ML20128L674

Safety Evaluation Accepting Third ten-year Interval for Pump & Valve Inservice Testing Program for Facility

ML20128L674
Person / Time
Site:	Three Mile Island
Issue date:	10/11/1996
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML20128L671	List: ML20128L666 ML20128L674
References
NUDOCS 9610150160
Download: ML20128L674 (25)

Text

.___.

f* "*%

k UNITED STATES p

j NUCLEAR REGULATORY COMMISSION r

2 WASHINGTON, D.C. 20066-0001

,o SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATED TO THE THIRD 10-YEAR INSERVICE TESTING PROGRAM GPU NUCLEAR CORPORATION THREE MILE ISLAND NUCLEAR STATION. UNIT 1 DOCKET NO. 50-289 l

1.0 INTRODUCTION

The Code of federal Regulations, 10 CFR 50.55a, requires that inservice testing (IST) of certain American Society of Mechanical Engineers (ASME) Code Class 1, 2, and 3 pumps and valves be performed in accordance with Section XI of the ASME Boiler and Pressure Vessel Code (the Code) and applicable addenda, except where alternatives have been authorized or relief has been requested by the licensee and granted by the Commission pursuant to Sections (a)(3)(i),

(a)(3)(ii), or (f)(6)(1) of 10 CFR 50.55a.

In proposing alternatives or requesting relief, the licensee must demonstrate that:

(1) the proposed

.iternatives provide an acceptable level of quality and safety; (2) compliance would result in hardship or unusual difficulty without a compensating increase in the !avel of quality and safety; or (3) conformance is impractical for its facility. Section 50.55a authorizes the Commission to approve alternatives and to grani, relief from ASME Code requirements upon making the necessary findings. NRt. guidance contained in Generic Letter (GL) 89-04, " Guidance on Developing Acceptable Inservice Testing Programs," provides alternatives to the Code requirements determined acceptable to the staff. Alternatives that conform with the guidance in GL 89-04 may be implemented without additional NRC approval, but are subject to review during inspections.

Further guidance was given in Generic Letter 89-04, Supplement 1, and NUREG-1482, " Guidelines for Inservice Testing at Nuclear Power Plants."

The applicable Code used in developing the Three Mile Island Station (TMI),

Unit I third 10-year interval IST program was the 1989 edition of ASME Section XI, Subsections IWP and IWV, which reference the ASME/ ANSI Operations and Maintenance (OH) Standard (0Ma-1988), Part 6 and Part 10, respectively.

The TMI, Unit I third 10-year interval program was submitted in a letter dated September 21, 1995. Mechanical Engineering Branch staff had conference calls with the licensee on February 12, May 6, and August 22, 1996, discussing the licensee's proposed relief requests. The licensee submitted a revision to their third 10-year IST submittal dated May 23, 1996, which made the following changes to the licensee's 10-year program submittal:

revised Relief Requests P1, P2, P3, and PG3; withdrew Relief Requests PGI, PG2, VI, V2, V3, V4, and V5; and added Relief Request P4.

Evaluations of Relief Requests PI-P4, PG3, and VG1 are included in this Safety Evaluation (SE).

l 9610150160 961011 PDR ADOCK 05000289 p

PDR

i The NRC staff's findings with respect to the proposed alternatives and i

requested reliefs as part of the licensee's IST program are contained in this SE.

In addition, the staff has reviewed components in the reactor building spray system and technical positions taken by the licensee in their IST program to assess their compliance with the ASME Code requirements. The licensee should address the anomalies identified in Section 7 within 1 year of the date of this SE or the next refueling outage, whichever is longer, unless otherwise noted.

Because the recent IST audit at THI-1 (June 3, 1996 through June 7,1996) included a review of the deferred te:;t justifications for components which cannot be tested quarterly or during cold shutdowns, the staff has not included an evaluation of the deferred test justifications in this SE.

2.0 IST PROGRAM ISSUES 2.1 Scope Reauirements of Certain Cateaory C Check Valves The licensee originally submitted two separate relief requests to exclude certain Category C check valves which the licensee determined to have a closed safety function to prevent diversion of flow from their IST program. Upstream of the valves are closed manual valves. One manual valve is an ASME code class valve, one is outside the code class boundary. The licensee stated that all four valves are Seismic Category I.

These relief requests were withdrawn by the licensee after the staff inquired in a phone conversation as to the active safety function of each check valve.

If the manual valves are closed and the system function is not safety-related, then the manual valves should be verified closed by using plant procedures.

Check valves that do not have an active safety function to open or close need not be included in the IST program.

2.2 Extendina Check Valve Inspection Freauency Based on " Extreme Hardshio" The licensee originally submitted three separate relief requests to extend the sampling inspection interval of certain check valves inspected in accordance with GL 89-04, Position 2.

In a phone conversation with the licensee, the staff inquired if the valves in the three relief requests meet the guidance for extending the valve frequency due to extreme hardship as described in Position 2.

The licensee's May 23, 1996, submittal withdrew the three relief requests and stated that these valves did meet the guidance for extending the valve test frequency. No specifics were given in the submittal, however, the licensee stated that the relevant information will be documented in their IST program.

The licensee should be particularly aware of what constitutes extreme hardship. Two plant evolutions which are considered examples of extreme hardship, given in NUREG-1482, Page A-13 and A-14, are off-loading of the reactor core and operating at mid-level of the reactor coolant loops.

If a valve or group of valves are accessible for inspection, with minimal radiation exposure or generated waste, then this valve or group of valves would not be considered to meet the definition of extreme hardship and therefore one valve in the group should be disassembled and inspected on a sampling basis each

refueling outage as specified in Position 2.

The licensee's determinations on specific check valve inspection frequencies may be reviewed in a future NRC inspection.

2.3 Endina Date of Third 10-Year IST Interval The third 10-year interval was scheduled to start on September 22, 1994, however, the NRC granted an extension of the interval start date to September 22, 1995.

Extension of the 2nd 10-year interval is not intended to change the completion date of the third 10-year interval which remains September 22, 2004.

Therefore, the end date of September 22, 2005, stated in the licensee's submittal is incorrect and should be changed.

3.0 GENERAL RELIEF REQUEST 3.1 Relief Reauest Number PG3 (revised)

The licensee has requested relief from the Code acceptance criteria requirements of OMa-1988, Part 6 (OM-6), Paragraph 6.1, for all pumps in the licensee's IST program.

The licensee has proposed to use the action range requirements of ASME OM Code ISTB-1995, Paragraph 6.2.2.

In addition, the licensee has proposed to allow the analysis provision in ISTB 6.2.2 to apply to components in the alert range.

3.1.1 Licensee's Basis for Reauestina Relief The licensee states:

An analysis of the pump condition can demonstrate that the pump can furnish i;s design function especially for those pumps with i

large margins above their design requirement.

Doubling test frequency for pumps is only intended to get more data.

