ML20205E059

Safety Evaluation Supporting Amend 105 to License NPF-58

ML20205E059
Person / Time
Site:	Perry
Issue date:	03/26/1999
From:	NRC (Affiliation Not Assigned)
To:
Shared Package
ML20205E057	List: ML20205E059
References
NUDOCS 9904050040
Download: ML20205E059 (18)

Text

- - - - -

. [S g

2 UNITED STATES p

s NUCLEAR REGULATORY COMMIS810N WASHINGTON, D.C. 300eM001

• g*****

. $&ETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATED TO AMENDMENT NO.105 TO FACILITY OPERATING LICENSE NO. NPF-58 FIRSTENERGY NUCLEAR OPERATING COMPANY PERRY NUCLEAR POW W PLANT. UNilf DOCKET NO. 50-440 1.0 INTRODUCT!ON By letter dated September 9,1998, as supplemented by submittals dated January 6, March 4, and March 18,1999, the FirstEnergy Nuclear Operating Company (the licensee, formerly The Cleveland Electric illuminating Company and Centerior Service Comoany) proposed changes to the Perry Nuclear Power Plant, Unit 1 (PNPP) Technical Specifications (TSs) related to j

hydrostatic (water) testing of the containment isolation valves in the feedwater system. The proposed changes, which were submitted pursuant to 10 CFR 50.59 and 10 CFR 50.90, would revise the licensing and design basis of the feedwater isolation provisions.

The licensee has experienced extensive operational difficultiec; wrth regard to the containment isolation and leak rate testing of the feedwater system check valves. Past leak rate testing of these valves, pursuant to Appendix J to 10 CFR Part 50, have questioned whether adequate containment isolation would be attained following a postulated accident. The licensee has proposed changes to the licensing and design basis for the overall feedwater isolation system to improve and enhance the reliability of the containment isolation provisions.

' The licensee's letter of January 6,1999, included a proposed exemption to the leak rate testing requirements of Appendix J to 10 CFR Part 50, for the feedwater system check valves. The licensee proposed to conduct a visual examination of the check valves in lieu of leak rate testing. During the review process, the staff and the licensee concluded that performanc.e of a

. leak rate ic;t of the check valves was a superior method to demonstrate valve operability as opoosed to a visual examination. Performance of a leak rate test would satisfy the requirements of Appendix J and preclude the need for an exemption. In the licensee's letter of March 4,1999, the licensee committed to conduct leak rate testing of tte check valves consistent with Appendix J requirements and acknowledged that an exemption to Appendix J was no longer necessary.

The supplemental information in the licensee's letters of January 6, March 4, and March 18,1999, contained clarifying information and did not change the initial no significant hazards consideration determination and did not expand the scope of the original Federal Register notice.

WO4050040 990326 TWR ADOCK 05000440 P

POR x

E

-2

2.0 BACKGROUND

2.1 Current Feedwater Penetration Isolution Desian The feedwater penetration is a unique case for containment isolation. For the majority of system j

transients or loss of coolant accidents (LOCAs) other than a feedwater pipe break, feedwater

{

~ flow will be maintained in order to get cooling water to the reactor vessel. However, there are certain conditions when feedwater flow must be isolated. Isolation provisions must (1) eliminate containment atmosobere leakage in the feedwater piping for LOCAs inside containment, and (2) isolate reector coolant system leakage flowing in the reversa direction for feedwater line breaks outside containment.

The current Perry feedwater penetration and isolation provisions are shown in Figure 1. (it should be noted that there are two feedwater penetrations designated train A and B. However,

]

for simplicity, Figure 1 only lists the valve numbers and does not provide the train A and B designations.) Each feedwater line penetrating containment has three containment isolation valves. Two piston lift-style check valves, located both immediately inside and outside the containment penetration, are in each feedwater line for isolation of significant flow from a feedwater line break outside containment. These anti-waterhammer check valves are not designed for air tests at low pressures. A third valve (a remote manual motor-operated gate valve), located outside containment upstream of the outside check valve, is provided for long-term, high integrity leakage protection when, in the judgment of the operator, continued make-up from feedwater is unnecessary or is not available. There is no automatic isolation of the feedwater lines based on accident sign:Is, so that feedwater flow can be maintained to the reactor vessel. As noted in ANS-56.2/N271-76, ' Containment isolation Provisions for Fluid Systems," greater plant safety is maintained with a feedwater supply to the teactor.

l The feedwater leakage control system (FWLCS) is designed to eliminate containment atmosphere through-line leakage in the feedwster piping for LOCAs inside containment by providing a positive water seal between the isolation valves. The FWLCS consists of two independent trains. As shown in Figure 1, Division 1 of the FWLCS fumishes sealing water from the suppression pool using the low-pressure core spray (LPCS) waterleg pump to an outboard volume (between the outboard feedwateridolshn check valves, B21-F032, and the re mote manual gate valves, motor operated valves (M(Ns), B21-F065). The MOVs (B21-F065) are powered by Division 1 electrical power. Division 2 of the FWLCS fumishes sealing water from the suppression pool using the residual heat removal (RHR) waterleg pumps (B/C) te a n inboard V";lume (between the two feedwater isolation check valves, N27-F55g and B21-F032). The FWLCS system is designed and installed as a single failure proof, safety-related, seismicsily qualified system that is designed to withstand the dynamic effects of postulated piping fsliures in the steam tunnel including protection from intomally generated missiles.

