Text

Review of Fuel Pool Cooling During Postulated Off-Normal & Accident Events SSES Units 1 & 2.
	ML17157C140
Person / Time
Site:	Susquehanna
Issue date:	08/31/1992
From:	Brinkman K; PENNSYLVANIA POWER & LIGHT CO.
To:	;
Shared Package
ML17157C139	List: ML17157C138; ML17157C140; ML17157C141; ML17157C142; ML18026A248;
References
	NUDOCS 9301050145
	Download: ML17157C140 (137)
	v • d • e

REVIEW OF FUEL POOL COOLING DURING POSTULATED

' OFF-NORMAL AND ACCIDENT EVENTS Susquehanna Steam Electric Station Units l a 2 August l992 Prepared by:

Kevin W. Brfnckman Pro)ect Engineer-Nuclear Systems

'P39105Df45 921127 PDR ADOCK 05000387

Preface This paper presents the findings of a review of the adequacy of spent fuel pool cooling at Susquehanna SES during off-normal and emergency situations. The intent of the review is to determine if plant safety can be maintained. This paper is not a revie~ of the licensing requirements for spent fuel storage at Susquehanna SES.

The design of the spent fuel pool cooling system (FPC) was examined to determine to what extent its operability and integrity would be affected by postulated off-normal or accident conditions. The availability of other means of spent fuel pool cooling was researched, and the function and accessibility of the alternatives was evaluated. The design of the spent fuel pool instrumentation was reviewed to determine if spent fuel pool monitoring could be maintained in an off-normal or accident condition.

The consequences of a boiling fuel pool were considered. It was assumed that sufficient decay heat would exist in the stored spent fuel to produce pool heat-up and eventually a boil-off condition if cooling was not provided.

t No detailed calculations were performed as part of this effort.

Conclusions are drawn from current design and licensing information.

Analysis or re-analysis of issues such as post-LOCA radiation levels, probability and consequences of clad and fuel failure, and equipment qualification could alter some of the conclusions. Based on the research done to compile this report> it is the author's opinion that further evaluation of the concerns discussed in EDR 620020 and this report are warranted to assure spent fuel storage at Susquehana does not reduce plant safety.

TABLE OF CONTENTS Pacae Preface .............-.-.-.-.. ---. ~ .~- ~ ~ ~ ~ ~ ~ . ~

1.0 Introduction .. ..........-.---.--.. .- ~ ~ ~ ~ ~ ~ ~ ~ . ~ ~ ~ ~ ~ ~ ~ 1 2.0 Postulated Events 2 3.0 Safety Evaluation ....................................-.-. 5 4.0 Recommendations ................,........................ 17 5.0 Conclusions lf 6.0 References .............................................. 21

Page 1 1.0 Introduction The purpose of this paper is to provide a review of spent fuel pool cooling capability at Susquehanna SES following a postulated off-normal or accident condition. A series of events was considered to determine whether potential situations exist at Susquehanna SES which could result in fuel pool boiling. These events are discussed in Section 2.0.

Section 3.0 provides a safety evaluation of the postulated events in Section 2.0 and discusses what aspects could be further evaluated to assess the consequences of the respective event.

Potential problems with a loss of fuel pool cooling and subsequent alignment of the RHR fuel pool cooling assist mode are found. Access to the RHR valves necessary to manually align RHR in the fuel pool cooling assist mode would be hindered by an event which caused core degradation and fission product release to the primary coolant. The consequences of using ECCS equipment for RHR fuel pool cooling assist during an accident need to be considered. A PMEA is warranted to determine if ECCS can provide fuel pool cooling and long term post-LOCA containment cooling with postulated single failures and breaks. Operator guidance for implementing RHR fuel pool cooling assist post-LOCA should be developed.

If fuel pool boiling were to occur during a LOCA, the energy and moisture released to the reactor building would create a severe environment for which much of the safety grade equipment may not be qualified. The entrained water could overload the SGTS moisture removal equipment and reduce the iodine removal efficiency of the charcoal beds.

The effect of a loss of fuel pool cooling during refueling with a full core offload is evaluated. The lack of operable RHR equipment is the concern in this situation. A fuel pool boiling event along with a LOCA on the other unit would put the plant in an unanalyzed condition.

Page 2 2.0 Postulated Events In this section off-normal or accident events are discussed in which the normal spent fuel pool cooling system (FPC) is lost and alternate means are required to remove the decay heat from the fuel pool. The events chosen are considered to provide an envelope of situations where FPC is lost and the plant is in a degraded condition.

Events 1 to 5 are p'resented in order of what is considered highest to lowest probability of occurrence< although frequencies have not been researched as part of this effort. The last event considered is a loss of FPC with a full core offload to the spent fuel pool. It is the only event evaluated which is not concerned 'with decay heat in the reactor vessel.

2.1 Event41 : Loss of Offsite Power LOOP Event 41 postulates a dual unit LOOP, which causes a loss of power to the fuel pool cooling system (FPC) pumps and loss of service water for cooling of the FPC heat exchangers. The event assumes both reactors are stabilized and no fuel failure occurs.

2.2 Event 2 : LOCA Without Fuel Failure Event 42 postulates a LOCA on one unit. It is assumed that the emergency core cooling systems perform to provide sufficient core cooling to maintain fuel and clad integrity. No seismic event is postulated, however, hydrodynamic loads due to steam discharge to the suppression pool must be considered. Since portions of the FPC are not seismically designed it is indeterminate whether the FPC piping and equipment will be damaged by the hydrodynamic loads. It is assumed that FPC is lost for the duration of the accident. The non-1E fuel pool level and temperature indication is also assumed" lost due to structural damage resulting from the hydrodynamic loads. If still functional, the fuel pool cooling pumps should be shut down by the non-essential load shed 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months after the LOCA, if implemented per EP-IP-055 . To prevent fuel pool boiling, RHR fuel pool cooling assist must be initiated.

Page 3 2.3 Event¹3 : LOCA<<LOOP Without Fuel Failure Event ¹3 postulates a LOCA on one unit coincident with a LOOP.

It is assumed that the emergency core cooling systems maintain adequate core cooling to prevent fuel or clad damage. FPC is lost for the duration of the accident due to the mechanisms discussed in Events Nl and ¹2. It"is assumed that hydrodynamic loads will disable the fuel pool instrumentation'.

2.4 Event¹4 : LOCA With Fuel Failure Event ¹4 postulates a LOCA on one unit with fuel failure as described in the accident analysis of FSAR Chapter 15.6.5, which specifies release of radioactive material in accordance with the assumptions of Regulatory Guide 1.3 . Radiation dose rates within the reactor building for this scenario have 'been calculated in FSAR Chapter 18.1.20 per the guidelines of NUREG-0737 , which applies the assumptions of Regulatory Guide 1.3 to specify the release of radioactive materials to the primary coolant.

e FPC is lost immediately due to the service water LOCA load shed and it is assumed that the hydrodynamic loads during the LOCA blowdown cause the non-seismmic portions of FPC to fail. FPC is lost for the duration of the accident and RHR fuel pool cooling assist must be placed in service to remove the 'decay heat from the spent fuel pool.

2.5 Event 5 : LOCA-LOOP With Fuel Failure Event ¹5 postulates the LOCA discussed in Event ¹4 coincident with a LOOP. FPC is lost for, the duration of the accident due to the mech-anisms discussed in Events ¹1 and 2. It is assumed that hydrodynamic loads will disable the fuel pool instrumentation.

Page 4 2.6 Event¹6 : Loss of FPC With Full Core Offload This event postulates a loss of FPC during refueling when the entire core is offloaded. The emergency heat load as defined in FSAR Chapter 9.1.3 is considered, as well as a lesser heat load which would not normally require RHR fuel pool cooling assist.

Page 5 3.0 Safety Evaluation 3.1 Event01 : Loss of Offsite Power LOOP 3.1.1 Evaluation In the event of a LOOP, the spent fuel pool cooling system (FPC) is disabled due to loss of power to the FPC pumps and loss of service water for cooling of the FPC heat exchangers. Assuming the reactor is stabilized and no fuel failure occurs spent fuel pool cooling can be provided using the RHR fuel pool cooling assist lineup. Operations has at least 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months with the design heat load on the fuel pool to initiate RHR fuel pool cooling assist before fuel pool boiling is calculated to occur. Since no fuel failure results from the LOOP+ access is available to the reactor building and the manual operations required to establish RHR fuel pool cooling assist can be performed without exposing operators to doses above station limits. RHR fuel pool cooling assist has sufficient 1

cooling capacity to remove the pro)ected spent fuel pool emergency heat load. as defined in FSAR Chapter 9.1.

Per the action statement of Tech Spec 3.8.1.1, the reactor must be in cold shutdown within 108 hours0.00125 days 0.03 hours 1.785714e-4 weeks 4.1094e-5 months after the LOOP if at least one source of offsite power to the Class lE distribution system in not operable. The ECCS design contains sufficient redundancy and flexibility that the reactor could be brought to cold shutdown while operating RHR fuel pool cooling assist.

Fuel pool temperature and level indication are provided on panel OC211 located on the refueling floor. Panel OC211 is powered off of the diesel generators and will have power available during the LOOP.

The fuel pool instrumentation is non-1E and is not environmentally qualified. While the LOOP. environment is not expected'to be severe on the refueling floor and the instrumentation should remain functional,

Page 6 this cannot be guaranteed. A group alarm for OC211 exists in the control room, but the history of trouble with this alarm is extensive.

Operator access to the refueling floor could be limited due to airborne radioactivity from evaporation off of the fuel pool surface or elevated air temperatures due to the loss of HVAC. Therefore, the operators should be-given some guidance on the amount of time they have after a loss of fuel pool cooling to initiate RHR fuel pool cooling without having to rely on the OC211 instrumentation. ON-135-001 does inform the operators that they have at least 25 hours2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months until the fuel pool boils after loss of FPC. This time could be extended considerably by a cycle specific analysis.

Since the fuel pool will not boil for at least 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months , electrical loads required to support RHR fuel pool cooling assist will not have any impact on short term recovery from the LOOP.

3 1 2 ~Sllss&l' LOOP does not present a significant challenge to providing fuel pool cooling. RHR fuel pool cooling assist can be aligned and has ample heat removal capacity to handle the spent fuel pool heat load. However, adequate indication of fuel pool conditions may not be available to the operators and it would be advantageous to provide additional guidance to the operators on how long they have to establish RHR fuel pool cooling assist after a loss of spent fuel pool cooling.

3.2 Eventf2 : LOCA Without Fuel Failure 3.2.1 Evaluation If the recirculation discharge line break is postulated, one loop of LPCI is lost. With the other loop of LPCI functional a substantial

Page 7 makeup source to the reactor vessel is maintained and extended fuel uncovery and cladding damage can be avoided. To prevent fuel pool boiling RHR fuel pool cooling assist must be initiated. Since no fuel failure is postulated, access to RHR valves necessary to align RHR fuel pool cooling assist is available and can be accomplished without large dose rate exposures to the operator.

With the design'eat load on the fuel pool< Reference 2 calculates that operators have approximately 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months to provide fuel pool cooling to prevent fuel pool boiling. RHR fuel pool cooling assist is aligned with RHR loop A suction off of the fuel pool skimmer surge tank and discharge through the loop A RHR heat exchanger back to the fuel pool.

If the postulated break is in the reactor recircuation loop B discharge line, RHR loop A LPCI is required for vessel reflood. Therefore, to align RHR fuel pool cooling assist, RHR loop A discharge to the vessel would need to be terminated after 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months . PSAR Chapter 6.2 analyses for long term cooling after the containment design basis LOCA assumes at least one LPCI pump is available for vessel makeup or containment spray.

A potential problem exists here if the break in recirculation loop B disables loop B LPCI and a single failure of the loop B containment spray valve is postulated. With this scenario< loop A RHR flow may be needed for containment pressure/temperature control and would not be available to provide a closed loop of RHR fuel pool cooling assist.

If the hydrodynamic loads resulting from steam blowdown to the suppression pool do not damage the fuel pool temperature and level instrumentation, power is available from the diesel generators to keep the equipment operable. However, as discussed in Section 3.1, the operation of this instrumentation cannot be guaranteed. Therefore, it it is assumed that fuel pool conditions will be indeterminate during a a LOCA unless access to the refueling floor is possible.

Page 8 3.2.2 ~Susmar Without fuel failure resulting from the MCA, access to the reactor building and refueling floor is available and the necessary operations required to monitor fuel pool status and- initiate RHR fuel pool cooling assist can be performed. However< the availability of fuel pool temperature and level indication cannot be guaranteed and alternate means of fuel pool J monitoring or specific guidance on time to initiate RHR fuel pool cooling assist is recommended. Consideration must be given to the ability of ECCS'to provide long term cooling to both the containment and fuel pool with postulated pipe breaks and single failures. A more detailed PMEA is warranted for this event.

3.3 Event%3 : LOCA-LOOP Without Fuel Failure 3.3.1 Evaluation The consequences of this LOCA-MOP on fuel pool cooling are no different than those discussed for a MCA without fuel failure. Fuel pool cooling is not needed in the short~term when recovery from the LOCA is taking place. By the time the fuel pool approaches boiling

(> 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months per Reference 2), the MCA recovery should be in the long term cooling phase. RHR pumps will be operating with RHR heat exchangers valved in to remove decay heat from the containment. Zt is assumed that sufficient ECCS was available to maintain core cooling and preclude fuel failure. Therefore, the RHR equipment area will be accessible for aligning the RHR fuel pool cooling valves, and the fuel pool can be maintained in a safe condition. This evaluation is again contingent on sufficient ECCS equipment remaining available to provide both long term containment and fuel pool cooling.

The availability of fuel pool level and temperature indication is the same as discussed in Section 3.1.

Page 9 3.4 Event44 : LOCA With Fuel Failure 0 The postulated fuel failure for this event changes the situation considerably. In the previous events without fuel failure< the reactor building remains accessible. The occurrence and extent of fuel failure are the major factors in determining the options available for responding to a loss of fuel pool cooling. This event is evaluated below for several different s'ituations.

3.4.1 NUREG-0737 Postulated Fuel Failure For the postulated LOCA scenario, FPC is lost and RHR fuel pool cooling must be established to avoid fuel pool boiling. To align RHR for fuel pool cooling five valves must be manually opened (151060, 151070, 153021< and 153070A,B for Unit 1, or their counterparts for a Unit 2 event). Three of these valves are in room I-514 and two are in room I-202.

Access to the reactor building is restricted by the post-LOCA radiation dose rates which are a dependent upon the extent of postulated core degradation, system operations which transport the radioactivity throughout the reactor building, and shielding in place to isolate the sources.

A Susquehanna specific plant shielding analysis was performed in response to one of the action plan requirements of NUREG-0737 NUREG-0737 requires the use of Regulatory Guide 1.3 assumptions for release of fission products from the fuel. The shielding analysis, presented in FSAR Chapter 18.1.20, calculates radiation levels in the the reactor building and provides the results as radiation zone maps one hour after the LOCA in Figures 18.1-2 to 18.1-8. Per this analysis room I-514 is a radiation zone V (< 50R/hr) and I-202 is a radiation zone VIII (>5000R/hr) with the postulated fuel failure per Regulatory Guide 1.3 . These radiation levels make the valves inaccessible until sufficient decay has occurred. FSAR Figures 18.1-9 and 18.1-10

Page 10 provide decay factors as a function of time which can be applied to the dose rates calculated in the shielding analysis. These curves show that an order of magnitude decrease in radiation dose rates can be expected 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months after the LOCA< by which time RHR fuel pool cooling assist must be established to avoid fuel pool boiling assuming the design heat load. This reduction in dose rates makes zoom I-514 accessible for the valve operations required within the 5 Rem per activity guideline of 10CFR50'ppendix A GDC 19. However, an order of magnitude reduction in the room I-202 calcuated dose rate will not allow access to the RHR valves. Therefore, with fuel failure postulated per the guidelines of NUREG-0737 , the necessary manual valve alignments cannot be performed to establish RHR fuel pool cooling assist and if a sufficient decay heat load exists fuel pool boiling will occur.

A further examination of room I-202 shows that it houses much of the RHR piping running from the RHR Loop A pumps to the containment.

The DBA LOCA which is analyzed for peak clad temperature is the recirculation loop discharge line break with a failure of the LPCZ injection valve on the other loop. With the loss of both LPCZ loops

. core uncovery, results for a sufficient time to expect clad damage and release of fission products to the primary coolant. Transport of these fission products throughout the reactor building creates the radiation sources in the rooms housing the RHR valves. Since room I-202 houses the RHR Loop A pump discharge pipe< the water drawn from the suppression pool by Loop A will travel through this zoom. Fission products within the pipe will sustain the radiation source in the room. Valves 151060 and 151070 are in close proximity to the RHR discharge pipe and the operator would have no permanent shielding if required to manually open the valves. Therefore, without a reduced source term in room Z-202, access to RHR valves necessary to establish RHR fuel pool cooling is not possible in time to prevent fuel pool boiling at the design fuel pool heat load.

Page 11 3.4.2 Puel Pool Boilin and Makeu Without FPC or RHR fuel pool cooling assist available, fuel pool boiling will occur at a time dependent on the decay heat load in the spent fuel pool. Emergency makeup water is available by design from the ESW system by performing manual valve alignments in rooms I-105 and Z-514.

These rooms are reported as radiation zones VIIZ and V respectively in the PSAR Chapter 18 anaiysis discussed above and the same accessibility problem will exist with room I-105 as found with I-202. A better alternative for fuel pool'makeup with a degraded coze condition is via a firehose on the refueling floor, if it is established prior to the fuel pool reaching boiling temperature. This alternative could be implemented with minimal dose to the operator. However, the post-LOCA condition of the non~ fire protection system is uncertain and access to the ESW valves may ultimately be required for emergemcy makeup to the fuel pool.

Appendix 9A of the FSAR analyzes the consequences of a loss of spent fuel pool cooling to determine the time to boiling, capability to add makeup water, and off site releases in the event of fuel pool boiling. A problem with responding to.a loss of spent. fuel pool cooling with makeup only is that the decay heat generated by the spent fuel is ultimately transferred to Zone ZII and the LOCA unit's reactor building.

The design spent fuel pool heat load is 12.6E+6 Btu/hr which is two to three times the calculated post-LOCA heat load on the LOCA unit plus Zone IZI. The extreme amount of energy and moisture deposited into the reactor building from a-boiling fuel pool would undoubtedly create an environment which could )eporadixe equipment operability and make the reactor building inaccessible. Moisture carryover to the SGTS would reduce the iodine removal efficiency of the charcoal beds if the entrained water overloaded the moi.sture removal equipment in the SGTS trains. This would effect offsite releases and the assumed efficiencies used in offsite dose calculations.

Page 12 3.4.3 ECCS Desi n Basis IOCA The ECCS design basis LOCA considers the break in the primary coolant boundary along with the worst case single failure which results in the highest fuel cladding temperature. For Susquehanna SES, the ECCS DBA LOCA is the recirculation discharge line break with a failure of the LPCI injection valve on the other loop. Loss of both LPCI loops results in core uncovery for a sufficient time to expect clad damage and release of fission products to the primary coolant.

The LOCA analysis performed to determine the peak clad temperature and clad oxidation does not predict the fission product release since the 10CFR100 offsite dose analysis must be performed using the releases specified in Regulatory Guide 1.3. However, per NUREG-1228 6 , the fuel fission product release from the DBA LOCA would be substantially less than that specified in Regulatory Guide 1.3 . For example, if fuel clad temperature is maintained between 1300F and 2100F, NUREG-1228 6 assumes 2 percent of. the iodine in the core is released. This value is based on estimated gap release values derived in WASH-1400 for 7

situations of core degradation with cladding failure.'he position of NUREG-0737 is that 50 percent of the core iodine is transfered to the primary coolant.

