assessing runout margin, based on the specific pump configurations.
TECHNICAL EVALUATION i
In general, pumps manufactured by Dresser / Pacific wit'h low capacity, sandcast i
impellers have'10 gpm margin available above the design runout flowrate of j
550 gps. Previous assessments for this type of pump may have assumed a margin; i
of 15 gpa or higher. Operating a pump beyond.its i
challenge its operability, cause pump damage, and. actual' runout-limit may.
possibly result in a loss of
~
function. Pumps with high capacity and/or investment cast impellers rty have greater than 10 gpm margin available above their design-runout: value.
Westinghouse has not evaluated the applicability of the issue regarding the amount of available margin beyond the design runout flow for pumps made by other manufacturers since no other change in pump runout margin has been identified by other manufacturers.
l For more detailed information on Dresser / Pacific margin guidelines, see l
Appendix 1.
In addition, a summary of pump runout flow limitations for the various Dresser / Pacific Pump configurations is presented in Table 1 of g
i Appendix 1.
Note, these guidelines represent fire limitations unless a test program has conclusively demonstrated that higher flows are acceptabl:.
i During the injection mode of operation the high. head SI pumps inject to the i
cold legs by means of a common header. Excessive pump runout will not occur if both pumps start since the pumps share the same discharge header. Excessive pump runout should be evaluated for single pump operation during the injection mode. During some phases of.the recirculation modes, some of the high head ?
pump discharge paths are realigned such that each pump may have a separate -
1 discharge header even if more than one pump operates. Potential runout problems may occur with more than I pump operating in these configurations.
When evaluating this issue of available runout margin, there are several previously identified system issues that may increase pump runout, which should be considered as well. These issues have been described in various previous Westinghouse connunications. These issues are: miniflow, Reactor Coolant Pump (RCP) seal injection, Technical Specification verification ECCS (branch line :
and total system) resistance, suction boost during recirculation, and ECCS flow Page 2 of 5-TAM m er/ET-ust-W st-gl.22s/3:09/19/s!
~
11-23-1W3 07:3SR1 MDM Dr'L NtJULW ENGINEER!tG TO t>osa s t.
i EMERGENCY CORE COOLING PUMP RUN0tfi LIMIT ISSUES measurement inaccuracies. Some of these issues may not be accounted for in preoperational or periodic pump runout testing. All of these issues should be assessed, regardless of pump manufacturer, to assure that the ECCS configuration does not result in pump operation that exceeds the runout limit for the pump configuration.
In addition to affecting pump runout flowrates, these issues may affect the adequacy of the response of the CCCS to accident I
scenarios. Note, the highest ECCS pump runout flows are developed following large break LOCA scenarios, and these issues may result in higher flows than 1
predicted for these events. This could affect the conclusions and validity of the large break LOCA analysis found in the FSAR.
A brief discussion of each issue is provided below (detailed discussions are provided in Attachments 1 through 4):
MINIFLOW Many plants have utilized the interim fix of eliminating the auto-isolation of the miniflow lines to prevent damage to centrifugal charging pumps in response to IE Bulletin No. 80-18 entitled " Maintenance of Adequate Minimum l
Flow Thru Centrifugal Charging Pumps Following Secondary Side High Energy Line Rupture." The interim fix results in the use of miniflow until its
+ isolation is required by the emergency operating procedures. This use of mintflow may not have been accounted for in the ECCS analyses. The i
diversion of flow to the miniflow path may result in a higher runout flow and decreased ECCS flow. For more information, see Attachment 1.
RCP SEAL INJECTION i
The actual RCP seal injection line resistance may be less than assumed in the system analysis, which may result in increased charging pump runout flow and/or reduced flow delivered to the core. For more information, see, Issue 1.
TECHNICAL SPECIFICATION VERIFICATION Inconsistedtes may exist between several of the current plant Technical Specifications and the current Westinghouse analyses of record. Therefore, the use of the current Technical Specifications may not assure that all assumptions of the ECCS analysis, as well as the runout limitations of the pumps, are met.
For more information, see Attachment 2 Issue 1.
ECCS (TOTAL SYSTEM) RESISTANCE Utilities may have reduced the system resistance to restore the original runout flow after a degradation in the charging pump head. The reduction in system resistance may result in reduced flow delivered to the core for
~
some LOCAs. This issue does not affect pump runout as long as the reduced system resistance does not result in flowrates that exceed the runout limits of the pumps. However, the affcct on the ECCS analysis assumptions and validity of the ECCS analysis results should be addressed in a comprehensive ECCS perfonnance analysis. For more information, see, Issue 2.
Page 3 of 5 TAMt11er/(7-NSL.MFst-91-ZZs/4:09/19/91
n -zs-m s eu sssn raon we ruct.ess erweenma.
To san.um
- r. m EMERGENCY CORE COOLING PUMP RUN0UT LIMIT' ISSUES ECCS (BRANCH LINE) RESISTANCE Utilities may be allowing the flow imbalance between branch lines to be greater than that' assumed in the ECCS analysis.. Increased branch line imbalance may result in increased spill flow and decreased flow delivered to the core. This issue does not affect pump runout, but should be addressed in a cceprehensive ECCS perforiaance analysis.
For more information. see Attachment 2, Issue 3.
SUCTION B00ST DURINE RECIRCULATION The ECCS-is tested and balanced while taking suction from the RWST.
However, for most plants, during the post LOCA recirculation mode of safety injection the CH/SI and HHSI pumps are " boosted" by the RHRS pumps, which are aligned to the containment sump. This " boost" increases the suction pressure and causes the SI pumps to runout further, which may cause SI pump damage if the system balancing did not account for this " boost". For more:
information, see Attachment 3.
