ML18095A778

Application for Amends to Licenses DPR-70 & DPR-75, Consisting of LCR 87-07,Rev 1.Amends Modify Tech Spec Section 3/4.8.1 to Incorporate Guidance of Generic Ltr 84-15 to Improve Diesel Generator Reliability

ML18095A778
Person / Time
Site:	Salem
Issue date:	03/06/1991
From:	Labruna S Public Service Enterprise Group
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
Shared Package
ML18095A779	List: ML18095A778 ML18095A780
References
GL-84-15, NLR-N90068, NUDOCS 9103130128
Download: ML18095A778 (16)

Text

REQUEST FOR AMENDMEJ>!T REGARDING DIESEL GENERA.TOR

. RELIABILITY. "

REC'D W/LTR DID 03/06/91 .... 9103130128

-NOTICE-THE ATTACHED FILES ARE OFFICIAL RECORDS OF THE INFORMATION &

. REPORTS MANAGEMENT BRANCH.

THEY HAVE BEEN CHARGED TO YOU FOR A LIMITED TIME PERIOD AND MUST BE RETURNED TO THE RE-CORDS & ARCHIVES SERVICES SEC-TION P1-22 WHITE FLINT. PLEASE DO NOT SEND DOCUMENTS CHARGED OUT THROUGH. THE MAIL. REMOVAL OF ANY PAGE(S) FROM DOCUMENT FOR REPRODUCTION MUST BE RE-FERRED TO FILE PERSONNEL.

--NOTICE-

NLR-N90068 ATTACHMENT 1 LCR 87-07 Rev. 1 I. Description of Change Modify Section 3/4.8.1.1, 3/4.8.1.2 and B3/4.8 of the Salem Unit 1 and 2 Technical Specifications to address the changes to action statements and surveillance requirements suggested in NRC Generic Letter 84-15. Changes outside the scope of the Generic Letter, that are intended to clarify and/or improve the existing Technical Specifications, are also. included. The proposed change is intended to improve and maintain diesel generator reliability and achieve consistency in the Techn.ical Specifications for both Salem units. Specifically, the following changes are proposed:

1) "Offsite circuit" is being changed to. liind.ependent A.C.

circuit" throughout Specification 3/4.8.1

2) Revise the action statements of Specification 3.8.1.1.

a. Action statement 3.8.1.1.a is being revised to exclusively address one inoperable independent A.C.

circuit (it currently addresses either one inoperable independent A.C. circuit or one inoperable diesel).

The time required for performing Surveillance Requirement 4.8.1.1.2.a.2, to verify operability of the diesel generators, is being extended to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. The present requirement is to do the test within one hour and once per 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> thereafter.

b. New action statement 3.8.1.1.b is being added to address one inoperable diesel generator. It changes the requirement to demonstrate operability of the remaining diesels by extending the time required for performing Surveillance Test 4.8.1.1.2.a.2 to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, similar to the proposed change to 3.8.1.1.a.

c. Action statement 3.8.1.1.c, for one independent A.C.

circuit and one diesel generator inoperable, is being revised to require performance of Specification 4.8.1.1.2.a.2 for the remaining two diesel generators within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />. The present Technical Specifications require the Surveillance Test to be done within one hour and every 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> thereafter.

d. Action statement 3.8.1.1.d, for two inoperable independent A.C. circuits, is being revised to require performance of Surveillance Requirement 4.8.1.1.2.a.2 within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />, similar to the proposed change_ to 3.8.1.1.c

• NLR-N90068 e. New action statements 3.8.1.1.f and 3.8.1.1.g are being added specify the required actions if a fuel oil transfer pump or fuel storage tank becomes inoperable.

3. Revise the Surveillance Requirements of Specification 4.8.1.1.

a. Revise Specification 4.8.1.1.2.a to reference the test frequency of Table 4.8-1 (Unit 1 only).

Revise Specification 4.8.1.1.2.a.2 to delete "from ambient condition" The one hour test at continuous rated load (currently 4.8.1.1.2.a.3) is being made part of Specification 4.8.1.1.2.a.2, which is the operability test performed per the action statements.