If the postulated core damage is limited to cladding failure and the guidelines of NUREG-1228 are applied, the radiation levels within the reactor building due to activity within the primary coolant could be reduced by an order of magnitude. However, even at the reduced radiation levels it is uncertain whether access to the RHR valves is possible while remaining within the post-accident operation dose rate criteria of NUREG-0737 . Analysis of the source term expected from clad failure and the decay factor would have to be performed to determine radiation levels for this scenario. Conclusions on the accessibility of RHR valves could then be drawn.

Page 13 Again, consideration must be given to the ability to provide long term cooling of the containment. If RHR loop A could be placed into RHR fuel pool cooling assist mode< RHR loop B would be required for reactor/containment decay heat removal. Since the single failure has already been postulated for the LPCI injection valve, RHR loop B would be available for containment spray. Case C of the long term cooling analysis in -PSAR section 6.2.1.1.3.3.1.6 bounds this scenario.

3,4 4 ~SU@El&l If a LOCA is postulated which produces core degradation and fission product release, the valves required to be manually opened to align RHR in the fuel pool cooling mode and ESW for fuel pool makeup will be inaccessible unless the extent of core damage is small. Analysis does not currently exist to quantify what "small" is. The plant shielding analysis in PSAR Chapter 18 predicts LOCA radiation levels which are too high to permit access to the RHR valves in time to prevent fuel pool boiling with the spent fuel pool design heat load. Although, fuel pool boiling is analyzed in Appendix 9A of the FSAR for offsite dose releases, analysis of the consequences on reactor safety could not be found.

It is not sufficient to merely makeup the water evaporated from the pool surface. This mode of operation uses the reactor building and refueling floor as a heat sink for the spent fuel decay heat and reservoir for the condensate. In doing so, an environment is imposed on the plant for which it is not analyzed.

It is not certain if realignment of RHR loop A to fuel pool cooling mode will leave sufficient ECCS equipment to provide long term containment heat removal as analyzed in FSAR Chapter 6.2. . Further of this issue is warranted'i: ~ " ~ ~

" 'valuation

Page 14 3.5 Event45 : LOCA-LOOP With Fuel Failure The LOCA-LOOp with fuel failure causes the same concerns as the LOCA with fuel failure. The radiation levels currently analyzed in the rooms where operator access is required to manually align RHR fuel pool cooling assist are too high to allow the required action.

Since operators'have at least 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months to establish RHR fuel pool cooling assist before pool boiling, short term recovery from the LOCA is not an issud. However, the effect of removing a loop of RHR from the long term LOCA recovery and the additional impact of the LOOP on equipment availability should be evaluated.

3.6 Event%6 : Loss of FPC With Pull Core Offload 3.6.1 Emer enc Heat Load If the full core offload produces the emergency heat load (EHL) on the spent fuel pool as described in PSAR Chapter 9.1> then RHR

.fuel .pool cooling is required to remove the heat load on the pool.

In this situation, RHR would be aligned in the fuel pool cooling configuration and safety is maintained. Since all of the fuel is offloaded from the reactor vessel, RHR is not required for core cooling and use of the RHR equipment for fuel pool cooling is the priority.

3.6.2 Maximum Normal Heat Load If the full core is offloaded for refueling with decay heat loads at current levels< FPC has sufficient capacity to adequately cool the fuel pool. Techncical Specifications for refueling operations (3.9.11) require RHR shutdown cooling be operable when irradiated fuel is in

~ the reactor vessel. With a full core offload, applicability of this requirement is removed and both loops of RHR may be inoperable. If

Page 15 FPC is lost, operators have at least 19 hours2.199074e-4 days 0.00528 hours 3.141534e-5 weeks 7.2295e-6 months to get a loop of RHR (one pump and heat exchanger) into service for fuel pool cooling.

0 Nineteen hours is the time to boil with the maximum normal heat load2.

The emergency heat load need not be considered in this scenario since a lesser heat load which normal FPC could cool was postulated.

Off normal procedure ON-135-001 provides operators with actions for dealing with a loss of spent fuel pool cooling. They are first instructed to operate RHR in the fuel pool cooling assist mode if available. Zf all means of cooling are lost, the procedure instructs the operators to "ALLOW water in Fuel Pool to boil", and provides a caution that evacuation of the refuel floor may become necessary due to increasing radiation levels. A note is also included which informs the operator that boiling should not occur before 25 hours2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months after loss of cooling.

The major concern with this event is radiation release due to the boiling fuel pool. Since the core is, offloaded, degradation of equipment in the reactor building due to the severe environment imposed by the heat, moisture, .and radiation released from .the fuel pool. is not a concern provided a makeup water source can be maintained to the fuel pool. FSAR Appendix 9A analyzes the consequences of a boiling fuel pool and finds that the offsite releases are within 10CFR100 limits.

Another concern is the effect on the equipment of the other unit which if not also in refueling, may need to respond to a LOCA. A LOCA on the operating unit would initiate recircualtion with Zone IZZ and the LOCA unit's air space. The effects of the boiling fuel pool could then be imposed on the LOCA unit and create equipment qualification and operability questions.

%a ) r ~ ~

Page 16 3.6.3 ~Suauaae If both loops of RHR are rendered inoperable during refueling with a full core offload, the plant is not placed in an GCO. However, a loss of FPC would require a loop of RHR and RHR service water be placed back in service for RHR fuel pool cooling assist before the fuel pool boils. If fuel pool boiling occurs on one unit, equipment degradation on the other unit could result and place the plant in an unanalyzed condition.

Page 17 4.0 Recommendations This section provides several actions which are recommended to provide a better understanding of the consequences of a loss of fuel pool cooling and provide Operations with guidance on operator response to the event.

1) A LOCA which results in core degradation and releases of radio-activity to the primary coolant is an event which also )eporadizes the ability'o cool the spent fuel pool with systems currently intended for that function. If operators are required to implement RHR fuel pool cooling during a LOCA, analysis is warranted to determine what dose rates would be encountered to access the RHR valves for a range of degraded core conditions. With this information it would be possible to determine if alignment of RHR in the fuel pool cooling mode is feasible during a LOCA where core damage is postulated.

2) A PLEA should be performed to determing if alignment of, RHR in the fuel pool cooling mode during, a LOCA (or post-LOCA) would place the plant in an unanalyzed condition with respect to long term post-LOCA containment heat removal. Operator action should be developed to respond to a loss of fuel pool cooling with a LOCA.

3) If non~ equipment is assumed unavailable during a LOCA, then spent fuel pool temperature and level indication is lost. Therefore, it would be prudent to provide the operators with the anticipated time to fuel pool boiling on loss of fuel pool cooling on a cycle specific basis.

This would allow maximum time for source term decay prior to required access to high radiation zones.

4) The preferred response to loss of FPC is to establish an alternative method of decay heat removal such, that the, total energy does not get transferred to the Zone I, II< and III atmosphere. Otherwise, the k

consequences of fuel pool boiling on the plant's ability to safely

Page 18 shut down should be analyzed.

One alternative to be considered is crosstying the Unit l and 2 fuel pools by flooding the shipping cask storage pit and removing the gates to both fuel pools. Then by initiating RHR fuel pool cooling assist on the non-LOCA unit; cooling could be provided to both pools. Further evaluation of this alternative should address whether power would be available to the overhead crane to remove the gates in a LOCA or LOOP scenario, the ability of RHR fuel pool cooling assist to cool both pools> sources of water to flood the cask pit, and allowable operator response time to initiate this procedure.

Page 19 5.0 Conclusions Following is a summary of the major observations and conclusions resulting from the evaluations performed to compile this report.

1) In a LOCA condition, postulated pipe breaks and single failures can disable enough RHR equipment that operators,may not be able to align RHR Loop A'for fuel pool cooling assist and still maintain adequate core and containment heat removal. Currently, Operations has no clear instructions on how to handle this situation and would likely be hesitant"to remove equipment from core/containment cooling service following a LOCA. While alternate sh'utdown cooling and other means of decay heat removal may-exist, the emergency operating procedures contain no instructions directing the operators to redirect ECCS equipment to spent fuel pool cooling service. Removing all RHR pumps from either LPCI or containment spray duty would place the plant in an unanalyzed condition for long term post-LOCA cooling.

2) Two of the RHR valves required to manually align RHR fuel pool cooling assist are in a room which contains much of the RHR Loop A piping.

In order to acces the valves, operators would be in the direct line of site and in close proximity to RHR piping containing primary coolant. A LOCA which produces core degradation and fission product release will make the RHR valves inaccessible due to the source term from the primary coolant. Decay of this source term is not expected to be sufficient enough to allow access to the valves before the predicted time to fuel pool boiling.

3) If the fuel pool was allowed to boil, moisture and energy released to the reactor building during a LOCA would create a severe environment for which much of the safety-grade equipment may not be qualified.

This would put the plant in an unanalyzed condition.

Page 20

4) Fuel pool monitoring equipment is located on panel OC211 on the refueling floor. The instrumentation is non-lE but is powered off of the diesel generators. The non-1E equipment is not qualified to be operable in an accident condition. Therefore< the trouble alarm in the control room cannot be relied on. Access to the refueling floor is required to monitor fuel pool conditions.

5) Loss of FPC during refueling with a full core offload is a concern if both RHR loops are inoperable. RHR equipment must be placed back into service quickly enough to cool the fuel pool before boiling occurs. A boiling fuel pool on the refueling unit along with a LOCA on the other unit causes equipment qualification and operability concerns on the LOCA unit.

6) Off-normal procedure ON-149-001 notes that operators have 25 hours2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months before fuel pool boiling occurs on a loss of FPC, and instructs them to initate RHR fuel pool cooling assist if available. Otherwise<

the operators are told to allow the fuel pool to boil and provide makeup.

Page 21 6.0 References

1) Bechtel Power Corporation Drawing E10-1, "Single Line Meter a Relay Diagram 125VDC, 250VDC, 120VAC Systems Units 1 a 2",

Revision 18.

2) Pennsylvania Power a Light Co. Calculation M-FPC-009 Revision 0, "Spent Fuel Pool Boiling Analysis".

3) Pennsylvania Power a Light Co. Procedure EP-ZP-055 Revision 0, "Post Accident Response to LossofReactor Bldg HVAC (Reactor Bldg Non-1E Electrical Load Shed)".

4) U. S. Atomic Energy Commision Regulatory Guide 1.3, Revision 2, "Assumptions Used For Evaluating The Potential Radiological Consequences of a Loss of Coolant Accident for Boiling Water Reactors", June 1974.

5) U. S. Nuclear Regulatory Commission, NUREG-0737, "Clarification of TMZ Action Plan Requirements"< November 1980.

6) U. S. Nuclear Regulatory Commission, NUREG-01228, "Source Term Estimation During Zncident Response to Severe Nuclear Power Plant Accidents" > October 1988.

7) U. S. Nuclear Regulatory Commission, WASH-1400 (NUREG-75/014),

"Reactor Safety Study: An Assessmemnt of Accident Risks in U.S.

Commercial Nuclear Power Plants", October 1975.

~ ~ ~

September I, 1992 George T. Jones A6-2 SUSQUEHANNA STEAM ELECTRIC STATION REVIEW OF FUEL POOL COOLING P I-7 88 Fi A

Reference:

EDR-G20020 .

I have completed a review of the concerns over the adequacy of fuel pool c'ooling and the consequences of fuel pool boiling at Susquehanna brought about by the referenced EDR. The intent of the review is to determine the consequences of off-normal and accident events on the ability to maintain fuel pool cooling, and to evaluate the safety implications of a loss of fuel pool cooling. The review is not an evaluation of the licensing or design basis of fuel pool storage and cooling at Susquehanna.

Attached is a report which documents my evaluation and findings. It is my opinion that further evaluation of the concerns discussed in the referenced EDR and my 'report are warranted to assure spent fuel storage at Susquehanna does not reduce plant safety. Section 4.0 of the report provides recommendations and Section 5.0 details my conclusions. Provided below is a summary of these conclusions.

I) The use of RHR in the fuel pool cooling assist mode, post LOCA, occupies ECCS equipment which is accounted for in the long term cooling analysis of FSAR Chapter 6.2. I believe postulated break and single failure combinations exist where RHR equipment is needed for core/containment cooling to remain within the analyzed conditions of the FSAR.

2) A LOCA which results in core degradation and fission product release will make the RHR valves required for manual alignment of RHR fuel pool cooling assist inaccessible. This combined with a loss of normal fuel pool cooling will lead to fuel pool boiling given sufficient spent fuel decay heat.

3) If the fuel pool was allowed to boil, moisture and energy release to the reactor building during a LOCA would create a severe environment for which the safety-grade equipment may not be qualified. Degradation of the iodine removal efficiency of the SGTS charcoal beds due to moisture carry-over could effect off-site release calculations.

4) The fuel pool trouble alarm in the control room cannot be counted on for reliable indication. Access to the refueling floor is required to monitor fuel pool conditions.

5) Loss of fuel pool cooling during refueling with a full'ore off-load is a concern since by Technical Specifications, both RHR loops may be inoperable'. A boiling fuel pool on the refueling unit along with a LOCA on the other unit causes equipment qualification and operability concerns on the LOCA unit.

G. T. Jones Page 2 September I, 1992 PLI - 72288 If a large break LOCA occurred at Susquehanna, I am confident that our current procedures, equipment, and practices would maintain the plant in a safe condition. However, combined with a loss of fuel pool cooling, the operators would be put in a position where they would be required to make decisions on removing ECCS equipment from containment/core cooling service to cool the fuel pool. It is my opinion. that these evaluations need to be done beforehand so that methods are in'place to handle the situation ff it arises.

From the research I have done over the past several weeks, I have found that there are many issues which warrant a more detailed evaluation than I was .

capable of in this short time. I have tried to capture the major issues and provide a quick assessment of each.

Kevin M. Brinckman Attachment'C:

J. E. Agnew A6-3 w/a F. G. Butler '.

A6-3 w/a G. Byram A6-I w/a M. H. Crowthers A6-3 w/a G. D. Gogates SSES w/a G. J. Kuczynski SSES w/a D. A. Lochbaum Enercon w/a S. M. Hausman A6-2 w/a D. C. Prevatte A6-3 w/a G. D. Miller A6-3 w/a J. R. Miltenberger A6-I w/a M. R. Mjaatvedt A6-3 w/a C. A. Myers A2-4 w/a J. G. Refling A6-3 w/a T. J. Sweeney SSES w/a J. A. Zola A6-3 w/a Nuclear Records A6-2 w/a

Attachment 16 PP8L Memo from J. R. Mi ltenberger to G. T. Jones, "Spent Fuel Pool Cooling", September 9, 1992 (PLI-72367)

Note: This memo was prepared by the PP&L Manager of the Nuclear Safety Assurance Group at the request of the PP8L Manager of Nuclear Plant Engineering. The memo acknowledges the concerns in EDR G20020 need to be resolved and suggests that Engineering go beyond the EDR concerns and conduct an in-depth design review of spent fuel pool cooling operations. By this date,- the Manager of Nuclear Plant Engineering had two (2) independent, in-house evaluations of the concerns raised in EOR G20020 which did not refute the primary safety issues in the EDR.

cc: H. M. Kefser TM-N <</o

, September 9, 1992 R. G. Byraa Ad 1 <</o H. G. stanley SSES <</o C. A. Nyers AZ-4 <</o A. J. Oomfnguez w/o G'. Jones A5-2 MIW tSOIwwNI 'QRK/~

SSKS SuSqvEHANNA STae a.ZCTR1C STATioN SPENT FUKl. PGQL COOLlNG Ceorga:

On August 19 you requested that I provide yau copies of the <<ark that NSAG has dane an 'the fuel pool cooling issue and that I state precise'ly what my concerns are regarding the fuel pool coolfng system and its employment.

lfy concerns and issues are <<s folio<<s:

1. The campany does not have an official calculation of the decay heat loads to be expected under outage candftions. A calculation prospected ahead for several years fs needed for outage plannfng. The prospection fs necessary to account for the accumulat1an af fuel in the paal.

2. The station fs vulnerable to loss of decay heat removal capabi11ty during the service <<<<ter outage period of a refuel1ng outage. During this t<me <<e rely upon the operating unit spent fuel pool cooling system far decay heat removal. 8ack up methods of decay heat removal are: 1) use af an Operating Vnft RHR system in the Fuel Pool Caoling Assist Node and R) ba11ing. Neither back up method 1s 'attractive.

a Boiling, although safe, is nat a viable alternative for polftfcal reasons.

o Vse af an RHR system frms the operating unit in the fuel pool cooling assist mode is not attractive for operational reasans.

o RHR has not been tested in the fuel pool caal1ng assist made.

HSAG has argued and will contfnue to argue that the risk is acceptable.

Ha<<ever, a study shauld be made to determfne <<hether a mare viable method of backup decay heat removal during service eater autages is feasible.

3. Calculations should be made af the radioactive effects of loss oF water free the spent fuel pool. Ne currently do nat know the effects on the pa<<er plant or upon the public. Also, <<e do nat have a procedure in place dealing with the radfolag1cal consequences of losing water from the spent fuel pool.

lG-23-1892 18:83 V.sara;

'Hr. 8. T. Jones Page 2 September 4, 1992 PL)- j2361

~ ~

4. Indications are needed to en>>ble the operators fn the control room to directly monitor conditions in the spent fuel pools under normal operating conditions and under casualty conditions.

The above items are based upon a body of work done by NSAG over the years.

They are also based uuon review of NVNRC 91-05 NIDKLINKS FOR INDUSTRY ACTIONS TO ASSESS SNlfNN NANAtlENENT and upon aur partfcfpatfon fn various industry Ineetfngs - the latest of which was the IHPQ Outage hen>>ger/Operations Nanager tfeetfng fn August 92.

They para1'le'I saei but nat a'll of the concerns raised by Dan Prevatte and Dave Lochbaut>> fn thefr paper, SAFETY CONSKOUENCKS OF A BOILING SPENT FUEL POOL AT THK SUSqUEHANHA STENf KLKCTRIG STATION.

I believe that all of the NSAG items can be addressed during the process of resolvfna the Kngfneering Discrepancy Report and subsequent correspondence subaftta9 by Prevatte and Lochbaui.

I have attached the relevant NSAtl Reports and correspondence.

The fo1lowfng sections provide su>>i>>ries of the NSAl: Reports and othe~

pertfnant documents and some supporting arguments.

).0 NSAC REPORTS 1.) Report 13-84, Implfcat1ons of Loss of Mater fram the Spent Fuel Pool Due to Reactor Cavity Failure or Other Causes.

This repa~t was uenerated fn response to the Haddau Sick Cavity Seal Failure. The ma3ar recoim>>endatfons have been 1mplemented and HSAG considers the report to have been closed.

The following ftees ware not, dane and shau1d be reconsidered:

5.2.1 The implications of lass of spent fuel pool level upon radiation shielding be thoraugh1y analyzed. This analysis be added to the FSAR.

5.2.3 Provide additional fnstreaents that would f>>tprove the operator>s ability to respond to a loss ot'evel frat>> the spent fuel pool. l.evel, temperature and radiation fnstr~nts should be considered.

The essence of both of the above recaeiaendatfons was repeated in HSAG Report 1-88. See be1aw, 5.2.4 Insta'l1 a watertight door between the reactor building sump roN>> and the d1vfsfon II care spray roam.

10-23-'992 16-84 I ~ U4i'1 "

S

~ '4 4 I~

~~

~ ':.'; .

Hr,'.

Page'.3 T. Jones September 0, 1992 PL I-72361 NPE dfd an ana'lys1s and concluded that the rtsk of flaodfng dfd nat warrant installation of a door to protect the 91vfsfon II care spray roaal (Pl.I-46640 of August I, 1986). This was not pursued further by HSAG. A study was dane by competent elgfneers and HSAG accepted the results. The issue may be worth reconsidering fn light of the Prevatte concern.