FLOW MEASUREMEN" UNCERTAINTIES Actual flow coefficients for orifice plates in small lines, such as ECCS injection lines,'may be greater than those indicated by previous ASME standards, which may result in the actual system flow being higher than the indicated system flow. The higher flow may result in lower system resistance than assumed in the ECCS analysis, which would increase spill flow and decrease flow delivered to the core.
In addition, the higher system flow may exceed the pump runout limits. For more information, see.
SAFETY SIGNIFICANCE Operating a pump beyond its actual' runout limit may challenge its operability,:
cause pump damage, and possibly result in a loss of the safety injection function. A loss or change in the margin of SI flow delivered could result in~
LOCA peak clad temperature penalties. A higher injected flow may adversely affect the Low Temperature O'/erpressure Protection Systems ~setpoint analysis, Steam Generator Tube Rupture analysis, and Containment Integrity analyses.
Note, no Non-LOCA events are. penalized by higher SI flows during the recirculation mode.
OPERA 81LITY ASSE5SMENT Continued operation may be justified if the available runout margin of pumps, based on the guidelines of Appendix 1, is sufficient to accommodate the cumulative effect of the ECCS performance issues.
In addition, careful assessment of current test data may show that greater runout margin is available, even beyond what the generic runout guidelines predict.
If the predicted runout flows exceed the limits of Appendix 1. Table I, Westinghouse believes that continued operation may be justified until the Page 4 of '.
TAM d le r/E T-NsLMst 225/5: 09/19/91
11-ZS-1*MS 07140RM FROM - DPC NUCLEAR ENtilNEERitG lu BBSuivi t.Ac l
' (
I i
EMERGENCY CORE COOLING PUMP RUN0UT LIMIT. ISSUES
-l k
issues are resolved. The following may be ap~licable, depending on p
plant-specific findings, as input to a justification for continued operation:
Depending on the plant-specific pump full flow test data (e.g., during preoperational or subsequent maintenance testing), pump operability may be justified for a limited time of operation at flowrates that exceed the runout limits of Appendix 1. Table 1.
Following a large break LOCA1 probability of 5.0E-04/ year), where the ~
RCS backpressure is the lowest and the runout flowrates for the NHSI i
pumps are predicted to be the highest runout flow rate, the low head i
safety injection pumps may be shown to provide adequate core cooling.
Emergency procedures exist, References 1 and 2, to shutdown the plant following a loss of HHSI during accident scenarios with high RCS i
backpressure. These procedures present-a means to cooldown and depressurize the RCS so that RHRS cooling can be initiated. Note, the i
high RCS backpressure will limit the runout flow rate of.the ECCS pumps.
i, Emergency Response Guideline modifications could be devel' ped to give o
operators guidance to limit pump runout.
If the operator observes i
that pump runout flow exceeds the limits, the operator could then use -
i locally operated valves outside containment to throttle the system i
flow until it is within the runout limit. This throttled flow should i
not be decreased below the required injected flow based on the analysis of record.
.i i
RECOMDGED ACTIONS
[
Westinghouse recommends that each utility evaluate its current ECCS pump runout i
i limits and the cumulative effects of the system performance issues identified i
above to determine if the potential exists to exceed the runout limit for the particular pump / system configuration.
If the potential-to exceed.the limit-exists, then the utility should modify the system configuration-to atture that-pump operability will not be challenged.
i Also, in the event of sand cast impeller replacement, the use of investment cast impellers may yield greater runout margin.
l REFERENCE 4
I ECA-1.1, " Loss of Emergency Coolant Recirculation."
2.
ES-1.2, " Post LOCA Cooldown and Depressurization."
1 Page 5 of 5 TAMiMer/ET-Nst-MFSL 91-225/6:09/19/91
e 11-23-A"J53 epi 40Rn FROM DPC NUCLtees c)w4te m ait TU 88313131 P,13 l
i EMERGENCY CORE COOLING PUMP RUNOUT LIMIT ISSUES i
I i
APPENDIX 1 I
i ECCS PUMP MAXIMUM RUNOUT FLOW LIMITATIONS i
r i
i i
1 l
TAM 111er/ET-NSL M 5L-91-225/7:09/19/91
~
. - ~
11-23-1553 07 M lFrl PROrt DPC NUCLEM EBOINEEkiro lu
- 00510171 F=~ 14
' y.
APPENDIX l-ECCS PUMP MAXIMUM RUN0UT FLOW LIMITATIONS i
Page 1 of 7 BACKGROUIS 1
Recent developments with potential flow measurement inaccuracies and Emergency Core Cooling System (ECCS) flow balancing requirements have dictated that the Centrifugal Charging / Safety Injection (CH/SI) pumps and Safety Injection (SI) pumps may be required to operate continuously with flow conditions that exceed 1
the existing pump runout limitations. For Westinghouse 4-loop design plants, the CH/S! pumps have a design runout limit of 550 gpm and the SI pumps have a j
design runout limit of 650 gpm. For Westinghouse 3-loop plants, the CH/S!
pumps have a design runout limit of 650 gps. Operation with flow rates that exceed the design runout limits requires an evaluation to determine the effect of the increased flows on the pumps.
l In order to implement increased runout flows for the ECCS pumps,' an evaluation of the specific pum) and system parameters must be performed. The two major i
concerns that must se addressed in the evaluation are cavitation and motor horsepower capability. Cavitation will occur if the pump required NPSH at the increased runout flow rates is not satisfied by the system available NPSH.
Minor cavitation can lead to long-term pump degradation, while severe cavitation ami 2-phase flow can lead to short-term pump damage. Thus.
operation with cavitation should be avoided.