The FWLCS is a manually activated system, effective within approximately 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> after the onset of a LOCA. When the operator has determined that feedwater is either unavailable or not necessary, the FWLCS is actuated to provide a water seal in the feedwater penetration line to prevent through-line leakage of the containment atmosphere to the environment. The current licensing Ms assumes that operater action will be initiated within 20 minutes of a LOCA to

. (1) close the motor operated gate valve, and (2) initiate the FWLCS such that a water seal between the isolation valves will be established within approximately 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />. (It should be noted that it takes approximately 40 minutes to fill the piping volume between the three containment isolation valves. Thus, a water seal is assumed to be established within one hour of a LOCA.)

The FWLCS includes interlocks to ensure that the outboard FWLCS (Division 1) is not initiated without the feedwater MOV being closed thereby preventing the inadvertent discharge of suppression pool wter to the feedwater piping system. The inboard FWLCS (Division 2) system is not interlocked with the feedwater MOV.

2.2 Need for Chanae bMsgasina Basis

~

The current licensing basis for feedwater isolation relies on the FWLCS to establish a water seal between the three containtnent isolation valves approximately one hour following an accident.

Successful operation of the FWLCS relies upon operator action to close the MOV and initiate I

the FWLCS approximately 20 minutes following the accident. The feedwater check valves must be essentially leak tight so that the injected water from the FWLCS can establish a water seal.

The check valves are leak tested during every refueling outage and have an acceptance criteria of 1 gpm water leakage. If leakage exceeds this limit, a water seal may not be established due to limitations of flow from the FWLCS makeup pumps.

Operational experience has not always demonstrated that the feedwater check valves are leak tight. "As-found' testing over the last several refueling outages has shown that the check valves leak in excess of 1 gpm and actual values have been on the order of 4-14 gpm. Whenever leakage exceeds allowable limits, the licensee is required to take corrective actions to restore valve integrity to the licensing limus. This requirement has led to excessive costs and man-rem exposures. During the sixtn refueling outage, the licensee estimated that valve restoration cost approximately $880,000 and 5 man-rem exposure. More importantly, however, is the concem that due to excessive check valve leakage, the existing FWLCS may not be capable of performing its safety-related function.

The licensee explored a number of options to improve the reliability of the feedwater isolation system. As described in the next section, the licensee selected a design that would no longer inject the FWLCS in the piping volume between the containment isolation valves but would Y. ject the FWLCS through the stem of the MOV and thus establish a water seal between the double discs of tho MOV. Reliability of the MOV to close would be increased by introduc.ng an attemate electrical power supply to the MOV. Finally, the licensee proposed to perform leak rate tests for the check valves pursuant to Appendix J to 10 CFR Part 50 to demonstrate continued valve integtky.

2.3 Prooosed Feedwater Penetration isolation Desian I

in the proposed design, as shown in Figure 2, both independent trains of the FWLCS will be routed to the bonnet area of each existing feedwater MOV. To allow the FWLCS to weal the 20" gate valves, the FWLCS will be supplied to an existing %"-threaded connection in the packing area of the valve bonnet. Once in the bonnet area of the 20" gate valve, the FWLCS seal water flows around the wec'ge-shaped gate into the area between tho two hardfaced maineests in the I

C-l

valve body forming a water seal. Since FWLCS water is supplied at higher than accident i

pressure (P.), any air leakage path past the check valves would be sealed by the FWLCS.

The rerouted FWLCS subsystems will continue to be designed and installed as single failure proof, safety related, seismically qualified syatems and will be designed to withstand the dynamic effects of postulated piping failures in the steam tunnelincluding protection from intemally generated missiles, in the revised configuration, the Division I train FWLCS interlocks will be maintained to prevent actuation of the Division 1 LPCS waterleg pump unless the feedwater MOVs are closed. The Division il FWLCS RHR (B/C) waterleg pump operation will be govemed by plant procedures and instructions.

The proposed design change includes provisions for providing attemate power from Division lll to the feedwater MOVs if Division I power is lost. This reduces the possibility of a FWLCS failure upon concurrent loss of offsite power and loss of Division I power. The attemate power design approach to be taken is similar to that taken at PNPP for station blackout where Division 111 is backup for Division il through procedures. This design change enhances the likelihoed of MOV closure witnin a 1-hour time frame if Division i power is lost. The 1-hour time frame for MOV closure and the establishment of the water seal in the feedwater line penetra5on is consistent with the current licensing base, j

2.4 Prooosed Technical Soecification Chances The licensee's proposal includes the following changes to the TSs:

(1)

A note will to be added to Surveillance Requirement (SR) 3.6.1.3.11 of TS 3.6.1.3,

" Primary Containment Isolation Valves," that excludes the Gedwater check valves from tha hydrostatic test program. This char.ge will relieve the licensee from conducting leak rate tests of the feedwater check vailves with a 1 gpm acceptance criterion.

(2)

TS 5.5.2. " Primary Coolant Sources Outside of Containment,"is a program which provides controls to minimize leakage from those portions of systems outside containment that could contain highly radioactive fluids during a serious transient or accident, will be modified to include two new piping systems. The two new piping systems will be the feedwater system motor operated isolation valves end the reactor water cleanup system retum to feedwater lines. These changes are needed due to the rerouting of the FWLCS piping.

(3)

TS 5.5.12, " Primary Containment Leakage Rate Testing Program," will be modified to state that the containment isolation check valves in the feedwater penetrafons will be tested pursuant to the Inservice Testing Program (TS 5.5.6). This change documents that leak rate testing of the feedwater check valves will be performed. However, the acceptance cri'.erion for these tests will be relaxed.

e 1

i.