1.2 Report 2-86, Analysis of Operations Mfth Patantfal for Draining the Reactar Vessel.

This report was a follow on ta Report 13-84. It analyzed the Technical Specfffcatfons, potential drainage paths and industry events. It receenended that an 1nstructfon ba wrftten deiin1ng Operations with Potential'or Ora1nfng the Reactor Vessel.

AQ-gA-326, Operations with Potential for Draining the Reactor Vassal/Cavity, was written and want through e1ght revisions. It xas reissued on March 3D. I992 as OP-AO-326. It fs act1vely used durfng outage planning and execution.

There are na open liSAC issues.

1.3 Report, 1-88, Inadvertent Orafnfng of 'Nater from the Spent Fuel Poo'ls on September 12, 1987 via the Cask Storage Pft.

This was a response to a drainage fncfd!nt. The event of September 12, 1987 uncovered a vulnerability to fuel pool drainage that had not been cansidered in NSAG Reports 13-S4 and 2-86.

Hfscellaneous pragramuatfc recaslaendatfons were implemented including a policy that the gates w111 be kept installed as much as possible. This ~ ~

policy, by the way, has been followed fafthfu11y.

Two recoaeendatfons regarding the physical installation rema1n open.

6.1.3 Carrect the problems with the fuel pool cooling system alarms. These include:

b. The Unit 2 side of the DC211 panel has no reflash capability. (DC211 1s the lacal fuel pool coalfng panel.)

3.5.3.l Upgrade the fnfonatfan concerning the fuel poals and fuel pool cooling system that fs available to the control roam operator. resider the engineering work request submitted by Operations management that calls for the critical parameters to be <<ada available ta the operators.

10-23-"992 16-84 V 'Idb/l 1

!~ ~ >

4!;a, ~, , ~ ~

=.;::.", '"Nr;:.'Co. Te "Jones

. . Page 4

'. September 9, 199K

PLI.,723H h

! ~

Recommendation 3.5.3.4 ls essentially a repeat of recoamendation 5.2.3 of Report 13-84. IC is the subject of RSB~s memo to GTJ dated June 30.

Copy attached.

A Chird recommendation is of interest. It reads as follows:

5,2.2 Assess the adequacy of the fuH pool draining analysis in the FSAR considering the condition in which the cask storage it gates are removed. Prepare a safety evaluation and OCFR50.59 review before the gates are renaved, Consider the possibility of uncovering spent fuel by drain1ng and or siphon1ng the cask storage pit.

This recommendation ls a reattack on the recommendation for such an ana1ysis in Report 13-Sh. The safety evaluation was writCen and accepted by the PORC and by the NC (PLI-5453I). I took issue with the, fact, that the shine from the spent fuel pool was not addressed and wrote a memo to ayers (dated 55arch 10, 1988) strong1y reconsending that a shine analysis be done. (See attached.) However, no action was taken.

I closed the 1tem. There was no question in my mind that Che evolution is safe and other Chan the shine question, which seemed to bothe no one but me, the safety evaluation answered the mail.

Report 4-90, Outage Planning informat)on.

The intent of this report was to consolidate the knowledge gained by HSAG during the course of reviewing outage safety. The principa1 user was intended to be NSAG. However, the informat1on was made ava11able to everyone concerned with outage planning.

The various heat removal paths, including spent fuel pool coo11ng, have been analyzed and the capabilities and 11aitations of each have bean 11sted. Every effort was made to cite design base references, focal test results and approved calculations. Nen no calculations existed, NSAQ did their own.

A detailed analysis has been made for each milestone af an outage with respect to decay heat removal. A table has been developed showing the requirements for each condition, the primary and Che back up cool1ng-methods available. A discussion of the service water ouCage and the imp'llcations of using the spent fuel pool cooling systems ls found on pages 28 - 25. The capabilities of the spent fua1 pool cooling systems are found in hppendix C.

There are no receeendations per se.

The principal problem with Che report is that the design heat loads are out of date. ht the time the report was written HPE did not have an approved heat load calculaCion. The heat load figures were taken from a

18-?.5-199? 18: AS I

~ ~

'r. C.

Page 5 Ti Jones S<<ptaiber 9, 1992 PLI-1 2351 calculatfon made by Jack Ref ling for 8 X 8 fuel. HSAG intends to update the report. Me have'gained knowledge sfnce the r<<port was written, particularly about the eff<<cps of boflfng, and we have found some mfnor errors that n<<<<4 to be corrected. Also, we want to include a chapter on containment, which was not ready wh<<n the report was publfshed.

The principal obstacle to upgrading HSAG 4-90 is the absence of a definitive decay heat load calculatian. The Reflfng calculation is a ood enough approxfmatfon to permit general outage planning. However, t understates the heat loads, partfcular ly fn the early days of an outage I would like to update Report 4-90. It is useful to NSAG and to other planning entities within PPCL. It has been distr1buted to the NRC, to INPO and to interested part1es in the industry. It has been cited by both the NRC and by INPO as evidence of PPAL<s leadership in the outage stanagement field. I was invited to address the recent INPO Outage/Operatfons tfanagers~ Heetfng largely because I had gone over this report in deta11 with the INPO Outage folks last swor.

I do not 1ntand to update the r<<port until I have a calculation signed by the appropriate angfn<<<<ring authority fn hand. NSAGts intent1on Is not to do original angfnearfng work. R<<port 4-90 consolfdates existing knowledge in a fore that is useful to persons planning and reviewing outage safety. I want to be sure that any inferences fn 4-90 are bas<<d upon off1cial PPQ. heat load calculations. NSAG, af course, is aware that a calculation fs done by Nuclear Fuels before each outage. Such a calculation was used as the basis for NSAG Report 5-90, ANALYSIS OF ALTERNATE SHUTNNN COOLINO. However, the company does not have an ~ ly official calculation of the decay heat loads to be expected under outage A study pro)ect<<4 ahead for several years 1s ne<<dad for

~ ~

conditions. ~ ~

outage planning.

2.0 NUNRC 91-06 As you know, HN headed a NUNARC task force to address outage 1ssues The result was NNfARC 91-06, GVIOELINKS FOR INUSTRY ACTIONS TO ASSESS SWTDOXN NNAGENENT. HNARC 91-06 is, fn le judgment, an excellent piece of work and 1t b<<hooves us to fmpl<<a<<nt it exp<<dftiously.

NSAQ.has reviewed the status of our pol1cies v1s-a-vis NNARC 91-06. The finding fs that we are in general comp)fance with the guidelines. However, a large number of our policies an4 practices are not captured in directives.

Also, a number of the issues are not well covered by our procedures. Oetafls are found fn PLIS-39784 of July 24, 1992.

%NARC 91<<06 contains three sections whfch are pertfnant to this discussion.

They ares

lO-23-'S92 16:86 P

"..:.;- .:Nr..Q. T; Jones Page 6 September 9, 1992 PL'l-72367 4.1.I Loss of Oecay Heat Roloval 4.1.3 Loss of Spent Fuel Pool Cooling I.2.5 Reactor Cavity Seal Failure I will say a few words about each.

2.1 Loss of Decay Heat Reeoval The guideline reads as follows:

A procedure should be established to address the loss of normal OHR capability during shutdown conditions. The procedure should prioritize the alternate cooling methods available (e.g qrav1ty feed and bleed, low pressure pump feed and bleed, high pressure pump bleed and feeds.

reflux cooling, etc.) and that would be employed for a given set I-that are planned for the outage. The procedure should have oi'onditions a

sound technical basis that includes the following:

o initial magnitude of decay heat o tioe to boiling o time to core uncovery o initial RCS water inventory condition .....

o RCS configurat1ons I ~ ~ ~ ~ ~ ~ ~ ~ ~

The SSES procedure for alternate decay heat raoval is DH-149-00I, l.OSS DF RHR

.SHUTDNN COOLINQ NDE, Th1s procedure was in part based upon NSAG Reports, 2-90, 4-90 and 6-90. It was recently rev1ewed by Engineering and found to be acceptab1e {PL!-71670 of June 19, 1992).

NSAQ agrees that N-149-001 is technically correct. The necessary 1nformation has been included to enable Operators to execute the various heat removal paths. However, our review showed that the procedure does not ref1ect all of the guidance of NNRC 91-06. It needs to be updated to include the 1nitial magnitude of decay heat, time to boiling, capab111ties of the various heat reIOVal oathS, OtC. HuCh Of the neCeSSary infarmatiOn 1S fOund in NSAQ RepOrt 4-90, {NONAGE PLAHNlKG IKFDRNTIN. However NSAG l-90 is based on a decay heat calculation assuming 8 X 8 fuel. This calculation needs to be updated to

~

reflect current conditions, In order to properly implement NUHARC 91-06 a decay heat calculation pro)ected ahead for several years is needed for 1nput into OH-li9-001.

18-23>>i993 i8:87 Hp G. To Jones Page 7 September 9, 1992 PLl 72367 2.2 Lass af Spent Fuel Pool Coo1$ ng The gu1del1ne states:

Hany ut1llt1>>s have chosen to off-load ih>> care to the spent fuel pool (SFP} during the1r refueling autages. Th1s practtce shffts decay neat removal requ1reeents from the RCS to the sFP. An event that results 1n the lass af SFP cool1ng may have the sam undeslrah1e eff>>cts as a loss of DHR event tf appropr1ate compensatory measures are not taken.

Gu(d>>lan>>s I) The autag>> schedule should prov1de a DEFENSE ?N DEPTH ceaaensurate w1th the r1sk associated w(th loss of SFP cool1ng.

.2) A procedure should be establ1shed for respons>> to a loss of SFP caol1ng even'to 2.2.1 Pracedure Me wil1 constder the procedure first. ON-135-001, LOSS OF FllKL POOL COOLIN/COOLANT INENTORY prov1des guidance on haw to restor>> fuel pool cooling 1n the event of var1aus casua'lt1>>s. If fuel pool cao11np can nat be r>>stored, the options ar>> to place RHR 1n the Fuel Pool Caol1ng res1st Nod>> 1n accordance with OP>>lh9-003, RHR OPERATION IN FUEL POOL COOt.NG ASSIST or ta a11ov the iaol to batl. The procedure states that bofl1ng shau1d nat occur before 25 fiours after lass of cool1ng and <t specff1>>s var1ous methods of adding water.

The procedure does not d1scuss cross connect1ng the fuel paols v1a the cask t

storage pit and remov1ng the heat us1ng the other un$ fuel pool canl1ng system. Heat can b>> removed by forced canvecUan (The procedure 1s 1n OP-135-001.} or by natural canvect1on. Cool1ng by natural convect)an was done for an ~ 1ght day per1ad durkni the Vnft 2 2RIO. (Sea NSAG Repart 4-90). It 1s alsa poss1ble to cool the unaffected fuel pool us1ng RHR 1n fuel Paal Coo11ng Ass1st and remve heat frea the affected pool by natural convect1an via the cask storage pit. This method $ s nat discussed 1n ON-185-001.

In sumaary a viable aracedure addr>>sses the lass of spent fuel pool cool1ng.

Her er, all of the leod bac'kup methods are not 1ncluded.

2.2.2 Defense 1n Depth

~

During outag>> olann1ng reviews NSAG takes credit for the follox1ng methods af spent fuel poo) cool1ng:

I) Normal op>>rat(on af th>> outage un1t fuel pool cool1ng system.

Io-ay3-:SBoy 18: 8>

g , C. T. Jones Page 8 September g, lgg2 PLI-F23N'se of both fuel pool cooling systems in parallel. Flow between pools is by natural canvectfon via the cask storage pit. This was tested during the Vnft 2 1RIOn See NSAG 4-N, Cooling both fuel pools using the operating unit fuel pool coalfng system per OP-135-001.

Cooling the outage fuel paal using an outage unit RHR system in fuel paal coaling assist.

Cooling ihe operating fuel pool using an operating unit RHR system 1n fuel pool coaling assist and cooling the outage unit by natura) convection vfa the cask storage pft.

6) Soiling 0 u ring refueling outages the station routinely removes the entire station d y e lo sing he op a 1 g unit fuel pool caoling system. e ian 1s that the care has been off loa e an 4 of service. In o er a d o e s e rvfce water outage and to elean the caaling towers, the outage unit fuel pool coolin99 system must be disab a 1 ed.. 'A rationalize doing this based upon the fo)lawfngf

1) Prior to disabling the SM system we run a test and prove that the operating fuel pool cooling 'system wf'll handle the heat loa , e test results are approved by the PORC.

2) The appropriate PHs have been dane to the fue1 pool caoling systems prior to the outage.

3) The operating fuel paol cooling system can hold fuel pool temperature below 200'F Hth one heat exchanger aut of serv ce.

4) The operating unit RHR system can be used fn the fuel pool cooling assfst mode if necessary 'ta cool the pool.

case ba111 wfl'l accur. If this happens there wfl) be no s1gnificant g damage to th e s tatton and na haTmful effect upon the public.

5) Ample time exists to initiate corrective action. The FSAR states that avet'l hours wfll elapse before boflfng begins.

s stems are out of service for about 10 days n this period the back up cooling methods are f,n,,f th,,th,n, unit RHh Loss in fuel pool uooiinu assist or hoi'iinu.

ve. Soilfn violates the Technical Sp~fifcations and makes the 815 ele a on uninhahitahfu. Use op tha oparatinu unit huh 1 lies takin a ti hour Lco on thee operating ~ft h sa fe t y ssyntone needed to support operations.

a~ it reu~s the capability af the

Hr. G, T. Jones Page 9 September 0, 1992 PLI~72367 A point to consider 1s that at SSES the RHR system has never been fully tested in the fuel pool coolfno assist mode. Ouring the test prograts flow xas established at about 2000 gpe. Higher flows were not attained because the skfssser surge tank kept running dry. OP-149<05 calls for a max1mum flow of 6000 gpm in the fuel pool cooling ass1st mode. RHR xas actually used in the fuel pool cooling assist mode at Brunswick 1n 1983. However, they were unable to maintain flaw above 2000 gpss because at higher flow rates the pool overflowed. h flow of 2000 gpss will carry away the heat load during service water outage conditions. Fuel pool cooling flaw is about l800 gpm and the delta Ts would be siailar.

Based upan the above I believe that a study should be made to determine whether w sore viable method of backup decay heat removal dur1ng service xater outages is feasible.

3.0 HRC Input On August 27 GTJ issued PLI-72267 ta Glenn Miller addressing the Prevatte/Lochbauw concern. The letter stated, There are twenty-eight open itetss resulting fros NAG Review of Fuel Pool Cooling. These need to be included in this review.

I presume that this statement is based an dim Kennyts Nemo of August 25 documenting his conversation with Scott Barber and Jim Raleigh or the HRC.

On September 2 Andre Oaminguez discussed the NSAC items with Scott Barber.

Scott su¹ssarized his concerns as follows')

PPEi. has reeva'iuated the seal lives frea five years reconeended by the vendar to 12 years far the lower and 24 years for the upper.

2) t.ack oi fuel pool cooling instrumentatian )n the cantrol roots.

3) Adequacy of operator training. Is it adequate? Is it still be1ng dane?

l) 'Ihe effects of radiation fallowing a drain down event have not been addressed.

Oolinguet discussed Barber>s concerns with Joe Zola, Jim Agnew and Hark Kfaatvedt later on September 2. They will be addressed. Agnew and Oeminguez plan to discuss the items with Barber on October S.

Barber~s item 2 1s the sub)act of NSAG's only offic1al open items in the spent fuel pool area,

}.0-23-"998 16-89 P a 11/1'

\

r

~, '...

~

I I~

',' ", V

~ ~

'I

}Q,, ~ ~

g p,'}t...w, ~; r ~ ~~ ~

Hr. Q. T. Jones Page ]0 Septeaber 9, 199?

PL)-72367 41 0 CNCLVSIONS Thfs napar ~as wftten fn response to yaur request Chat I sunn}arfze the ftes}s ralathe to fue'I paal coolfng ChaC are af cancern to NSAL The senary fs found fn the apenfn'g paragraph. The succaedfng pages pravfde background and gustfffcatfan, 1 suagestad above that the HSAS ftems can be addressed durfng the process af resobfng the KDR sub}afttad by .Prevatte and Lochbaui. I ln}used alsa lfke to sugaest that Kngfneerfng go hayand resolvfng Che spacfffc concerns. The fuel poof'oolfng systa}ss are a vftal part of the SSES outage process. Engfnaarfng should take thfs opportunfty to do an fn<epth analysfs of the fuel paal coolfng sftuatfon fncludfng revfnrfng our Ixfstfng practfcas and procedures.

Enofnearfng should then pravfde Cha Cechnfcal fnforaatfon necessary to fnEellfgantly operate the fuel paol coolfng ca}aplex under normal and upset condftfons.

Very respectfully, Attac

Attachment 17 PP8L Letter from James E., Agnew to David A. Lochbaum, "EDR G20020, Spent Fuel Pool Design Discrepancies",

October 7, 1992 (ET-0785)

Note: This letter transmitted the formal evaluation performed by the PPSL Engineering Discrepancy Management Group for EDR 620020. This evaluation determines the issues in the EDR to have minimal safety significance with no affect on plant operability using technical reasons which contradict those expressed in the EDR and in the independent PP8L engineering report (Attachment 15).

Pennsylvania Power &. Light Company Two North Ninth Street ~Allentown, PA 18101-1179~ 215I774-5151 October 7, 1992 Mr. David A Lochbaua Enercon Services, Inc 4115 William Penn Highway One Franklin Centre Murrysville, PA 15668 Engineering Discrepancy Management Group EDR G20020, Spent Fuel Pool Design Discrepancies Please find enclosed completed copies of the'Screening Evaluation, Reportability Evaluation, and Operability Evaluation for the subject Engineering Discrepancy Report (EDR).

The evaluation results and disposition are summarized herein:

lmml1aa determined iito ilmi i*

be minimal.

iii However, i W the priority of implementation has been elevated to ensure prompt resolution of the discrepancies.

i reportable. However, I have specifically requested an independent review of reportability by Nuclear Licensing.

i ii impact on the operation of Susquehanna SES.

The discrepancy evaluation function is considered a continuous process. A re-evaluation of the Safety Significance (Screening),

Reportability, and/or Operability status will be performed at any additional stage of EDR processing, including implementation, as information becomes available, in accordance with the Discrepancy Management Program.

If you have not hesitate any comments or questions to contact me.

on the attached, please do James E. Agnew (215) 774-7777

cc: G.T. Jones A6-2 G.D. Miller A6-3 C.A. Myers A2-4 EDR File A6-3 ET Memo File A6-3 NR File A6-2

Attachment 18 PP8L Hemo from G. D. Miller to G. D. Hiller, "Assignment of EDR", October 7, 1992 (ET-0780)

Date: 10/07/92 G D Miller A6-3 Engineering Discrepancy Xanageaent Group kssignILent of BDR This is to assign you EDR No. G20020 Rev. 0 for implementation in accordance with EPM-QA-122, Revision 3.

SUBJECT:

LOSS OF SPENT FUEL POOL COOLING EVENT DESIGN DISCREPANCIES ACTION ITEM: Establish original design basis for fuel pool fuel cooling sys, determine appropriate design basis for spent pool coolong sys, compare the design basis E resolve as necessary INITIAL UNIT ACTION DATE: 11/19/93 (Cycle U17}

The EDR CLOSURE DATE:. 05/20/94 (Cycle U26)

The Priority Classification is: g The INITIAL UNIT ACTION DATE is for EDRs written against both units and reflects the date and cycle for which action is required. All activities associated with this issue for that unit shall be completed before the end of the cycle.