For pumps that are normally standby, the degrading mechanism should not be active when the pump is off. Doubling test frequency may not establish any additional information.

Based on a successful operating history with no step changes or abnormal trends, data which may occasionally fall in the alert range may not be unacceptable. Modifying or replacing the pumps to reduce vibrations only to assure that they do not occasionally fall in the alert range could be unnecessary.

Further, doubling the test frequency would result in running the pumps more often at low flow conditions and would provide no useful information.

The deviation of pump hydraulic and/or vibration data which falls in the required action range of Table 3, " Ranges for Test Parameters," (consisting of Figure 1 and Tables 3a and 3b of the errata to OMa-1988 contained in OMb-1989) is an indication of degradation of pump performance. However, such a deviation does not address the ability of the pump to perform its intended safety function or the rate of degradation.

It may be possible through analysis of past performance data from the subject pump or other

\\

i

.- I j

similar pumps to ensure that the pump remains capable of performing its intended safety function until the next scheduled surveillance test.

Such an analysis could prevent the unnecessary shutdown of the unit to perform repairs to the pump.

l These provisions were permitted by earlier editions of ASME Section XI, Subsection IWP, and are currently permitted by the i

ASME ON Code ISTB-1995, Paragraphs 6.2.2 and 4.6.

f 3.1.2 Alternate Testina The licensee proposes:

In lieu of the requirements in Paragraph 6.1, " Acceptance Criteria" in OM-6 for pumps whose hydraulic and/or vibration data falls into the required action range of Table 3, " Ranges for Test Parameters," the requirements of ISTB 6.2.2, " Action Range," of the ASME ON Code will be implemented for the inservice testing of safety related pumps.

The related requirements of ISTB 4.6, "New Reference Values," will also be implemented for the inservice i

testing of safety related pumps.

If measured pump parameters fall within the alert range the test frequency will be doubled until the cause of the deviation is found and corrected or an analysis of the pump is performed and new reference values established.

If measured test parameters fall within the required action range, the pump shall be declared inoperable until either the cause of the deviation is determined and the condition corrected or an analysis of the pump is performed and new reference values established.

i The analysis will include a comparison of the test results with the required design parameters, an evaluation of previous data to establish a trend, an investigation into the reason for the parameter change, and if necessary the collection of additional data. To be successful, the evaluation must conclude that the condition does not impair the ability of the pump to perform its safety function. The evaluation will be maintained in the test records.

3.1.3 Evaluation The corrective action requirements of ASME Section XI, Paragraph IWP-3230(c),

allowed licensees to perform an analysis to demonstrate that the current mechanical or hydraulic performance levels of the pump do not impair the pump operability and that the pump would still perform its safety function.

Further, this section also allowed the licensee to establish new reference values after the analysis was performed. Changes to ASME/ ANSI OMa-1988, Part 6, " Inservice Testing of Pumps in Light-Water Reactor Power Plants," were made to address the concern that repeated establishment of new reference values would allow the pump to operate in a significantly degraded condition from the original pump reference values while meeting the design basis flow

__-_.-m j.*

j,

and pressure requirements of the system.

In addition, there were concerns as i

to the ability to perform an " analysis" of the pump to demonstrate that the pump was operating in an acceptable range. This issue is further discussed in l

NUREG-1482, Section 5.6.

i OM-6, Paragraph 6.1, " Acceptance Criteria," specifies actions required to be taken if any of the measured pump parameters fall within the alert or required action ranges.

For the alert range, the test frequency is required to be doubled until the cause of the deviation is determined and the condition is corrected. An entry into the required action range specifies that the licensee must declare the pump inoperable until the cause of the deviation is determined and the condition corrected.

The 1995 Edition of the ASME OM Code, i

Subsection ISTB 6.2.2, " Action Range," which is not currently endorsed by the l

NRC, allows that "[i]f the measured test parameter values fall within the required action range of ISTB 5.2.1.1, Table ISTB 5.2.1-2, Table ISTB 5.2.2-1, a

or Table ISTB 5.2.3-1, as applicable, the pump shall be declared inoperable until either the cause of the deviation has been determined and the condition is corrected, or an analysis of the pump is performed and new reference values are established in accordance with para. ISTB 4.6."

This paragraph clarifies that, if a pump can be shown to be capable of performing its safety function, i

it may be returned to service with adjusted reference values. This reflects that there are pumps that have a significant margin over the safety i

l requirements that might degrade from their initial performance, but still be capable of meeting their safety function.

Pumps which do not have margin would not be returned to service without repairs or replacement.

The analysis must justify that the degradation mechanism will not cause further degradation l

such that, before the next pump test or before repairs can be performed, the j

pump would fail. The licensee has also stated in their proposed alternative j

that they would meet the related requirements of ISTB 4.6.

As such, the l

alternative will provide an acceptable level of quality and safety for monitoring the pumps. This analysis should be used cautiously as it is not intended to be used regularly to evaluate the operability of all pumps that i

fall into the required action range in order to declare the pump operable and define new reference values where significant degradation is occurring. The 1

use of this analysis is expected to be a rare occurrence.

Repeated application of analysis could lead to " stair stepping" the Code alert and required action range limits downward to the safety limit of the pump. The licensee should have an understanding of the margin of each safety-related 1

pump above its design-basis hydraulic requirements. The analysis, which j

should include detailed justification and any change in the pump reference i

values, must be documented in accordance with the Code requirements and the regulations.

In addition, it is not foreseen that this section in the Code i

could be used for pump bearing vibration readings, as there are no defined i

safety margins related to pump vibration.

1 I

The licensee also requested to apply ISTB 6.2.2 to the portion of OH-6, Paragraph 6.1, that applies to pumps in the required action range. Subsection 6.2.1, " Alert Range," included in the OM Code ISTB-1995, does not allow the licensee to perform an analysis of the test data to exit the alert range.

1 Therefore, approval of this portion of the proposed alternative would be j

contrary to what has been approved through the ASME consensus process in i

ISTB-1995 of the ASME OM Code. The licensee's arguments for performing an i

4 4

analysis when pumps enter the alert range include:

1) doubling the test frequency in the alert range may not establish any additional information for standby pumps; 2) modifying or replacing the pumps to reduce vibrations only to assure that they do not occasionally fall-in the alert range could be unnecessary; and 3) doubling the test frequency would result in running the pumps more often at low flow conditions. These arguments do not demonstrate a unique burden or hardship on the licensee and do not provide an acceptable level of quality and safety. Therefore, this portion of the relief request is denied.

3.1.4 Conclusion The alternative to use the 1995 edition of the ASME Code, Section ISTB 6.2.2, for pumps in the required action range is authorized pursuant to 10 CFR 50.55a(a)(3)(i) based on the acceptable level of quality and safety that will be provided by the alternative.

The alternative to use the 1995 edition of the ASME Code, Section ISTB 6.2.2, for pumps in the alert range is denied.

