View source for 3F1096-02, Submits Partial Response to RAI Re Potential Overpressurization of Nuclear Svc Closed Cycle Cooling Sys. Rept 0920-208-01,Rev 0, Evaluation of SW Overpressurization Overpressurization, Dtd 920730,encl

{{Adams
| number = ML20138F506
| issue date = 10/08/1996
| title = Submits Partial Response to RAI Re Potential Overpressurization of Nuclear Svc Closed Cycle Cooling Sys. Rept 0920-208-01,Rev 0, Evaluation of SW Overpressurization Overpressurization, Dtd 920730,encl
| author name = Beard P
| author affiliation = FLORIDA POWER CORP.
| addressee name = 
| addressee affiliation = NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
| docket = 05000302
| license number = 
| contact person = 
| document report number = 3F1096-02, 3F1096-2, NUDOCS 9610170121
| package number = ML20138F511
| document type = CORRESPONDENCE-LETTERS, INCOMING CORRESPONDENCE
| page count = 5
}}

=Text=
{{#Wiki_filter:.
Florida Power CORPORATION 500 October 8, 1996 3F1096-02 U. S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, D. C. 20555-0001

==Subject:==
Potential Overpressurization of the NSCCC System Piping - Request for Additional Information

==Reference:==
: 1. NRC to FPC letter, 3N0896-19, dated August 27, 1996
: 2. NRC Information Notice 89-54, 3N0689-20, dated June 23, 1989

==Dear Sir:==

Florida Power Corporation (FPC) is submitting this letter as a partial response                                    l to the subject request for additional information (RAI) concerning a postulated failure of a Reactor Coolant Pump (RCP) Seal Area Heat Exchanger which could cause an overpressurization of the Nuclear Services Closed Cycle Cooling (SW)
.           System. RAI questions 1, 5, and 6 are addressed in this submittal. Responses to RAI questions 2, 3, and 4 will be provided on or before February 28, 1997.
FPC is unable to completely meet the 45-day response request in Reference 1 because of higher priority issues which limit the engineering resources available to resolve this issue. We believe a delay in the submittal of the complete RAI response is reasonable because our evaluation of the risk of this potential failure shows it to be a non-risk-significant event.                               This evaluation was performed using FPC's updated Individual Plant Examination (IPE) model for Crystal River Unit 3 (CR-3). Further discussion of this risk evaluation is presented in the response to RAI question 6. FPC's original position that this postulated failure is outside the licensing / design basis for CR-3 has not changed. We will provide the requested justification with the complete submittal in 1997.
I I
9610170121 961008 PDR                                                                                                         /
ADOCK 05000302 P                    pon CRYSTAL RIVER ENERGY COMPLEX: 15760 W. Power Line St . Crystal River, Florida 34428-6708 e (352) 795-6486 A Florida Progress Company

U. S. Nuclear Regulatory Commission
~ 3F1096-02 Page 2 of 5 RAI OVESTION 1 Provide a copy of ABB Impell Report No. 0920-208-01, Revision 1, " Evaluation of SW System Overpressurization."
FPC RESPONSE There is no Revision 1 to this report.           We have attached Revision 0 to this letter.
RAI OUESTION 5 Discuss the transients that could result from a thermal barrier heat exchanger failure as an initiating event and related operator actions to mitigate them.
Please provide applicable emergency operating procedures for operator action times and plant protection features. Also, please discuss plant response during these transients and cooldown using natural circulation.
FPC RESPONSE TRANSIENTS THAT COULD RESULT FROM A RCP THERMAL BARRIER HEAT EXCHANGER FAILURE AS AN INITIATING EVENT This type event has been previously evaluated by B&W (now Framatome Technologies, Inc.) in a "RCP Cover Leak Consequence" analysis performed in 1986.                The following is an excerpt from the introduction to this report:
         "As supported by field measurements of pump cover crack depths and demonstrated analytically in previous sections of this repcrt, the pump cover cracks are calculated to reach a depth of 160 mils and then arrest.
Field measurement of the crack depth has confirmed that all existing RCP cover cracks are 90 mils in depth or less.... The narrowest portion of the Reactor Coolant System (RCS) is shown ...to be 343 mils. It has therefore been concluded that propagation of an RCP cover crack leading to discharge of RCS fluid to either the containment or the Nuclear Services Closed Cycle Cooling System is not expected, ...would be detectable with existing equipment and instrumentation, ...that existing procedures are sufficient to control the event, and that the event is bounded by existing safety analysis."
RELATED ACTIONS TO MITIGATE THE EVENT Any leakage from the RCS would manifest itself as a net loss of RCS inventory.
This would be identified by the plant operators initially as an unexplained decreasing Makeup Tank (MUT) inventory.
Procedures currently exist that provide direction for identification of the leakage. The actions to be taken whenever any unidentified leakage from the RCS is suspected are contained in Operating Procedure OP-301, Operation of the Reactor Coolant System, Section 4.12, RCS Leakage Guidelines. The operator actions to be taken whenever leakage into the SW system is suspected begin at Step 4.12.9. Annunciator alarms on differential SW cooling water flow (inlet vs.