Revise the one hour test at continuous rated load to change the loading rate from a fast (< 60 seconds) to a gradual load increase, typically 500 kw per minute.

Instead of a load greater than or equal to 2600 kW for Unit 2 and greater than or. equal to 1400 kW for Unit 1, a range of 2500-2600 kW will be used as the continuous rated load for this test.

b. Add new Specificatio~ 4.8.1.1.2.b to Unit 1, identical to that of Unit 2. This specification will require inspection for and removal of water from the day tanks, at least once per 31 days and after each period of diesel operation greater than or equal to one hour.

c. Add new Specification 4.8.1.1.2.c to require a fast start and load, and one hour operation at a range of 2500-2600 kw, at least once per 184 days.

d. Delete Unit 2 Specification 4.8.1.1.2.c.5, which requires simulating loss of a diesel following a blackout, and verifying that the blackout loads cannot be auto-connected to a loaded bus.

e. Revise the 18 month testing requirements (proposed Specification 4.8.1.1.2.d). For Unit 1, the test is being expanded to simulate various plant transients (e.g., loss of offsite power by itself, ESF signal by itself, etc.) and verify proper diesel performance and load sequencing, consistent with Unit 2. Additional revisions provided for Unit 1 and 2 test requirements consist of: ~

o Addition of a footnote stating that manufacturer's recommendations will be followed for surveillance tests unless loading times are specified in the individual surveillance requirements.

NLR-N90068

• o Revise Specification 4.8.1.1.2.d.6(c), for verifying diesel trip signal bypass. The signal used is being changed from loss of voltage and/or safety injection to loss of voltage and safety injection. The Surveillance Requirement is being reworded to clarify that the test is intended to verify the nonessential trips are bypassed. Verification that the essential trips (engine overspeed, low oil pressure, bus differential and generator differential) are not bypassed is not part of the test.

• o Modifying the loads used for the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> test. For the first two hours of the test, the maximum load and duration will be based on the worst_: case loading as determined by the Diesel Generator Load study (Ref. 1).

Where the load study results yield a load greater than 2860 kw for a specified duration, the test will mimic the load study profile in the region of exceedance.

For the remainder of the first two hours of the test (i.e., load study results are less than 2860 kw), a range of 2760-2860 kw will be used.

For the remaining 22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br /> of the test, a range of 2500-2600 kW will be used. An explanatory footnote is being added to state that the ranges are being used to avoid routine overloading of the diesels.

o Revise Specification 4.8.1.1.2.d.7 to change the auto-start test, immediately following the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> run test, from a simulated ESF actuation signal to a simulated ESF concurrent with a loss of offsite power signal. Add a note to state that if the auto-start test fails, the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> test need not be repeated.

o Revise Specification 4.8.1.1.2.d.8. Instead of verifying the auto connected loads do not exceed the 2000 hour0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> rating, the test will verify that the load profiles of the diesel generator load study are not exceeded. *

f. Revise Specification 4.8.1.1.3.b, regarding fuel oil testing, changing "ASTM 0975-68 11 to "ASTM 0975-77 11 *

g. For Unit 2, revise the reporting requirements of specification 4.8.1.1.4 to refer to specification 6.9.2 for Special Reports. Add new Specification 4.8.1.1.4 to Unit 1 to provide reporting requirements identical to those of Unit 2.

h. For Unit 1, revise Table 4.8-1, Diesel Generator Test Schedule, to be identical to the one proposed for

• Unit 2 via letter dated July 27, 1990 (NLR-N90156), and approved via Amendment 93. Test frequency will be based on the number of failures in the last 20 valid

NLR-N90068 tests, and will be determined on a per diesel generator basis.

4. Revise Specification 3/4.8.1.2, regarding A.C. power requirements during Shutdown.

a. Provide editorial clarification to Specification 3/4.8.1.2.

b. Revise the Surveillance Requirements of Specification 4.8.1.2. For Unit 1, surveillance Requirement 4.8.1.1.3 (except for 4.8.1.1.3.a.2) is added to the shutdown test requirements. This will require that the Unit 1 fuel is sampled and tested in a manner identical to that of Unit 2.