4.0 PUNP RELIEF REQUESTS 4.1 Relief Reauest Number P1 (revised)

The licensee has requested relief from the Code pump run time and test procedure requirements of OH-6, Paragraphs 5.6 and 5.2, respectively, for the makeup and purification pumps MU-PIA, MU-PIB, and MU-PIC.

The licensee has proposed to use the pump curves in tb final safety analysis report (FSAR) for each individual pump along with the guidance on pump reference curves in NUREG-1482, Section 5.2, to perform hydraulic pump testing during refueling outages.

In addition, the pump run time for some or all pumps may be below the Code minimum requirement of two minutes.

4.1.1 Licensee's Basis For Reauestina Relief The licensee states:

The amount of time that the Makeup Pump injects at full flow to the Reactor Coolant System (RCS) must be limited.

Pumping to the RCS will raise the level in the pressurizer and a plant transient can occur. Run time, therefore, must be limited.

Pumping time is limited to a total of approximately five minutes for all flow configurations.

Because of the short time available for a test run, throttling to a specific reference point can not be accomplished. The pump is run with several different valve lineups to verify that flow rate and head are equal to or higher than accident design requirement.s.

Flow rate and pressure measurements for each lineup are compared to previous test data.

Acceptance is based on meeting or exceeding accident flow and head

,' requirements. This meets the intent of the code. The test is simihr to that described in NUREG-1482, Paragraph 5.2 except for the following:

l 1)

A manufacturer's curve is not used. Comparison is with the FSAR Safety Analysis curve and previous full flow tests, 2)

A five point curve is not used. The pump will operate at several different points.

These tests demonstrate pump operability and meet the intent of the code.

4.1.2 Alternate Testina The licensee proposes:

As permitted by NRC GL 89-04, Position No. 9, the pumps will be full flow tested at a refueling outage frequency (see justification). The refueling outage test will include measurement of stable flow rate and DP [ differential pressure].

Pump testing will be performed with the system lined up to pump to the RCS through different flow path combinations to provide pump data at various flow rates.

Run time through each flow configuration may be less than the two minutes required by OM-6.

4.1.3 Evaluation The Code requires that, for variable resistance systems, a fixed reference value (either flow rate or DP) be established and maintained for 2 minutes after the hydraulic conditions have stabilized to anticipated operating conditions. Only after this 2-minute period can valid pump flow, DP, and bearing vibration measurements be recorded for purposes of inservice testing.

The licensee states that the full-flow run time of each pump during a refueling outage test is limited to approximately 5 minutes because of the RCS capacity. One of the primary functions of the makeup and purification system is to provide makeup flow to the reactor coolant system.

Flow control valves are included in the system to provide throttling capability. However, because of the limited pump run time, the licensee states that throttling to a fixed reference value cannot be performed within the limited time period. Meeting both Code requirements could potentially be a hardship for the licensee because setting the fixed reference value and waiting the required 2 minutes may exceed the approximate 5 minute test limit.

If this were to occur, the licensee would be required to rerun the test until all the data could be collected.

The licensee has proposed to test the pumps during refueling outages using the pump curve in Figure 6.1.2 of the TMI-1 FSAR. The licensee states that a number of points will be taken and compared to the FSAR curve. The guidance in NUREG-1482, Section 5.2, specifies that a reference curve should be either established or verified prior to performing the inservice test when the pump

is known to be in good operating condition. An acceptable method to determine pump degradation using pump reference curves is to establish pump operation within the boundaries of the curve and then record pump flow and DP. The flow point recorded during the test h then plotted on the pump characteristic curve (in this care, the FSAR curve).

If the measured DP is within 0.9 to 1.1 times the referen:e DP, which is taken from the characteristic curve of each makeup pump, then the hydraulic performance of each pump is acceptable. While the licensee states that they will conform with the guidance in Section 5.2 of NUREG-1482, their proposed alternative testing appears to revalidate their FSAR curve for all three pumps every refueling outage without establishing any acceptance criteria or monitoring for degradation.

Figure 6.1.2 in the licensee's FSAR is the characteristic curve for all three makeup pumps.

The licensee stated in their May 23, 1996, submittal that this curve has been validated. However, the accuracy of the curve may not be adequate for purposes for verifying the hydraulic acceptance criteria. The thickness of the pump curve line represents approximately 1.6% of the vertical axis scale or 125 feet of total developed head. A reading within this inherent band of uncertainty could mask the determination of pump hydraulic performance in a degraded condition.

The licensee states that testing will be performed under stable flow conditions. The time required to establish stable flow conditions varies for each pump depending on the pump design and conditions during testing. During a refueling outage test, stable operating conditions should be achieved quickly because of the lower operating pressures and temperatures in the RCS and the pump is being operated at substantial flow conditions.

It would be a hardship on the licensee if the Code test could not be completed because of the 2-minute run time requirement even though the pump was operating in a stable flow regime.

Therefore, the proposed alternative would provide a reasonable assurance of operational readiness provided that the Code-required measurements are taken after each pump achieves a stable operating condition.

The licensee stated that throttling to a specific reference point cannot be accomplished because of the limited time available to perform the inservice test at refueling outages.

It could potentially be a burden for the licensee to meet the Code pump test procedure requirements. However, NUREG-1482, Section 5.2, states that the licensee must demonstrate the impracticality of achieving the reference conditions for IST. Suspension of the 2-minute run time requirement would provide additional time to establish the fixed reference value.

In addition, the guidance in NUREG-1482, Section 5.3, states that relief from the variance of the Code fixed reference value by *2% is not required.

Therefore, it may be possible for the licensee to establish a single reference value and record all the hydraulic and vibration data within the limited test duration.

The licensee should review their IST program and test procedures to determine the adequacy of their proposed hydraulic acceptance criteria and the accuracy of the pump reference curve.

In addition, the licensee should evaluate the justification of impracticality to establish a fixed reference value within the limited test run time and either revise or withdraw this portion of their relief request. An interim period should be allowed for the licensee to

. = _ -

perform these evaluations and determine the practicality of establishing a fixed reference value. The proposed testing provides reasonable assurance of operational readiness during the interim period because the pumps are tested at power quarterly ~ in accordance with the guidance of GL 89-04, Position 9.

4.1.4 Conclusion The proposed alternative to the Code pump run time requirements of OM-6, Paragraph 5.6, for the makeup and purification pumps MU-PIA, MU-PIB, and MU-Plc is authorized pursuant to 10 CFR 50.55a(a)(3)(ii) based on the l

determination that compliance with the specified requiremente results in a hardship without a compensating increase in the level of quality and safety.

The alternative is authorized with the provision that the Code-required measurements are taken after the pump achieves a stable operating condition.