The EDR CLOSURE DATE reflects the deadline for complete EDR closure, includin the disposition of all related actions required to resolve the deficiency.

Enclosed is a copy of the subject EDR and copies of the appropriate evaluations.

Please ensure timely implementation of this EDR and keep me current of developments leading to EDR closure.

The EDMG Planner will visit your assigned engineer/planner on a routine basis for status updating.

You will beifalerted at regular intervals (90, 60, 30 and 10 days to closure), applicable, to allow you time for an orderly implementation of the EDR.

Upon implementation of the EDR, please notify EDMG of completion of required work and of any documentation generated.

D Miller Attachments cc: J E Agnew A6-3 w/o M R Mgaatvedt A6-3 w/a J A Zola A6-3 w/o D Lochbaum A6-3 w/EDR form D F McGann SSES w/a Engr Tech File (JW) A6-3 w/o EDR File (MSS) A6-3 w/a NR File A6-2 w/a

Attachaent 19 Letter from David A. Lochbaum and Donald C. Prevatte to George T. Jones, "Reportability of Boiling Spent Fuel Pool Concerns", October 9, 1992 Note: This letter was hand delivered in a meeting requested by the authors. The authors had not received the PP&L formal evaluations (Attachments 17 and 3) for EDR G20020 until minutes before the meeting. This letter declared the authors intentions to report this matter to the NRC if PP&L did not properly evaluate the concerns by November 2, 1992.

On July 27, 1992, the authors had escalated their concerns to the PP&L Manager of Nuclear Plant Engineering (Attachment 8). On this date, the authors escalated their concerns to the Senior VP, Nuclear, the SSES Plant Manager, and ~a members of the Safety Review Committee by copy of this letter.

George T. Jones October 9, 2.992 Pennsylvania Power & Light, Company Two North Ninth Street, A6-2 Allentom, PA 18101-1179 SUMECT~ REPORTABZLZTY OP BOZLZNG PUEL POOL CONCERNS

Dear Mr. Jones:

On April 16, 1992, Engineering Discrepancy Report EDR G20020 (Attachment 1) was initiated in accordance with PP&L engineering procedure EPM-QA-122 by the two signatories to this letter, Mr.

David A. Lochbaum and Mr. Donald C. Prevatte, to address nine nuclear safety concerns relating to the boiling spent fuel pool event at the Susquehanna Steam Electric Station. This letter is being written to formally reiterate our belief that these concerns are very real and very significant nuclear safety issues, and to convey our concern that they are not being addressed in accordance with either the letter or the intent of PP&L procedures and Federal regulations, and to express our determination that they must be acknowledged, reported, and resolved in a manner commensurate with their significance.

These concerns developed from Mr. Prevatte's evaluation of the reactor building ventilation systems for power uprate, Mr.

Lochbaum's evaluation of the spent fuel pool cooling system for power uprate, and our technical reviews of each other's work.

These concerns were initially documented in our memo dated March 19, 1992 to our supervisor, Mr. Mark Mjaatvedt (Attachment 2). Mr.

Mjaatvedt routed this memo to his supervisor, Mr. Glenn Miller. On April 15, 1992, Mr. Mjaatvedt directed us to initiate an EDR based upon Mr. Miller's review of our memo.

It was not Discrepancy until June 11 or 12, 1992 that the Engineering Management Group (EDMG) engineer, Mr. Joe Zola, contacted one of us, Mr. Lochbaum, to report that from his preliminary assessment of EDR G20020, the concerns had no safety significance. Mr. Lochbaum indicated at that time that he sincerely considered each of the nine concerns to have adverse nuclear safety significance both at the present time and in the future.

Mr. Zola arranged a meeting on June 18, 1992 to discuss the concerns. Attendees at this meeting were Mr. Zola, the EDMG supervisor Mr. Jim Agnew, Mr. Charlie Brown, Mr. Kevin Browning, Mr. Dave Pai, and both of us. We felt the meeting was successful in that the system engineers (Mr. Brown and Mr. Browning) conceded that the spent. fuel pool would boil following reasonable scenarios

within the SSES design bases. Mr. Pai reported that the standby gas treatment system was not designed for the conditions resulting from a boiling spent fuel pool and would isolate on high temperature. We issued a supporting document for EDR G20020 on June 22, 1992 (Attachment 3) providing additional information on the nine issues. This document explicitly stated the regulatory requirements which are not being satisfied for each of the nine concerns in EDR G20020 along with the associated adverse safety implications.

An onsite meeting was held on July 8 or 9, 1992 to discuss EDR to G20020. We learned about. the meeting the day before and asked attend, but this request was denied. We were told that our position at the meeting would be represented by the EDMGMr. engineer Chris (Mr. Zola) . Based upon feedback from Mr. Mjaatvedt, Boschetti, and Mr. paul Weaver, this meeting was not productive.

After Mr. Mjaatvedt told us about the onsite meeting, we looked at the EDMG file on EDR G20020. Other than the EDR itself, the only contents in the file were a page from a memo dated April 23, 1992 (Attachment 4) in which EDR G20020 was determined not to affect Unit 1 operability due to "no apparent impact on plant. DBD issue" and a draft screening worksheet for EDR G20020 prepared by Mr. Art White (Attachment 5). The screening worksheet, which stated "this discrepancy has no basis in fact" and other blatant falsehoods, convinced us that EDR G20020 was being dismissed improperly. We protested immediately to Mr. Mjaatvedt, who promised to consult with Mr. Agnew on the status of EDR G20020.

On July 10, 1992, Mr. Miller came to us and asked G20020 if we had in problems with the EDR process in general and EDR particular. After a discussion lasting several hours in which we expressed our strong concern with both, Mr. Miller pulled EDR G20020 from its file and "Verified" the identified concerns. Mr.

Miller also directed the EDMG to arrange a meeting with all concerned parties to review EDR G20020 and obtain its resolution.

This large meeting was conducted on July 15, 1992. We prepared supplemental information which was distributed at this meeting (Attachment 6). This information addressed the regulatory requirements, licensing requirements, and design evolution history for the items in EDR G20020. Attendees at this meeting included Mr. Mike Detamore, Mr. Miller, Mr. Jim Kenny, Mr. Rocky Sgarro, Mr.

Zola, Mr. John Bartos, Mr. Mjaatvedt, Mr. Tony Roscioli, Mr. White, Mr. Agnew and both of us. The meeting was not productive.

Basically, we were told that our concerns were unfounded because the SSES design bases were not required to handle a LOCA/LOOP event coupled with a loss of spent fuel pool cooling event, and operators would take the necessary appropriate corrective measures anyway.

We strongly disagreed with both of these positions.

On July 17, 1992, Mr. Lochbaum was informed that his contract would not be extended beyond July 31, 1992. On July 23, 1992, Mr.

Lochbaum asked Mr. Mjaatvedt for a final meeting with Mr. Agnew to update him with the status of EDR G20020 prior to his termination.

When the scheduled meeting was cancelled and not. rescheduled, we came to you on July 29, 1992 to express our concerns with the EDR process and EDR G20020. We prepared another summary of the boiling spent fuel pool concerns which we provided to you at that time (Attachment 7). This summary covered the four major safety concerns in EDR G20020 with their requirements and consequences.

Mr. Lochbaum met with Mr. Jim Miltenberger later that day at your request.

On August 18, 1992, Mr. Miller issued a letter addressing EDR G20020 (Attachment 8) and stating that its safety significance was minimal because the NRC had reviewed and approved the fuel pool cooling system design at SSES. Mr. Prevatte responded to Mr.

Miller's letter (Attachment 9) and Mr. Lochbaum called you to register disagreement with the position outlined by Mr. Miller. On August 27, 1992, you issued a letter (Attachment 10) to Mr. Miller directing him to reconsider the safety classification for EDR G20020 and provide a schedule by August 31, 1992. On August 31, 1992, Mr. Miller responded to your letter (Attachment 11) by clarifying his position and reporting that a schedule was still under development. In the meantime, Mr. Kevin Brinckman completed his independent appraisal of the boiling spent fuel pool issues and released his report (Attachment 12). Mr. Brinckman's study essentially endorses every concern identified in EDR G20020 some and even points out that the. probability and/or consequences of concerns may be greater than presented in the EDR.

A number of informal discussions between Mr. Miller and Mr.

Prevatte and Mr. Lochbaum on October and 6, 1992 addressed PP&L's 5

reportability and operability determinations. Mr. Miller stated that these determinations were about to be formally issued. .Mr.

Miller indicated that PP&L determined the concerns in EDR G20020 not to be reportable under 10CFR50.72 and not to affect operability. Both Mr. Prevatte and Mr. Lochbaum registered for strong objections to the justification offered by Mr. Miller these determinations. Mr. Miller indicated the issue would be discussed with the NRC at the quarterly meeting on October 7, 1992. Mr.

Lochbaum asked Mr. Miller was told it if he could attend this NRC meeting and would be inappropriate. Mr. Lochbaum additionally requested that Mr. Miller in his presentation before the NRC clearly state that the originators of EDR G20020 have not yet been provided with documentation of PP&L's operability and reportability determinations and have strongly disagreed with the justifications offered informally by Mr. Miller. Mr. Miller told Mr. Prevatte that he would provide the NRC Resident Inspector with a copyandof EDR all G20020, its reportability/operability determination, related correspondence.

Page 3

0 Mr. Lochbaum discussed this matter with Mr. George Jones by telephone on October 8, 1992. Mr. Jones stated the fuel pool concerns had been discussed with the NRC during the previous day' meeting and that the SSES NRC Resident Inspector would be given information on EDR G20020 that day.

While there have been some informal discussions since August 31, 1992, there has not been any documented progress made on this issue since the end of August 1992. In EDR G20020 and the supplemental information, we provided PP&L with a comprehensive package detailing nine problems with the boiling spent fuel pool event at SSES along with their associated adverse nuclear safety implications. To date, none of these nine items has been formally refuted by PP&L. It appears to us that PP&L is unwilling to concede these are problems until it has completely defined the measures required to resolve the problems. This course of action does not satisfy the procedural requirements 'of EPM-QA-122 or the reporting requirements of 10CFR50.70. Nuclear safety concerns must be formally reported to the NRC in order for other sites with similar conditions to be alerted.

We consider the problems identified in EDR G20020 to have significant adverse nuclear safety implications. A design bases event at SSES is a LOCA with a concurrent LOOP. Even in the case of a LOCA without a LOOP, SSES procedures may initiate a shedding of non-Class 1E loads inside the reactor building in order to limit room temperatures. Since it is a non-safety related system not powered from Class 1E sources, the fuel pool cooling system will not operate. Without the fuel pool cooling system, either the fuel pool cooling assist mode of RHR (a non-safety related, non-single failure proof function) must be initiated to provide fuel pool cooling or ESW makeup to the fuel pools must be initiated to maintain water level in the boiling spent fuel pool. Either operation would require operator entry into radiation fields significantly higher than reported in FSAR Chapter 18 and permitted by 10CFR50. If the spent fuel pool boils, the effects of the latent heat load on reactor building room temperatures and of the condensation/overflow on reactor building equipment operation have not been evaluated. The boiling spent fuel pool therefore represents the potential for providing the means for the common mode failure of all ECCS and safety related equipment in the reactor building. If ESW makeup to the boiling spent fuel pool cannot be achieved, there is also the potential for the meltdown of irradiated fuel outside of the primary containment with the concurrent failure of the standby gas treatment system.

The nine concerns identified in EDR G20020 indicate that PP&L has not performed an integrated engineering evaluation of the boiling spent fuel pool design event. The resolution to EDR G20020 must Page 4

include a thorough engineering assessment of the loss of normal spent fuel pool cooling event on component, system and plant ensure that adverse consequences such as pressurization within levels'o the reactor building due to spent fuel pool boiloff do not jeopardize secondary containment integrity and ECCS performance.

We are very concerned that the nuclear safety issues raised in EDR G20020 are not being evaluated for operability and reportability in a proper manner. You have stated that the NRC Resident and Region I personnel have been informally notified about this issue.

However, we have not yet seen a document prepared by PP&L, other than Mr. Brinckman's study, which presents this issue in a complete and accurate manner. To date, we have seen no action taken on Mr.

Brinckman's study. Mr. Miller and the EDMG have repeatedly attempted to narrow the scope of EDR G20020 by focusing solely on the fuel pool cooling system design bases. During our only contact with Nuclear Licensing, Mr. Kenny claimed our concerns were unfounded since SSES was not required to design for both a LOCA/LOOP and a loss of spent fuel pool cooling event. Therefore, we doubt that PF&L has presented the NRC with a thorough understanding of the issues raised in EDR G20020. Consequently, we discount any claim by PP&L that the NRC has indicated that these issues are not reportable.

Because EDR G20020 was initiated over five months ago, because we meet with you over nine weeks ago, because Mr. Lochbaum is no longer working at PP&L and because Mr. Prevatte may not be working at PP&L for much longer, we respectfully recpxest that PP&L provide to us in accordance with EPM-QA-122 Sections 4.4 and 4.5 written documentation of:

1a) A technical justification for ~eac of the nine items identified in EDR G20020 indicating why each item is not a nuclear safety concern, OR-1b) A copy of the report made to the NRC addressing the items raised in EDR G20020.

2) A final approved screening worksheet for EDR G20020 per EPM-QA-122 and EPM-703.

3) A final approved Reportability Evaluation for EDR G20020 per EPM-QA-122 and EPM-704.

4) A final approved Operability Evaluation for EDR G20020 per EPM-QA-122 and EPM-705.

We. request your response by no later than November 2, 1992.

Page 5

If wereport NRC consider the technical to be incomplete, justification to be inadequate, the or if PP&L fails to respond by the specified date, we intend to proceed with our own report to the NRC on this subject. This report would cover the nine concerns identified in EDR G20020 and indicate that several of these concerns had been raised numerous times in the past but never resolved by PP&L. It would also express our concerns that PP&L does not have an effective program for handling and resolving questions of nuclear safety as evidenced by PP&L's treatment of EDR G20020 and other recent EDRs and safety issues. We acknowledge your statement that you are not satisfied with the EDR program, but the existing program is severely faulted and no substantive corrective measures have been instituted since our meeting with you on July 29, 1992.

We deeply regret having this issue reach this stage, but we know of no legitimate alternate actions we could have taken to avoid this point. In fact, we sincerely feel we have been extremely patient, professional and open-minded in our dealings with PP&L on this issue. We urge PP&L to properly resolve this issue so that our next step .need not be taken.

Sincerely, David A. Lochbaum Donald C. Prevatte Attachments:

1) Engineering Discrepancy Report G20020, "Loss of Spent Fuel Pool Cooling Event Design Discrepancies", April 16, 1992

2) Memo from Dave Lochbaum and Don Prevatte to Mark Mjaatvedt, "Susquehanna Steam Electric Station Spent Fuel Pool Boiling Issues", March 18, 1992 (ET-0149)

3) Memo from Dave Lochbaum and Don Prevatte to Joe Zola, "Supplemental Information for EDR G20020 on Boiling Spent Fuel Pool", June 22, 1992 (ET-0471)

4) Operability Statement Page 63, EDR gG20020, April 23, 1992

5) Screening Worksheet, EDR G20020, Draft by Art White

6) EDR G20020 References, July 15, 1992 Page 6

7) White Paper prepared by David A. Lochbaum and Donald C.

Prevatte, "Safety Consequences of a Boiling Spent Fuel Pool at the Susquehanna Steam Electric Station", July 27, 1992

8) Memo from G. D. Miller to G. T. Jones, "Fuel Pool Cooling Deficiencies", August 18, 1992 (ET>>0586)

I

9) Memo from D. C. Prevatte to G. T. Jones, "Fuel Pool Cooling Deficiencies", August 20, 1992 (ET-0587)

10) Memo from G. T.-Jones to Glenn D. Miller, "Fuel Pool Cooling EDR's G20020, G00005", August 27, 1992 (PLI-72267)

11) Memo from Glenn D. Miller to George T. Jones, "Fuel Pool Cooling EDR's G20020, G00005", August 31, 1992 (PLI-72297) from Kevin W. Brinckman to George T. Jones, "Review of

'2)

Memo Fuel Pool Cooling", September 1, 1992 (PLI-72288)

Copies: H. W. Keiser (w/a)

R. G. Byram A6-1 (w/a)

H G. Stanley SSES (w/a)

'W. R. Corcoran 21 Broadleaf Circle (w/a)

Windsor, CT 06095 J. S. Kemper 115 Polecat Road (w/a)

Glenn Mills, PA 19342 R. L. Doty A9-3 (w/o)

A. F. Iorfida SSES (w/o)

A. R. Sabol A2-5 (w/o)

W. R. Licht A6-1 (w/o)

J. S. Stefanko A9-3 (w/o)

J. R. Miltenberger A6-1 (w/o)

C. A. Myers A2-4 . (w/o)

J. M. Kenny A2-4 (w/o)

F. G. Butler A6-3 (w/o)

G. D. Miller A6-3 (w/o)

J. E. Agnew A6-3 (w/o)

J. A. Zola A6-3 (w/o)

M. R. M)aatvedt A6-3 (w/o)

G. J. Kuczynski SSES (w/o)

C. A. Boschetti SSES (w/o)

T. J. Sweeney SSES (w/o)

G. D. Gogates SSES (w/o)

M. J. Manski Enercon (w/o)

J. D. Richardson Enercon (w/o)

Page 7

Attachment 20 PP8L Memo from D. A. Lochbaum and D. C. Prevatte to George T. Jones, "EDR System Concerns", October 13, 1992 (PLI-72365)

Note: This memo followed up on the concerns voiced by the authors in a meeting with the PP8L Manager of Nuclear Plant Engineering on the overall handling of nuclear safety issues by PP8L.'

~ ~ ~

October 13, 1992 George T. Jones A6-2 SUSQUEHANNA STEAM ELECTRIC STATION EDR SYSTEM CONCERNS PLI-72635 FILE A45-1A This letter is being written to follow up on the meeting on Friday, October 9, 1992 between yourself, Mr. Chuck Myers, Mr. Glenn Miller and the signatories to this letter to discuss the evaluation of EDR G23323.

In this meeting and other previous conversations with you and Mr.

Miller, we have expressed our concerns that the EDR process is not working as required by the procedures and federal regulations. The following is a listing of the most significant concerns we have developed in attempting to work with the system along with some suggestions on how we feel the system could be improved:

CONCERNS MITH CURRENT EDR SYSTEM AND APPROACH The EDR Group is a part of the same organization which has the primary responsibility for the plant design, thereby creating a basic conflict of interests. The EDR Group should be independent.

2. Inadequate resources are earmarked for the disposition of EDRs.

3. The "presumption of operability" philosophy is carried to the illogical extreme, to the end that, for many EDRs an adve s rial cia ions sip results ot>>'

. 8 ED in' cr and the system. Although the "presumption of operability" should exist until proven otherwise, hand-in-hand with this philosophy should be a "presumption of validity" in any concern raised. The EDR process should be a search for the truth, not a process to make problems go away. The discoverer of a problem should not have to make an air-tight case for the process to work. All he should have to do is have reasonable indication and belief that a problem exists, and the process should then research the issue on both sides, and then let the issue stand or fall on its own technical merits or lack thereof.

4. Every step in the process should be performed to a clock.

EDRs must not be allowed to languish as they sometimes do today.

G. T. Jones Page 2 October 13, 1992 PLI-72635

5. The solutions to the reported problem, the anticipated solutions, or the lack thereof should be entirely divorced from the evaluations of the EDRs.

to It shouldEDRs,not only resolve be the to responsibility of the EDR Group evaluate them..

6. Although not sanctioned by procedure, the present presumption seems to be that if an item is reportable, true. J.I.O.

it is not operable.

be generated to .