Specification 4.8.1.2 is also being revised, for both units, to reflect the incorporation of present specification 4. 8 .1.1. 2. a. 3 into proposed specification 4.8.1.1.2.a.2 (i.e., one hour run will be part of operability test) .

For both units, add a note to clarify when a surveillance performed during shutdown is a "valid test", per Regulatory Guide 1.108.

5. Revise Bases Section B 3/4.8 to provide editorial clarification and, for Unit 1, add references to Regulatory Guide 1.9 and 1.108.

II. Reason for the Proposed Changes Salem Unit Nos. 1 and 2 each have three ALCO 18 cylinder V-type, medium speed diesel engines. Each engine is connected to a 4160 volt generator. The generators are used as an emergency power supply for each of the station vital buses. The diesel generator units have a 2600 kwe continuous rating and a 3100 kwe one-half hour rating. The diesels are designed to start and be up to required voltage and frequency within 13 seconds of either a Loss of Offsite Power condition and/or an ESF actuation signal.

The diesel generator testing program differs between the two Salem units because of the differences in the applicable Standard Technical Specification surveillance requirements at the time of issuance of the Facility Operating License for each Salem unit.

However, the monthly surveillance tests are conducted under the same conditions. Most diesel generator starts (including starts for surveillance purposes) take place from ambient conditions.

There is one feature available on the Salem diesels that differs

• • from certain others used in the nuclear industry. This feature is a prelubrication system and lube oil heating system which serves to circulate warmed oil through the diesel when it is at rest. The temperature of the oil is maintained between

NLR-N90068 approximately 110°F and 120°F with the prelubrication system in operation. This system is necessary for this type of diesel as the manufacturer does not support f a~t starting of the diesel when the temperature falls below 100 F.

The sample technical specifications provided in GL 84-15 (herein referred to as the "sample TS") were reviewed_ in detail by PSE&G.

Many of the suggestions contained in the sample TS were determined to provide a preferable alternative to the present testing program. The proposed program will give a more accurate indication of diesel generator reliability. There are several factors which lend themselves favorably to this conclusion.

These are:

1) Implementing the proposed diesel generator test frequency on the basis of individual diesel generator reliability will help to avoid unnecessary diesel starts, while promoting the identification and correction of specific sources of diesel unreliability. Determining test frequency according to performance on a per nuclear unit basis penalizes diesels with acceptable reliability.

2) The extended time criteria for testing the remaining operable diesels when an offsite circuit or another diesel is inoperable is beneficial. This will serve to prevent overtesting of diesel generators with good reliability histories. Eliminating repetitive testing while in an Action Statement will also help minimize unnecessary starts.

3) The reduction in frequency of tests involving fast loading, to once per 6 months, will reduce unnecessary wear and minimize thermal transients in the diesel engines.

For these reasons, PSE&G requests a license change to adopt many of the changes suggested by GL 84-15 in an effort to contribute to diesel generator reliability and to provide for more realistic monitoring of diesel generator performance on an individual basis.

Technical specification consistency between both Salem units through this proposed license change was accomplished in two phases. The first phase was an upgrading of the Salem Unit 1 testing requirements to those additional requirements addressed during the licensing of Salem Unit 2. This consisted of adding the more detailed Salem Unit 2 testing requirements to comparable sections of the Salem Unit 1 Technical Specifications and represents a conservative change. The second phase was the consistent application of appropriate sections of the GL 84-15 Sample TS for both Salem units.

PSE&G is also proposing changes to the diesel generator Technical

• Specifications that are 84-15, but are intended program. These changes outside the scope of Generic Letter to improve the diesel generator testing consist of new Action statements to

NLR-N90068 address inoperability of a fuel oil transfer pump or fuel storage tank; revision of the acceptance criteria for continuous rated load and 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> load testing; referencing a more current ASTM standard for fuel oil testing; and revision of reporting requirements for diesel failures to be consistent with current industry practice and reporting regulations.

III. Justification for the Proposed Changes Changing "offsite circuit" to "independent A*C* circuit" more accurately describes the circuit required by Technical Specification 3/4.8.1 (i.e., the circuit between the offsite power source and the onsite vital busses).