Interim relief is granted for a period of 1 year or the next refueling outage, whichever is longer, from the Code test procedure requirements OM-6, Paragraph 5.2, pursuant to 10 CFR 50.55a(f)(6)(i) based on the impracticality of performing the test in accordance with the Code requirements, and in consideration of the burden on the licensee if the Code requirements were imposed on the facility. During the interim period, the licensee should review their IST program and test procedures to determine the adequacy of their proposed hydraulic acceptance criteria and the accuracy of the pump reference curve and evaluate the justification of impracticality to establish a fixed reference value within the limited test run time and either revise or withdraw this portion of their relief request.

4.2 Relief Reauest Number P2 (revised)

The licensee has requested relief from the Code pump run time and test procedure requirements of OM-6, Paragraphs 5.6 and 5.2, respectively, for the turbine-driven emergency feed water (EFW) pump EF-P1 and the two motor-driven EFW pumps EF-P2A and EF-P28. The licensee has proposed to use reference curves and record pump hydraulic and mechanical data immediately after conditions have stabilized.

4.2.1 Licensee's Basis for Reauestina Relief The licensee states:

The EFW pumps are only used for emergency use.

They are not used for startup, shutdown, or normal plant operation.

EFW flow to the once through steam generators (OTSGs) is limited by the cavitating venturies.

Since the pumps operate only for test, no significant degradation is expected.

l Since the refueling interval tests transfer lower quality water to the OTSGs, the number and duration of tests must be limited to minimize routine exposure of the OTSGs to lower quality water.

Minimizing test duration is necessary tc limit the amount of water injected into the OTSGs where corrosion damage promoted by 0, can occur. Throttling to a reference value of flow / differential

I

{:f pressure as well as waiting out a minimum of two minutes run time i

lengthens the amount of time the EFW pumps are running while pumping oxygenated water into the OTSGs.

1 l

Justification for (the licensee's] request is based on limiting j

the amount of oxygenated water that is pumped onto the OTSGs to minimize the potential for steam generator tube degradation. The feed water oxygen concentration limit for normal plant operation is 5 parts per billion (ppb); and TMI-1 normally maintains feed water oxygen concentration less than 1 ppb. At the lower temperatures during shutdown, hydrazine is unable to react and scavenge the oxygen. So to help lower the oxygen content, prior to the test a nitrogen blanket is applied to the condensate storage tanks through spargers.

[The licensee] can typically lower the oxygen concentrations of the water to approximately 200-300 ppb. This range is still many times higher than the limit recommended by the Electric Power Research Institute [ reference EPRI Final Report TR-102134, "PWR Secondary Water Chemistry Guidelines - Revision 3," dated May 1993].

[Thelicensee]

believes that performing a shorter test will not compromise [the]

ability to adequately demonstrate EFW pump operability and meets the intent of the Code.

The EFW pump quarterly tests verify that the pumps are operational, start on demand, and can generate the required discharge pressure. During the quarterly test, vibration data will be taken on each bearing while pumping through the recirculation line.

4.2.2 Alternate Testina The licensee proposes:

TMI-1 Tech Specs [ Technical Specifications] requires a test each refueling to demonstrate that the motor driven EFW pumps can pump water from the CSTs (Condensate Storage Tanks] to the steam generators.

Run time while pumping to the OTSG may be less than two minutes as required by OM-6.

During the next refueling outage (12R), GPU Nuclear will verify the pump curve to be the valid accident design curve in accordance with the guidance in NUREG-1482, Section 5.2, except that only thrse points on the curve will be taken. Thereafter, each refJeling outage, a full flow test of all three pumps will be performed in which accident design flow and differential pressure for iach pump will be verified by running the pumps only long enough to take stable flow, differential pressure, and vibration J

datt.; pressure and flow will be compared to the reference curve.

1 This test verifies acceptable flow rate and differential pressure

{

of the pumps.

I

l',

4.2.3 Evaluation j

The Code requires that, for variable resistance systems, a fixed reference value (either flow rate or DP) be established and maintained for 2 minutes after the hydraulic conditions have stabilized to anticipated operating conditions. Only after this 2-minute period can valid pump flow, DP, and i

bearing vibration measurements be recorded for purposes of inservice testing.

The licensee states that the full-flow run time of each pump during a refueling outage test must be minimized to limit the amount of lower quality water added to each OTSG. The primary functions of the EFW system are to provide a reliable source of feed water to the OTSGs during a loss of reactor l

coolant flow caused by a loss of offsite power, a loss of main feed water, a pipe rupture in the secondary system, a steam generator tube rupture, or a small loss of coolant accident. Flow control valves are included in the j

system to provide throttling capability to meet all these functions. However, because of the licensee's desire to minimize pump run time, the licensee i

states that throttling to a fixed reference value cannot be performed.

I Meeting both Code requirements could potentially be a hardship for the licensee because setting the fixed reference value and waiting the required 2 minutes may lead to an unnecessary addition of lower quality water to each OTSG.

It is understood that limiting the amount of oxygenated water into the OTSGs is important to their long-term condition. The licensee's justification for meeting the Code fixed reference value and run time requirements is that the J

oxygen concentration limit of each OTSG is 5 ppb under normal operating conditions and the source of water from the condensate storage tank is approximately 200-300 ppb. However, the proposed alternative testing is taking place during refueling outages. EPRI Final Report TR-102134, Revision 3, Table 3-1, recommends action be taken when the oxygen concentration exceeds 100 ppb with the OTSG in cold shutdown / wet layup at an RCS temperature s200*F. Although the water being added into the OTSG may be above the value cited in the EPRI report, the overall OTSG oxygen concentration may be well below the recommended action value due to the substantial volume of water in the OTSG.

In addition, according to TMI, Unit 1 FSAR Section 10.6.2, the motor-driven pumps come up to full flow in 10 seconds and the turbine pump comes to full flow in 40 seconds. The process of starting each pump, achieving a fixed reference value, and taking the required data should add only a few thousand gallons of water into both OTSGs. The licensee has not addressed how this testing will effect the overall concentration of the OTSGs as compared with the EPRI report guidelines.

The licensee states that data collection will be performed under stable conditions. The time required to establish stable flow conditions varies for each pump depending on the pump design and conditions during testing. During a refueling outage test, stable operating conditions should be achieved quickly because of the lower operating pressures and temperatures in the OTSG and the pump is being operated at substantial flow rates.

It would be a hardship on the licensee if the Code test could.not be completed because of the 2-minute run time requirement even though the pump was operating in a stable flow regime. Therefore, the proposed alternative would provide a

reasonable assurance of operational readiness provided that the Code-required measurements are taken after each pump chieves a stable operating condition.

The licensee stated that throttling to a specific reference point cannot be accomplished because of the desire to limit the addition of lower quality water into each OTSG.

It could potentially be a burden for the licensee to meet the Code pump test procedure requirements. However, NUREG-1482, Section 5.2, states that the licensee must demonstrate the impracticality'of achieving the reference conditions for IST. Suspension of the 2-minute run i

time requirement would provide additional time to establish the fixed reference value.