Operation at increased runout flows also can increase the brake horsepower required by the pump. The motor must be capable of operating satisfactorily at the new horsepower level.
l HORSEPOWER CONSIDERATIONS The CH/SI and SI pumps are designed such that the pump developed' head falls i
sharply as the flow rate approaches and surpasses. the design runout flow.
The falling head curve causes the brake horsepower curve to become very flat at flow rates beyond the design runout point. Because of this horsepower curve-characteristic, the horsexneer at the increased runout flows will-be 6ssentially the same as tte horsepower at the design runout flows.
It is, therefore, expected that the required. horsepower at increased runout flows will remain within the horsepower capability of the motor. The increased runout.
flows should not affect the qualified life of the motor insulation.
Additionally, the increased runout flows are not expected to change. the electrical load requirements for the emergency diesel generators. However, l-each of these factors should be checked on a pump specific basis before -
increased runout flows are approved.
i s
tam t 11er/ET-NsL.W st-91-225/8: 09/19/91
u-zs-ms mme mon D C ttKLERR EPOltiEERItO TO w
Ah r
APPENDIX 1 i
ECCS PUMP MAXIMUM RUNOUT FLOW LIMITATIONS l
Page 2 of 7 l
i
\\
PUMP CONFIGURATIONS L
CH/SI Pumps 1
The large majority of CH/SI pumps currently in service are the Dresser Pump Division (Pacific Pumps) model 2-1/2' RL-IJ. This pump model has been supplied with several different impeller configurations that result in different runout '
flow limitations. All of the 2 1/2" RL-IJ CH/S! pumps have a high capacity-suction impeller. ~ However, there are two different types of radial impe11ers.
available and two different impeller casting processes.
The radial impe11ers are available in high-capacity (650 gpm design runout) and I
low-capacity (550 gpm design runout) configurations. The high-capacity impe11ers are primarily used in Westinghouse 3-loop plant designs.
Westinghouse 4-loop plants use primaril The impeller capacity for a particular y the low-capacity radial impe11ers..
i pump can be determined pump outline drawing or the original vendor performance curve.from either the 1
The suction impellers and both the high capacity and low capacity radial impe11ers have been manufactured with either a sand casting process or an 1
investment casting process. The sand cast and investment cast impe11ers have a distinct difference regarding the pump perfomance at runout flow conditions.
The sand cast 1 spellers inherently have somewhat rough surface finishes and' have minor irregularities in the vane shapes. The investment casting process provides smoother surface finishes and improved vane shapes.. As a result, the.
3 investment cast impe11ers have somewhat lower NPSH requirements and are able to operate at higher runout flows than the sand cast impe11ers. The investment castings also will provide slightly higher developed head than the sand cast impe11ers without a significant increase in brake. horsepower. The two impe11er'.
types must be evaluated independently for determining the acceptability of the increased runout flows. The high capacity impe11ers were upgraded to the investment cast design in the mid-1970's. The low capacity impe11ers were upgraded to the investment cast design in the early 1980's.
Impe11ers supplied after these time frames will generally be the investment cast type.. However, the impeller casting type for any specific pump rotor should be confirmed oy plant maintenance records and by manufacturing records before assessing the-runout limitations for that specific rotor.
~
r TM111er/ET-k$L4st-91-225/9:09/19/91
man a n - " * "* '
~~"
L.
. 22-zs-ms eazm l
APPENDIX 1 ECCS PUMP MAXIMUM RUNOUT FLOW LIMITATIONS Page 3 of 7 SI Pumps The large majority of SI pumps currently in service are Dresser Pump D1 vision (Pacific Pumps) model 2-1/2" JTCH (JTCHM) and model 3" JHF. All JTCH (JTCHM) i pumps and all replacement rotors for these pumps were manufactured with sand 1
cast impellers. There is only one impeller design in terms of capacity J
(650 gpm design runout). The JHF pumps were manufactured in both 10 stage and i
11 stage configurations, depending on plant s>ecific hydraulic requirements.'
i The 10 stage JHF pumps were supplied with ettser sand cast or investment cast-impellers. The 11 stage JHF pumps were supplied only with investment cast tmpellers. Both of the JHF models were designed for the same runout flow i
capacity (650 gpa design runout). Due to the nature of the 10 stage JHF pump i
impeller design,- the use of sand cast and investment cast impellers will not have a significant effect on pump performance and'NPSH requirements.
Thus, it i
is not necessary to determine the type of impeller casting in order to evaluate l
increased runout flows for either SI pump model.
CAVITATION CHARACTERISTICS The characteristics-of centrifugal pump impellers often result in a specific flow capacity at which the NPSH required to. suppress cavitation increases in an asymptotic manner. Thi4 condition can occur in the suction impeller of the pump or in the subsequent radial impe11ers, dependent-on the particular characteristics of the specific impeller designs.
If an attempt is made to j
operate at or beyond this critical flow capacity, cavitation will occur i
regardless of the NPSH available at the suction of the pump. The developed head of the pump will degrade until it matches the system, and the pump will operate in a state of partial cavitation.
Cavitation in a pump can be associated with both long-ters and short-ters degradation mechanisms.
Long-term degradation mechanisms would include impeller level cav,itation energy. diffuser and wear ring erosion due to ~the continual pre Short-term degradation mechanisms would include wear ring rubbing, mechanical seal face wear and bearing wear due to high levels of i
rotor vibration and deflection resulting from high level-cavitation energy.
High levels of cavitation can be similar in effect to running a pump dry.
The point at which the asymptotic increase in required NpSH occurs in a particular pump is critical in evaluating the increase of pump runout flows.
3 For the 4-loop CH/SI pumps with low-capacity sand cast impellers.-Dresser Pumps has determined that the asymptote 1's located beyond 560 gps.
