NRC-95-0077, Forwards Revised Pump Relief Requests PR-6,PR-7 & PR-8 for IST Program for Pumps & Valves & New Pump Relief Request PR-11 for EECW Pumps,As Result of Util 950407 Response to Insp follow-up Item 95002-01
| ML20086G978 | |
| Person / Time | |
|---|---|
| Site: | Fermi  |
| Issue date: | 07/14/1995 |
| From: | Gipson D DETROIT EDISON CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| CON-NRC-95-0077, CON-NRC-95-77 NUDOCS 9507170360 | |
| Download: ML20086G978 (15) | |
Text
l Dougjag R. Gipson Sensor Wce Prescent Nuclear Generation l
Detroit re-2 l
e 6400 North Omie Highway I
Newport. Michigan 48166 (3I3) 586-5249 i
July 14,1995 l
' U. S. Nuclear Regulatory Commission Attn: Document Control Desk Washington, D. C. 20555 i
References:
1)
Fermi 2 NRC Docket No. 50-341 NRC License No. NPF-43 2)
Detroit Edison letter to NRC, " Response to Inspection Follow-up Items 95002-01 and 95002-02," NRC-95-0034, dated April 7,1995
Subject:
Relief Requests for Inservice Testing Program for Pumps and Valves i
Detroit Edison is submitting four (4) Relief Requests required as a result of the Detroit Edison response to Inspection Follow-up Item 95002-01 (Reference 2).
Three (3) relief requests (PR-6, PR-7, and PR-8) are revisions of pump relief l
requests already contained in the Inservice Testing Program for Pumps and Valves
}'
for Fermi 2, however, they require NRC approval as new relief requests.
Additionally, one new pump relief request (PR-11) for the Emergency Equipment Cooling Water (EECW) pumps is being submitted for NRC approval.
PR-6, PR-7, and PR-11 request relief from the Code to use pump reference curves j
over a narrow range versus a single fixed reference as re ;uired by Articles IWP-1500 and IWP-3000 of ASME Section XI.
PR-6 and PR-7 have been revised to provide additional information in support of testing each system by running pumps in parallel and for the use of variable -
reference curves. Additionally, Detroit Edison has obtained the services of two outside consultants, Mr. Brad Stockton and Mr. Ned Zeman, both of VECTRA, to-review the positions taken by the company. They have independently concluded that testing as required by the Code is impractical. They have further concluded that the testing methods and acceptance criteria employed in these two Relief Requests are adequate to assess system and pump performance. These
~(
[1j 9507170360 950714 PDR ADOCK 05000341 Q
(
UShiRC July 14,1995 NRC-95-0077 Page 2 conclusions are consistent with the Basis for Relief provided in the Relief Request.
The consultants report is available for NRC review.
The Standby Liquid Control Pumps already are tested to a fixed reference value.
Except for the Core Spray, Residual Heat Removal Service Water and Emergency Equipment Cooling Water pumps, which are covered by the relief requests, the remainder of the pumps in the IST Program are to have fixed reference values developed for them. Since Reference 2 was submitted, Detroit Edison determined that fixed reference values can be developed for more pumps than originally believed to be feasible. Due to the significantly larger scope, the fixed reference value development will not be completed by July 21,1995 as originally committed.
Detroit Edison intends to develop new fixed reference values for each pump during its normal surveillance test. These fixed reference values will be developed and validated through the conduct of the normal quarterly surveillances. While these are being developed, Detroit Edison requests interim relief from the requirements of Articles IWP-1500 and IWP-3000 for fixed reference values until Refueling Outage 5. Until the fixed references are developed, the variable references will be used as the acceptance criteria.
PR-8 has been revised to propose alternate vibration criteria for overall velocity of the Residual Heat Removal (RHR) pumps. The new criteria is being adopted from ASME OMa-1988, Part 6. Reliefis required because Detroit Edison intends only to adopt the acceptance limits at this point but not take measurements at the upper motor bearing housing.
Detroit Edison requests prompt review and approval from the NRC for relief requests PR-6, PR-7, PR-8 and PR-11 and for the interim relief on testing other pumps to a fixed reference value to complete the resolution of this issue.