 !       U. S. Nuclear Regulatory Commission
     ,   3F1096-02 Page 3 of 5 1
                                                                                            \
outlet) and/or high radiation alarm in the SW System provide an indication that a leak in the RCP/SW piping could be.in progress. Guidance for operator response    l
.        to these alarms is provided in Annunciator Response (AR) Procedure AR-303, ESC       '
]        Annunciator Response.
EMERGENCY OPERATING PROCEDURES RELATED TO OPERATOR ACTION TIMES If an event such as the postulated RCP seal area heat exchanger tube rupture
:        scenario were to occur, the various operating procedures described above would provide guidance for operator actions.          If these procedural steps were
:        unsuccessful, eventually the operators would be directed to Emergency Operating
 .       E0P-08, LOCA Cooldown.       Except for the provisions of OP-301, no separate procedures exist which provide guidance for managing a RCS/SW 1eak condition.
Abnormal Procedure AP-330, Loss of Nuclear Service Cooling, deals specifically     i
!        with the loss of the entire SW system; however, this procedure does not address i        the RCS/SW leak concern since that guidance is provided in OP-301.
PLANT PROTECTION FEATURES As discussed in the first part of this response, the design of the system is such that any leak of consequence is unlikely. The design also has shown itself capable of sustaining small cracking that does not propagate even through a relatively small dimensional wall thickness. FPC does not expect gross failure of the heat exchanger tubing as discussed in Information Notice 89-54 (Reference 2), but rather if some failure did occur, we would expect the reactor coolant to
         " weep" into the SW System.
Leaks that could exist would be scall in nature. This provides adequate time for an orderly plant shutdown.        Sn.all leakages would manifest themselves as     '
in-leakage to the SW system surge tank (SWT-1) and would be indicated en the Radiation Monitor (RML-3) associated with this system. Very small leak rates may   l not show up as an increased level in SWT-1 (normal SW System leak rate may         i overshadow), but would be indicated as an increase on RML-3. Higher leak rates would be evidenced by an increase in the level in SWT-1, an increase in RML-3,     1 and a decrease in the level in the MUT.                                            l 1
PLANT RESPONSE DURING THESE TRANSIENTS If an RCP tripped as a consequence of a RCS/SW leak, and the unit was above approximately 75% power, a plant runback would result. This condition is adequately handled by an existing plant procedure AP-545, Plant Runback.
If an RCP tripped as a consequence of a RCS/SW leak, and the unit was operating at less than 75%, only an automatic re-ratio of feedwater to the once-through       j steam generators (OTSGs) would result (this assumes the OTSG levels remain above    :
low level limits).
                                                                                              ~
If the leakage was significant enough to be treated as a loss of coolant accident    ,
(LOCA) and cause a plant trip, E0P-8, LOCA Cooldown, provides guidance and the       '
needed strategies.