Specification 3.8.1.1 Action Statement 3.8.1.1.a The reason to perform diesel operability tests following the loss of one independent A.C. circuit is to ensure that a backup power source is available and capable of performing its intended function. The present action statement requires verification of diesel starting capability within one hour and then once every 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> thereafter.

Demonstration of diesel starting capability within one hour of a loss of an offsite power source and subsequent testing every 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> thereafter is both excessive and unwarranted.

If the diesel has been surveillance tested within the previous 31 days, an operability test provides little further assurance of diesel availability. The loss of an independent A.C. offsite circuit does not suggest that any diesel has become less reliable. Consistent with these statements, operability testing within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> is more reasonable than testing within one hour. Diesel testing within the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> will provide assurance that the diesels are operable. Furthermore, the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> time period permits sequential testing of the diesels rather than simultaneous testing of the diesels and is consistent with the statements of GL 84-15.

The operability test of 4.8.1.1.2.a.2 is being revised to complement the reduction in test frequency per the action statements. The test is being changed from a start without loading to a start with gradual loading and one hour run at continuous rated load. This change is based on industry operating experience and vendor recommendations, which suggest that repetitive starts and operation at no load contributes to engine degradation. Testing in the proposed manner also recognizes that starting problems should not necessarily be treated as the dominant failure mechanism, and testing under loaded conditions yields a more meaningful demonstration of operability.

NLR-N90068 Repetitive operability testing every 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> is both unwarranted and counterproductive in providing assurance of diesel starting capability. Inasmuch as excessive fast starts have been identified as contributing to premature diesel engine degradation, repetitiye testing is contrary to the stated purpose of providing continued assurance of starting capability. Consistent with GL 84-15, one operability start per diesel should provide assurance of capability to start and run at rated load in the event of a loss of an independent A.C. circuit.

Action statement 3.8.1.1.b The reason to perform a diesel operability test following the loss of a diesel is to assure that the remaining diesels will be available and capable of starting as designed.

Specifically, an operability test of the remaining diesels provides assurance that the remaining operable diesels are not subject to the same failure (i.e., common-mode failure).

Rather than relying on previous surveillance testing, operability testing within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> is proposed consistent with GL 84-15.

The sample TS presented in GL 84-15 allows for diesels to be inoperable for a time period to be specified by the licensee but not to exceed a certain number of cumulative days per year. The changes proposed herein retain the 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> allowable outage time instead of the sample TS cumulative limit. It has been the experience at Salem that, with proper planning, most maintenance can be performed on the diesels within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />. Using the "maximum cumulative inoperability time" suggested in the sample TS may cause maintenance activities to be deferred to avoid exceeding the time limit. A cumulative inoperability limit would also require the redefinition of Technical Specification 3.0.5 because of the impact of emergency power sources on other Technical Specification related equipment and would therefore require a reevaluation of Salem's Probabilistic Risk Assessment. PSE&G feels that this approach to diesel inoperability will unnecessarily complicate the Technical Specification in addition to creating another. administrative tracking function.

The sample TS for this action statement include a footnote defining inoperability of a diesel. That is, a diesel generator is considered inoperable until it satisfies the Surveillance Requirement performed to demonstrate diesel operability. PSE&G has not included this footnote since it is understood that successful completion of Technical Specification 4.8.1.1.2.a.2 is required to restore operability to a diesel .

NLR-N90068 Action Statement 3.8.1.1.c This action statement covers the appropriate responses when declaring one offsite circuit and one diesel generator inoperable. Consistent with the proposed changes to Action Statements 3.8.1.1.a and 3.8.1.1.b, proposed Action Statement 3.8.1.1.c will provide adequate assurance of diesel availability by one time testing within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />, eliminating extra diesel starts. Additionally, performance of the test within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> provides added time for inspection prior to test and a more orderly test sequence.

Proposed Action statements c, d and e differ from the sample TS in one substantive respect. The sample TS require entry into Action Statements a or b as diesel generators and/or offsite circuits are restored to operability. The proposed Action statements c, d and e are self-contained; they do not refer back to Action statements a or b. This approach eliminates the confusion associated with going from one Action statement to the other, particularly with regard to retesting diesels (per Action Statement a or b) that have either been verified operable (per Action statements c, d or e) or have just been restored to operability. Because the shutdown requirements of the proposed Action Statements are effective from the time of initial loss of an AC power source(s), PSE&G believes that this change is consistent with GL 84-15.