1 The following resised commitment is being made in this letter:
- 1. Develop fixed reference values for IST pumps (except for Standby Liquid Control, Core Spray, Residual Heat Removal Service Water, and Emergency Equipment Cooling Water pumps) by the startup following Refueling Outage 5.
1 s erw-
US'NRC July 14,1995 NRC-95-0077 Page 3 i
Ifyou have any questions on this matter, please contact Mr. Hari O. Arora at (313) 586-4213.
Sincerely, Enclosure cc: T. G. Colburn H. J. Miller M. P. Phillips A. Vegel l
)
Enclosure to NRC-95-0077 Page1 RELIEF REQUEST NO. PR-6-R1 PUMP (S):
Core Spray PUMP NO.s:
E2101C001 A, E2101C001B, E2101C001C, and E2101C001D CLASS:
2 FUNCTION:
Provide low pressure emergency core cooling SECTION XI REOUIREMENTS: Anicles IWP-1500 and IWP-3000 require that a set of fixed reference values be established for each pump l
in order to detect deviations in pump performance.
BASIS FOR RELIEF: Reliefis requested to deviate from the Code in I
two areas-i 1.
To test both Core Spray Pumps in a Disision in parallel. That is, both pumps are to be run together and treated as a single component. Differential pressure and developed head reference values represent a combined pump flow characteristic. Vibration data will continue to be monitored on each pump. Since both j
pumps are run in parallel, acceptance criteria for flow rate and differential pressure are set at more restrictive criteria than that provided in Table IWP-3100-2.
2.
For flow rate and differential pressure, a flow reference curve, rather than a single fixed value of differential pressure and corresponding flow, will be utilized. This reference cune will be developed utilizing linear regression with four or more flow versus differential pressure data sets over a limited range of flow.
Each Core Spray Division is designed to have two pumps operating in parallel in accordance with Technical Specifications Paragraph 3.5.1.a.
If one pump is determined to be inoperable, then the division is declared inoperable. The system test line configuration is not designed to allow for single pump testing as system operation with a single pump is
i I;'
I Enclosure to j'
N1(C-95-0077 Page 2 f
j RELIEF REQUEST NO. PR-6-R1 (Continued) l l
considered inoperable. It is designed only to accommodate full flow i
i testing required by Technical Specifications Surveillance Test requirements.
j.
i It is not practical to run the Core Spray Pumps one at a time in the test j
i lineup configuration. The test line flow control valves are throttled j
approximately 13% open for Division 1 and 9% open for Division 2 to i
i control two pump test flow ( Pumps A and C are in Division 1, and l
Pumps B and D are in Division 2 ). The existing flow control valves are not capable of throttling low enough(less than 5% open) to
{
4 q
accommodate single pump operation without experiencing unstable operation, cavitation, and severe vibration.
I i
L Significant system modification would be required to enable testing of the Core Spray Pumps in accordance with Code requirements. These i
modifications would be costly. They are costly, both in terms of l
resources and increased radiation exposures during installation. Since j
the system performs adequately, there is little benefit, other than J<
compliance with the Code, for installing any of them.
The use of pump reference curves is further necessitated by the fact that existing test line and flow control valves are over sized. The flow control valves are opened to a point in the span of travel in which small l
changes in valve position result in relatively large changes in flow rate.
This would present an unnecessary challenge to both the equipment and the Plant Operators to attempt to return to a fixed reference value. The combined referer.ce pump curves were developed using five to seven i
data points over a 600 gpm range of flows. The data was curve fitted using linear regression, which is a sensible method considering that the pump curve is essentially linear over this very small range.
A review of preservice test data and inservice test results obtained prior to establishing the reference curves confirmed that the pumps were in good operating condition when the curves were developed. A review of the test results obtained using the reference curves shows that the data l
. ~.
~
Enclosure to NiiC-95-0077 Page 3 RELIEF REQUEST NO. PR-6-Ri (Continued) is consistent and trendable. Additionally, the individual pump vibration data is extremely stable and indicates no signs of degradation on any of the Core Spray Pumps. Ifinvalid data were used to generate the punip reference curves, or if the curve fit was poor, the test results would be erratic, and 10 years of historical performance demonstrate that this is not the case for the subject pumps.