3.0 EVALUATION The staffs review focused on the following five areas:

(1)

Imomet on Containment isoletion Provisions - By rerouting the FWLCS piping from the volume between the ' isolation valves to the stem of the MOV, leak testing requirements pursuant to Appendix J to 10 CFR Part 50 may change due to changes in the test j

medium. A determination must be made regarding the adequacy of the licensee's leak rate testing program.

(2)

Imoact on Existino Pioe Break Analvsis - A determination on whether the proposed changes to the feedwater isolation provisions impact any of the existing pipe break analysis.

(3)

Feedwater Leakaoe Control System Reliability - The proposed design change is not single failure proof and the mechanical failure of the MOV to close on demand could compromise containment isolation. The licensee's risk-informed discussion must support the proposed design and licensing changes.

(4)

Electrical Interface for Proposed Attemate Power Suoolv - The MOVs of both feedwater trains are powered from Division I Power. The proposal enhances the reliability of electrical power by introducing an attemate power supply fed from Division Ill. The electrical interface must be accomplished in an acceptable manner.

(5)

Evaluation of Manual Operator Actions - Similar to the existing licensing basis, operator actions are relied upon to initiate the FWLCS. An evaluation of the new operator actions for the proposed design chen;y,s must be found acceptable.

3.1 Imoact on Containment Isolation Provisions Leak Rate Testing of the Feedwater Check Valves

' The proposed testing change is based on design and licensing basis changes proposed for implementation to improve functioning of the FWLCS. Generic Letter (GL) 89-04, " Guidance on Developing Acceptable inservice Testing Programs," noted that teste need to be performed on check valves that perform a safety function in the closed position to prevent reverse fbw as stated in Position 3 of Attachment 1 to GL 89-04, "Back Flow Testing of Check Valves."

Category C tests on such " safety function check valves" were described as needing to prove that the disc closes promptly on its seat on cessation or reversal of flow. As stated in the GL, verification that a Category C valve is in the closed position can be done by visual observation, by an electrical signal indicated by a position-indicating device, by observation of appropriate pressure indication in the system, by leak testing, or by other positive means. Main feedwater header check valves were listed in GL 89-04 as an example of ASME code class check valves that are frequently not tested.

a

! The licensee proposes that the feedwater containment isolation check valves be Category C tested for their safety function at an appropriate frequency as determined in the Inesrvice Testing Program. Yo be consistent with GL 89-04, as discussed in NUREG-1482, " Guidelines a

for Inservice Testing at Nuclear Power Plants,' and as addressed in Supplement 1 to GL 89-04, the test interval for check valves verified closed by leak testing may be extended to the refueling l

outage. Therefore, the Inservice Testing Program test interval is consistent with the current Appendix J test interval. Testing will meet the " exercised closed" test and the " exercised open" l

test. The " exercised closed" test will require a hydrostatic (water) leak rate test, with an j

acceptance criterion of s200 gallons per minute (gpm) per feedwater penetration, when tested at 21.1 P,. This test is to be identified in TS 5.5, " Programs and Manuals," section 5.5.12, " Primary Containment Leakage Rate Testing Program," by reference to the Inservice Testing Program (TS 5.5.6) for the containment isolation check valves in the feedwater penetrations.

The limit of s200 gpm per feedwater penetration, when tested at 21.1 P., is used as the method to test for proper check valve closure (Category C

• exercised closed") and will also ensure no "significant leakage" (Category A leak testing). Hydrostatic testing is accepteble to the staff since that is the medium expected to be acting on these check valves when they are performing j

their safety function to prevent a feedwater line break outside containment from becoming an uncontrolled LOCA. A specific leak rate limit, established by the licensee to be s200 gpm per l

feedwater penetration, is consistent with the Appendix J acceptance criterion (Option A, lli.C.3(a)) that the fluid leakage rates do not exceed those specified in the technical i

specifications or associated bases. In addition, the Inservice Testing Program test ensures no

• significant leakage" for the feedwater line break outside containment and therefore the staff concludes that the Appendix J acceptance criterion (Option A, Ill.C.3(b)) concoming an j

adequate valve seal-water inventory is also met, in this case, the reactor coolant makeup is sufficient to maintain the sealing function for at least i hour, by maintaining the feedwater line full, at which time credit for remote-manual operator closure of the MOV has been previously accepted by the staff. For this case, the feedwater break outside containment, the specific

)

requirement for a 30-day inventory is not considered to be necessary as the accident may be terminated by closure of the MOV within i hour.

While the licensee's acceptance criterion for the feedwater check valves will be s200 gpm por 1

feedwater penetration, it should be noted that the current configuration of the PNPP feedwater penetration line and the available taps into the piping limit the range of the leak rate tests that i

can be performed. Based on discussions with the licensee, it was determined that the existing taps in the feedwater line cannot pass more than 19 gpm at the expected test pressurs of 21.1 i

P,. Therafore, for the upcoming refueling outage, the acceptable leak rate for which no check valve inspection or refurbishment is needed will be less than 19 gpm for each check valve. The licensee is considering changas to the feedwater penetration line to allow for tests at higher leak

)

rates in the future, or to continue to pursue the visual inspection option by developing a means I

- to ensure that the leakage is within the 200 gpm allowable.

As described above, the staff finds the proposed leak rate testing of the feedwater check valves acceptable because it is consistent with the requirements of Appendix J to 10 CFR Part 50.