Action Statement 3.8.1.1.d This action statement covers the appropriate responses when two offsite power circuits are declared inoperable. The justification for the proposed action statement is consistent with that presented above for Action statement 3.8.1.1.a. As previously noted, loss of an offsite circuit does not suggest that a diesel has become less reliable since its previous surveillance test. Given the significance of losing both offsite circuits, one operability test per diesel within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> provides adequate assurance of diesel generator availability. This change is also consistent with GL 84-15.

Action statement 3.8.1.1.e This action statement covers the appropriate responses when declaring two or more diesels inoperable. There are no substantial differences between this action statement and Action Statement 3.8.1.1.d of the existing Salem Unit 2 Technical Specifications.

Action Statements 3.8.1.1.f and 3.8.1.1.g

• The purpose of these action statements is to separately address the inoperability of a fuel transfer pump and a fuel

NLR-N90068 storage tank, respectively. A 72 hour8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> time period is allowed to repair the equipment before requiring entry into at least hot standby within 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> and cold shutdown within the next 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, which is consistent with the action statement for one inoperable diesel generator.

Inoperability of a single transfer pump or storage tank is not indicative of a failure of any of the diesel generators themselves. Therefore the performance of specification 4.8.1.1.2.a.2 to demonstrate operability of the diesel generators should not be required. Since one fuel transfer pump supplies sufficient fuel to all three diesels at their maximum ratings, the loss of one fuel transfer pump and/or fuel storage tank should not result in declaring a diesel inoperable.

Section 4.8.1.1 The proposed Technical Specification testing frequency is based on the number of valid failures in the last 20 valid tests, on a per diesel generator basis consistent with GL 84-15. PSE&G feels that the testing of a redundant diesel based on the failure of another diesel is excessive and not technically justifiable.

such testing adversely affects the performance and reliability of the other diesels. Changing the specification to a per diesel basis addresses individual diesel reliability and enhances the overall reliability by increasing the test frequency of only those diesel generators which are experiencing failures.

Salem diesel generators are designed to start within 10 seconds and are required to be ready for loading within 13 seconds from the start signal. Failure of the diesels to reach 900 rpm within 10 seconds of a start signal is indicative of a problem with the starting capability (e.g., controls, starting air system).

Modifications to the existing control system would be required to perform a slow start test. Per the vendor survey on Table 4.5 of NUREG/CR-4557, "A Review of Issues Related to Improving Nuclear Power Plant Reliability", a general vendor comment was that 11 10 second starts are acceptable, but loading time should be increased beyond 60 seconds **. DGs should not be run unloaded".

Therefore, the present acceptance criteria are being retained for routine surveillance testing. Verification of fast loading capability will be performed during the 184 day test. The phrase "from ambient condition" is being deleted because of the continuously operating prelubrication/warming system used for the diesel generators.

Specification 4.8.1.1.2.a.3, which requires the diesel generator to operate for at least 60 minutes while loaded, is being incorporated into 4.8.1.1.2.a.2, which is the test performed pursuant to the action statements. This change is being proposed to reduce the number of diesel runs with no load, which tends to increase engine degradation. Testing at rated load will also provide a greater assurance of operability than the test

NLR-N90068

• presently required by the action statements, which only verifies capability to start and reach synchronous speed.

The 60 minute test at rated load will undergo two substantive changes. First, the loading rate is being: changed to "gradual" to reduce the degradation associated with the frequent fast loads. The addition of specification 4.8.1.i.2.c (see below) will retain the requirement to perform the fast loading test, with the frequency reduced to once per 184 days. This reflects the GL 84-15 philosophy that the monthly tests to verify starting and load handling capability should be performed in accordance with manufacturer's recommendations and simulation of design basis conditions should be reql.lired less frequently.

The second change proposed to the 60 minute test is to use a range of 2500 to 2600 kW as an acceptable continuous rated load for test purposes. The present requirement to load the diesel to greater than or equal to 2600 kW (for Unit 2) potentially leads the operator to routinely exceed continuous rated load in order to meet the test criteria.