As pump performance degrades, the actual combined two pump performance curve will degrade in a manner similar to a single pump's performance curve. The only difference is that if only one pump in a combination curve for two identical pumps suffers a 10% degradation, then the combined two pump curve would result in a 5% indicated degradation (assuming no degradation for the other pump). Therefore, the imposition of more restrictive limits than those contained in Table IWP-3100-2 assures that equivalent degradation detection sensitivity is achieved for two pump testing versus single pump testing.
ALTERNATE TESTING: The following acceptance criteria will be established for the Core Spray Pumps:
1.
For each Core Spray Division, differential pressure and flow is being measured with both pumps running. A reference curve for pump performance monitoring has been developed over a range of 600 gpm (this represents only 8.5% of the total flow range) at flows which exceed the Technical Specifications minimum flow and discharge pressure requirements. This reference curve is being utilized to evaluate current pump performance for degradation.
2.
In order to enhance the ability to detect the equivalent of one pump's degradation to the minimum acceptable level of performance per Table IWP-3100-2, the following acceptance criteria will be utilized, which are more stringent than Code limits:
i.
Enclosure to NkC-95-0077 Page 4 RELIEF REQUEST NO. PR-6-R1 (Continued)
Acceptable Rance Alert Rance Reauired Action Rance
~
AP Low Values Hich Values Low Values High Values g
0.965 to 1.02 0.95 to
>l.02 to
<0.95
>l.03
<0.965 1.03 1
!)
i
___.______-..m_
Enclosure to NRC-95-0077 Page 5 RELIEF REQUEST NO. PR-7-R1 PITMP(S):
RHR Service Water PtiMP NO.s:
El 151C001 A, El 151C001B, El 151C001C, and El151C001D CLASS:
3 FITNCTION: Provide ccoling water to the RHR Heat Exchangers SECTION XI REOtIIREMENTS: Articles IWP-1500 and IWP-3000 require that a set of fixed reference values be established for each pump in order to detect deviations in pump performance.
BASIS FOR RELIEF: Reliefis requested to deviate from the Code in two areas:
1.
To test both RHR Service Water Pumps in a Division in parallel.
I That is, both pumps are to be run together and treated as a single component. Differential pressure and developed head reference values represent a combined pump flow characteristic.
Vibration data will continue to be monitored on each pump.
Since both pumps are run in parallel, acceptance criteria for flow i
rate and differential pressure are set at more restrictive criteria than that provided in Table IWP-3100-2.
2.
For flow rate and differential pressure, a flow reference curve, rather than a single fixed value of differential pressure and corresponding flow, will be utilized. This reference curve will be developed utilizing linear regression with four or more flow
)
versus differential pressure data sets over a limited range of j
flow.
Each RHR Service Water Division is designed to have two pumps operating in parallel per Technical Specification 3.7.1.1. Design basis flow rates and cooling capacities cannot be achieved with only a single pump operating. Technical Specifications paragraph 3.7.1.1 requires i
i i
i
Enclosure to NRC-95-0077 Pace 6 RELIEF REQUEST NO. PR-7-R1 (Continued) that two pumps in each RHR Semce Water Subsystem be operable. In this configuration, one RHRSW pump in a Division is not simply a backup for the other. Both are required to fulfill the subsystem's safety function. There is no surveillance requirement to verify a specific flow rate.
To run a single pump in an RHRSW Division would require that the test flow rate be established at a point between the flow at which cavitation at the flow control valve occurs and pump runout. When running only one pump in a Division, flow rate must be throttled to prevent pump runout. However, if flow rates are throttled below 5,250 gpm when discharging to the cooling towers, or 5,600 gpm when discharging to the cold weather bypass, extreme valve cavitation will occur. In addition, flow induced vibration in the RHR Heat Exchangers can occur in the 4,200 to 4,500 gpm flow range. Thus, the range of permissible testing flows for a single pump test are very restrictive and the risks of allowing sustained flows in the restricted flow ranges are too great to adopt a single pump test procedure.