' Isolation Provisions for Branch Unes into the Feedwater Pipe Once the FWLCS is rerouted to the stem and bonnets of the MOVs in lieu of the feedwater l

pip lng volume, branch lines off of the feedwater line need a different licensing besis for leakage mitigation (See Figure 2). These iines no longer have a water seal since the FWLCS will no longer be used to fill the feedwater piping between the isolation valves.

The RHR branch line off of the feedwater line will be treated as a closed system outside of containment similar to the lines discussed in Note 4 to USAR Table 6.2-33 and Note 7 to Table 6.2-40. These notes explaln why leakage in these lines is not considered to be bypass leakage.

Leakage from the systems listed in USAR Notes 4 and 7 as " closed" are controlled by the Primary Coolant Sources Outside Containment program. A safety-related globe valve (1E12-F053) in this branch line will be treated as a high integrity containment isolation valve, similar to the feedwater MOVs. The 1E12-F053 valves will be added to the containment isolation valve listings. These valves meet the qualifications of a containment isolation valve. An air test will be performed on 1E12-F053 and the air leakage will be added into the Type C totals and limited by 0.60 L, in accordance with Appendix J to 10 CFR Part 50. Also, the leakage from the F053 valves will be added into the Type A integrated leak rate test (ILRT) since the feedwater penetrations will not be drained during the ILRT. This RHR branch pathway will consist of an air leak rate tested containment isolation valve and a closed system outside of containment. In i

addition, a high-to-low-pressure interface water test is performed on the 1E12-F053 globe valve and the check valve inboard of the F053 (1E12-F050)in accordance with ASME Section XI.

These valves are tested to water leakage limits of s5 gpm.

A reactor water cleanup (RWCU) branch line also exists. This line retums the filtered RWCU water to the reactor vessel through the feedwater lines. The piping " outboard" of the RWCU branch line check valve (1G33-FO52) leads directly back to containment penetration, and is ASME Code Class 2, Seismic Category 1, protected from pipe whip, missiles and jet forces, and analyzed for " break exclusion." This closed system outside containment contains only mechanical joints, including the packing on the outboard containment isolation valve (1G33-F039). This outboard valve, including the stem and bonnet, is already part of the air leak rate test program. The remainder of the RWCU line between the feedwater line and the containment penetration will be added to the TS 5.5.2 Primary Coolant Sources Outside Containment program with a specific leskage acceptance limit of zero (0) leakage when tested at RWCU operating pressures (>1,000 psig). Zero water leakage outside the piping when operating at over 1,000 psig ensures that there will be no air leakage from those mechanical joints at P. (7.8 psig for PNPP) and the RWCU branch line check valves (1G33-F052) are not added into the leak rate testing program. This approach meets Branch Technical Position CSB 6-3

" Determination of Bypass Leakage Paths in Dual Containment Plants," Item B.g. Item B.g spechies the criteria for when a closed system may be used as a leakage boundaty to preclude bypass leakage. This approach is also confirmed by PNPP leak test program results where joints that showed water leakage at full system operating pressures did not exhibit measurable airleakage when tested at P.

}

_ g The piping of each FWLCS subsystem, which will connect to the bonnets and seats of the MOVs, currently contains two existing isolation valves. These valves receive a high-to-low pressure interface water test since they connect back to the RHR/LPCS waterleg pumps. These tests will continue to be performed.

As described above, the staff has concluded that containment isolation valves in branch lines leading to the feedwater piping will be appropriately treated through Appendix J to 10 CFR Part 50 and the licensee's Primary Coolant Sources Outside Containment program. Therefore, the staff concludes that the containment isolation provisions for the branch lines are acceptable.

3.2 Imonet on Existing Pios Break Analysis FeedwaterLine Break Outside Containment For a feedwater line break outside containment, there will be a 1,000 psid pressure acting to close the feedwater check valves to prevent significant reverse flow through the line. The analysis of this accident is presented in Section 15.6.6, "Feedwater Line Break - Outside Containment," of the PNPP USAR. Closure of the feedwater check valves is assumed to occur shortly after the postulated break and 1.454x10' Ibm of condensate comprise the inventory used i

for the radiological consequences analysis. The resulting doses are calculated to be well within 10 CFR Part 100 guidelines and are bounded by the doses resulting from either the main steam line break outside containment or the feeciwater line break inside containment.

At PNPP, the current FWLCS leakage test performed at 21.1 P is used by the licensee to demonstrate that closure of the check valves at reactor coolant pressure (1,000 psi) would l

occur, consistent with the USAR 15.6.6 analysis for a feedwater line break outside ca.kainment.

Hydrostatic leak rate testing at low pressures (i.e.,1.1P.) will continue to be used at PNPP to j

demonstrate proper closure. A sensitivity study was performed by the licensee to determine the I

amount of leakage from the feedwater penetrations that would result in consequences similar to the limiting main steam line break (MSLB) outside containtnent. It was determined that 200 gpm per line (400 gpm total) leakage for 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> would have to be exceeded for the consequences to exceed the current value in USAR Table 15.6-11. The results of this study will be included in USAR Section 15.6.6.5.2.4, " Sensitivity Analysis," as identified in the licensee's letter dated January 6,1999. The " exercised closed" test, a hydrostatic (water) leak rate test, with an j

acceptance criterion of s200 gpm per feedwater penetration when tested at 21.1 P., will verify i

proper closure of these valves to prevent significant lernkage of this order of magnitude.

i Therefore, based upon the above information, the staff concludes that the proposed changes conceming the FWLCC and the leak testing of the feedwater penetration do not affec' the licensing basis for the feedwater line break outside containment.