In addition, the guidance in NUREG-1482, Section 5.3, states that relief from the variance of the Code fixed reference value by 12% is not required. Therefore, it may be possible for the licensee to establish a single reference value and record all the hydraulic and vibration data while minimizing the addition of lower quality water into each OTSG.

The licensee should review their IST program.and test procedures to determine the ability to establish fixed reference values for the motor and turbine-driven EFW pumps while minimizing the addition of lower quality water into each OTSG.

In addition, the licensee should evaluate the justification of impracticality to establish a fixed reference value and either revise or withdraw this portion of their relief request. An interim period should be allowed for the licensee to perform these evaluations and determine the practicality of establishing a fixed reference value. The proposed testing provides reasonable assurance of operational readiness during the interim period because the pumps are tested at power quarterly in accordance with the guidance of GL 89-04, Position 9.

4.2.4 Conclusion The proposed alternative to the Code pump run time requirements of OM-6, Paragraph 5.6, for the turbine-driven EFW pump EF-P) and the two motor-driven EFW pumps EF-P2A and EF-P2B is authorized pursuant to 10 CFR 50.55a(a)(3)(ii) based on the determination that compliance with the specified requirements results in a hardship without a compensating increase in the level of quality and safety. The alternative is authorized with the prevision that the Code required measurements are taken after the pump achieves a stable operating condition.

Interim relief is granted for a period of 1 year or the next refueling outage, whichever is longer, from the Code test procedure requirements OM-6, Paragraph 5.2, pursuant to 10 CFR 50.55a(f)(6)(1) based on the impracticality of performing the test in accordance with the Code requirements, and in 1

consideration of the burden on the licensee if the Code requirements were imposed on the facility. During the interim period, the licensee should review their IST program and test procedures to determine the adequacy of their proposed alternate testing and evaluate the justification of impracticality to establish a fixed reference value and either revise or withdraw this portion of their relief request.

• ]

4.3 Relief Reauest Number P3 (revised) l The licensee has requested relief from the Code test procedure requirements of OM-6, Paragraph 5.2, for the nuclear service river water (NR) pumps NR-PIA, NR-PIB, and NR-PIC. The licensee has proposed to measure flow rate for each l

pump individually every refueling outage.

4.3.1 Licensee's Basis for Reauestina Relief The licensee states:

The test flow instrumentation for this system is located in the common discharge from all three pumps.

The piping configuration i

does not facilitate installation of individual pump flow measuring devices.

GPU Nuclear his not been successful in attaining acceptable accuracy or regatability using individual annubar flow instruments.

To read total NR pump flow (NR-FI-290), for any pump combination

)

including two pump operation, it is necessary to isolate makeup water to the circulating water flume by closing the 30" butterfly valves (NR-V4A and NR-V48). This directs all NR pump flow to the nuclear service heat exchangers and reduces the temperature of nuclear service closed cooling water which cools the reactor coolant pump (RCP) seal return coolers. This results in a decrease in makeup system supply water temperature including the supply for RCP seal water. The resulting fluctuations in RCP seal leakoff flow rate reduces the performance of the pump seals and adds to the risk of eventual RCP seal damage.

During normal plant operation, at least two of the three pumps are in operation. Operation of only one Nuclear Service River Water Pump could jeopardize plant equipment due to heat loads.

If all but one NR pump were secured for the purpose of testing, this would cause significant temperature variations in safety related components, add significant operator burden to assure adequate cooling of the many plant components that would be affected, and result in some operational risk.

Individual flow rate measurement is impractical during plant operation or during Cold Shutdowns of short duration. The quarterly test to measure differential pressure and vibration as well as the refueling test to verify head and flow rate greater than the accident design will continue to assure operability of the NR pumps.

4.3.2 Alternate Testina The licensee proposes:

Flow rate for individual pumps will be measured at refueling outages.

i-14 _

i 4.3.3 Evaluation j

The Code requires that an inservice test shall be run on each pump quarterly.

During this test, either the pump flow or differential pressure shall be varied until the fixed reference value is reached. After the pump hydraulic and vibration values stabilize and the required 2-minute run time is 1

completed, the variable hydraulic reference value is then recorded and compared with the acceptance criteria to determine pump degradation.

4 The three nuclear. service river water pumps are vertical centrifugal pumps.

The licensee states that individual pump flow measurement is not possible for any of the three pumps during normal plant operation. TMI, Unit 1 Drawing Number 302-202 indicates individual flow elements Fell 52A-C are installed on j

each pump.

In a phone conversation on August 22, 1996, the licensee explained j

that there were significant efforts made to use these annubar flow elements to i

measure the individual NR pump flow.

Because of the discharge piping geometry of each NR pump, the accuracies of the instruments proved to be neither within the Code accuracy requirements nor repeatable. The licensee stated there is no location where ultrasonic flow meters can be utilized.

In addition, pump flow using any combination of NR pumps cannot be taken using flow indicator FI-290 because cooling flow would have to be isolated to the reactor coolant pump seals.

Redesigning the NR system to accommodate individual pump testing would require the installation of new piping which would be a hardship on the licensee without a compensating increase in the level of quality and safety.

The licensee has proposed to measure the flow rates of each pump individual every refueling outage.

During power operation, pump vibration and differential pressure would be taken on a quarterly frequency.

This is similar to the guidance provided in GL 89-04, Position 9, which allows the i

measurement of pump flow rate to be deferred to refueling outages for pumps which can be tested only on recirculation flow quarterly but can be tested at substantial flow during refueling outages. The licensee stated in the August 22, 1996, phone conversation that the quarterly pump test for all three pumps would be performed at a substantial flow rate. The licensee's proposal provides reasonable assurance of operational readiness because the pumps would be operated at substantial flow quarterly, vibration and differential pressure measurements would continued to be collected quarterly, and the Code hydraulic testing would be performed on all three pumps, albeit on a refueling outage 1

frequency.

4.3.4 Conclusion 1

)

The proposed alternative to the Code test procedure requirements of OH-6, Paragraph 5.2, for the NR pumps NR-PIA, NR-PIB, and NR-Plc is authorized pursuant to 10 CFR 50.55a(a)(3)(ii) based on the determination that compliance with the specified requirements results in a hardship without a compensating increase in the level of quality and safety.

4.4 Relief Reauest Number P4 The licensee has requested relief from the Code absolute alert vibration value of OH-6, Table 3a, referenced in the acceptance criteria requirements of Para-

graph 6.1, for decay heat pump DH-PIB. The licensee has proposed to establish a new alert value for the vertical direction of 0.400 inches /second (ips).

4.4.1 Licensee's Basis for Reauestina Relief The licensee states:

OH-6 requires doubling test frequency whenever the overall vibration amplitude reaches 0.325 ips during quarterly testing.

The code assumes that the equipment has degraded to the point where more frequent monitoring and possibly a repair are warranted. There is no consideration for test conditions, vibration history, or equipment maintenance history. These pumps are tested each refueling at both medium and high flow rates where the vibration levels have always been lower with the majority of vibration occurring at the vane pass frequency.