New specification 4~8.1.1.2.c is being added, consistent with GL 84-15, to require a fast start and one hour operation of each diesel generator once per 184 days. This test is currently performed at least once per 31 days. As discussed above, it is proposed that the monthly test be performed per manufacturer's recommendations, to verify capability to start and operate under load. Simulation of design basis conditions via fast loading will be performed less frequently in order to minimize wear and thermal transients.

The proposed changes to the 18 month test requirements of specification 4.8.1.1.2.d may be divided into two categories:

changes to make Unit 1 consistent with Unit 2 and changes to upgrade both units. The first category is justified based on the fact that Unit 2 currently has a more extensive and thorough test program than Unit 1. The second category includes:

0 Emphasizing that manufacturer's recommendations will be followed, unless otherwise specified by individual surveillance requirements.

0 The range of 2500-2600 kW is being proposed for the use as the continuous rated load. The justification used for specification 4.8.1.1.2.c above also applies here.

The two hour test load (i.e., first two hours of the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> test per Specification 4.8.1.1.2.d.7) is being revised to account for the results of Salem's diesel generator load study. The present Unit 2 two hour test value of greater than or equal to 2860 kW is open ended and increases the

• likelihood of overloading the diesel generators. PSE&G is proposing to use a range of 2760-2860 kw, except where the study results yield a load greater than 2860 kw. In these

• NLR-N90068 cases the generator will be loaded in accordance with the load study for the period of time that the load exceeds 2860 kw. The magnitude and duration of the exceedances are such that testing in the proposed manner will not be detrimental to diesel performance. The only exceedance of 2860 kw given by the present load study is 2872 kw for approximately 15 minutes.

For the remaining 22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br /> of the test, a range of 2500-2600 kw, with a target value of 2550 kw, will be used.

This range will prevent routine exceedances of the continuous rated load (2600 kw). The highest long term (beyond one hour) loading as given by the load study is 2542 kw. Therefore, the proposed loading method for the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> test will provide a meaningful test of load capability without resulting in overloading the diesels.

o The test following the 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> run will utilize an ESF plus loss of offsite power test signal, instead of the ESF test signal by itself. This is consistent with the sample TS.

o Changing Specification 4.8.1.1.2.d.8 to reference the maximum design load instead of the 2000 hour0.0231 days <br />0.556 hours <br />0.00331 weeks <br />7.61e-4 months <br /> rating is consistent with the proposed practice of testing based on Salem's EDG load study results.

o Diesel fuel oil surveillance testing is presently based on ASTM 0975-68. The proposed change does not affect the actual acceptance criteria, but would incorporate the same standard used by the Hope Creek Generating Station and Revision 4 of the Westinghouse Standard Technical Specifications (ASTM 0975-77). The original fuel oil testing program meets the intent of Regulatory Guide 1.137 Revision 1, as acknowledged by the NRC in its letter to PSE&G dated October 6, 1980. Therefore, PSE&G is not proposing to incorporate the additional surveillance Requirements of the sample TS. The Updated Final Safety Analysis Report (UFSAR) will be updated to provide a more detailed description of the fuel oil testing program.

Unit 2 Specification 4.8~1.l.2.c.5 requires verification that upon a simulated loss of a diesel generator, with offsite power not available, the diesel cannot be auto connected to a loaded bus, and subsequent loading is in accordance with design requirements. This specification is part of the original Unit 2 Technical Specifications; it is not contained in the Unit 1 TS.

It is similar to Specification 4.8.1.1.2.d.6 of Revisions 1 through 4 of the Westinghouse Standard Technical Specifications.

On July 21, 1983, the NRC issued Generic Letter 83-30, which removed STS 4.8.1.1.2.d.6, stating that the surveillance

• Requirement is inconsistent with (i.e., beyond the scope of) 10 CFR 50 General Design Criterion 17, Regulatory Guide 1.108, and Standard Review Plan 8.2 and 8.3.1. The test is not included

NLR-N90068 in the sample TS of Generic Let~er 84-15* Therefore, PSE&G is proposing to delete the test from the unit 2 TS, and leave it out of the Unit 1 TS. The present test sequence for both Salem Units includes manual actuation of the diesel overspeed trip function while the diesel is powering blackout loads. Diesel trip and opening of the diesel output breaker are verified. This test verifies proper breaker operation following loss of a diesel.