Significant system modifications would be required to enable individual testing of each RHRSW division. These modifications would be costly and would also unnecessarily complicate normal RHRSW System operations. They are costly, both in terms of resources and increased radiation exposures during installation. Since the system performs adequately, there is little benefit, other than compliance with the Code, for installing any of them.
The use of pump reference curves is funher necessitated by the fact that the system is tested with the flow control valves in tne full open position. There is no flow control ability in the system with the valves in the full open position (the flow control valves do not throttle in the closing direction). The actual flow achieved is dependent upon whether the tested subsystem is in cooling tower bypass or standby mode, and on i
what other systems are in operation that draw from the RHR reservoirs at the time of testing. A review ofinservice test results obtained using the reference curves shows that the data is consistent and trend G.
I 2
Enclosure to NRC-95-0077 Page 7 RELIEF REQUEST NO. PR-7-R1 (Continued)
Additionally, the individual pump vibration data has been, and is, extremely stable and indicates no signs of degradation on any of the l
RHR Senice Water Pumps. Ifinvalid data were used to generate the pump reference curves, or if the curve fit was poor, the test results would be erratic, and this has not been the case in over 10 years of testing experience with these pumps.
As pump performance degrades, the actual combined two pump performance curve will degrade in a manner similar to a single pump's performance cun'e. The only difference is that if only one pump in a combination curve for two identical pumps suffers a 10% degradation, then the combined two pump cun'e would result in a 5% indicated degradation (assuming no degradation for the other pump). Therefore, the imposition of more restrictive limits than those contained in Table IWP-3100-2 assures that equivalent degradation detection sensitivity is hieved for two pump testing versus single pump testing.
P ALTERNATE TESTING: The following acceptance criteria will be established for the RHR Senice Water Pumps:
[
1.
For each RHR Senice Water Subsystem, differential pressure and flow is being measured with both pumps running. A l
reference curve for pump performance monitoring for each subsystem has been developed over a range ofless than 500 gpm j
(this represents only 5.5% of the total flow range). This
{
reference curve is being utilized to evaluate future pump l
performance for degradation.
l 2.
In order to enhance the ability to detect the equivalent of one pump's degradation to the minimum acceptable level of performance per Table IWP-3100-2, the following acceptance criteria will be utilized, which are more stringent than Code l
limits:
i 4
3 h
I 1
i i
Enclosure to NRC-95-0077 Page 8 1
i RELIEF REQUEST NO. PR-7-R1 (Continued)
Acceptable Range Alen Range Reauired Action Range AP Low Values liigh Values Low Values High Values r
0.965 to 1.02 0.95 to
>l.02 to
<0.95
>l.03
<0.965 1.03 l
r l
l l
l I
Enclosure to NRC-95-0077 Page 9 RELIEF REQUEST PR-8-R1 PLIMP:
RHR Pumps PUMP NO.:
E1102C002A, E1102C002B, El 102C002C, E1102C002D CLASS:
2 FUNCTION: Provide low pressure coolant injection SECTION XI REOUIREMENTS: Vibration amplitude displacement shall be measured quarterly during insenice testing (Table IWP-3100-1 and IWP-3400 (a)).
BASIS FOR RELIEF: During preoperational testing, the vibration velocity measurements for the RHR pumps were high (.19 to.27 ips) relative to the vibration velocity ranges given in Relief Request PR-1. Three of the four pumps fell into the alert range.
The preoperational vibration velocity measurements were part of a series of vibration tests which included vibration amplitude. The results -f the tests are F
summarized in Detroit Edison Engineering Research Report 84L;8-2 which concludes:
"The vibration measurements which were taken during the preoperational test indicate that the pumps are operating in a satisfactory condition."
Given that the pumps are operating acceptably at the higher vibration velocities, a new set of vibration velocity ranges were developed with the assistance of the vendor as described in Detroit Edison Engineering Research Report 85D15-5, Revision 1.
Report 85D15-5 recommends the measurement of overall velocity (filter out),
overall amplitude (filter out) and running speed amplitude (filter in). All three quantities will be measured at the pump bearing and compared to the ranges given in Alternate Testing.