For the 4-loop i
.f TAMiller/ET-N$L-MF5L-91 225/10:09/19/91 e
n-zs-m s on 4ssn rei rc h
-~~ ~
w t
i
[..
s I
i i
APPENDIX 1 1
i
.ECCS PUMP MAXINUM RUNOUT FLOW LIMITATIONS l
Page 4 of 7
'l
(
CH/SI pr r.
..th low-capacity investment cast impe11ers, Dresser Pumps has j
i determinou Nat the asymptote is located beyond 580 gpa. For both of these
}
impeller types, the NPSH asymptote occurs on the second stage of the pump. For j
the 3-loop CH/SI pumps and all SI pumps, Dresser Pumps has determined that the asymptote is located b6 yond 675 gpa.
For these cump models, the NPSH asymptote 4
3 probably occurs on the suction impeller of the pump. These flow rates should' j
be treated as maximum allowable flows for acceptable ~ pump operation..These i
flow rates are acceptable only if the NPSH available satisfies the identified -
l j
pump requirements.
j PWP RUNOUT FLOW ~ LIMITATIONS i
4-Loop CH/SI Pumps i
i i
Dresser Pumps has identified that the originally supplied sand cast
]
low-capacity impellers have the potential for.second stage cavitation occurring
)
at increased runout flows. The second stage cavitation can occur despite the i
head developed by the first stage of the pump due to the NPSH characteristics j
of the sand cast low-capacity impellers. This second stage cavitation l
condition is independent of the suction. impeller NPSH margin. Dresser Pumps l
l has determined that the low-capacity impellers should not start to cavitate until the flow rate exceeds 560 gps. Thus,.the sand cast low-capacity l
3 I
impellers should not be operated beyond 560 gpa in order to preclude j
cavitation. To support operation at 560 gpa, the available NPSH at the pump j
suction should be at least 30 feet.
Dresser Pump Division has determined second stage cavitation is not a concern i
for the 2.1/2" RL-!J pump model with investment cast low-capacity radial.
1 j
impe11ers at flow rates up to 580 gpe.
Dresser has estimated based on test i
data at 550 gpm that the investment cast low-capacity impe11ers will require l'
NPSH of approximately 30 feet at 580 gps. The low-capacity investment cast 1
impellers should not be operated beyond 580 gpe-in order.to preclude 1
cavitation. To support operation at 580 gps, the available NPSH at the pump t
suction should be at least 30 feet.
f 3-Loop CH/SI Pumps i
Dresser Pumps has identified that the controlling NPSH limit for the CN/SI
[
pumps with the high-capacity impe11ers is probab ;y in the-suction impeller.
Thus, these pumps are probably limited by first stage cavitation rather than
[
second stage cavitation. The HPSH requirements for the sand cast and i
l i
TAM 111er/ET-NSL M R-91 225/11:09/IS/91
.i I
l j
q a
22_z3-2-,33 em3m emn ore Wuctem enm.-~
v eestsist c.~
APPENDIX 1 ECCS PUMP MAXIMUM RUN00T FLOW LIMITATIONS Page E of 7 investment cast suction impellers are nearly identical.. Dresser Pumps has determined that the high-capacity impellers should not. start to cavitate until
{
the flow rate exceeds 675 gpm. Thus, the high-capacity impe11ers, sand cast-and investment cast, should not be operated beyond 675 gpm to preclude cavitation. To support operation at 675 gpm, the available NPSH at the pump suction should be at least 30 feet.
i SI Pumps Dresser Pumps has identified that the controlling NPSH limit for the SI punms, both 2-1/2" JTCH (JTCHM) and 3" JHF, is in the suction impeller. Thus, these pumps are limited by first stage cavitation rather than second stage cavitation. Dresser Pumps has determined that these pumps should not start to cavitate until the flow rate has 4xceeded 675 gpm. Thus, the SI pumps should not be operated beyond 675 gpm in order lo preclude cavitation..To support operation at 675 gpa, the 2-1/2" JTCH (JTCHM) pumps should have an available NPSH of at least 30 feet. To supnort operation at 675 gpe, the 3" JHF 10 stage pump with sand cast or investment cast impellers should have an available NPSH of at least 35 feet. To support operation at 675 gpm, the 3" JHF 11 stage pump with investment cast impellers should have an available~ NPSH of at least l
18 feet.
j sHORT-TDtM 4-LOOP CH/SI PUMP OPERATION WITH CAVITATION It has been identified that the 4-loop plant CH/SI pumps for some specific plants are currently aligned to operate at a flow rate of 580 gpe. This is an' acceptable operating point with *.he f avestment cast impellers. However with i
the sand cast impellers and a flow rate of 580 gps, it is probable that cavitation will occur in the inlet of the second stage impeller even though no cavitation occurs in the first stage. This is due to the fact that the second stage impeller and subsequent radial impellers are designed for lower capacity 4
than the first stage impeller. Because the fluid in the inlet of the second stage has already been elevated ane visately 40 psi above the pump-suction pressure, the level of cavitati".c ew c is expected to be relatively low in the second stage. This cavital w *A J1d then be suppressed by the additional pressure increase that occurs t W the second stage impeller, thus.the third stage and subsequent stages shoo, se relatively free of cavitation.
It is not possible to quantify the rate or extent of pump degradation that can occur due to cavitation. However, based on the level of cavitation energy.
expected while operating the low-capacity sand cast CH/SI pump. impellers at j
1 I
TAM 111er/ET-NSL-MFSL-91 225/12:09/19/91 i
~.