This is not necessarily A can continue plant operati" n in most cas.= whe e an item .is.

reportable. The two determinations must be independent.

7. The determination of operability and reportability should have less reliance on the legalistic aspects of the issue and more on the technical validity or lack thereof. If there is a conflict between the legal and the technical evaluation, the technical should prevail. There is overwhelming evidence that this is the intent of the NRC in all of the CFR reporting requirements.

8. Potential cost of resolution should have absolutely zero consideration in operability/reportability determinations.

9. Responses to concerns in EDRs should be made on a point-by-point technical basis, not motherhood type statements that are more appropriate for press releases than technical documents.

10- The following reasons and other similar reasons for dismissing an EDR concern should be absolutely disallowed:

a ~ The NRC approved it,.

b. We have (had) an understanding (undocumented) with the NRC.

c ~ There are (x) number of backups. Therefore, the weakness in this item will be made up for in the backups.

d. We'e allowed a single failure.

e. The operators will take whatever action is necessary to make up for the weakness.

The EOPs, EPs and other similar features of our defense-in-depth make up for the weakness.

g, PRA (appropriate for J.I.O.s, not for reportability/operability determination).

h. The issue will be addressed in an upcoming DBD.

The EDR process is too complex and convoluted. The process and forms should be simplified. There are too many gates that EDRs must pass through.

G. T. Jones Page 3 October 13, 1992 PLI-72635

12. The effectiveness of the EDR process should not be judged on the volume of EDRs that pass through the system, but rather on the technical depth and quality of the evaluations.

13. The argument has been raised that the NRC does not want us reporting too. much, and this position has been used to rationalize not reporting conditions which probably should have been reported. It should be borne in mind that the penalty for reporting too much is essentially nothing; the penalty fo- reporting too l'ttle 'an be catastrophic, for the company, for the individual, for the customer, and in the extreme case for the communities around the plant. We have everything to lose and nothing to gain by not reporting when we should. From every point of view cost, reputation, safety, ethics, credibility we should err on the side of reporting.

14. A step in the EDR evaluation process should be the written concurrence or non-concurrence of the originator with the evaluations.

We are submitting this critique with the hope that it can be used constructively to improve the quality and effectiveness of the system. If we can provide any further input, please do not hesitate to call upon us.

D. A.. Lechbaum~ D. C. Prevatte J. E. Agnew, Jr. A6-3 G. D. Miller A6-3 C. A. Myers A2-4 EDR File A6-3 N R File A6-2 EDRCON.DCP/kbw

Attachment 21 PP8L Memo from George T. Jones to G. D. Miller, "Spent Fuel Pool Issue", October 14, 1992 (PLI-72640)

Note: In this memo, the PP8L Manager of Nuclear Plant Engineering directs a coordinated engineering effort to be initiated to address the concerns in EDR G20020. This action comes approximately six (6) weeks after receipt of an in-house engineering report (commissioned solely to assess the concerns in EDR G20020) which did not refute the primary safety issues raised in EDR G20020.

10-26-f992 09:58 P.02/03 OCtober X4, 1992 Nr. C. D. Miller SUSQURSRMI OTNM KECIRIC 5%hTIN OPSIS ÃJNL POOL ZSSUI You are to immediately fora a team othioh inc)udaa representation from Technology, Modifications, Fue)s, Systaas Engineerinq, Operations and XJ.censing.

1. Develop an4 update of the operability and the reportability deteriination that apecifica)ly address all of the issues raised in the sub)oct EDR. This action is to be ccapleted by October 21, 1992.

2. Develop a )ustification fox interim operation that clearly addresses all of the issues raised in SDR-20020'his activity is to be completed by October 21, 1992.

3. Develop any long tera actions needed to ccaapletsly resolve the issue to the point of iaplaaentation, i.e. procedures revision ready for PORC review, modifications ready for start of design. The initial scope of any actions is to ba completed by October 2$ , 1992, This action is to be completed by Novaaber 11, 1992.

By copy of this letter to C. A. Hyers, J. 8. Stefanko, C. Z.

Kuosyneki, H. C, Stanley and N. W. Simpson, you are rayxestod to supply individua) participants in this effort.

C aortae . ones RECEtV<>

cc! Q. 8. Xeasynaki - SSES C. k. ayers A2-i 15 > 9~

N. I. Shannon

- )El 2 I OCT H. C. Stanley

-SSES gu-~r m Z. 8. Stefanko N-3 Nuclear Records - A5-2

Attachaent 22 PP&L Memo from George T. Jones to G. 0. Miller, J. S.

Stefanko and M. M. Simpson, "Spent Fuel Pool Cooling Issue", October 14, 1992 (PLI-72641)

18-26-1992 89: S9 P. 83i83

- W

~

~ ~ ~

octoh>>r 14, 1992 Mr. a. D. Miller Kr. J. S. St>>fanko Mr. H. sr. 8&peon IQOQUXRLlQR STYX IIICTRXC STATXOS SPkÃ% ÃUSf AXLE COOLZÃ0 78$ UE While we hav>> analys>>4 the heat loads in the Spent Fuel Pools due to new r>>loads and operating strat>>gies, we have not updated certain design analys>>s for Spent Fuel Pool Cooling System D>>sign Basis Events. T also need to know if there ar>> otQ~ similar situations. You ar>> assigned the following actionsi

1. Do'cumont the h>>at loads that ne>>d to be considered during normal operation, single unit (fu11 care offload) outag>>s, and, a 2 unit (2 full core offloads) outag>>s for use by Nucleai Technology in updating design basis analysis. Lead

- A S. it>>texaco.

R. Update our d>>sign basis analyses for the Spent Fuel'Pools (and the appropriat>> FBAR s>>ctions) using th>> abov>>

information. Lea4 - I. O. Xi11>>r.

3. Updat>> our radiological analysis for the gent Fuel Pools (and the appropriat>> FShR sections) using th>> abov>>

information. Lea4 - 0 D. Mill>>r~

Mentlfy all cases were changes in fu>>l d>>sign or core loading strategy may Qapaot substantially th>> d>>sign of the plant including changes in design analys>>s, operation proaedur>>s or bardwar>>. Xdentify which items are not handled as a normal part of r>>load analysis and design.

heao - J. I. Nt>>fauRo S. Provid>> proc>>inures revision that provide clear instruction for review of modification hy Fu>>ls Croup and ada@cate input plant design modif ications. Lead x. 1. slayaoa Z>>rry Stefenko has the ov>>rail 1>>ad for this>>ffort and for aasuring that PPK takes all of the measurea repaired to ref 1>>ct the hgect ot fuel reloads into our design, proc>>dure and r>>gulatory system.

.REQEl YEN ocT 16 $ 9Z

Attachment 23 PP8L Hemo from George T. Jones to All Nuclear Engineering Hanager s and Supervisors, "Engineering Discrepancy (EDR) Program", October 14, 1992 Note: This memo directs PP&L supervisors and managers to review the PP8L EDR program with their personnel. It states that for EDRs initiated, "the presumption is Validity until proven otherwise." The PP&L Supervisor of the Engineering Discrepancy Hanagement 6roup told one of the authors (Lochbaum) that the reason he did not actively pursue the EDR from April 1992 to July 1992 was that the consequences of EDR 620020 were so large that it that they could have been missed during design.

was incomprehensible

~ ~ ~

Date: October 14, 1992 To: TO ALL NUCLEAR ENGINEERING MANAGERS AND SUPERVISORS From: GEORGE T. JONES g

Subject:

ENGINEERING DISCREPANCY EDR PROGRAM I would like to share some thoughts with you on the Engineering Discrepancy Program and I want you to share them with your people within the next two days and supply me with the attendees and resulting comments.

It is important to keep in mind that an essential element of any safety culture is ensuring that conditions adverse to quality, plant safety and reliability are promptly identified, reported and corrected.

Our purpose in establishing the Engineering Discrepancy Program and the Group which supports to this essential element.

it is to assure continued attention In addition to establishing the group, we established a periodic review of the progress in close out of EDRs with both PORC, SRC and ERC. I also review our progress in this area with Senior Management. It is an item of their co tinuin interest and will which most of you attended. If you or your people have not attended training, please contact Walt Rhoades and schedule this training. It is your responsibility that your people are trained.

We have achieved success in identifying potentially safety significant items and have improved our performance in resolving these issues in a timely fashion. We have also achieved success at identifying and separating those items not considered to have safety significance but important to do.

It is important to understand everyone's obligation to this Program. In Policy Letter 90-003, signed by Harry Keiser, it states:

"Our expectations of Nuclear Department Personnel are that they will:

A clods Identify and report any known or perceived deficiency in the design or operation of Susquehanna, or any fghu4 DIIICcf RPPokf5 significant event which occurs at Susquehanna in accordance with established procedures."

~~ t=)g pgoaL pals

'S

~pgp g )/oct=$

~the C

~A<p,p5 fo pRo6~~ Came

In section 4.2 of EPM-QA-122 it describes your responsibilities as an originator of an EDR.

Initiate the EDR form, making an initial assessment of whether the potential discrepancy is a Safety Concern or impacts a Technical Specification Action Statement (TSAS),her determining if a SOOR is warranted, and notifying his or Supervisor of the potential engineering discrepancy.

Provide a clear, concise statement of the discrepancy by describing both the requirement and the existing condition so that the difference is clear. The discrepancy is to be described in a concise manner such that it may be understood familiar with the by an individual who is not intimately or is task, special process, item, etc., which constitutes associated with the cited discrepant condition. The description shall provide for direct reference back to the material, equipment, systems, activities or services associated with the discrepant condition.

Inherent in the responsibility of the originator is the obligation to review any potential adverse condition with his supervisor, or the EDMG. If possible this should to be done before initiating an EDR. This permits information be exchanged when it is most current in the who need to resolve the issue.

originators mind It should and is an aid to those not be a cause of delay in preparing the EDR. The originator is in the best position to fully describe the condition, and thus the organization can benefit from the research already performed and allow forbemore timely evaluations of significance. The originator will contacted by EDMG and requested to participate in EDR evaluations to ensure the concern is adequately addressed.

Once the originator completes the form it is brought to their supervisor. The supervisor will, within one day, determine the validity of the EDR (process is calledEDRvalidation) and document Continuation Sheet. The the basis of the determination on an supervisor will then perform the following:

Explain his/her determination to the Originator and assure the individual fully understands the basis for the determination.

Legibly sign, date and print full name.

Transmit to NE-Engineering Discrepancy Management Supervisor (James E. Agnew). The presumption is "Validity" until proven otherwise.

The Engineering Discrepancy Group is responsible to:

Verify the accuracy of the validation process (called Verification).

Accept and document the EDR.

Track, screen, perform initial and subsequent operability/

reportability determination with Licensing, SSES.

compliance and plant support, prioritize, evaluate the EDR.

Assure assignment and acceptance of that assignment, of evaluation and resolution.

Prepare SOORs when appropriate and notify the Originator of the results of validation and resolution.

Review and report status of outstanding EDRs commensurate with their safety significance and age.

Monitor the accuracy of the Engineering Discrepancy List and Database.

Drive the implementation and close out of EDRs.

Transfer EDR records to SRMS.

The process has recently been revised to:

Provide an appeal path whenever the originator disagrees with the results of the validation evaluation, the operability evaluation or the reportability evaluation.

Each originator and supervisor is expected to utilize this appeal process. Whenever the appeal process is utilized, I expect the supervisor to assist and support the originator in making the appeal.

Let me emphasize my commitment to this program. It is important for all personnel to recognize the use of this procedure (EPM-QA-122, Rev. 3) is not optional. Our values demand we utilize procedures because they represent the consensus best way to do the job. Please reinforce with your personnel this procedure is expected to be used in identifying discrepancies.

Attachaent 24 Letter from David A. Lochbaum and Donald C. Prevatte to George T. Jones, "Disagreement with Screening, Reportability and Operability Evaluations for EDR G20020", October 14, 1992 Note: This letter transmits the authors'oint by point rebuttal of the technical reasons formulated by PP8L in determining that the concerns in EDR 620020 had minimal safety significance.

Mr. George T. Jones October 14, 1992 Pennsylvania Power & Light Company Two North Ninth Street, A6-2 Allentown, PA 18101

SUBJECT:

Di.sagreement with Screening, Reportability and Operability Evaluations for EDR G20020

Dear Mr. Jones:

In the meeting on October 9, 1992 between yourself, Mr. Glenn Miller, and Mr. C. A. Myers of PP&L and the signatories to this letter, we provided PP&L with several technical problems with the screening, reportability and operability evaluations completed by PP&L for EDR G20020. This purpose of this letter is to formally st'ate transmit our comments on these evaluations and to clearly that we consider the technical justifications offered in these documents to be inadequate with respect to the items outlined in our letter dated October 9, 1992 to you.

PP&L's position to date has been that restoration of offsite power can. be accomplished in time to permit operator actions to provide adequate cooling to the spent fuel pools and'revent fuel pool bpjl jng ~ EDR G20020, its supplemental information and Mr. Kevin Brinckman's report all clearly identify existing SSES design conditions in which normal fuel pool cooling is lost and when manual valves in the reactor building used to initiate ESW makeup or RHR fuel pool cooling assist are inaccessible. Loss of offsite power is only one of these cases. PP&L has not yet shown that the NRC has reviewed and approved a duration of <20 hours for the loss of offsite power design event. By comparison, PP&L defined the design basis LOCA/LOOP event to be the LOCA with a concurrent LOOP and consistently applied this design assumption in calculations, reports, FSAR discussions, and licensing correspondence.

PP&L's position also relies upon operator actions to prevent fuel

. pool boiling if normal fuel pool cooling is lost. While alternate fuel pool cooling methods may be utilized under certain conditions, PP&L has not yet shown that the NRC has reviewed and approved methods to cover all of the operating and postulated accident conditions required within the SSES design basis. The licensing basis for SSES in the event of loss of fuel pool cooling as described in the SSES FSAR and in the NRC's SER is to permit fuel pool boiling and use ESW makeup to maintain water level. As detailed in EDR G20020 and in Mr. Brinckman's report, the boiling spent fuel pool condition represents an unanalyzed state with potential severe adverse consequences.

In addition, the use of RHR fuel pool cooling assist in case normal spent fuel pool cooling is lost is not described in the SSES FSAR or NRC's SER and could adversely affect core and containment cooling following an accident. The boiling spent fuel pool condition is clearly not adequately analyzed in the SSES design.

Therefore, we consider the concerns identified in EDR G20020 to warrant a "Considerable" safety significance determination and to be reportable under 10 CFR 50.72. In addition, the operability of SSES is presently adversely affected by the problems. We are confident that a justification for continued operation could be written to permit SSES operation until necessary modifications are implemented.

We remain hopeful that the ongoing reviews by the SRC and Nuclear Licensing will result in a proper resolution to EDR G20020. We are available to respond to any questions regarding our position on this issue.

Thank you for your continued personal attention to this matter.

Sincerely, David A. Lochbaum Donald C. Prevatte Attachment Distribution: C. A. Myers (w/a) A2-4 G. D. Miller (w/a) A6-3 J. E. Agnew (w/a) A6-3

October 13, 1992 Attachment Document Sectio Comment General The screening, reportability and operability evaluations lack references to cited information which makes verification difficult. Considering the level of detail that went into documenting the concerns in EDR G20020, it is inappropriate to justify determinations on nuclear safety issues with vague, uncited references.

General The screening, reportability and operability evaluations place undue emphasis on the fuel pool cooling system design. Eight of the concerns identified in EDR G20020 had nothing to do with the design requirements of the fuel pool cooling system. PP&L has repeatedly attempted to divert attention from the real problems identified in EDR G20020 by focusing solely on the fuel pool cooling system design.

General The screening, reportability and operability evaluations rely primarily upon timely restoration of offsite power and operator actions to prevent fuel pool boiling. The basis for limiting the duration of the LOOP has never been reviewed and accepted by the NRC. Operator actions cannot be performed for the pos ulated DBA LOCA with the radiation levels reported in SSES FSAR Chapter 18, which do not include airborne contributions. If airborne activity is considered, the conditions are significantly worse.

Screening/Page 2 The characterization of the concerns in EDR G20020 is too simplistic. Rather than a 'lack of suitable documentation', EDR G20020 reported several cases in which existing documentation was non-conservative.

Screening/Page 2 The five questions in the screening are not addressed for each of the nine concerns raised in EDR G20020. The five questions in the screening are not even addressed for the most severe or most limiting of the nine concerns raised in EDR G20020 ~

Screening/Page 2 Although no selection is specified, it is assumed

, that PP&L's answer to Item I is "NO".

Screening/Page 2 Last paragraph states that the most common failure mode for a complete loss of spent, fuel pool cooling is a loss of offsite power. SSES FSAR Appendix 9B analyzed a complete loss of spent fuel pool cooling due to a seismic event.

What is the basis for a LOOP being the 'most Page 1

October 13, 1992 Attachment common failure mode?'n addition, even if the LOOP case is the most common failure mode, all other failure modes must be covered in the design.

Screening/Page 2 Last paragraph states that "the estimated time to restore offsite power ranges from 15 minutes to 20 hours2.314815e-4 days 0.00556 hours 3.306878e-5 weeks 7.61e-6 months ...". No basis is provided for this estimate. In reality, numerous examples can be cited of LOOPs that have lasted longer than 20 hours2.314815e-4 days 0.00556 hours 3.306878e-5 weeks 7.61e-6 months , most recently Turkey Point as a result of Hurricane Andrew. Additionally, three credible causes of LOOP which are required by Federal regulations to be designed for can cause LOOPs which could last longer than 20 hours2.314815e-4 days 0.00556 hours 3.306878e-5 weeks 7.61e-6 months . They are tornado, earthquake and sabotage.

Screening/Page 2 Last paragraph specifies a 25 hour2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months time to boil for the spent fuel pool. EDRs G00005 and G20020 and Kevin Brinckman's report (PLI-72288) all challenge the validity of the 25 hour2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months time.

Screening/Page 2 Last paragraph concludes that if offsite power is restored in a timely manner, then spent fuel pool boiling will not occur. This response does not address any of the other failure modes for the non-safety related, non-seismically designed fuel pool cooling system. A seismic event, random failure of the non-safety related equipment in the system, or common mode failure of the non-safety related equipment in the system has the potential to incapacitate the fuel pool cooling systems for longer than the time required for the pools to achieve boiling. In addition, current SSES, emergency procedures following a LOCA require the operators to de-energize the non-1E reactor building electrical loads (which includes the fuel pool building cooling temperatures equipment) are as if currently the reactor analyzed.

Screening/page 3 First paragraph misrepresents the concerns of EDR G20020. The severe core damage which renders the reactor building inaccessible is required by Federal regulations to be included in the design basis as documented in SSES FSAR Chapter 18.

With the reactor building inaccessible for days after a LOCA, spent fuel pool boiling will occur following a loss of fuel pool cooling caused by a LOOP, a non-1E reactor building load shed required under emergency procedures, a seismic event, or a failure due to the consequences of the LOCA itself (PLI-72288). This discrepancy can then adversely affect cooling of fuel in the core and in the spent fuel pools as well as secondary containment integrity.

Page 2

October 13, 1992 Attachment Screening/Page 3 Although no selection is specified, it is assumed that PP&L's answer to Item II is "NO". The correct answer is "YES".

Screening/Page 3 The response to Item ZI is technically incorrect.

The discrepancy has the potential for causing the common mode failures of ECCS equipment, the standby gas treatment system, and other safety related equipment required to mitigate the consequences of design basis accidents.

Screening/Page 3 The last paragraph in Item II misrepresents the concerns of EDR G20020. EDR G20020 does not state that fuel pool boiling will cause fuel damage in the core. However, EDR G20020 does point out that for the postulated design basis LOCA with the core damage conditions documented in SSES FSAR Chapter 18, the operator actions required to align either ESW makeup or RHR fuel pool cooling assist to the fuel pools will not be possible and damage to the fuel in the fuel pools may result. Kevin Brinckman's report (PLI-72288) supports the position in EDR G20020.