_=
f.*'
Enclosure to l
NIlC-95-0077 i '
Page 10 i
Testing data over 10 years has revealed that the vali!es for overall velocity l
have remained fairly constant and that the action levels recommended in Report 85D15-5 are not conservative. Therefore, the vibration limits j
contained in ASME/ ANSI OMa-1988, Part 6, Table 3, (ranges for Test j
Parameters) are being adopted as the vibration limits for overall velocity for the l
RHR Pumps.
?
l Reliefis required for the use of the vibration limits without implementation of
]
all of the requirements of OMa-1988, Part 6. Specifically, Detroit Edison does i
not intend at this time to implement Paragraph 4.6.4(h). This paragraph i
j requires that on vertical line shaft pumps ".. measurements shall be taken on i
the upper motor bearing housing in three orthogonal directions, one of which is i
the axial direction." The measurements will continue to be taken at the pump bearing. This is acceptable because the long term hydraulic and vibrational test results indicate that all four pumps are performing acceptably and have displayed very little, if any, degradation.
ALTERNATE TESTING: Pump vibration overall velocity, overall amplitude and running speed amplitude measurements will be taken. The acceptance criteria for the readings will be as follows:
Overall Velocity i
Acceptable Range:
s0.325 in/sec 1
Alert Range:
0.326 in/see to 0.700 in/sec Required Action Range:
>0.700 in/sec Overall Amplitude The ranges given in Table IWP-3100-2 will be applied.
l Runnine Speed Amplitude The ranges given in Table IWP-3100-2 will be applied.
l
.o 4 Enclosure to NRC-95-0077 Page1i RELIEF REQUEST NO. PR-11 PUM P(S):
Emergency Equipment Cooling Water, EECW PUMP NO.s:
P4400C001 A and P4400C001B CLASS:
3 i
FUNCTION: Proside cooling water to ECCS components and other essential equipment in the Reactor Building in the event that RBCCW is lost.
i SECTION XI REOUIREMENTS: Articles IWP-1500 and IWP-3000 require that a set of fixed reference values be established for each pump in order to detect deviations in pump performance.
BASIS FOR RELIEF: Reliefis requested to desiate from the Code requirement for a fixed reference value for flow and differential pressure. For flow rate and differential pressure, a flow reference curve, rather than a single i
fixed value of differential pressure and corresponding flow, will be utilized.
This reference cmve will be developed utilizing linear regression with four or more flow versus differential pressure data sets over a limited range of Aw.
The use of pump reference curves is necessitated by the fact that both Ei CW Systems are tested by placing the EECW systems in operation and then restoring cooling water flow to the isolated (non-essential) loads (sump heat exchangers, containment penetration coolers, etc ). The test loop then consists of the Division 1 or Division 2 EECW Distribution with the non-essential loads restored for EECW Pumps P4400C001 A and B, respectively. The system flow is then adjusted by throttling the manually operated pump discharge gate valve to obtain the desired flow rate. The Division 1 Discharge Valve, P4400F004 A, and the Division 2 Discharge Valve, P4400F0048, are normally Locked Open (LO) valves to allow maximum flow in the EECW mode. The ability to control flow with a manual gate valve is very limited, and it is difficult to always return to a precise flow. To require that Plant Operators test the system at such a fixed flow rate represents an undue hardship on Operators and an unnecessary challenge to the EECW system, since it requires a greater amount of time be spent in this abnormal lineup.
d Enclosure to NkC-95-0077 i
Page 12 A review ofinsenice test results obtained using the reference cun es shows that the data is consistent and trendable. Additionally, the individual pump vibration data is extremely stable and indicates that pump performance has not degraded for either of the EECW Pumps. Had invalid data been used to generate the pump reference curves, or if the curve fit was poor, test results would be erratic, and this has not been the experience for these pumps.
ALTERNATE TESTING: The following acceptance criteria will be established for the Emergency Equipment Cooling Water Pumps:
For each EECW Pump, differential pressure and flow is being measured over a narrow range of test flows. A reference curve for pump performance monitoring for each pump has been developed over a limited flow range. This reference curve is currently being used to evaluate pump performance for degradation.
I 1
i i
l v
i
.. I