Consideration of vibration amplitudes was not part of the original acceptance criteria for pumps procured for many of the earlier nuclear plants. As a result, some of the pumps purchased in the late 60's and early 70's had inherently high vibrations. During low flow conditions, typical of IST testing, vibration amplitudes are at their highest.

TMI plant data, shop testing by GPU Nuclear, and conversations with several pump vendors indicate that it is not unusual to experience vibrations in excess of 0.325 ips with these older pumps, especially at low flow conditions.

Provided there is a successful long term operating history and provided there is no significant change in vibration amplitude or spectra, there is no reason to suspect equipment degradation at these vibration levels. TMI's Decay Heat Removal (DHR) pumps are one specific example and show the type of evaluations that we perform for pumps that exceed the alert limit.

TMI's DHR pumps are early edition American Petroleum Institute (API) 610 process pumps.

They have operated with occasionally high but untrending vibration since 1974. This includes extensive operating time between 1979 and 1985 (TMI's extendad shutdown).

The pumps have not failed, there is no unusual degradation in hydraulic performance, and seal and bearing life are normal.

Vibration amplitudes average 0.293 ips (standard deviation of 0.1) with the highest vibration occurring at the lower flow IST conditions.

Because of normal variation in vibration response and measurement, measured vibration exceeds 0.325 ips about once per year during low flow IST operation. However, there is no upward trend in the data and vibration has always been at vane pass frequency.

GPU has discussed these relatively high vibration readings with several vendors who manufactured API pumps. The vendors stated that high vibrations are expected with early edition API 610 pumps, particularly at the low flow rates encountered during inservice testing. Additionally, GPU has shop tested TMI's spare

. DHR pump and found that its vibration readings were almost identical to TMI's two inservice pumps. During the shop test, vibration data were recorded at many different flow rates. At flow rates equal to and below the IST flow rate, vibrations occasionally exceeded 0.325 ips. This pump was inspected prior to and after the shop test to assure no degradation had occurred.

The spare pump is identical to the inservice pumps and has never been used.

Figures 1 and 2 [ attached to submittal) show vibration data from tests of the DH pumps (DH-PIA and DH-PIB, respectively) since 1987.

Points shown represent the highest vertical overall amplitude since amplitudes in the vertical direction are higher than the horizontal vibration amplitudes. The vibration spectrum is essentially all at vane pass frequency with no IX or 2X pump i

harmonics that could indicate pump problems. These data present a i

clear case for raising the allowable vibration levels for DH-PIB.

Several attempts have been made to rcduce the vibration levels for the decay heat pumps, including pump motor alignment, strengthing the backfoot, and adding lead weights to the back end of the pump.

i Based on the successful operating history of DH-PIB, no step changes or trends in vibration data as shown in Figure 1, extensive vibration analysis, shop testing, and vendor input, GPU Nuclear does not consider the vibration amplitudes of THI's operating decay heat pumps unacceptable. Replacing or modifying the pumps to reduce vibrations only to assure they do not occasionally exceed 0.325 ips is unnecessary.

Further, doubling the test frequency would result in running the pumps more often at low flow /high vibration conditions and would provide no useful information. Therefore, this request to allow raising the alert range to greater than 0.400 ips (vs greater than 0.325 ips) is justified for DH-PIB.

4.4.2 Alternate Testina The licensee proposes:

The alert range for DH-PIB will be raised to vibrations greater than 0.400 ips (vs. greater than 0.325 ips) in the vertical direction.

The alert range for the horizontal direction will remain at vibration levels greater than 0.325 ips.

4.4.3 Evaluation l

OM-6 requires that, if a pump vibration velocity level exceeds the alert range l

of 0.325 ips, the testing frequency shall be doubled until the problem with the pump is determined and corrective action taken.

Previous Code editions calculated the vibration alert and required action range values as a function of a vibration reference value determined by the licensee through its IST program. OM-6 also uses vibration reference values to determine vibration limits. However, this edition of the Code also instituted absolute alert and

required action range vibration limits for centrifugal pumps, regardless of the magnitude of the vibration reference value.

The 1icensee has proposed to raise the alert range vertical vibration limit for the DH-PIB pump bearing to 0.400 ips. A review of Figure 2 included in the May 23, 1996, submittal shows that the DH-PIB pump vibration level averages 0.311 ips from data taken in late 1988 to early 1996. Three spikes in the da ta above 0.400 ips are shown at 1 year intervals between 1993 and 1995. Although the licensee has not specifically addressed these spikes, it is noted that the subsequent test returns to previous vibration levels. The i

data show no discernable trend upward.

The licensee stated that they have consulted with several pump manufacturers and stated that the level of vibration that these pumps were experiencing was consistent with pumps of similar age and design.

It was also noted that i

vibrations will be higher at low flow rates where IST testing is performed at power. The licensee also stated that they have attempted to lower the pump vibration by performing a pump motor alignment, strengthing the backfoot, and adding lead weights to the back end of the pump. A spectral analysis of the pump vibration modes shows vane pass frequency mode to be the dominant contributor with no IX or 2X pump harmonics that could indicate pump problems.

In addition to the above analysis and evaluations, the licensee has shop tested their spare decay heat pump, which is identical to the two installed decey heat pumps, at varying flow rates and' recorded similar vibration values.

Disassembly and inspection of the shop pump after bench testing did not reveal any degradation in the pump.

The licensee's evaluation of the DH-PIB pump vibration, coupled with the historical pump vibration data and shop testing of their spare decay heat pump, show that the pump normally runs at high levels of vibration and has not experienced any failures to date. Requiring the licensee to meet the Code requirements by increasing the frequency of the~DH-PIB pump testing would be a hardship without a compensating increase in safety because of the additional testing that would have to be performed on a pump that typically operates at elevated vibration levels. The proposed testing provides reasonable assurance of operational readiness because the pump will continue to be tested quarterly and the licensee will maintain the Code alert ranges for the axial and vertical components of vibration.

In addition, the required action range will 1

be adhered to as stated in the Code.

4.4.4 Conclusion The proposed alternative to the Code acceptance criteria requirements regarding raising the vibration level for the vertical component of decay heat pump DH-PIB from 0.325 ips to 0.400 1ps is authorized pursuant to 10 CFR 50.55a(a)(3)(ii) based on the determination that compliance with the specified requirements results in a hardship withcut a compensating increase in the level of quality and safety.

l 1

1

l 5.0 VALVE RELIEF RE0 VEST 5.1 Relief Reauest Number VG1 1

The licensee has requested relief from the Code leak rate testing frequency requirements of non-containment isolation valves and the frequency requirements for position indication testing as specified in OMa-1988, Part 10 (0M-10), Paragraphs 4.2.2.3(a) and 4.1 respectively, for the valves listed below. The licensee has proposed to perform the Code testing each refueling outage.