Screening/Page The response to Item ZIZ is overly restrictive in 3

that system.

it only addresses the fuel pool cooling With the exception of temperature and level instrumentation in the fuel pools, EDR G20020 does not question or challenge the design and operation of the fuel pool cooling system.

The concerns in EDR G20020 are that the boiling spent fuel pool is an inadequately analyzed event with the real potential for causing the failure of every safety related system in the reactor building. These systems are explicitly listed in the Technical Specifications. Therefore, the answer to Item ZZZ should be "YES".

Screening/Page 4 The response to Item ZV does not address the concerns identified in EDR G20020. With the exception of temperature and level instrumentation in the fuel pools, EDR G20020 does not question or challenge the design and operation of the fuel pool cooling system. EDR G20020 specifically questions the ability of the ESW system to provide makeup flow to the boiling spent fuel pool and of other safety related systems in the reactor building to operate in the conditions resulting from a boiling spent fuel pool. Therefore, the answer to Item ZV shoul'd be "YES".

Screening/Page 4 It should be noted that EDR G00005 1990 and has yet to be resolved.

was A

written in timely and complete resolution to EDR G00005 would have Page 3

October 13, 1992 Attachment removed some of the 'complexity'f the issues surrounding EDR G20020.

Screening/Page 4 The response to Item V is incorrect. The SSES FSAR and the NRC's SER do not discuss the use of RHR fuel pool cooling assist, to cope with a loss of fuel pool cooling. The SSES FSAR and the NRC's SER only discuss the use of RHR fuel pool cooling assist to handle the condition of a fuel core offload during a refueling outage. As stated in Kevin Brinckman's report (PLI-72288)g the use of RHR fuel pool cooling assist for fuel pool cooling following a LOCA is an unanalyzed event. The SSES FSAR and the NRC's SER state that the design provision for the loss of spent fuel pool cooling event is for the ESW system to provide adequate makeup to the boiling fuel pools. EDR G20020 identifies concerns that such makeup may not be available, and available may adversely affect performance of if made other safety related equipment. Therefore, the answer to Item V should be "YES".

Screening/Page 5 The response to Item VI is based on operator actions to prevent fuel pool boiling. The response states "it is not reasonable to assume that the operators will take no corrective action and allow the pool to boil." The regulatory basis for designing to satisfy postulated accidents is to provide assurance that actual plant responses are bounded. EDR G20020 does not state or imply that every loss of spent fuel pool cooling event must result in spent fuel pool boiling. However, the only safety related design provision to cope with a loss of spent fuel pool cooling at SSES is use of the ESW system to provide makeup to the boiling fuel pools. The other non-safety related methods of cooling the spent fuel pool have not been fully analyzed for use under all required conditions. Therefore, the concerns identified in EDR G20020 represent at least 'moderate'afety significance

'considerable'afety significance.

if not Reportability/Pg 2 The last paragraph of Section IZ states that each of the nine discrepancies in EDR G20020 are addressed separately and in greater detail in the "EDR Evaluation". If this document is the screening document, the nine concerns were not addressed separately in Revision 1 of the screening document and we did not see Revision 0 of the screening document.

another document, we have not If this document is seen it.

Page 4

October 13, 1992 Attachment Reportability/Pg 3 The response to Item technically inaccurate.

III EDR is misleading and G20020 raised questions concerning the radiological release analysis for a boiling spent fuel pool and identified several non-conservatisms in this analysis. Therefore, the 10 CFR 100 limits are challenged.

Reportability/Pg 3 The response to Item III states that the spent has "two (2) safety fuel pool cooling system grade independent sources of water systems for make up and cooling, the ESW and RHR systems."

The RHR fuel pool cooling assist mode is non-safety related, non-single failure proof. As detailed in Kevin Brinckman's report (PLZ-72288),

use of RHR fuel pool cooling assist following a LOCA is an unanalyzed condition with the potential for adversely affecting core and containment cooling. As detailed in EDR G20020, use of ESW makeup to a boiling spent fuel pool is an inadequately analyzed condition with the potential for adversely affecting performance of safety related equipment in the reactor building.

Reportability/Pg 4 The response to Item III describes an alternate cooling method using ESW supply to the fuel pool with draindown of water to the refueling water storage tank. The recent ESW flow balance concluded that sufficient flow was available to makeup to compensate for boiloff, but did not indicate the flow margin necessary to maintain the fuel pool below boiling. Additionally, there is no evaluation indicating the RWST can handle the heat rejected by this means. It is highly inappropriate to justify away concerns about an unanalyzed condition by relying upon another

.equally unanalyzed condition.'he Reportability/Pg 4 response to Item III states that the fuel pool instrumentation are expected to be available during a loss of offsite power. As detailed in Kevin Brinckman s report, this instrumentation would probably not be available following a LOCA.

This instrumentation is non-lE, non-safety related and non-seismically designed.

Reportability/Pg 4 The response to Item IV is technically inaccurate. The response states that the "analyzed design basis accident (DBA) is a LOCA with a concurrent loss of offsite power." Since either a LOOP or the consequences of the LOCA can produce a loss of spent fuel pool cooling, the combined effect of a DBA LOCA and a loss of spent fuel pool cooling must be analyzed. Hence, EDR G20020 was written.

Page 5

October 13, 1992 Attachment Reportability/Pg 5 In the response to Item IV, the use of RHR fuel pool cooling assist is discussed. As detailed in Kevin Brinckman's report (PLI-72288), the use of this RHR mode following a LOCA is an unanalyzed conditions with potential adverse impact on core and containment cooling functions. Reliance upon this unanalyzed RHR function is inappropriate.

Reportability/Pg Zn the response to Item IV, it is stated that the post-LOCA radiation levels reported in SSES FSAR Chapter 18 are for EQ requirements. However, these Chapter 18 requirements are for DBA LOCA conditions covering personnel access. The ESW and RHR system manual valves in the reactor buildings on both units would be inaccessible post-DBA LOCA due to the contained radiation dose which is reflected in SSES FSAR Chapter 18 and the airborne contribution which is not addressed in SSES FSAR Chapter 18.

Reportabil ity/Pg The paragraph at the top of Page 6 stated that "an analysis has been performed that concludes the equipment can withstand the temperature effects of a loss of fuel pool cooling."

However, no such valid analysis has ever been found. Some of the existing analyses of reactor building temperatures following a ZOCA do assume a fuel pool temperature of 212 F and account for the sensible heat from a boiling pool, but none of these calculations account for the significant latent heat transferred from a boiling fuel pool.

Zn addition, there are no analyses which show the safety related equipment in the reactor building can withstand effects such as humidity, flooding, condensation, etc from a boiling spent fuel pool.

Reportability/Pg The third paragraph on page 6 is misleading. As stated above, the existing calculations at best only considered the sensible heat from a boiling spent fuel pool. The neglected latent heat represents almost 5 times the calculated reactor building heat load and would clearly adversely impact reactor building room temperatures.

Reportability/Pg 7 The response to Item VZZ is inaccurate and misleading. The SSES plant design for the boiling spent fuel pool condition does not meet

,.all the Federal regulations if systems beyond theIn fuel pool cooling system are considered.

addition, design evolutions at SSES since the initial design, such as for highanddensity spent fuel storage racks, 9x9 fuel, the non-1E shedding of reactor building loads, should have been opportunities to detect and correct the deficiencies identified in EDR G20020.

Page 6

October 13, 1992 Attachment Reportability/Pg 7 Zn the response to Item VIZ, the Emergency Plan is relied upon heavily. Heroic operator actions are appropriate to mitigate events outside the plant's design basis such as ATWS. The post-LOCA scenarios are analyzed and documented, but they are deficient as detailed in EDR G20020. It is to

'inappropriate to rely on operator actions fulfillrequirements imposed on the plant design.

Reportability/Pg 7 In the response to Item ZZX (sic VIZZ),

stated that "as part of the design requirements it is the equipment located in the secondary containment must be able to withstand the effects of a loss of fuel pool cooling." As stated previously, EDR G20020 was written to report several conditions in which these design requirements are not satisfied. Once again, operator actions cannot be relied upon to satisfy the design requirements under postulated DBA LOCA conditions.

Reportability/Pg 9 EDR G20020 is determined not to be reportable based primarily upon reliance on operator action and restoration of offsite power within the time to boil. Sufficient technical justification has not been provided which would support this determination for the DBA LOCA with the radiation levels reported in SSES FSAR Chapter 18.

Operability The comments for the screening and reportability evaluations apply to the operability evaluation as well.

Page 7

Attachwent 25 Memo from Charles A. Myers to George T. Jones, "Fuel Pool Cooling Issues - Reportability / Operability",

October 20, 1992

4u gran.v~9e.

H~ 3 28 October 1992 To: George T. Jones AS-2 Copy: Glenn Miller Rocky Sgatra Jbn Kenny Jim hgaee M-3 Daa McGNI SSES S&k&

From: Charhs A Myers AZ-4 Suhject: Fuel Pool Coolhs hmes ReyortabNiy/OpersLSNty Fer your request, I have reviewed the subject natter. I dM not identify any single matter that, at this time with the information availabla, appeared to meet the requirements for reportina to NRC. Sonic of issues are, however, of safety significance; they shouM be pursued more nyeditiously than has been the case and they should be formally brought to NRC's attention.

Scope The scope of my review included review'f docunuadation assochted with the subject and discussions with my staff {Rocky S8arro and Dan McGann). l did not perform any in<epth interviews but did obtain some additional data and input from some of the people involved..

Since I wanted the effort to have a degree of from the work done by your staff, I did not discuss their assessment with them nor rely on the logic they had used in their determinations. Rocky Sgarro has reviewed their determinations h detail and wi11 provide feedbadr. directly to Jim Agnew. My focus was on whether the matters involved were reportable or not and their impact on the operabilityof pIant systems.

I used EDR G20020 as the basis for the scope of my review; I recognize that certain matters witMn that EDR are being handled under another EDR (and appropriately so). I thaught it prudent to con.'Neer the entire issue here.

In detemining reportability, I have bas64 my review on the rephements contained in 10CFRK.72 ani.73 and on 10CFR60.9 (references 1-7).

Zl/28'd T0:Q8 2667-Z~T

in doing this review, I have not given specific consideration to the impacts of the Fewer Uprate project since ve have not yet received NRC approval. I did conclude that eh/le the project mould have some impact on the numbers produced by various analyses, it @as unlikely that the changes would change the basic course of any design basis events. That is the impact appears to be one of slightly reducing time margins or small increases in dose consequences on the order of 5%.

Overviev EDR 620020 and supporting documentation give a generally accurate identification of the issues. 1've grouped these hms into 5 general concerns as described below (parenthetical rel'erences are to the specific paragraphs in section 9 of the EDR).

1, The design basis for the plant and the FSAR have not been sufBciently updated to reflect changes in fuel design and operation, particularly in regards to the Spent Fuel Pool Cooling system design analysis (Reference i1) an6 the Spent Fuel Pool events such as described in FSAR Appendix 9A (Reference 12). There may ba additional consequances because of these fuel design an4 operational changes that have not been adequately analyzed. The principal changes in the fuel design and operation include increased burnup, changing from a 12 morrth to an 18 month fuel cycle, and using 9X9 fuel designs {EDR Section 9E, 9F, 9G and 9H)

2. The design basis for the plant and the FSAR have not been updated to reflect the changes in how the Sperrt Fuel Pool is operated during outages. There may be additional faUure modes and/ot additional ccnseguences, because of these changes, that have not been ader@ately analynxL The principal dueges are routinely oEoading the entire core each outage, intertieing the two fuel pools {using one fuel pool cooling system for both during service vater oilers), and using the RHR system in decay heat. removal mode as an assist to fuel pool cooling to handle poo) heat loads greater than the fuel pool cooling system design. NDR Section 9F., 9F, 9G, 9H, and 9l)

3. The designbasis of the plant and the FSAR (for normal operation) do not address the effect of Spent Fuel Pool boiling on the ertuipmerrt, principally the Standby Gas Treatment System as identified in NRC Regulatory Guide $ .29, that might be used to reduce the ccnsetymces of loss of Spent Fuel Pool cooling systems. (KDR Section 98).

4. The design bash of the plant and the FSAR (for normal operation) do not address the effect of Spent Fuel Pool boiling on the equiprrient needed to achieve safe shutdown. particularly in regard to moisture content in the air, handling of condensation (including Qooding protection), and the additional heat load on the seccndary coetainment ventQation systems. (EDR Section 9A, 9B)

0

5. The design bash of the plant and the FSAR address Mlther the effect of Spent Fuel Pool bailing an equipment needed to mitigate the design h&s LOCA/LOOP event (same hazards as above) nor measures to restore cooling to preverrt the Spent Fuel Pool from boiling during the I.OCA/LOOP. (The Spent Fuel Paol Cooling Systems is one of the loads shed during the event.) (EDR Section 9A, 9B, 9C, 9D)

Discussion In my review, I carefully considered each of'he 9 items (EDR G20020, itAMns 9A through 91). Far ease of handling, I will address then in tbe 6 groups identified above.

As indicated in the EDR items 9G and 9H a number of clMeges in the design of the fuel ar how it operates have been made since the tables associated with FSAR Section 9.i.3 and FSAR Appendix 9A were prepared. These include a change to 9X9 fuel, higher burnup, a longer fuel cycle, and fixing of the outages in spring and fall. Clearly these FSAR portions require updating to accommodate these changes. My review, however, hdicated that none of these drarges have a substantial irrrpact on the safety of the spent fueL The principal isotope of concern is I-131, This isotope buiMs to an equiBbriurn wlQQn a few weeh af the start of a cycle and the inventory in. the spent fuel is relatively insensitive to burnnp, cycle length, or the size of the thai pellet. There may be soms smaH chejnges in production rate due to additional Plutonium fisslan with increased burnup, or same changes in the thearetlcal release of fission pretexts from leaking fuel rods (the escape rate coefTicient) but, here too, 1 expect the impact to be small. There is goad reason to believe that the existing analysis results are stiQ bounding.

Longer refueling cycles lead to larger discharges of fuel each outage; this is addressed below with other Eel paal operational issues. The retool applications made each cycle appear to address properly the other issues (i.e. issues other than the impact on the sperrt fhsl pool).

QDa&iaa- While the slec5c fuel draractetistics are not the same as what is addressed in the FSAR in regard to spent fuel storage, the results af the FSAR anaiysas are judged to remain accurate and the condltianis judgedto ~signiQcantly compromise plant safety.

Items 9E, 9F, 96 and 9H discuss changes in how the spmt arel is operated that have not been addressed in FSAR Section 9.i..3 and AppentBx 9A. These chN~ hclude more fuel assemblies discharged per fuel load. because af the longer cycle, fall oEaad of the core each refbeling outage, routine operation in outages with the kai pools intertied with ane Fuel Pool Cooling System in servtce, and routine operation in outages with RHR in decay heat removal mode to assist the Fuel Pool Cooling System during 5di aEaad.

My review indicated that the normal practice of fully oNoading the core each outage results a condition more severe than the basis for the "Maxmrum Normal Heat Load" case addressed in FBAR 9.1.3 and more severe than the Initll conditions assumed in the PSAR AppeuBx 9A. Using both actual numbers for the current outage (see Reference S) and review of the operational changes, I estimate that the actual heat load is about 4 times that used in the FSAR analysis for these two situations.

This increase in heat lead increases the total heat load given in FSAR tab1e 9. 1-2a and -2b by increasing the heat load due to the most recently discharged fhel (the amount of heat due to previously discharged assinblies is low and is relatively unaffected by the concerns Identified in this KDR.) These tables are provided for information, only.

My review indicated that we currently do not exceed the "Emergency Heat Load" conditions identified in FSAR Table 9,1-2c and -2d but are only a small amaunt lower.

FSAR Appendix 9A analyzes radiological conssquences of the loss of fuel pool cooling during normal operation. 'its approximiteiy four fold increase in heat load due to full core offload would lead to an increase h evaporation rate with a consequential increase In the release of I-$31 from the boiling pool. M FSAR analyses provides a parametric study but the largest calculated dose is 0.096 REM (Thyroid Dose at the LPZ for 30 days). The NRC in their SZR for SSKS (Reference 9) identifies the limit they use as i,5 REM. Using a factor of four increase in releases wou16 lead to a total dose of aber OA REM which is still a sma6 fraction of the value used hy NRC in the SER as a yardstick.

Gmzlusiau- The changes in operation of the fuel pool, principally routinely offioading the core, are not analyzed in the FSAR end do lead to an increase in radiological col~lsices.

The magnitude of the release is still small, however, and is still well below NRCs guidelines as identified in our SER. Under the current ruies this does not appear to be.a condition that Wa~y +amming plW army.

The various modes of operation of the 5a9 pools and the cavities are not identified in the FSAR. One analysLI has been submitted to the NRC (Reference 10) which provides us with some of the results we might get when the analysis is completed, Usmg the values i haitiQsi for the analysis in Safety Evaluation NL-86405 (attached to Reference i

i0), my wavier indicated that either the canBguratian pool and 1 cavity or 2 fuel pools (at the appropriatetima post shutdown, would probably result in a thne to boil not significantly different from the 25 hours2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months used in the Appendix 9A analysis, and therefore the mode of operation would not chmge the analysis substantially.

Cgzlmgn, - The changes in mode of operation of the Suei pools (other than the complete core offload strategy discussed earlier) probably does not substantially affect the results of the FSAR anafyses. There is a need ta do the appropriate analyses of the current operating modes. Such a review might uncover additioed failure modes not efdt¹. sad in the FSAR or Night give different results based cn the IQOdeiing of movement of water bet%Ã9l the $ lools.

This analysis will also veriiy the adequacy oi'administrative limits and tests currently done to ensure the safety of the spent fuel. Based on the information available at this time, I conclude that there is no basis for concluding that this condition sianificant1y compromises safety.

The conditions that make up the two items above do not have consepaaces that exceed current reportability steuhtds, however, the existerxe of these errors over a long period of time with many people knowing of the changes that had taken place constitate a concern larger than night exist if a single oversight had gone undiscovered. %die the provisions of 10CFRK.9(b) do not appear to apply because of the lack of a significant bnplication for public health and, sake or cammn defense and security', it may be prudent to FomaQy notify the NRC of both the condition and our corrective action so that they can confima our judgment of significem.

NRC in their SER (which refers to Regulatory Guide 1.29 and 1.52) identifies the need for the SCABS to meet certain criteria as a basis for approving the use of non-Seismic Cfass I 5ml pool cooing systetns. VhUe no credit appears to have been taken in FSAR Appendix 9A for the beneGts cf Etration through SGTS, it is reasonable to assume that it vas NRC's intent to have SGTS available for mitigation of releases Rom the Pml pool. The SGTS is apparently designed for the envtromnent (using preheating to reduce humidity). 'Re KDR indicates that the impact of candensation buQdup on the structural integrity of the ductwork has not been properly analyzed. It is not char whether the ductvotk would or muhi not fail, and if it Med, it is not clear that it mould prevent SGIS frtzn Sanctioning for this event.

Baseii on these uncertainties, I can not make a judgnmd; in regard to the impact on plant safety.

QgngJlgigg - The required analysis should be completed expeditiously (or an applicable existing analysN found) and the reportability of this item be re-assessed. It appears, from the evidence that I reviewed, that this matter is reportable if it is concluded that the condensate buildup prevents SGTS Rom performing its function.