Feedwater Line Break (LOCA) Inside Containment For a feedwater line LOCA inside containment, the operator trst verifies feedwater uncvailability through low feedwater pressure (approximately 30 psig), then closes the outboard MOVs with i

keylock switches, and opens the motor operated FWLCS valves from the control room. The current licensing basis assumes that this operator action will take place within the first m

.m.._

e 9

j 20 minutes following an accident. Sealing water is provided from the suppression pool via the residual heat removal (RHR) and the low-pressure core spray (LPCS) waterleg pump (s). Since the source of sealing water is the suppression pool, a 30-day water supply is assured. Who')

the FWLCS is initiated following a LOCA, there should be no demand for keep-fill water in th e RHR and LPCS systems since these systems will be operating. Therefore, the waterleg putnps should be totally dedicated to provide sealing water to the FWLCS.

Operator actions to recognize feedwater unavailability and initiate the FWLCS remain unchanged under the proposed design. A water seal in the feedwater piping (i.e., the double disc of the gate valves) will be established and a 30-day supply from the suppression poci will still be available. Under the proposed design changes, the time necessary for the FWLCS to establish the water sealis approximately g minutes as opposed to the previous 44 minutes thus allowing additional time for control room operators to take action. Therefore, the staff concludes that the propossd changes concoming the FWLCS do not affect the licensing basis for the

]

feedwater line break inside containment.

3.3 Feedwater Leakaae Control System Reliability The licensee discussed the risk impact of the proposed change and provided adequate information for comparing the proposed change in risk to acceptance guidelines consistent with the Commission's Safety Goal Policy Statement, as documented in Regulatory Guide (RG) 1.174 entitled, "An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis." An acceptable approach to risk-informed decision making is to show that the proposed change to the licensing basis meets several key principles (RG 1.174). One of these principles is to show that the proposed change results in an increase in risk, in terms of core damage frequency (CDF) and large early release frequency (LERF), which is small and consistent with the Commission's Safety Goal Policy Statement. Information submitted by the licensee indicated that the plant CDF would not change and the plant LERF would actually decrease once the proposed modification is implemented.

The proposed feedwater line isolation change does not have any impact on the plant's CDF because it is related to containment isolation followag core damage (Level 2 PRA). Since PNPP has not performed a Level 2 PRA, the results of the Level 1 PRA were used in conjunction with a reliability study (which compares the reliability of the current feedwater line isolation des;gn to the reliability of the proposed modification) to show that the proposed modification would most likely decrease the already low contribution to the plant LERF associated with a 23dwater line isolation failure. Even though the plant CDF is not affected, a Level 1 CDF discussion is relavant in determining the overall acceptability of the proposed modification because the feedwater line penetration does not need to be sealed by the FWLCS unless core damage has occurred. Furthermore, the feedwater line penetration will most likely be sealed even without crediting the FWLCS if the core damage scenario does not involve a feedwater line break at a low elevation of the system inside containment but is associated with a pressurized vessel providing a strong seating force on the check valves in the feedwater line.

Based on the PRA submitted by the licensee as part of its individual plant examination (IPE) and the current 'living" PRA, the CDF from intemal events for the PNPP is less than 2x104 per year.

e an-

+ -

-v

.,w...

. ~. -

,---vu:

- n -

-e-o s.,

'I 10 -

l i

_ This CDF is dominated by sequences which do not involve a feedwater line break and are

{

associated with a pressurized vessel at the time the feedwater line stops feeding the vessel (for i

example, sequences initiated by various transients, ATWS events, loss of offsite power and station blackout). This pressure provides a strong seating force on the check valveu in the line at the beginning of the event, which is how the valves are designed to seal well. Many of these events would also provide the water seal on the feedwater penetration since the feedwater line l

would not be broken, and water would remain from the initial injection feedwater or the refloed j

water.

l l

The only core damage scenarios requiring successful operation of the FWLCS to seal the l

feedwater penetration lines involve a feedwater line break at a low eluation inside containment.

l The frequency of such core damage scenarios is a small portion of the CDF from loss of coolant l

accidents (LOCAs) which is s'oout 2x104 per year based on the IPE results and 5x104 por year based on the current "living" PRA results. This shows that the contribution of the FWLCS in preventing large release following a core damage event is very small (the LERF would most l

likely increase by less than 1x104 per year if the FWLCS was assumed to always be l

unavailable). A comparison of the reliability of the current feedwater line isolation design to the l

reliability of the proposed modification showed that the proposed modification would most likely decrease the already low contribution to the plant LERF associated with a feedwater line isolation failure.

Impact of the Proposed Change on the Reliability ?J the Containment Isolation Provision The licenseo submitted information from a reliability study which compared the reliability of the current feedwater line isolation design to the reliabilities of two attemative desigr.s. One of the attemative designs was the proposed modification whereas the second attemative design was j

tile two MOV alternative. Tne two MOV attemative assumed a second MOV gate valve would be installed in series with each of the existing outboard gate valves and each division of FWLCS would also be routed to each division's respective MOV. Although this modification is " single

]

failure proof" with respect to active component failures, manual action is still required, and j

preferred, to close the MOVs and initiate the FWLCS. The time needed to pump in the water sealis less than 9 minutes (as for the proposed design), allowing a much longer time for

)

operator diagnosis and response. The staff review of the manual operator actions is provided below.

l PNPP assessed the conditional probability of failure to provide the required water seal in both feedwater lines including the wrrent design as well as the two attematives for two cases: one j

assumed no loss of offsite power (LOOP), the other assumed a LOOP. In comparing these probabilities, it was conservatively assumed that the current design worked as assumed in the licensing basis (i.e., the check valves close and estab!!sh a water seal).