AH-VIA-D HM-V3A/B MU-V36 RC-V42 CA-V4A/B HM-V4A/B MU-V37 RC-V43 CA-V5A/B HR-V22A/B MU-V51 RC-V44 CA-V189 HR-V23A/B NR-VIA/B/C RR-VIA/B CF-VIA/B IC-V2 NR-V2 RR-V3A/B/C CF-V2A/B IC-V3 NR-V4A/B RR-V4A/B/C/D CF-Vl9A/B IC-V4 NR-V6 RR-V5 CF-V20A/B IC-V6 NS-V4 RR-V10A/B CM-V1 MS-VIA-D NS-V15 WDG-V3 CM-V2 MS-V2A/B NS-V35 WDG-V4 CM-V3 MS-V8A/B NS-V52A/B WDL-V49 CM-V4 MU-V3 NS-V53A/B WDL-D50 DH-V1 MU-V10 RB-V2A WDL-V61 DH-V2 MU-V12 RB-V7 WDL-V62 DH-V4A/B MU-V14A/B RC-RV2 WDL-V303 DH-VSA/B MU-V16A-D RC-V2 WDL-V304 DH-V6A/B MU-V18 RC-V4 WDL-V534 EF-V2A/B MU-V20 RC-V28 WDL-V535 HM-VIA/B MU-V25 RC-V40A/B MU-V2A/B HM-V2A/B MU-V26 RC-V41A/B 5.1.1 Licensee's Basis for Reauestino Relief The licensee states:

The refuel cycle for THI-1 is nominally two years.

Several of the valves requiring leak testing cannot be tested with the plant operating.

If, due to an intermediate outage, the refueling cycle exceeds two years, the code requirement could require a shutdown simply to test the certain valves. This is impractical. Testing each refueling is a reasonable alternative.

Typically, valve position verification is done more frequently than once every two years.

Some valves must be stroked to verify position. Of these, several cannot be stroked with the plant operating. As delineated above, the refuel cycle may extend beyond two years.

Position verification using the code specified frequency, could cause the plant to be shut down.

Position verification at least every refueling is a reasonable alternative to at least every two years.

Refueling interval testing has been extended to accommodate the 24-month refueling interval. This change was made by a license amendment for all safety related equipment required to be tested each refueling interval.

In some cases, the regulations have been changed to accommodate 24-month refueling intervals without the need for an exemption. However, for IST position verification tests, the ASME Code has not been changed and therefore written specific relief is required. The fact that this does not repre-sent a significant change is sufficient to justify this relief.

5.1.2 Alternate Testina The licensee proposes:

These tests will be performed each refueling outage instead of the Code specified frequency "at least once every two years."

5.1.3 Evaluation The Code requires that valves with remote position verification that are within the scope of a licensee's IST program shall be observed locally at least once every 2 years.

In addition, Category A valves which perform a function other than containment isolation shall be seat leakage tested at least once every 2 years.

The licensee states that the refueling cycle at TMI, Unit 1 is nominally 2 years. Adherence to the Code requirements may result in the plant shutting down prior to the completion of the cycle to perform the position verification and leak rate testing.

The licensee has proposed to perform the Code testing on these components once every refueling outage. The licensee's proposed alternative does not define the length of the refueling outage cycle. THI, Unit 1 Technical Specification (TS) Section 1.25, " Frequency Extension," specifies the frequency of surveillance intervals.

" Refueling Interval" is defined in TS Table 1.2 as once per 24 months (2 years). TS Section 1.25 also allows a 25% maximum frequency ex-tension for operational considerations that applies to the refueling interval frequency. The refueling outage interval in the proposed alternative appears to be equivalent to the refueling interval defined in the TS. However, a definitive frequency extension needs to be defined in order to account for operational considerations. Therefore, the proposed alternative provides an acceptable level of safety provided that the licensee defines their refueling outage interval as once every 24 months and adopts the 25%

interval extension specified in TS Section 1.25 for cycle lengths which exceed 24 months.

5.1.4 Conclusioji The alternative to the Code leak rate testing frequency requirements of non-containment isolation valves and the frequency requirements for position indication testing as specified in OMa-1988, Part 10 (0M-10), Paragraphs 4.2.2.3(a) and 4.1 respectively, for the valves listed above is authorized pursuant to 10 CFR 50.55a(a)(3)(1) based on the acceptable level of quality and safety that will be provided by the alternative. Therefore, the proposed

alternative provides an acceptable level of safety provided that the licensee defines their refueling outage interval as once every 24 months and adopts the 25% interval extension specified in TS Section 1.25 for cycle lengths which exceed 24 months.

6.0 IST PROGRAM SCOPE SYSTEM REVIEW An IST program scope review was performed on the TMI, Unit I reactor building spray system. System piping and instrumentation diagrams (P& ids) were reviewed and valves determined to have safety functions were compared to the licensee's IST program listing for the respective systems.

Individual valve testing and safety attributes proposed by the IST program were also reviewed for consistency with applicable codes and regulatory guidance. As a result of this review, the staff determined that certain components and component safety functions may have been omitted from the scope of the licensee's IST program.

The licensee should review the following components identified by the staff against the requirements of ISTC 1.1 and revise their IST program as necessary.

6.1 Reactor Buildina Soray Relief Valves The reactor building spray pump suction relief valves, BS-V45A&B, and the decay heat pump suction header relief valves, BS-63A&B, are not included in the THI, Unit 1 IST program. These valves have a safety function, as described in OM-10, Paragraph 1.1, to protect systems or portions of systems which perform a required function in shutting down a reactor to the cold shutdown condition, in maintaining the cold shutdown condition, or in mitigating the consequences of an accident.

These valves should be added to the licensee's !$T program.

6.2 Outboard Reactor Buildino Isolation Valves The outboard reactor building isolation valves, BS-1A&B, are included in the licensee's IST program, however, they are stroke-time tested in the open direction only. The valves may have a closed safety function and may be required to be closed post accident. The licensee should evaluate whether these valves should be stroke-time closed and modify their IST program as necessary.

7.0 ANOMALIES The following anomalies were noted during the course of the IST program review. The licensee should review these items and make changes to their IST program, testing procedures, or other plant documentation as necessary.

Items which require a response to the NRC should be completed within 1 year or the l

next refueling outage, whichever is longer, unless otherwise stated.

)

7.1 The licensee should change their IST program documentation to reflect the correct third 10-year interval end date of September 22, 2004 (see Section 2.3 of this SE).

I

7.2 The licensee's request to apply the 1995 Edition of the ASME OM Code, Subsection ISTB 6.2.2, " Action Range," to the portion of OH-6, Paragraph 6.1, that applies to pumps in the required action range was denied.

Subsection 6.2.1, " Alert Range," included in the OM Code ISTB-1995, does not include the provision to perform an analysis of the test data to i

exit the alert range. Therefore, approval of this portion of the proposed alternative would be contrary to action by the ASME Code Committee. The licensee's basis for requesting relief did not demonstrate a unique burden or hardship on the licensee and does not provide an acceptable level of quality and safety (see Section 3.1 of 2

this SE).