The issue h the previous three questions involved changes in the design or operation of the plant or the adequacy of an analysis speciQcaHy done to assure acceptab0ity. The remaining two questions involve whether or not the phenomena associated with the fuel pool boiling even need to be considered as a hazard to other safety equipment during normal operation and accidents. PPRL may well decide that these conditions are ones that thiy believe neet to be addressed to meet its standards, but the reportabiHty question is whether or not the

.matter is within our licensing design basis. The record is not perfectly dear.

Ve can find no evidence in the form of NRC guidance, questions to us during liceming, or matters on our docket that indicate that NRC considered these phenomena as part of our licensing basis. There is discourse about the seismic capability of the fuel pool cooling systens, about the methods and assumptions of analysis of the raHological cansepmces of the event, and about the design of the SGTS, but not about the impact of boUing on the other safe shutdown or accident mitigation equipment.

Historicaily, conditions that might devalop substantially after the initiation of the design basis events were frcqm5y not considered part of the Bcensing basis. This was because the focus was on immediate actions end direct consequences, and the presenption that other matters could be handled by the emeryncy staff tdanlge control). Increasingly with time, NRC has started to address such matters adding 1onger term nuheup of cooling water, 30 day analysts of spray ponds, etc. None the less most of the consequences of design basis events outside the safety related areas are stiH not routinely, formally evaluated during initial licensing activities.

There are elements in the record that do raise some questions:

~ During the licensing of SSES, the NNP-2 pltmt was dealing with a concern for the impact of fuel pool boiling condensate flooding on ECCS equipment; after substentia1 time, they upgraded their fuel pool coolin8 systen (Reference 13). Despite the fact that SSES was not yet Bamsed, I found on evidence that the VifNP-2 issue was raised on SSES.

~ $

Our Reactor Building HVAC analysis in the late 80's (identified as the COTTAP 212'F anaiysh in the KR) took Into accotmt the "sinsible heaV from a fuel pool at but not the 'latent heat'. It is my uruierstanding that this was done for extra conservatism. Note that the model usedat the time apparentiy couM not address the effects of moisture released to the air from the pool, so such conservatism On the form of higher heat hput) might have, therefore, been appropriate.

~ Reportedly, earlier (1970's) heat load calculates have been purported to include heat inyuh from a boiling pool. (I have not, inspected the perUnent documents myself.j 2T/L8 d ~:88 266T-22~T

- Vie there are some factors that raise doubt, the preponderance of the evidence says that the effect of the boiling environment on both normal end post accident environments ms not considered to be a aquirement for the licensing of SSES, A complicating factor that gives me substantial concern is that the effect of Interest falls in the category of consequential failures; that is, the danege is a direct result of evaluating the deterministic progress oF the event and not due to assumption of arbitrary Failures, I therefore have reviewed generaUy vhat the cccmquences might be for loss of fuel pool cooling during normal operation.

For this situation, the Reactor Building Ls sufficiently accessible so that ESV can be aiigld to provide the needed ma&up and initiallythe pool temperature and level can bc determIned by direct observation. Analysis is required to verify that suFuient Insttmnentation is available or a safe strategy Is available to maintain vater level is available vithout instmnentation after the pool temperature Increase prectudes direct observation. It appears that a successful strategy is possible vithout ma]or changes.

During this event, RB HVAC Zone 3 (refueling floor) should. remain isolated fsccn the remahdar of the Reactor Building (or be capable of being isolate4. This should prevent the hot, moist envimmaat ftomdirectly affecting most ECCS ejiipment. Condensation In the zone, however, vill enter the drain sysbms, and might, therefore, be able to impact the ECCS systems due ta flooding or sectary evaporation and heating due to hot vtter In the sumps. Noni the less, given the time involved, it appears feasible to tab damage control steps to accoauaodate the impaats.

Qgg1~jgg, - VMe substantial analysis is required, and, proceiures migS need to be developed, I do not currently beHeve that this constitutes a matte that s~icently compromises plMltsafety.

The Introduchey paragraphs ln the previous section discuss the impact of this matter on the licensing basis of'he plant. The limiting event appees to be the LOCAJLOOP because In this event, bath hat pool cooling systems are mnoved horn service thr~ load shehHng.

InitiaL review indicahs that, practically, offsite pover. is kepdred (for the service vater system) to restore fidel pool cooling. RHR is fully occupied with coo}Ing the cores andlor suppression pools on beth units.

ZTiBS'd

An additional concern exists in regard to the effect af bath direct radiation and radiation in the building atmosphere on operation of essential equipment. Specifically, if substantial amounts of radiatian are released to certain plant systems ar to the building atmosphere, damage control actions wiiibe Hmlted and access to the EST ta initiate makeup to the pool may by precluded ar extremely difficult. The safety significance af this condition is therefore dependent on the release af fission products during the event. Use af more realistic analyses may be appropriate in handling matters like this that are outside the bounds considered in original licensing. The FSAR Chapter 15 provides such a "reaiistic" analysis. Recent GK work (being used in aur Paver Uprate applicatian) indicates that the potential for damage fram a LOCA (and LOCAtLOOP) is much loves than originally estimated (due to lover peak clad temperatures) so the FSAR 'realistic" analysis may be overly conservative. None the less, the information available does nat provide a basis for concluding vliether ESV is sufficiently accessible ar what access is available for damage control efforts.

Conclusions md Reeammemhtieas:

i. With the emption of EDR 20020 item 9I, the concerns about the analysis end phenomena appear to be valid.

2. The matters identified in EDR 20O20 items 9E, 9P, 96 and 9H mvolve chliges in the design end operation of the plant that have nat been properly reflected in the FSAR. They must be properly analyzed (including iOCFR50.59 safety evaluations vhere not currently available or complete) and the PSAR ulxhited. My reviev indicated that the calculated radiological conseqeaices (FSAR Appendix 9A) vouhi be increased but stiE vlthin the limits identified by NRC in our SER. As such the coiiditlon does not appear to meet existing criteria far repartab0lty.

3. Not vitlistamilng the information in 2 above, l believe that it is prudent to officiallyend formally inform NRC af this matter because it appears to be associated with a program meahuss that alloved changes to occur that are not properly documented in the FSAR.

Given my current that Nuclear Fnaineerlng has determlnei that the consideratian af ha6 poal boiling iinpacts an safety systems (other than SGTS and the RefueHng Bee enclosure) is not part of the design basis of the plant and my reviev indicating that it vas not part of the Hcensin8 basis of the plant, I do not Gnd a basis for the.

matter to be reportable. Nate that this matter appears typical of shat others have faced during design basis recoiistitutlan efforts and that ve vtli face in the ftlture an the DBD project.

B I tt t t tttt tt ~

The above (paragraph g) does notmean that the matter may not be significant ttt t tt tMttt LOCA/LOOP event and considering the dif6culties that the operators might, in fact, experience to safety.

trying to cope vlth those consequences, l heHeve the matter merits prompt attention. Those'anetlyses may AH uncover conseguences that viQ be determined to be reportable so cceitinuai review of reportabiHty is requiretL t tt 6..The impact oF heat eei moisture (from the spent hei pool during loss of fuel pool cooHna) on the SGTS should be evaluated immediately; if SGTS is not detcenined to be the capaMe of performing its htlans, 1 beHeve that this condition vouM be reportable es an umealyzed condition that signiQamtly compromises plant safety,

7. 1 suggest use of the voluntary LER approach to address the comnnmicatian re@unmended in paragralih3 above. If you chose to follow thh approach, I recommend that a discussion of the concerns in regard to loss of fuel pool cooling during the LOCA/LOOP be discussed and the general approach to addressing the matter be identified, There is suEcient connection betvsen the two matters that addressing only one veuM constitute an incomplete coauaunicatice. NRA is ready to assist you in report preparation it you vish.

8. During my review, the actual seismic design basis for the spent fuel cooHng system ms not clem". Vhile it is clearly not Seimic Qass I, the cooHng portion should be capable of maintaining its pressure boundary Ercun at least ihirmg an earthquake and preferably during the loading due to the suppression pool dynamIcs during a LOCA. Plan-Seismic Category I, Quality Group C).

9. The record does not provide any evidence that NRC has determined ot suggested that this matter is either reportable or not reportable. VhHs NRC has received information from PPM. about this matter, they have not to the best af my knovledge provided any Feedback in regard to reportabiHty, nor do I exyect than to whee ve are still processing the issue.

i0. I auld note that I have subshntiai sympathy for the EDR originators. I beBeve that they have pefanned us a service by ideaUfying tha issues and maintaining the pressure to ensure that the issues are ahhmed. ! recognime the feeHng* that a matter of safety importaae uast, as a mitter of course, be "reportable'. The requirements for reportaMty, ho%aver, are not written in that @ay. They are inteniied to ensure that oen operates vithh the bounds auttlrhed by lMls Hcense. I was unable to identify a condition that put us outside our Hcense that met the reportablUity criteria, Not withstmdhig this, I beHeve ve have certain moral responsibilities to not only address these safety issues but to formally notify NRC of our concern and our actions.

f1. I recogni2e that you have recently re-directed your stat! to attack. the technical issues hvolved. I beHeve that @as appropriate and that the urgency should remain.

ZTr 8T'd C,tr:88 ZGGT-ZZ~T

10 References (All of the following references were used in my review even though not ail are cited in the above memorandum. Additional information was loohd at during this review but did not provide specific information on the issues at hand different Rom that incfuded in the references below.)

1.. 10CFR50.72(b}(1)(ii)(A), (B), and (C)

2. 10CFR50.72(b)(2)(i)

3. '10CFR50.72(b)(2)(ill)(A), (B), (C), and (0)

4. 10CFR60.73(a)(2)(ii)(A), (B), and (Cj

5. iOCFRK;73(a)(2)(v)(A), (B), (C), and (D)

6. 10CFR50.73(a)(2)(vi)

7. 10CFR60.9 {b) ii SEW ii YM ~ldellll ii tddUilhliii4petMI ii H~

8. PU-72230, A. Dyszel to T. C. Daipbu, dated 2i August: 1992, "U2 RI05 Fuel Pool Decay a full core oEoad in the Unit 2 Pool. It aho identiBes administrative Lhnits that, if observed rifi provide at least 25 hours2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months to bo0fng (FSAR value) lf the fuel pool coo11ng it systcmis fest while is providing the only cooHag to the spent fuel. The limit also assures that requirett mahuy (barn the ESV is net greater than 60 gym (FSAR value).

9. NRC SER for SSES.

10. PLA-2720, H. V. Keiset to NRC, dated 12 September 1986, 'Proposed Am'/ment 41 To decease No. NPF-22, Supplemental Information.". hcfehs Safety Evaluation NL 005. This submittal eMresses an Emergency Technical SpadQcation Qmnge Request to tab the second train of RHR out of service during an outage with the core oE loaded and the cavity and beth @el poach connected. The analysis showed that with the larger water volme avaUable the thae to boil was increased even though the heat Load was subshntlally higher than FSAR AppeuHr 9A analysis used,

11. FSAR Section 9.1.3, Spmt Fuel Pool Cooling and Cleans System, including associated tables.

12. FSAR Apllendix 9A, Analysis For Non Seimnic Spent Fuel Pool Cooling Systems

13. SERCH licensing information, dated April $ 987, entitled "Non-Category I Spmt Puef Pool Cooling Systen (SFPCS)"

14 ET-OH9, dated 19 March 1992, Steam Electric Station Spmt Fuel Pool Boiling Imues. Documents the original concerns that were Later ~iocume&d h EDR 20020.

15. EDR 20020, Rev 0, dated 16 Apri) 1992, "Loss of Spent Fuel Pool Cooling Event Design Discrepancies"

16. ET-S}71, dated 22 June 1992, Dave Lochhmm and Don Prevatte to Joe Zola, entitled "Supplemental Information for EDR GZXZO on Boiling Spent Fuel Pool"

17. Not numbered, dated 26 June 1992, presutned to be frol Dave Lochbaum and Don Prevatte, entitled "EDR 620020 References"

.18. Report, dated 27 July 1992, by David A. Lochbaum and Occeld C. Prevatte, entitled "Safety Consequences of a Boiling Spent Fuel Pool at the Steam Electric Station" PU-72288, dated 1 September 1992, Kevin V. Brinchnan to George T. Jones, "Reviev 19.

th 8ty~dth 10IIIW Mtl . Yh I 'th of Fuel Pool Cooling". Contains a study by Kevin and commissioned by Qeorae, to evaluate dg' 620020, Kevin Brhxhnan, and myself are not substantially diNerent in regard to the fKR consequences should boiling occur.

20, PU-7E%7, dated 9 September 1992, J.R. Milbaherger to George T. Jones, "Spent Fuel Pool Cooling". Identifies NShG concerns and actions associated with spent fuel pool cooling. The comms ideNQed closely parallel those Mentified in the EDR end in my own reviee. This referee also goes in to the most date ahoy spent fuel pool operating modes.

21. Unnumbered, dated 9 October 1992, memo from David A. Lodhaum and Donald C.

Prevatte to George T. Jones, "ReportabNty of Boiling Fuel Pool Concerns'. This referee concerns about the handling of the subject matter, principally the amount of time 'aised involved, but also inctuties sama technical discussions pertinent to the mAer.

22. ET-07M, dated 7 October 1992, memo from J. E. Agnew to J. M. Kenny. Transmits EDR 620020, end tha most recent raportaMity and operability assessa~nt from Nuclear Tecbnology.

23. Umnnnbered, dated 14 October 1992, memo from David A. Lodlbaum an4 Donald C.

Prevatts to George T. Jones, "Disagreemnt vith Scremhtg, Reportability and OperaMlity Evaluations for EDR G20020'. Mentifies specific problems with Reference 22.

these operettingmodes.

hyISEtituaat yt WM t~

24. Draft analysis, dated 9 October 1992, by David G. Kosteln8r., entitled "Loss of Fuel Pool CooHng Eveat, Evaluation for KDR ~620020'. Tbis.draft paper ideatiGes the operatin8

~yl prdt

Attachsent 26 PP&L Memo from Glenn D. Miller to George T. Jones, "Evaluation of EDR G20020 - Spent Fuel Cooling Issue",

October 21, 1992 (PLI-72711)

Note: In this memo, PP8L for the first time addresses the concer ns in EDR G20020 individually. In doing so, PP8L agrees that seven of the nine concerns are valid, yet determines that the operability of the plant is not affected and it is not reportable.

~ ~ ~

I October 21, 1992 George T. Jones A6-2 SUSQUEHANNA STEAM ELECTRIC STATION

'VALUATION OF EDR G20020 - SPENT FUEL POOL COOLING ISSUE PL I-72711 FILE A45-1A The attached evaluation of EDR G20020 is provided in response to your memo PLI-72640. This evaluation was prepared by myself and a team of engineers working on the action plan to resolve the subject EDR.

In the course of reviewing this issue in detail, we have concluded that seven of the nine identified discrepancies are not valid deficiencies. This is explained in detail in the evaluation. The two remaining discrepancies are valid deficiencies but are not considered safety significant nor reportable.

I believe that the technical basis on several of these issues has been clarified considerably in the course of the past week. Therefore, I suggest providing this evaluation to Nuclear Regulatory Affairs for reconsideration of the reportability aspects.

We are continuing with the remaining action items as requested. A detailed engineering design report and justification for interim operation will provide more detail than contained in the attached evaluation. We are scheduled to meet with PORC on Honday October 26, 1992 at 2:30 pm to review this issue. We are working with Systems Engineering, Operations and NRA-Compliance with respect to potential compensatory measures.

.ue Glenn D. Hiller CC: G. J. Kuczynski SSES J. E. Agnew A6-3 C. A. Myers A2-4 M. R. Mjaatvedt - A6-3 H. W. Simpson Al-2 D. F. Roth SSES H. G. Stanley SSES J. H. Kenny A2-4 J. S. Stefanko A9-3 F. G. Butler A6-3 J. R. Miltenberger - A6-1 D. C. Prevatte - A6-3 Nuclear Records A6-2 D. A. Lochbaum - Enercon

October 21, 1992 Page 1 Evaluation of EDR G20020 This document contains an evaluation of the discrepancies documented in EDR G20020, "Loss of Spent Fuel Pool Cooling Event Design Discrepancies."

Conclusions of the author with respect to reportability of these concerns and operability impact on SSES are also provided.

Desi n Basis The design bases for the Fuel Pool Cooling System are found in FSAR section 9. 1.

The portions of the design basis relevant to EDR G20020 are as follows:

1. Maintain the fuel pool water temperature below 125F under "normal maximum heat loads" defined as 12.6 HBtu/hr (equivalent to a typical fuel cycle discharge schedule which fills the fuel pool, last quarter core offload at 6.7 days after shutdown).

2. Maintain fuel pool water temperature at or below 125F during the "emergency heat load" condition of 32.6 HBtu/hr (equivalent to a full core offload 10.5 days after a shutdown following a typical fuel cycle discharge schedule which fills the fuel pool) utilizing the RHR system (with or without normal fuel pool cooling) for fuel pool cooling. This mode of operation applies "during periods of higher than HNHL generation in the fuel pool, eg., storing of a full core of irradiated fuel shortly after shutdown". The RHR system is used under these conditions to assist the FPCCS in dissipating the decay heat. Thus, any heat load in excess of 12.6 HBtu/hr is considered to be within the design basis for the RHR FPC assist mode of operation.

3. Redundant Seismic Category I ESW connections to each pool are provided to allow for makeup of evaporative losses in the event of failure of the FPC system. The conditions are bounded by a fuel pool time-to-boil analysis based on the same typical fuel cycle discharge schedule as in basis ¹1 except the time after shutdown is 10.5 days instead of 6.7 days resulting in a heat load of 9.8 MBtu/hr. (This explains the difference between the two different heat loads, ie., 12.6 HBtu/hr for basis ¹1 and 9.8 HBtu/hr for basis ¹3. This is not a discrepancy.) The ESW makeup line is sized on the basis of this calculation (Reference FSAR section 3. 13).

4, The cause of the Loss of Spent Fuel Pool Cooling event is stated to be a seismic event.

5. All piping and equipment shared with or connecting to the RHR intertie loop are Seismic Category I and can be isolated from any piping associated with the non-Seismic Category I fuel pool cooling system.

Evaluation of Discre ancies Noted in EDR G20020 EDR G20020 describes nine discrepancies relating to the loss of spent fuel pool cooling event. This discussion will summarize each issue. The reader is

October 21, 1992 Page 2 referred to the complete text of the EDR.

General Statement The introductory paragraph of the EDR states: "...the design provision for the loss of spent fuel pool cooling event is to permit the fuel pool to boil and maintain its water level above the fuel through makeup from the ESW system. This design provision is necessary because the fuel pool cooling system used for normal operation and the RHR fuel pool cooling assist mode used for abnormal heat loads are not designed to satisfy seismic category I and single failure criteria."

As stated in design basis ¹5 above the RHR fuel pool cooling assist portion of the piping is designed to seismic category I requirements. No credit however is taken for this mode of operation in the fuel pool boiling analysis in the FSAR.

Credit is taken for this mode for emergency heat load situations as defi'ned by basis ¹2.

Discussion of EDR Items A throu h I In order to discuss and evaluate each of the nine discrepancies listed in the EDR it will be more logical to review them in a different order. Items E & F both relate to the time-to-boil calculations and will be reviewed first followed by items G through I, which are related to the time-to-boil concern. Items C 8 D involve operator action considerations and will be discussed next. Finally items A 8 8 relating to the evaporation effects will be discussed.

October 21, 1992 Page 3 Item E; Anal tical Time-to-Boil "The analytical 25 hour2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months time-to-boil for the spent fuel pool is nonconservative for the maximum normal heat load in the spent fuel pool."

As stated in basis ¹I the maximum normal heat load is 12.6 HBtu/hr. As stated in basis ¹2 the time-to-boil analysis is based on a heat load of 9.8 HBtu/hr.