]

Conditional probability of failure of feedwater leakage control FWLCS design Without LOOP With LOOP assumed Total HEP

• Total HEP

• Current design 0.267 0.26 0.28 0.260 Proposed design 0.0419 0.036 0.0689 0.067 Two MOV attemative 0.0362 0.036 0.0491 0.036

• Human Error Probability The results for the case which assumed availability of offsite power indicate the following:

1.

The relatively low reliability of the current design (i.e., the relatively high failure probability) is mainly due to the relatively high probability of operator failure to close the MOV and initiate the FWLCS (human error probability [ HEP] 0.26). This is due to the relatively short time (20 minutes) available for recognizing the need to isolata the feedwater system and for taking appropriate operator actions.

2.

The fact that the " current design" (assuming it works as assumed in the licensing basis) is " single failure proof" with respect to active component failures is not significant in terms of reliability. This is due to the fact that the human error probability dominates the reliability of the current design (0.26 human error probability versus 6x104 hardware failure probability).

3.

The improved reliability of the " proposed design," as compared to the " current design," is primarily due to the reduced human error probability because of the significarn increase in the time available for operator diagnosis and response (the human error probability tvould decrease from 0.26 to 0.036 while the hardware failure probability would remain essentially the same).

4.

There is no sign

• ant increase in reliability for the "two MOV alternative" as compared to the " proposed desi n" (the failure probability decreases from 0.0419 to 0.0362). This is t

due to the fact that the human error probabil.4y remains essentially unchanged (0.036) while the hardware failure probability of the " proposed design" is already low (about 6x104).

Similar insights were drawn about the reliability of these designs assuming no offsite power is available.

The staff reviewed the reliability study performed by the licensee and found it to be reasonable in addition, a sensitivity study performed by the staff, indicated that the reliability of the

" proposed design" remains comparable to the reliat;ility of the "two MOV alternative" even when the human error probability values are significantly smaller (up to an order of magnitude) than those assessed by the licensee.

,, s. _

.y a

12 -

Conclusions Regarding the Ucensee's Risk and Reliability Assessments The staff reviewed PNPP's submittal which included a discussion of the risk impact of the proposed change as well as information from a reliability study which compared the reliability of the current feedwater line isolation cesign to the raliabilities of two attemative designs, one of which was the proposed modification. The major findings of the staff's review are summarized below:

4 1.

The proposed feedwater line isolation change does not have any impact on the plant's CDF because it is related to containment isolation following core damage (Level 2 PRA).

2. -

A comparison of the reliability of the current feedwater line isolation design to the reliability of the proposed modification shows that the proposed modification would most likely de,rease the already low contribution to the plant LERF associated with a feedwater line water seal failure.

3.

The frequency of core damage scenarios requiring a feedwater line water seal is very small (most likely smaller than 1x104 per year).

4.

All three feedwater line water seal designs require operator action. The values of the human error probability dominate the reliability of all three designs. This implies that none of the three designs is " single failure proof" with respect to human error.

5.

There is no significant difference in reliability between the "two MOV attemative," which is a fully " single failure proof" design with respect to active component failures, and the

" proposed design."

A comparison of the reduction in the conditional failure probability to estabWsh feedwater leakage control (establish a water seal) for the propord design change to a change which would add an additional, independent MOV to each feedwater penetration line shows that the proposed design is comparable to a design which would add a second barrier to containment atmosphere leakage te the environment through the feedwater penetration line. The proposed design accounts for more than 90% of the available reduction in conditional failure probability, based on operator remote-manual closure of the MOVs and initiation of the FWLCS.

The likelihood of establishing a water seal in the feedwater penetration to prevent containment atmosphere leakage from a LOCA inside containment to the environment for the proposed design, based on a single barrier, is comparable to that cf a design which would include a second MOV in the line as a second barrier. The propmd design also improves the likelihood of establishing the water seal, when compared to the current design, as a result of the increased time available for the operator actions needed to start the FWLCS and establish the seal within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />.

..r,-

r--.-

n amn

-w<<

. 4 1

l Therefore, the staff concludes that the licensee's risk-informed discussica is consistent with the Commission's Safety Goal Policy Statement as documented in Regu! story Guide 1.174 and supports the proposed modifications.

l 3.4 Electrical Interface for Proposed Altemate Power Supply The motor operated gate valves in each of the feedwater trains are currently powered from the Division i electrical power supply. In order to provide greater assurance that these MOVs will be available for closure following a LOCA and a total loss of both the normal and emergency Division I electric power supplies, the licensee proposed to install an attemate power supply -

from the Division lli electrical power supply. Operator actions would be relied upon to manually connect the Division lll power supply to the MOVs.

Evaluation of Electrical Connections The licensee proposed to provide this attemate power supply by insta; ling a power cable l

between Division lli (motor control center EF1E1, Compartment V) and Division I (MCC EF1 A07, Compartment XV). The cable will be terminated at the load side of a fusible disconnect switch in the MCC compartments at each end.