7.3 The proposed alternative to the Code pump run time requirements of OH-6, Paragraph 5.6, for the makeup and purification pumps MU-PIA, MU-PIB, and MU-Plc is authorized with the provision that the Code-required measurements are taken after the pump achieves a stable operating condition (see Section 4.1 of this SE).

7.4 Interim relief is granted for a period of 1 year or the next refueling outage, whichever is longer, from the Code test procedure requirements OH-6, Paragraph 5.2, for the makeup and purification pumps MU-PIA, MU-PIB, and MU-Plc.

During the interim period, the licensee should review their IST program and test procedures to (1) determine the adequacy of their proposed hydraulic acceptance criteria and the accuracy of the pump reference curve and (2) evaluate the justification of impracticality to establish a fixed reference value within the limited test run time, and then either revise or withdraw this portion of their relief request (see Section 4.1 of this SE).

7.5 The proposed alternative to the Code pump run time requirements of OM-6, Paragraph 5.6, for the turbine-driven EFW pump ZF-P1 and the two motor-driven EFW pumps EF-P2A and EF-P2B is authorized with the provision that i

the Code-required measurements are taken after the pump achieves a stable operating condition (see Section 4.2 of this SE).

7.6 Interim relief is granted for a period of 1 year or the next refueling j

outage, whichever is longer, from the Code test procedure requirements OM-6, Paragraph 5.2, for the turbine-driven EFW pump EF-P1 and the two motor-driven EFW pumps EF-P2A and EF-P28. During the interim period, the licensee should review their IST program and test procedures to i

(1) determine the adequacy of their proposed alternate testing and (2) evaluate the justification of impracticality to establish a fixed reference value, and then either revise or withdraw this portion of their relief request (see Section 4.2 of this SE).

7.7 The alternative to the Code leak rate testing frequency requirements of non-containment isolation valves and the frequency requirements for position indication testing as specified in OMa-1988, Part 10 (OM-10),

Paragraphs 4.2.2.3(a) and 4.1, respectively, for the valves listed in Section 5.1 of this SE is authorized with the provision that the

licensee defines their refueling outage interval as once every 24 months and adopt the 25% interval extension specified in TS Section 1.25 for cycle lengths which exceed 24 months (see Section 5.1 of this SE).

7.8 The reactor building spray pump suction relief valves, BS-V45A&B, and the decay heat pump suction header relief valves, BS-63A&B, have a safety function, as described in OH-10, Paragraph 1.1, and should be added to the licensee's IST program (see Section 6.1 of this SE).

7.9 The outboard reactor building isolation valves, BS-1A&B, may have a closed safety function and may be required to be closed post accident.

The licensee should evaluate whether these valves should be stroke-time closed and modify their IST program as necessary (see Section 6.2 of this SE).

8.0 CONCLUSION

The staff concludes that the relief requests and other portions of their IST program as evaluated and modified by this SE will not compromise the reasonable assurance of operational readiness of the pumps and valves in question to perform their safety-related functions.

Relief Request PG3 was partially denied.

In addition, portions of Relief Requests P1 and P2 which requested relief from the Code test procedure requirements were granted on an interim basis for a period of 1 year or the next refueling outage, whichever is longer.

Finally, provisional authorization was given for portions of Relief Requests P1 and P2, which proposed alternatives to the Code pump run time requirements, and provisional authorization was given for Relief Request VGl. The staff has determined that approval of relief requests and alternatives pursuant to 10 CFR 50.55a (f)(6)(i), (a)(3)(i), or (a)(3)(ii) is authorized.

In making this determination, the staff has considered the impracticality of performing the required testing and the burden on the licensee if the requirements were imposed. Any information provided in submittals or during telephone conferences by the licensee is subject to confirmation during NRC inspections.

1 Principal Contributor:

J. Colaccino, DE/EMEB Date: October 11, 1996 i

APPENDIX A

SUMMARY

OF SUBMITTED RELIEF REQUESTS Relief SE Code EgJipment Alternate NRC-RegJest Section Paragraph & ~

Identification Method of.

Action

' Nuiber Requiramonts Testing PG3 3.1 QMa-1988, All pumps in IST Use the action range Alternative Part 6 program requirements of ASME OM authorized (OM 6),

Code ISTB 1995, (a)(3)(1)

Paragraph 6.1 Paragraph 6.2.2.

Acceptance criteria Allow the analysis Proposed provision in ISTB alternative Paragraph 6.2.2 to apply denied (See to components in the Section 7.2 stort range.

of this SE)

P1 4.1 OM-6, Makeup and Use pump curves in the Interim Paragraph 5.2 Purification Puups final safety analysis relief Test MU P1A report for each granted Procedure MU-P1B indivichael puup and in (f)(6)(1)

MU Pic NUREG 1482, Section 5.2, (See Section to perform hydraulic 7.4 of this pulp testing during SE) refueling outages.

OM 6, Pump run time may be alternative Paragraph 5.6 less than 2 minutes authorized Duration of depending on the levet (a)(3)(li)

Tests in the pressurizer, with provisions (See Section 7.3 of this SE)

P2 4.2 OM 6, Turbine Driven use reference curves and Interim Paragraph Emergency Feed Pulp record pump hydraulic relief 5.2(c)

EF P1 and mechanical data granted Test lammediately after (f)(6)(i)

Procedure conditions have (See Section stabilized to minimize 7.6 of this lower quality water SE) pumped into the once OM 6, Motor Driven Emergency through steam generator.

Alternative Paragraph 5.6 Feed Pumps authorized Duratien of EF-P2A (a)(3)(ll)

Tests EF P2B with provisions (See Section 7.5 of this SE)

P3 4.3 OM-6, Nuclear service river Measure flow rate for Alternative Paragraph 5.2 water (NR) pumps each pump individually authorized Test NR P1A, NA-PIB, and each refueling outage.

(a)(3)(li)

Procedure NR-P1C A-1

4 l

e Relief SE Code Equipannt Attornate INtC d

Request Section Peregraph &

Identification Method of Action Nunber Requirements Testing P4 4.4 QM-6, Decay heat puup DM-P1B Establish a new alert Alternative Table 3a, value for the vertical authorized referenced in direction of 0.400 (a)(3)(ll) 4 the inches /second.

ecceptance criterie 4

requiramonts of Paragraph 6.1, VG1 5.1 OMa-1968, Numerous valves in the Perform the code testing Attornative Part 10 ticonsee's IST program each refueling outage authorized (OM-10),

(a)(3)(1)

Paragraph with 4.2.2.3(a) provisions leek rate (See Section 4

testing 7.7 of this frequency SE)

OM-10, Paragraph 4.1 fregancy requiraments for position etternative indication authorized testing (a)(3)(1) d with provisions (See section 7.7 of this SE)

I I

J f

a 1

4 A-2 l