I 11-Z3-1"73 CT344RN r' ROM Df C FQJCLCRR CIC31Nt!"MITC nj 55313151 P.13 i
)
APPENDIX 1 4
l ECCS PUMP MAXIMUM RUN0UT FLOW LIMITATIONS Page 6 of 7 i
a 580 gpm, the pump should be capable of operation or a period of 15 minutes l
without loss of function. With the pump operating n this cavitating. state for i
longer than 15 minutes, the cavitation could endan tr the integrity of, critical i
pumpcomponentssuchasthemechanicalseals,impeIlerwearringsandshaft bearings.
If the pump flow rate is reduced from 580 gpa to 560 gpa or lower within 15 minutes, the pump should be capable of resuming normal operation.
CONCLUSION l
Table I summarizes the runout flow limitations for the all of the CH/SI and SI l
pumps that were manufactured by Dresser Pump Division (pacific Pumps).. The runout flow limit is dependent on the pump model and the type of impeller
{
casting that is used.
It has been established that the pumps should not experience significant cavitation and the motors should not overheat during extended operation under the conditions identified in Table 1.
The flow i
limitations in Table 1 are based on test data from a limited number of pumps of the specific models and configurations-identified and also on analysis of test data from pumps of similar capacity.
These limitations have been conservatively chosen so that they are applicable to all pumps of. the specified models and configurations. The NPSH requirements in Table 1 are based on
-i 4
testing that was performed with water at ambient conditions (<100 *F) and l
are also applicable to all pumps of the specified models.and configurations.
The CH/SI and SI pumps typically operate at runout only for full flow testing i
once each refueling outage and potentially for one year post-accident.
Because i
of this limited operation at runout, the increased runout flows should have no effect on the long-tors mechanical and hydraulic performance of the pumps.
Acceptable equipment operation can be adequately demonstrated on a pump specific basis by proper verification of the specific pump configuration,.the NPSH margin and the horsepower requirements.
l 1
If it is necessary to operate a CH/SI pump with low-capacity ' sand cast.
j impe11ers at 580 gps, it is likely that some cavithtion will occur in the pump second stage impeller. However, this cavitation is expected to be relatively low in energy and the pump should be capable of operating acceptably for a' period of 15 minutes as a one time occurrence. Operation at 580 gpa for longer than 15 minutes is not recommended since the cavitation would endanger the integrity of the pump mechanical seals, wear rings and shaft bearings.
Operation of the other pump configurations at higher than identified flow rates j
is not recommended even under short-term conditions.-
1 I
j tam 11e r/ET45L-W$L-91 -225/13 : 09/18/s1 5
11-23-1753 07 3 45RT1 FROrt DPC NUCLEFR EN31NEFRIK3 TO tsca131W1 P.20 APPENDIX 1 ECCS PUMP MAXIMUM RUNOUT FLOW LIMITATIONS l
Page 7 of 7 TABLE I t
SUMMARY
OF PUMP RUNOUT FLOW LIMITATIONS I
PUMP TYPE IMPELLER CURRENT' MAXIMUM 15 MINUTE NPSH CASTING DESIGN CONTINUOUS-LIMITED REQUIRED TYPE RUN00T RUNOUT RUNOUT (FT)
(USGPM)
(USGPM)
(USGPM) 2-1/2" RL-IJ SAND 550 560 580 30 4-LOOP CH/SI i
2-1/2" RL-IJ INVESTMENT 550 580 30 4-LOOP CH/SI 30 2-1/2" RL-IJ SAND 650 675 3-LOOP CH/51 5
f 2-1/2" RL-IJ INVESTMENT.
650 675 30 3-LOOP CH/SI 2-1/2" JTCH SAND 650 675 30 i
SI 3" JHF 10 STAGE SAND 650 675 35 SI 3" JHF 10 STAGE INVESTMENT 650 675 35 SI 3" JHF 11 STAGE INVESTMENT 650 675 18 SI eu TAM 111er/CT-NSL-NFSL-91-225/14:09/19/91
t
,11-Z3-A M3 DT34JRFI f'ftOrt M FUCLEM EPC3rCEttirG3 TO 5531.1151 N.cl 1
1 E
I i
L i
i t
i,
- E f
ATTACHMENTS l
i RELATED EMERGENCY CORE COOLING SYSTEM
- i
~
i PUMP RUN00T ISSUES f
i i
j i
i 1
t i
I i
i
~
i i
1
{
l s
4 TAM 111er/ET-NSL-NFSL-91-225/15:08/19/91 1
9 y-,,
-y--
r..
.,.-.7
,.,.,,rm.
e
,,.., -..,w, -,, -..,
-,c.,
d- -
,w.
m.l. -
11-23-15773 UT3 45R1 I"Ror1 DPC KUCLERR ENDINEERING TO 55313131 F. c.s.
i RELATED EMERGENCY CORE COOLING PUMP RUNOUT LIMIT ISSUES i
ATTACHMENT I MINIFLOW (IE80-18]
l Page 1 of 2
[
Westinghouse identified a concern for plants that utilize the Centrifugal Charging Pumps (CCPs) as Emergency Core Cooling System (ECCS) pumps, whert the miniflow for the CCPs is automatically isolated upon Safety Injection (SI) initiation. The particular circumstances that could have resulted in d: image vary from plant to plant, but involve unavailability of the power operated relief valve, with one or more CCP re-pressurizing the reactor during SI following a secondary system high energy line break. Since the SI signal automatically isolated the CCP miniflow return line, the flow through the CCP was determined by the pump characteristic head vs. flow curve, the safety valve setpoint, and the flow resistances and pressures losses in the piping and the reactor core. That resulting flow may not have been adequate to insure pump cooling, and may have resulted in pump ' damage before the SI termination criteria was met.