Differences From Generic Letter 84-15 Sample TS The following provides justification for differences from the sample TS that are not discussed above.

Surveillance Requirement 4.8.1.1.2.a.4 of the GL 84-15 sample TS specifies four different signals that may be used to start a diesel for its monthly test. PSE&G does not feel that listing these options is necessary and has not included them. Simulating a loss of offsite power and/or ESF actuation test signals for the purposes of starting the diesel generators, followed by manually loading the EOG to continuous rated load, introduces the potential for inadvertent ESF actuation. Performance of the test using simulated signals would require blocking out Safeguard Equipment Controller (SEC) outputs, with the exception of the diesel start. After the diesel is automatically started through the SEC, manual control would be required to allow loading, .to prevent inadvertent actuation of equipment that is automatically loaded by the SEC in its various modes of operation. Therefore, PSE&G believes this test to be a complicated evolution during plant operation and proposes using the auto-start signals only for the 18 month shutdown tests.

Sample TS 4.8.1.1.2.d.2 requires that voltage and frequency be maintained within the specified limits during rejection of the single largest emergency load. PSE&G is not adopting this change, but is retaining the provision to allow a 4 second recovery period for voltage and frequency to return to their specified values. This prevents the brief transient following load rejection from invalidating the test.

Sample TS 4.8.1.1.2.d.3 requires rejection of continuous rated load (2600 kW for Salem) without a diesel trip. During initial EDG acceptance testing (September, 1973), 2800 kW load rejection tests were performed, resulting in turbocharger surging. Surging occurs when compressor throughput is reduced to a point below design conditions such that erratic performance results. The head-flow curve is flat at the point where surging occurs.

Pulsations in flow and discharge pressure produce high frequency reversals in axial thrust on the compressor shaft. Turbocharger compressor blade failures, experienced during initial Salem EDG testing, were attributed to fatigue due to the stresses that resulted from surging. Corrective action for the turbocharger failure included a change in diffuser blade geometry in order to move the operating range further from the surge line. However,

• NLR-N90068 *._

• the transient imposed on the diesel engine by instantaneously rejecting a 2600 kW load is still likely ~o result in turbocharger surging. In order to avoid fatigue failure of the turbocharger, PSE&G has not included the continuous rated load rejection test recommended by the Sample TS. The capability to withstand rejection of the single largest load (785 kW) is demonstrated via surveillance Requiremen~-~-8.1.1.2.d.2.

The change to Unit 2 specification 4.8.l*l*4 Reports will clarify the reporting requirements for diesel generator failures. The revised reference to specification 6.9.2 for Special Reports is consistent with the changes proposed to Technical Specification Chapter 6 via PSE&G letter dated May 31, 1990. That is, reference to Specification 6.9.1 reporting requirements are no longer appropriate for diesel failures, based on the provisions of the Licensee Event Report (LER) rule as described by Generic Letter 83-43. A new specification 4.8.1.1.4 is being added to Unit 1, so that both units will have identical reporting requirements consistent with Regulatory Guide 1.108, Revision 1.

Unit 1 Table 4.8-1, Diesel Generator Test Schedule, is being revised to be identical to Unit 2 Amendment No. 93. The proposed change specifies EDG test frequency based on the test results during the past 20 tests, on a per EDG basis. The additional criteria presented in the GL 84-15 sample TS (test the diesel at least once per seven days until seven consecutive failure-free demands have been performed, and the number of failures in the last 20 valid demands has been reduced. to one or less) have not been adopted. PSE&G believes the test frequency approved in Unit 2 Amendment 93 provides adequate assurance of operability without resulting in excessive testing. Revision o of this change request.included sample TS Table 4.8-2, Additional Reliability Actions. Based on discussions with NRC staff, the NRC is no longer requesting that Table 4.8-2 be incorporated into the Technical Specifications. Additional reliability goals and actions are under consideration as part of the implementation of the Station Blackout rule.