U. S. Nuclear Regulatory Commission
, 3F1096-02 Page 4 of 5 C00LDOWN USING NATURAL RECIRCULATION The entry conditior for E0P-9, Natural Circulation Cooldown, is natural circulation being raquired where no other E0P is applicable. No additional procedural guidance related to this issue is warranted for the conditions of concern.
Question 5 requests that applicable emergency operating procedures for operator action times and plant protection features be provided. We have discussed this response with NRR project management and it was agreed that the above discussior, meets the intent of question five and that full copies of all applicable procedures are not required at this time.
RAI OVESTION 6 Provide a discussion of risk insights, including frequencies, probabilities, and applicable operator actions, for this scenario, If this scenario was screened out in your Individual Plant Examination (IPE), please provide justification.
FPC RESPONSE The postulated scenario is a rupture of a tube in the RCP seal area heat exchanger which causes an overpressurization of the SW System which is the cooling water source in the RCP seal area heat exchanger. The tube side of the RCP seal area heat exchanger operates at RCS pressure. The occurrence of a small-break LOCA in one of the RCP seal area heat exchangers has its most significant effect on the IPE event sequence designated        "SU"  which is a small-break LOCA, failure of HPI scenario. For this analysis, FPC assumed that such a LOCA would result in an overpressurization of the SW System to the extent that the system would be rendered inoperable.
Loss of the SW System would result in the eventual failure of the normally running Makeup Pump MVP-18. However, the Makeup System design for CR-3 provides the flexibility to mitigate cooling water failures. While MVP-1B can only be cooled by the SW System, the other Makeup Pumps, MVP-1A and MVP-lC, can be cooled by either the SW System or the Decay Heat Closed Cycle Cooling (DC) System. The DC system is independent of the SW system and has redundant trains to account for single failures. Makeup Pump MVP-lC, which is the normal Engineered Safeguard (ES) 'B' HPI pump, is normally cooled by train 'B' of the DC System (the SW System serving as backup cooling) and, therefore, would not be affected by the loss of the SW System. Makeup Pump MVP-1A is normally cooled by the SW System, but can also be cooled using train 'A' of the DC System.       This transfer of cooling water sources requires operator actions which are proceduralized.
Historical data show that, upon loss of cooling, the Makeup Pumps do not immediately fail. As evidenced by actual CR-3 event data, it can take up to 3 hours for a Makeup pump to fail due to loss of cooling water. Therefore, there is ample time for switching the cooling of MVP-1A to train 'A' of the DC System after the LOCA and subsequent loss of the SW System, but before MVP-1A fails.
The frequency of occurrence of a small-break LOCA in any location is 2.0x10'3 per year. FPC conservatively assumed for this quantitative analysis that the

  ~
I V. S. Nuclear Regulatory Commission 3F1096-02 Page 5 of 5 frequency of a small-break LOCA in one of the RCP seal area heat exchangers is one-tenth of this frequency, or 2.0x10 per year. The normal probability of the operators not recognizing tgat the cooling source for MVP-1A needs to be switched                  l to the DH System is 1.0x10' . Operator reliability for this action is high due                     I to the relatively large amount of time available.               A sensitivity study was             l performed to calculate the effect on the total core damage frequency of a                          i small-break LOCA in one of the RCP seal area heat exchangers using different operator failure probabilities. For all of these cases, except the baseline (the updated IPE model for CR-3), the SW System was assumed to fail due to the LOCA.
The results were generated using the latest version of the CR-3 IPE model, and are shown in the table below.
Operator      Core Damage Frequency            %        PSA Applications Guide Error           (CDF) (per year)          Increase             Region Probability                                   in CDF Baseline              8.2 x 10-6               -                   -
i 0.001                8.6 x 10'6              4.9        Non-Risk-Significant 0.01                8.6 x 10''              4.9        Non-Risk-Significant                   '
O.1                9.5 x 10~6              15.8       Non-Risk-Significant 1.9 x 10 4 1.0                                       131.7        Further Evaluation                   i Required l
If even marginal credit is given for the operating crew to switch cooling to                       '
MVP-1A, the risk increase resulting from consideration of RCP seal area heat exchanger LOCAs is non-risk-significant according to the EPRI PSA Applications Guide (EPRI TR-105396, August 1995). It should be noted that these calculations assume that one out of every ten small-break LOCAs occurs in a RCP seal area heat exchanger, an assumption FPC believes to be extremely conservative.
This initiating event was not included in the original IPE submitted in March, 1993.      If it had been, the initiating event frequency would have been significantly less than the 2x10'' per year used in                                and the operator failure probability used would have been 1x10'ghis             analysis,
                                                                        , given  the relatively large amount of time the operator has to perform the task. The increase in core damage frequency due to the inclusion of the RCP seal area heat gxchanger LOCA would have been less than the IPE reporting criterion of 1x10' per year or greater, and, therefore, non-risk-significant.
Sincerely,
       .   . Beard, Jr.
k Senior Vice President Nuclear Operations PMR/JWT Attachment xc:     Regional Administrator, Region II            Senior Resident Inspector NRR Project Manager
.}}