These disconnect switches will remain open and the fuses will remain out of their holders until such time as the attemate power supply is needed. The fuses will be stored in the bottom of the l

MCC compartments so as to be accessible as needed. Labels will also be applied to each compartment's door describing the purpose of the compartment and directing that the fuses are not to be removed from the compartment.

Prior to installing the fuses and closing the breakers for the attemate power supply, operators will be required to open MCC EF1 A07's feed breaker it Bus EF-1-A, and all the breakers on i

MCC EF1 A07. This will prevent potential back feed of power to other circuits. Specific plant i

procedures cover how to implement this e'temate power supply. Only the feedwater MOVs will i

be operated using this alternata power supply. The cable and its cc6duit will be seismically qualified and classified as safety-relt ted. The licensee has determined that the additional lead for Division lli under a LOCA condition is within the capabilities of the Division 111 diesel j

generator.

The staff was concerned about the potential of losing both electrical divisions when the Division 111 diesel is being used to close the MOVs and the offsite power was restored thus potentially powering all of Division 1. The licensee states that once it has been determined that the Division i diesel is inoperable, procedures include steps to disconnect all loads upstrearn of the valves in Division 1. After the lines have been disconnected from Division I power, the valves will be connected to Division lli to close. This will provide electricalindependence between Divisions I and Ill. If offsite power is restored during the 68-second period when the MOVs are being closed, it will not cause an adverse impact since the valves have been electrically isolated from Division 1.

L

.o Conclusions Regarding Electrical Connections The staff has evaluated the proposed design and procedure changes as follows:

1.

During power operation, the power cable between Division lil and Division I will not be used. This addresses any concems regarding the potential for losing both divisions if they were tied together.

2.

This special supply of Division 111 power will be limited only to circuits for MOVs B21-F065A/B and then only for approximately 68 seconds.

3.

Altemate power supply (Division lil) will be connected to the MOVs only when there is a complate loss of Division I offsite and onsite power sources. If the offsite power is restored during the 68-second period when the MOVs are being closed, it will not cause an adverse impact since the valves have been electrically isolated from the rest of Division 1.

4.

Existing physical separation and electrical indspendence between Divisions I and lll will be maintained.

5.

Additional loading for Division lli under a LOCA condition is within the capabilities of the Division Ili diesel generator.

The staff concludes that connecting Division ill to Division I can provide greater assurance that the feedwater motor-operated valves will be available for closure in the long term following a LOCA ar#t can be accomplished in an acceptable manner. Therefore, the staff finds the electricalinterface acceptable.

3.5 Evaluation of Manual Ooerator Action 3 The staff used the following guidance on manual operator actions and the time required to perform those actions to complete its evatushn:

1.

Generic Letter (GL) 91-18, "Information to Licensees Regarding Two NRC Inspection Manual Sections on Resolutk,n of Degraded and Nonconforming Conditions and on Operability (1991).*

2.

American Nationa! Standards institute /American Nuclear Society (ANSI /ANS)-58.8,

' Time Response Design Criteria for Safety-Related Operator Actions (1984).*

GL 91-18 states: "The consideration of manual action... must include the ability and timing in getting to the area, training of personnel to accomplish the task, and occupational hazards to be incurred such as radiation, temperature, chemical, sound, or visibility hazards." ANSI /ANS-58.8 provides guidance on estimafng response times for operator actions and allows licensess to use time intervals derived from independent sources, provided they are based on task analyses or empirical data. Based on these guidelines, the NRC staff evaluated the licensee's evaluation of the new operator actions, as detailed below. Operaior actions previous!y approved under the

_: n _.

n-4 b

i L current licensing basis (i.e., closing the MOV and initiating the FWLCS) are still considered acceptable and are not being further addressed.

i Specific Operctor Actions Required l

Operator action would be required to provide a method to use Division ill power to operate the feedwater MOV's (normally powered by Division l} in the event that Division I power is lost following a LOCA. This altemative power supply is provided by the temporary installation of a power cable between Division lll and Division 1. This action was evaluated againat the considerations in ANSI /ANS 58.8-1984, " Time Response Design Criteria for Nuclear Safety Related Operator /setions," to verify that the proposed contingency action can be accomplished.

A draft procedure was prepared by the licensee, developed from several existing procedures to conduct a walkdown.

l The walkdown showed that the time to perform all of the steps necessary to provkie Division lli power to the MOVs and to start the FWLCS without Division I power is approximately 19.5 minutes. This included the time for the operator to travel to the required area, obtain the required too;s, and perform the required actions, plus time for the control room actions to take place. A Shift Supervisor reviewei the results of the walkdown and all the assessed times. The operator action was assumed to begin 30 minutes after the start of the design basis LOCA, based on the guidance in ANSI /ANS 58.8-1984. The operator action was then estimated to take approximately 19.5 minutes to complete. Therefore, the action of initiating FWLCS and establishing a water seal at the MOV can be completed within the current licensing baisis period of 64 minutes, following the occurrence of a design basis LOCA.

Potentially Harsh orInhospitable Environmental Conditions Expected F.nvironmental conditions in the area in which the operator actions will occur are not expected to be inhospitable or harsh. The time considered for the dose evaluation used the time to perform j

the entire evolution, even though some of the actions take place in the control room. The areas accessed are on elevations other than the control room. All areas requiring access are outside l

the Radiologically Restricteo Area. Therefore, there are no components in the travel path containing radioactive materials that would result in radiation levels that would preclude access to the areas required to perform the proposed action.