Proposed Surveillance Requirement 4.8.l.l.2.d.6(c), for verifying that automatic diesel generator trips are bypassed is different than the comparable GL 84-15 sample TS due to the design of the diesel generators and ESFAS at Salem. The Salem diesels are designed to have all but the following trips bypassed; (1) engine overspeed, (2) low lube oil pressure, (3) 4 Kv bus differential, and (4) generator differential. The nonessential (bypassed) trips are in parallel to each other, and include the manual emergency stop pushbutton. In order to verify the appropriate trips are bypassed, the pushbutton is depressed while the diesel is operating in the ESF plus Loss of Voltage mode (the mode specified by the sample TS). If the diesel does not trip, the test is satisfactory. By using the manual stop pushbutton to stop diesel operation when the trips are not bypassed (at some point in the 18 month test sequence), the operability of the pushbutton will be verified, thereby validating the circuit

NLR-N90068 bypass test. The circuit design is such that the bypass feature does not affect the essential trips. The proposed rewording of the surveillance requirement is intended to clarify that the test does not apply to the essential trips.

Also, Surveillance Requirement 4.8.1.1.2.d.9 of the GL 84-15 sample TS has not been included in the proposed amendment. This surveillance requires that the diesel generator is capable of being synchronized to the offsite power source while the diesel is loaded. This cannot be done because the vital bus infeed breakers do not have synchronizing capabil~ty, as identified in Section 14.0 of supplement 4 to the Salem Generating Station Safety Evaluation Report. A diesel generator may be synchronized to a vital bus, but synchronization of the offsite power source back to a diesel-powered bus is not part of Salem's design. In a Loss Of Offsite Power scenario, when a diesel generator is loaded and offsite power becomes available, the Safeguards Equipment Controller (SEC) is reset and the vital bus is stripped of its loads. The diesel generator output breaker is opened, the vital bus infeed breakers are closed, and loads are restored to the bus. Diesel operation is manually terminated.

Technical Specification 3/4.8.1.2 A footnote has been added to the surveillance requirements to include clarification of valid successful tests and failures while in MODES 5 and 6. This is consistent with the Regulatory Guide 1.108 definition of valid tests and failures. Also, some editorial changes have been made in the wording of the action statement in order to make it consistent with similar items in Action Statement 3.8.1.1.

Bases Section 3/4.8.1 Changes to Bases section B 3/4.8.1 are editorial for Unit 2. For Unit 1, the changes will make the section identical for both units.

IV. DETERMINATION OF NO SIGNIFICANT HAZARDS CONSIDERATION The proposed changes to Technical Specifications 3/4.8.1.1 and 3/4.8.1.2 for both Salem 1 and 2:

(1) do not involve a significant increase in the probability or consequences of an accident previously evaluated. Reducing the test frequency while in an action statement and modifying Emergency Diesel Generator (EDG) starting and loading requirements is intended to enhance diesel reliability by minimizing severe test conditions which can lead to premature failures. Using the proposed basis for determining test frequency according to individual diesel

• generator performance will prevent overtesting of the diesels. The changes proposed to make the Unit 1 EDG surveillance requirements identical to that of Unit 2 is a

• NLR-N90068

• c conservative change; it will provide Unit 1 with a more comprehensive testing program. The proposed changes will continue to assure availability of the diesels and should serve to enhance EDG reliability and consequently the overall safe operation of the Salem Generating Station.

(2) do not create the possibility of a riew or different kind of accident from any accident previously evaluated. The proposed change affects testing frequency, starting and loading practices only and has no impact on the accident analysis. No new operating modes or equipment are introduced which could initiate or affect the progression of an accident.

(3) do not involve a significant reduction in a margin of safety. The changes in the testing requirements do not adversely affect the capability of the diesels to perform their required function. Rather, the purpose of the proposed changes is to increase the overall reliability of the diesels. In adopting many of the suggestions identified in GL 84-15, the requested change would implement actions which have been determined by the NRC to reduce the risk of core damage from station blackout events.

Therefore, PSE&G has concluded that the changes proposed herein do not involve a Significant Hazards Consideration.

V. REFERENCES

1) Salem Generating Station Diesel Generator Load Study, Report No. AEI-PSEG-8811-DR-001.