Therefore, Westinghouse issued two Information Letters both entitled
" Centrifugal Charging Pump Operation Following Secondary Side High Energy Line Rupture," issued in May 1980 and August 1980.
In addition, the Nuclear-Regulatory. Commission issued IE Bulletin No. 80-18, entitled " Maintenance of Adequate Minimum Flow Thru Centrifugal Charging Pumps following Secondary Side High Energy Line Rupture," dated July 24, 1980.
The Westinghouse Information Letters recommended that plant specific calculations be performed to determine if adequate minimum flow is assured under all conditions.
If adequate minimum flow was not assured, then Westinghouse recommended specific equipment and procedure modifications, which would result in adequate minimum flow. The modifications were termed " interim"
)
(i.e., intended for the short-ters), since they were not up to safety-related design standards and took credit for timely operator action.
The implementation of this interim fix may not have been incorporated into the analysis of record. The interim fix eliminated the auto-isolation of the miniflow lines until isolation is required by the emergency operating procedures. During the period when the miniflow path is open, injected h
T AM111er/ET-nst-stst-31 225/16: 09/19/91
11-23-1953 07245Rn PROM DPC NLCLERR EBOINEERIFO TO SS313151 P.23 I.
RELATED EMERGENCY CORE COOLING PUMP RUNOUT LIMIT ISSUES ATTACHMENT I MINIFLOW Page 2 of 2 core flow would be less than the flow calculated with miniflow path isolated.
In addition, for a large break LOCA, the Reactor Coolant System pressure may have decreased quickly.
If the runout flowrate of the ECCS pumps was established with miniflow isolated, then excessive pump runout may have occurred before the operator could have taken action to isolate the miniflow path.
The use of the interim fix should have been. evaluated to assure that the injected core flow is bounded by the analysis of record and that the runout conditions do not exceed the limits established by the pump vendor.
If the
" interim fix" is the " permanent fix", then the analyses should include this impact.
l I
1
)
TAMiller/ET-NSL-if5L-91-22s/17:09/19/91
11-zs-tess er:4esn enon vec rerum ereireemre To eesisist e.za
,f..
RELATED EMERGENCY CORE COOLING PUMP RUNOUT LIMIT ISSUES i
ATTACHMENT 2 RCP SEAL INJECTION / TECHNICAL SPECIFICATION VERIFICATION ECCS fBRANCH LINE AND TOTAL SYSTEM 1 RESISTANCE
[PI-88-012/038]
Page I of 4 An inconsistency was found between the plant Emergency Core Cooling System (ECCS) configuration assumed in the input to the Westinghouse supplied Loss-of-Coolant Accident (LOCA) analyses used to demnstrate compliance with the -equirements of 10CFR50.46, and the configuration allowed in the Technical Specifications.
In the event of a LOCA, the Safety Injection (SI) flowrates from the ECCS may have been lower than the flowrates assumed in the LOCA l
analyses. The review of the potential effects on the calculation of the safety injection flow used in the large break and small break LOCA ECCS analyses indicated that the analysis results may incur a penalty for the peak cladding temperature calculation.
Westinghouse did not believe that a substantial safety hazard (as defined in 10CFR21.3) existed. However, an unreviewed safety question may have existed, i
since it could not be determined if the margin of safety as defined in the basis for the technical specification had been reduced until each plant's ECCS configuration and analysis assumptions were reviewed.
Therefore, Westinghouse issued an Information Letter entitled "ECCS Flow Inconsistencies," in December 1989.
The Westinghouse Information Letter stated that based on the review of-the 3
assumptions made for the analysis of-record, three issues were identified that are discussed below:
ISSUE 1 - RCP SEAL INJECTION / TECHNICAL SPECIFICATION VERIFICATION A Reactor Coolant Pump (RCP) seal-injection flow at charging pump runout i
conditions was assumed, which corresponded to a specific seal injection itne system resistance that was used in generating ECCS flows for LOCA analyses. As long as the actual resistance was equal to or greater than that used in the
?
calculation, then the RCP seal injection assumptions were met or exceeded.
The RCP seal injection assumption was verified by Standard Technical Specification TAM 111er/ET-NSL M SL-91-225/18:09/19/91
1,1-23-1993 07347RM FROM DPC MJCLERR EPOINEERlr0 TO 88313191 P.25 RELATED EMERGENCY CORE COOLING PUMP RUNOUT LIMIT ISSUES f
ATTACHMENT 2 RCP SEAL INJECTION / TECHNICAL SPECIFICATION VERIFICATION ECCS (BRANCH LINE AND TOTAL SYSTEM) RESISTANCE Page 2 of 4 surveillance requirement 3.4.6.2.e, which compares measured seal injection flow at normal RCS pressure to a' calculated value based on the ECCS analysis resistance.
In reviewing the technical specifications and Westinghouse ECCS flow calculations, it was determined that the Technical Specification values for several plants were inconsistent with the ECCS flow calculations. The seal injection piping resistance associated with the Technical Specification basis was less than that assumed in the calculations; therefore, additional Charging /SI flow would have been pumped through the seal injection li.ne.
For a large beak LOCA, this potentially could have resulted in pump runout and/or a reduction of flow injected into the core.
For a small break LOCA, a reduction of flow injected into the core would have resulted.
The Information Letter stated that based on a parametric study the additional runout would be no higher than an additional 15 gpm for the Charging /SI pump, which would have operated without any safety-related problems. However, the amount of margin depends on the pump manufacturer, model, impeller type, and impeller casting type.