Ingress / Egress Paths Tal'en by Operators to Perform their Functions The areas required to be accessed to perform this accon are readily accessible. The areas where the actions are being performed all have adequate normal lighting and also have battery

" (ed emergency lighting.

Procedural Guidance for Required Actions Discussions with the licensee indicated that the final plant procedure for the proposed manual operator action will be developed and verified prior to conducting operator training.

d 4

6 Specific Operator naining Necessary to Cany Out Actions including any Operator Qualificatlans Required to Cany Out Actions The licensee stated that operator training on the proposed manual operator action for all operators will be completed prior to the end of the outage in which this design change wlll be installed.

Additional Gupport Personnel and Equipinent Required by the Operator to Cany Out Actions The licensee stated that fuses needed for the proposed manual action will be accessible i

because they are stored in the bottom of the MCC compartments. Labels will also be applied to each compartment's door describinC the purpose of the compartment and directing that the fuses are not to be removed from the compartment.

Description ofInformat;on Required by the ControlRoom Staff to Determine Such Operator Action is Required, including Qualified Instrumentation Used to Diagnose the Situation and to Verify that the Required Action has been Successfully Taken The licensee stated that successful diagnosis of the event could include either high radiation r

alarms or low feedwater pressure indication. The operators have 30 minutes to reach this diagnosis.

The staff concludes that the information discussed above is acceptable because it is consistent with Standard Review Plcn guidance, ANSI /ANS 58.8-1984, " Time Response Design Criteria for Nuclear Safety Related Operator Actions," and Generic Letter 91-18. On the basic of the s,bove information, the licensee has provided assurance that the required operator actions can be performed and therefore, the staff concludes that the licensee's responses related to the newly proposed operator actions, are acceptable.

~

4.0

SUMMARY

The licensee has determined that enhancements to the licensing and design basis of the feedwater isolation provisions are necessary to improve the reliability of the FWLCS to perform its safety-related function following a postulated accident.

The r taff has determined that the proposed changes to the FWLCS to provide a water seal on the MOV seat to eliminate air leakage would perform the same function as the current FWLCS which is to fill a portion of the feedwater piping between the containment isolation valves. While the proposed design relies upon closure of the MOV, the licensee has shown that the reliability of the proposed design is comparable to a design which would include a second MOV in each feedwater penetration line. The proposed design also improves the likelihood of establishing a water seal within 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> as a result of an increase in the time available for the operator to take action.

j in conclusion, the staff finds the proposed changes to the licensing and design basis of the feedwater isolation provisions to be an improvement over the existing design. The staff has

. ~.

,4 7.

.wc-m._,

m

e a a conducted r.n extensive review that focused on the physical modifications and continued compliance with all applicable regulations. As previously discussed, the staff has concluded that the proposed modifications meet the appropriate acceptance criteria with respect to feedwater pipe breaks, containment isolation, and leak rate testing. In addition, the staff reviewed and approved the licensee's risk-informed discussion supporting the proposed modifications, the introduction of an alternate electrical supply for the MOVs, and the additional operator actions.

Therefore, based on this review, the staff concludes that the proposed changes are acceptable.

5.0 STATE CONSULTfG in accordance with the Commission's regulations, the Ohio State official was notified of the proposed issuance of the amendment. The State official had no commsnts.

6.0 ENVIRONMENTAL CONSIDERATION

This amendment changes requirements with respect to installation or use of a facility component located within the restricted area as defined in 10 CFR Part 20 and changes surveillance requirements. The staff has determined that the amendment involves no significant increase in the amounts, and no significant change in the types, of any effluent that may be released offsite and that there is no significant increase in individual or cumulative occupational radiation exposure. The Commission bss previously issued a proposed finding that this amendment involves no significant hazards consideration and there has been no public comment on such finding (63 FR 56262). Accordingly, this amendment meets the eligibility criteria for categorical exclusion set forth in 10 CFR 51.22(c)(g). Pursuant to 10 CFR 51.22(b), no environmental impact statement or environmental assessment need be prepared in connection with the issuance of this amendment.

7.0 CONCLUSION

The staff has concluded, based on the considerations discussed above, that: (1) there is reasonable assurance that the health and safety of the public will not be endangered by operation in the proposed manner, (2) such activities will be conducted in compliance with the Commission'? regulations, and (3) the issuance of this amendment will not be inimical to the common defense and security or to the health and safety of the public.

Principal Contributors:

Edward Throm Nicholas Saltos Clare Goodman NarinderTrehan Douglas Pickett Date: March 26, 1999

'~ : :

^

.. a

'"'"i INBOARD OUTBOARD Mi Feedwaterflow mm s.[ ]x f

s b

l' _

l I

I lI @[ @ Steam

[B21-F032 N27-F559 Tunnel B21-F065 Outboard (Div.1)FWLCS Drywell Containment

,% RWCU Retum Line

=C G33-F052 fnboard (Div. 2) FWLCS

?

~

L l

a f y

\\

RHR Shutdown Cooling Retumline E12-F053 FIGURE 1 - EXISTING FWLCS CONFIGURATION

?N '

lT JNBOARD 0UTBOARD 4

3 h

_i-1 i-i W 1.. w d

i hh y,d, N21=F660 gr.ypsg ygg,yggy CONTAINMENT RWCU cleau 2, greak Esetudent.

ARYFELL retume to containment C33-Fost II

,N,.

t-C..,mi a, ac.,

FIGURE 2 - PROPOSED FWLCS CONFIGURATION

~

w-w

_