Based on recent guidelines presented in Appendix 1, this'15 gpm runout may be acceptable for continuous operation for all except the low capacity, Charging /SI pumps with sand cast impellers. These pumps have a design runout flow value of 550 gpa. The runout flowrate limit, as defined by Oresser/ Pacific and Westinghouse, is 560 gps. However, a runout flow as high as 580 gpa may be acceptable for these pumps for a limited period of operation (15 minutes). Note, Westinghouse recommends that a plant-specific evaluation be performed if the potential exists to exceed the recommended runout limit for the particular pump.
The effect of this issue is dependent on plant-specific characteristics.
Therefore, to resolve this issue, Westinghouse recommended that the Charging /SI ECCS performance be recalculated based on the current technical specification, or that the Technical Specifications be revised.
l
+
T AM 111er/E T-NSL-MF st-91-225/19 : 09/19/91 t
u-zs-im crruesn enon pec nuccess enoinezalna To ist:entw1 r>. ze RELATED EMERGENCY CORE COOLING PDMP RUNOUT LIMIT ISSUES ATTACHMENT 2 RcP SEAL INJECTION / TECHNICAL SPECIFICATION VERIFICATION ECCS (BRANCH LINE AM) TOTAL SYSTEM) RESISTANCE Page 3 of 4 ISSUE 2 - ECCS (TOTAL SYSTEM) RESISTANCE The system resistance is calculated based on the actual original Charging /SI pumps runout head / flow. Some utilities have informed Westinghouse that this resistance is periodically decreased. This is done to restore the original runout flow after a degradation in the pump head.
By reducing overall system resistance, a resulting decrease in the flow delivered to the core could result for some LOCAs.
This reduction in flow delivered to the core occurs when the ruptured branch injection line is spilling to a containment pressure that is significantly lower than the pressure of the Reactor Coolant System. The flow in each branch line is proportional to the square root of differential head divided by branch line resistance. The spilling line has a significant differential head (essentially pump discharge pressure less containment pressure). Thus, as the injection line resistance is decreased, flow out of the ruptured line increases, and the flow through the injection line decreases, relative to the original system resistance.
To insure the ECCS flow performance assumptions of the analysis of record are met, the hydraulic resistance associated with the Charging /SI pump runout conditions must be compared to the analysis resistance.
If the resistance is lower, it should be restored or the effect on the LOCA analyses should be determined.
In addition, the. Charging /SI pump operating conditions should be verified against the runout limit for the pump, as defined by the guidelines in Appendix 1, to assure the operability of the pump is not challenged.
ISSUE 3 - ECCS (BRANCH LINE) RESISTANCE The current Westinghouse ECCS flow calculations assume that the flow through each Charging /SI pump branch line is within a specific limit of each other with no backpressure. However, utilities may operate their plants with higher imbalance in these branch lines.
l T AM 111er/E T-kSLMEL-91 -225/ 20 : 09/19 /91
l 11-1"3-1993 U r..+or-u a cru,x s Lrv ic,tCLtM ENGINEERIfda eo
_..,u.
i, f, y.'
I RELATED EMERGENCY CORE COOLING PUMP RUNOUT LIMIT ISSUES ATTACHMENT 2 5
RCP SEAL INJECTION /TECM ICAL SPECIFICATION VERIFICATION
, ECC$ (BRANCH LINE AND TOTAL SYSTEM) RESISTANCE Page 4 of 4 This higher imbalance may be due to the reduction in resistance in a branch line below the value assumed in the analysis.. If a branch line with the reduced resistance ruptures, then. the spilling flow may be greater than assumed in the analysis. This issue does not affect pump runout, but should be addressed in a comprehensive ECCS performance analysis.
i t
TAM 111er/ET-lest-W sL.91 225/21:og/19/g1
j n-zs-1=f3s er:49AM FROM DPC' NUCLEAR ENGINEERING TO 441v1 t'. 26 -
v.r-t i
i e
RELATED EMERGENCY CORE COOLING PUMP RUNOUT LIMIT ISSUES
[
?
ATTACHMENT 3 i
1 e
SUCTIDN B00ST DURING RECIRCULATION (PI-88-026]
I i
i i
)
Page 1 of 2 f
The significance of the effect of Residual Heat Removal System (RHRS) to l
Emergency Core Cooling System (ECCS) boosting was partially' addressed in the' Westinghouse Information Letter entitled " Potential for Inadequate NPSH During.
Post-Accident Recirculation," issued in November, 1988. The Westinghouse Information Letter provided information for use in evaluating the potential for j
~
inadequate ECCS pump Net Positive Suction. Head (NPSH) during the post-accident j
recirculation phase of a Loss-of-Coolant Accident (LOCA),
t l
Nuclear Regulatory Commission issued Information Notice No. 88-74, entitled
-l "Potentially Inadequate Performance of ECCS in PWRs During Recirculation Operation Following a LOCA," on September 14, 1988. The Information Notice was issued to alert licensees to potential concerns that could result in inadequate performance of the ECCS during the recirculation phase of operation following a LOCA.
i The aforementioned Westinghouse Information Letter recommended that each utility verify that adequate NPSH is available to the RHRS, Containment Spray, i
and High Head Safety Injection (HHSI) pumps during the post-accident j
recirculation phase.
i Further to this recommendation, while the RHRS pumps supply suction to the HHS!
pumps during the recirculation mode, this " boost" increases the HHSI suction pressure and may cause the HHSI pumps to runout further than during the injection mode. Systems are typically configured to preclude pump runout based on the injection mode alignments. Therefore, runout flowrates, as well as NPSH i
should be addressed for each plant. Although abundant NPSH may be available I
during the recirculation mode, pump cavitation may occur.
See Appendix 1, for a detailed discussion of the potential for pump cavitation.
Westinghouse believes that continued operation can be justified until the
" boost" issue is resolved, if the predicted runout flows exceed the limits of Appendix 1, Table 1.
The following may be applicable, depending on plant-specific findings, as input to a justification for continued operation:
l l
i TAM 111er/ET.nsL.Mrst.gl.22s/22:09/1g/91 i
j 11-zs-zess er:4 san FRon pec r<a.tesr< c;a.niczroru To
- essist, r..
)
l i
]
RELATED EMERGENCY CORE COOLING PUMP RUNOUT LIMIT ISSUES i
ATTACHMENT 3 SUCTION 800ST DURING RECIRCULATION Page 2 of 2 Pump operability may be justified for a limited time of operation at flowrates that exceed the runout limits of Appendix 1, Table 1, depending on full flow test data of the pump;.
Following a large break LOCA (probability of 5.0E-04/ year), where the RCS backpressure is the lowest and the runout flowrates for the HHSI pumps are predicted to be the highest runout flow rate, the low head safety injection pumps may be shown to provide adequate core cooling; Emergency procedures exist to shutdown the plant following a loss of HHSI during accident scenarios with high RCS backpressure. These procedures present a means to cooldown and depressurize the RCS so that RHRS cooling can be initiated. Note, the high RCS backpressure will limit.the runout flow rate of the ECCS pumps.
Emergency Response Guideline modifications could be developed to give operators guidance to limit pump runout.
If the operator observes that pump runout flow exceeds the limits, the operator could then use locally operated valves outside containment to throttle the system flow until it is within the runout limit. This throttled flow should not be decreased below the required injected flow based on the analysis of record.
i i
i
~
I 4
i i
TAM 111er/ET-NSL-MFSL-91 -225/23 : 09 /19/91 i
l
11-23-1993 07:50AM FROM DPC f<UCLERR EfG1recmN 3y 88313191 P.3e
.m
,3 RELATED EMERGENCY CORE COOLING PUMP RUN0UT LIMIT ISSUES 1
s ATTACHMENT 4 i
FLOW MEASUREMENT UNCERTAINTIES (PI-90-022]
Page 1 of 2
{
Prior to initial plant startup and plant startup following Emergency Core Cooling System (ECCS) modification, utilities are required to perform tests to assure adequate system performance. One of these tests assures adequate Safety i
Injection balance, or(SI) performance, which might include total pump flow, branch line verification of system flowrate distribution.
Technical Specification surveillance requirement 4.5.2.h defines the balance test acceptance criteria.
This specification provides a requirement for the minimum total flow through all SI branch lines, excluding the highest flow line. The highest flow line is assumed to have ruptured and will spill its flow into the containment or spill its flow against reactor coolant backpressure, depending upon the postulated size and location of the pipe break.
In addition, the specification provides a requirement for the maximum flowrate to preclude pump j
- runout, r
The SI flowrate is typically measured by using orifice plates in the SI branch lines, although, some plants may use pipe friction taps or elbow taps.
Various considerations must be taken into account in order to develop the minimum and maximum required balanced SI flow. These considerations include seal i
injection, line resistance, and flow measurement inaccuracies.
l Westinghouse issued an Information Letter entitled " Emergency Core Cooling System Flow Measurement Potential Discrepancies," in October 1990.
The Information Letter addressed the notential for the flow coefficients for the j
i orifices in the SI lines to be gre'ater than those indicated-using previous ASME standards. The higher flow coefficients may result in an actual flowrate that is significantly higher than indicated and a potential inability to meet 1
Technical Specification surveillance requirement 4.5.2.h.
i Having an actual flowrate higher than the indicated flowrate may result in a violation of the maximum flowrate requirement of the ECCS.
Exceeding this requirement can result in lower flow injected into the core for small break Loss of Coolant Accidents (LOCAs) for SI performance based on the flow spilling 1
i i
TAM 111e r/ ET - NsL-MF SL 225/24 : 05/19/91
11-23-1333 07350Rr1 PRCF) DP"C KUCLERR EK31NEERifG ne C0013131 I'*31
'~
t RELATED EMERGENCY CORE C0OLING PtmP RUNOLE LIMIT ISSUES.
ATTACHMENT 4 i
FLOW MEASUREMENT _ UNCERTAINTIES page 2 of 2
\\
i to containment pressure.
In addition.. exceeding this requirement could result in higher pump runout.
Increased pump runout could result in damage to the pump and a loss of safety injection function.
Lower injected core flows or a loss.of $1 flow could result in LOCA Peak Clad
'k Temperature (PCT Temperature overp)ressure Protection Systems setpoint analysis, Stea penalties.
l Tube Rupture Margin to Overfill, and Containment Integrity analyses.
l j
Justification to continue operating for issues such as' increased PCT, steam.
generator tube rupture overfill, containment pressure analysis, and HHSI pump
}
~
runout flowrates have been provided on a case by case basis depending i
upon the magnitude of the potential. discrepancy and known margins available '
within or outside of the current licensing basis.
i Westinghouse recomended that all utilities that use orifice plates to calculate ECCS flowrates during initial or subsequent system tests assess this potential discrepancy,for their plant.
In addition,.if-a' utility used orifice i
plates to calibrate another method (such as friction or elbow taps) to assess i
ECCS performance, then they also should assess this flow measurement potential Westinghouse also recommends that the utility review the ECCS
- l discrepancy.
perfonsance analysis of record to assure that the appropriate flow measurement discrepancies have been applied to the minimum and maximum allowable flowrates.
i In addition to these recomendations, each significance of this potential flow measureutility should. assess the ment inaccuracy / bias to assure that i
operating conditions do not result in pump runout limits, as defined by the pump vendor, being exceeded.
.{
)
i 1
l 4
8 2
TAMillerKT st$LMgt.gg-gg3/g3:gg/gg/gg i
.Sd58 % JEL5V