These two design bases are in fact consistent and are based on the same "typical fuel discharge schedule" and refueling outage scenario. The difference in the heat load is due solely to the time after shutdown assumed for purposes of establishing the design basis.

Focusing on the time-to-boil analysis, a time after shutdown value of 10.5 days is used. This is the'time at which it is assumed that refueling is completed and the reactor cavity to fuel pool gates are reinstalled. Prior to that point the additional water stored in the reactor cavity is also available as a heat sink and the RHR system is available for fuel pool cooling. For times greater than 10.5 days the appropriate heat load will be even lower than the analyzed value of 9.8 HBtu/hr,. For the SSES Unit 2 5RIO the mme from reactor shutdown to fuel pool gates installed was 38 days. The decay heat in the Unit 2 pool at that time is calculated to be 5.65 HBtu/hr (Reference SEA-HE-405). The corresponding time-to-boil is 45 hours.

The EDR goes on to discuss other calculations which result in different heat loads using various assumptions. Calculation NFE-B-NA-053 was performed by Nuclear Fuels to account for actual fuel discharge history and future offloads accounting for power uprate conditions. The fuel pool heat load versus time curves obviously will increase subsequent to power uprate, however, these curves do not apply to the existing design. As long as the calculated decay heat is less than 9.8 HBtu/hr at the point where the fuel pool gates are reinstalled the original design basis time-to-boil calculation is still valid.

Calculation H-FPC-009 determined time-to-boil conditions post power uprate. This calculation shows that the time-to-boil for the design basis heat load of 9.8 HBtu/hr is slightly greater than 25 hours2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months .

In conclusion, the design basis for the time-to-boil condition is established by the 9.8 HBtu/hr value used in the original calculations. This design basis is met by planning the outage so that the fuel pool is not isolated from the reactor cavity or the RHR system prior to a point in time where the actual heat load is 9.8 HBtu/hr or less.

This discrepancy is not a valid deficiency, is therefore not reportable and has no impact on the operability of the plant.

October 21, 1992 Page 4 Item F: Time-to-Boil for Emer enc Heat Load "The analytical 25 hour2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months time-to-boil for the spent fuel pool does not account for the emergency heat load in the spent fuel pool."

As discussed above, the time-to-boil conditions apply to configurations where the spent fuel pools are isolated from the reactor cavity (ie., non-refueling configurations). As is correctly stated in the EDR, current practice is to fully offload the core during each refueling outage. Specific calculations are performed by Nuclear Fuels to determine the ability of the FPC system to remove the combined decay heat of the cross-tied refueling pools. Tests are also conducted to determine that the actual heat removal capability exceeds the actual fuel pool heat loads during the outage (Reference TP-235-011). Normally the reactor cavity is maintained flooded and cross-tied to the fuel pools. One loop of Core Spray is always operable in this configuration. One division of RHR is maintained in shutdown cooling mode except for a brief period required for the common RHR system outage window.

Design basis ¹2 states that heat loads in excess of the HNHL are considered to be emergency heat loads. The design of the RHR system to assist the FPC system during emergency heat load conditions assures that fuel decay heat is removed.

No time-to-boil calculation for this configuration is required since the RHR system will be in operation or available. At any rate, such a calculation should consider the effect of the additional water inventory available from the flooded reactor cavity, cask storage pit and dryer and separator storage pool which are all cross-connected during this time. makeup inventory is also available from Core Spray and the RHR system is normally in-service except for the common RHR system outage window.

In conclusion, no time-to-boil analysis is required for the emergency heat load design basis. Single failures of the RHR system are not required for this design basis for the emergency heat load (Reference SRP 9. 1.3).

This discrepancy is not a valid deficiency, not reportable and has no impact on plant operability.

October 21, 1992 Page 5 Item 6: Radiolo ical Release Calculation for Boilin S ent Fuel Pool "The radiological release analysis for a boiling spent fuel pool uses nonconservative evaporation rates."

This discrepancy is directly related to the heat load assumed for the time-to-boil analysis. The evaporation rate used in the dose calculation is based on a heat load of 9.8 MBtu/hr which is the design basis heat load for the time-to-boil cal'culation. Heat loads in excess of 9.8 MBtu/hr obviously result in higher evaporation rates. Since the discussion under Item E above establishes that 9.8 MBtu/hr is the correct original design basis and still bounds current operation there is no discrepancy in the offsite dose calculation. It uses an evaporation rate consistent with the design basis heat load:

This discrepancy is not a valid deficiency, not reportable and has no impact on plant operability.

October 21, 1992 Page 6 Item H: Nonconservative Activit Terms "The radiological release analysis for a boiling spent fuel pool uses nonconservative activity terms. The original design calculation (200-0048) assumed 12 month operating cycles and 184 bundle equilibrium reload sizes to determine the activity terms for failed fuel in the fuel pool. SSES currently has 18 month operating cycles with approximately 230 bundle reloads which will increase to approximately 254 bundles after power uprate. Since the calculation implied that most of the activity results from the most recent discharge batch, the effect of increasing the discharge size from 184 bundles assumed in the calc to 230 and 254 bundles would appear to be nonconservative with respect to the radiological release analysis."

The original radiological release analysis as referenced above is conservative for the following reasons:

(1) the activity levels used as a source term are based on 1X failed fuel.

All of the failed fuel rods are assumed to be in the offloaded batch of 184 fuel assemblies. Therefore, increased batch sizes will not increase the amount of the source term used in this analysis.

(2) the activity levels used for the iodine source term are based on saturation level inventories for a core operating at 3440 NHt for one thousand days. Therefore, the fuel cycle length will not affect the source term.

In conclusion, the offsite dose calculation remains valid.

This discrepancy is not a valid deficiency, not reportable and has no impact on plant operability.

October 21, 1992 Page 7 Item I: Anal sis f'r Hax Time Prior to Hakeu "The analysis for maximum time prior to makeup to a boiling spent fuel pool is based upon nonconservative assumptions. The original design calculation (175-14) determined the time using evaporation of the entire fuel pool water inventory.

The maximum time should be based upon a minimum fuel pool water level which is sufficiently above the top of the fuel to provide the shielding required to allow corrective operator actions."

The purpose of the referenced calculation was to determine refueling floor atmosphere conditions under various operating modes. The evaporation rates and assumptions used in the cited portion of the calculation were used solely to determine if condensation could be expected under fuel pool boiling conditions.

The conclusion of the calculation regarding time to boil the pool dry is not relevant to any operator action. Operator actions are based on maintaining normal pool level and temperature conditions. In any case, the cited nonconservatism would have a minor effect on the calculated 19 days to boil the pool dry, a result which is not used elsewhere in the design.

This discrepancy is not a valid deficiency, is not reportable and has no impact on plant operability.

October 21, 1992 Page 8 Item C: Manual ESW Valve Actions "The manual valve manipulations required to provide ESW makeup flow to a boiling spent fuel pool may not be possible."

In-plant post-accident radiation levels are analyzed for SSES to the requirements specified in NUREG-0737. This document requires that post-accident radiation levels be determined for purposes of vital area access by plant operators to perform short-term first priority actions. It specifies that radiation levels be determined on the basis of contained sources, and core damage source terms equivalent to those used for 10CFR100 calculations. These assumptions are clearly based on degraded core conditions which are beyond the design basis LOCA.

Airborne radioactivity sources from containment leakage are required to be analyzed for environmental qualification of equipment but not for personnel access.

A review of FSAR chapter 18 shows that access to the equipment necessary to provide makeup to the fuel pool from ESW is restricted for the approximately the first 30 hours3.472222e-4 days 0.00833 hours 4.960317e-5 weeks 1.1415e-5 months following the design basis event (Figure 18. 1-9). This analysis is based on a conservative source term equating to IOOX fuel damage resulting from core melt conditions as originally utilized for offsite dose calculations used to determine plant siting adequacy. These source terms were based on experiments involving heated irradiated uranium dioxide pellets.

An evaluation of actual fuel thermal response during design basis accidents results in no predicted fuel failures (Reference PLI-72696). Thus, the source term resulting from the DBA LOCA would only be equivalent to the radioactivity present in the reactor coolant as a result of normal operations (allowing for fuel defects as permitted by Technical Specifications). To bound the potential effects of a design basis accident, a realistic yet conservative analysis using an assumed 1X fuel damage resulting from core degradation under LOCA conditions was performed (Reference EP-548) and concludes that access to equipment necessary to mitigate the effects of a loss of fuel pool cooling following a DBA LOCA is assured.

In conclusion, post-accident operator actions are viable for all potential scenarios under consideration, for both the current design basis and those outside of the current design basis.

This discrepancy is not a valid deficiency, is not reportable and has no impact on plant operability.

October 21, 1992 Page 9 Item D: Instrumentation "The instrumentation available to the operator post-LOCA does not provide adequate indication of spent fuel pool temperature and level to allow proper response to a loss of fuel pool cooling event."

The instrumentation available to the operator is not required to be qualified since the design basis loss of spent fuel pool cooling is not coincident with the DHA LOCA conditions. This instrumentation is powered from an uninterruptible power supply and its'ssociated 1E AC source.

The minimum water level required per Tech Specs is below the weir elevation.

Since ESW makeup is provided to the pool the operators will know that when they see a rise in skimmer surge tank level the fuel pool level is at least as high as the weir. This provides a confirmation of adequate pool level without requiring access to the refueling floor.

the basis of the discussion in item C above, access to the Furthermore, on refueling floor is possible under all considered conditions. Therefore, 'it is possible to verify adequate fuel pool level visually from the refueling floor which is accessible from several locations.

While the available instrumentation is adequate for operator actions and meets the regulatory requirements of Reg Guide 1. 13, improvements to the instrumentation have been recommended in the past and should be implemented.

This would enhance plant safety.

In conclusion, the existing instrumentation is adequate for performance of required operator actions for the current design basis and for scenarios not included in the current design basis.

This discrepancy is not a valid deficiency, it is not reportable and has no impact on plant operability.

October 21, 1992 Page 10 Item A: Reactor Buildin Desi n Heat Loads "Reactor building design heat loads do not account for the boiling spent fuel pool event."

The reactor building temperature analysis is performed for OBA LOCA conditions.

The presumption of item A is that the spent fuel pool will reach boiling conditions prior to restoration of fuel pool cooling subsequent to a LOCA. The existing temperature analysis does account for the sensible heat load from the fuel pool at 212F.

A loss of spent fuel pool cooling event can result from several conditions. The design basis condition is a seismic event as analyzed in the FSAR. The Fuel Pool Cooling system is not designed for seismic loads. In this case, the Fuel Pool Cooling system is assumed to be lost. An evaluation of the plant response shows that several methods are available to assure that the spent fuel remains cooled.

These include: (1) the RHR system can be used to cool the fuel pools with alternate shutdown cooling of the reactor using Core Spray and RHR for suppression pool cooling; or (2) allow the fuel pool to boil with makeup supplied by ESW with consideration of either SGTS operating on Z'one III or providing a vent path from Zone III. The offsite dose analysis takes no credit for SGTS.

If available, normal reactor building ventilation would be used to provide III Under any of these scenarios cooling and venting of the Zone atmosphere.

transport of moist air to other portions of the reactor building would not occur.

This scenario is the design basis for loss of fuel pool cooling.

Other scenarios not included in the design basis include LOCA and LOOP events, and combinations thereof. The time frame for consideration of operator actions is based on reasonable expectations for the time-to-boil condition. As stated previously, for the current operating practice, the fuel pool, heat load prior to reactor restart is approximately 4.65 HBtu/hr. Time to boil under this condition is on the order of 55 hours6.365741e-4 days 0.0153 hours 9.093915e-5 weeks 2.09275e-5 months . Note that this is the shortest possible time-to-boil for the current fuel cycle. With the pools cross-tied the time-to-boil is greater than 100 hours0.00116 days 0.0278 hours 1.653439e-4 weeks 3.805e-5 months .

For a LOCA scenario, the FPC system will be lost initially due to the Aux Load Shed provisions. Although the Fuel Pool Cooling system and other non-safety related systems are not specifically analyzed for the effects of hydrodynamic loads it is expected that they will be able to perform their normal functions following a broad spectrum of design basis events. Credit for these systems is not needed to meet the design basis, however, plant operators will utilize any equipment available to them during emergency situations. Therefore, in the course of evaluating the effects of a OBA LOCA on the fuel pool cooling system, we acknowledge the availability of normal plant systems in responding to the emergency.

Independent of the LOCA condition, offsite power is needed to restore normal cooling systems. The SSES Individual. Plant Evaluation considered loss of offsite power (Reference IPE Appendix F). The IPE conservatively estimated the incidence of LOOP to be .04/year (plant related), .008/year (grid related), .00807/year

October 21, 1992 Page 11

{severe weather related), and .00066)year {extremely severe weather related).

The probability of recovery from the LOOP within specified times was also calculated as follows:

~Time hr P Recover within T hrs 12.0 97.96X 24.0 99.53X 60.0 99.923X 75.0 99.953X Thus, it can be reasonably concluded that offsite power will be available within 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months following the initiating event.

The remaining factor in addressing restoration of fuel pool cooling is access to the reactor building. This issue is discussed under Item C above. Except for degraded core conditions access to the reactor building is feasible after the first twelve hours of the initiating event (Reference EP-548).

Notwithstanding the above basis, we have also considered the scenario where offsite power is not available and access to the lower reactor building elevations is restricted (based on FSAR chapter 18 contained source terms).

Under these conditions (representative of a degraded core event) access to the refueling floor remains available. Provision for fuel pool cooling is made through use of the plant fire protection system. Venting of Zone III via the filtered exhaust system is also possible for this scenario. While access to level and temperature instruments would be questionable it is possible to verify adequate pool level visually .from the refueling floor which is accessible at several locations.

A Loss of Fuel Pool Cooling Event Tree is attached to this evaluation to help guide the reader through these various postulated scenarios.

In conclusion, for the design basis loss of fuel pool cooling the plant as currently designed and analyzed is acceptable. For other scenarios not specifically included in the design basis we have reasonable assurance that the effects of a loss of fuel pool cooling can be mitigated without adverse consequences on the plant.

This discrepancy is a valid deficiency. It is not a safety significant issue because we have established reasonable assurance that the effects of a loss of fuel pool cooling can be mitigated without adverse consequences on the plant and public health and safety and is therefore not reportable. The evaluation above also shows that this concern does not impact plant operability.

In consideration of this concern, additional analyses are warranted to further quantify the effects of evaporation and boiling conditions on the Zone III atmosphere and the potential transport of moist air to other locations in the reactor building for conditions outside of the current design basis.

'ctober 21, 1992 Page 12 Item B: Im act of ESW Hakeu Water "The impact of the ESW makeup water to the spent fuel pool on equipment in the reactor building has not been evaluated."

The analysis under item A above applies to this issue as well. This evaluation shows that with the current plant design and for existing design basis conditions the effects of a loss of fuel pool cooling are acceptable.

This discrepancy is a valid deficiency. As with item A it is not a safety significant issue and is not reportable. The evaluation above also shows that this concern does not impact plant operability.

In consideration of this concern, additional analyses are warranted to further quantify the effects of evaporation and boiling conditions on the Zone III atmosphere and the potential transport of moist air to other locations in the reactor building for conditions outside of the current design basis.

October 21, 1992 Page 13 En ineerin Re ort on Loss of S ent Fuel Pool Coolin A detailed report, SEA-ME-405, is being prepared to document this evaluation in further detail. This report contains technical input from several engineering groups and will provide a comprehensive set of references on this subject. The report will be completed within by October 28, 1992.

October 21, 1992 Page 13 En ineerin Re ort on Loss of S ent Fuel Pool Coolin A detailed report, SEA-HE-405, is being prepared to document this evaluation in further detail. This report contains technical input from several engineering groups and will provide a comprehensive set of references on this subject. The report will be completed by October 28, 1992.

Loss of Fuel Pool Cooling Event Tree Loss ct fPC.

Ycs SKsrnk fvsrlt Yss llIsobts4 wlh Ko LOCA frerst SOTS lrNnkrpt Uss fkS FPC assbl wlh Use ISS fPC assM wlh ss No Ak<<n>>e SOC ~ avaks64, Ak<<n>>s SOC I araksMe, Sakes T<<a Rsserit Otske Pow<<AvaksMet oth>>wls<<. oth<>ls4 ON eke Poe>> AvakaMet ONsks Poe>> AvshMet ltest<~~>4 Ak<<n>>s l SOC svakable.~~

sf vtersL oth<<<<br Yss fleet<<t INAC 4 Ma<<s Useft}SfPC sssM a Atow to bol wkh ESW Rr Suade Accsssbb t s IIISrfSWAccsssbbt FPC wlhna<<Mayst<<ns f skew to bol wkh SW makeup S bk vsrks4 I avalsbb. cth<<wb<<. aakcup d Ttk rsrled.

No Nest<<a FPC wkhn<<rest Akow to bol vrkh Re Use IIIS fPC assbt << Akow to bol wlh fke Uss fpS FPC assM a ctsl sall avaksMe. Rotsc5onrnak>>y S ttk aloe lo bol wkh ESW Raecuonrsak<<kr a tnt akowlobol wlhfSW OO>>wb<<. r>>t e4, ceakeup 4 tkt r<<led. rerks4 aakersr 4 tkt racked.

~~Uss FSS FPC assbt cc aloe to bol wlh ESW~~

~~~ rakerlr 4 ttlv>>ds4~~

~~\ 0 Attachaent 27 PPBL Memo from David A. Loclibaum and Donald C. Prevatte to George T. Jones, "Evaluation of EDR 620020 Reportability/Operability", October 26, l992 (PLI-72739)~~

~~~ ~~~

October 26, 1992 George T. Jones A6-2 SUSQUEHANNA STEAM ELECTRIC STATION EVALUATION OF EDR G20020 REPORTABILITY/OPERABILITY PLI-72739 FILE A45-1A We have read Mr. Myers'emo of October 20, 1992 evaluating the reportability/operability of EDR G20020. We would like to thank both you and Mr. Myers for your considerations in performing these evaluations.

Our overall impression is that, in this document, we have received another confirmation of the technical validity of the concerns that we raised. However, there are still areas of disagreement as to reporting requirements to the NRC, and misunderstanding as to some of the facts associated with this EDR. The following paragraphs address these disagreements and misunderstandings.

First, on reportability. For an operating plant, 10CFR50.72 requires licensees to report in one hour any instance of the plant being (a) in an unanalyzed condition that significantly compromises plant safety; (b) in a condition that is outside the design basis of the plant; or (c) in a condition not covered by the plant's operating and emergency procedures.

If any one of these reportable. 'e criteria applies, the condition is believe that all of these criteria are satisfied by the concerns described in EDR G20020 and other documents that have been communicated to you, to the EDR Group, and to others handling the evaluation of these concerns.

For the first condition, we have pointed out numerous areas where safety-related equipment in the reactor building is not analyzed for the conditions that would result from a boiling spent fuel pool, and where there is high potential that such analyses would show the conditions to be unacceptable, e.g.,

potential flooding of safety-related equipment, potential exceeding of EQ temperatures by large margins, potential pressurization of the reactor building, potential wetting of the charcoal in the SGTS, potential structural failures of ductwork due to condensation, etc..

These examples also illustrate areas where we satisfy the second condition of reportability. For every example, the probable consequences of a boiling spent fuel pool their design bases. are'utside The third condition of reportability is satisfied since we have no plant procedures which address how to cope with the conditions that would be generated by a boiling spent fuel pool, or even recognize that a boiling spent fuel pool would create these adverse conditions.

Legalistic literal interpretations of the CFR have been made that you must actually be in these conditions, that is, you must actually have an accident >n progress where these conditions exist, for,them to be reportable. This is, as we have discussed, a ludicrous argument. The obvious intent of the NRC is that we should never get into these conditions in the first place. Therefore, the common sense interpretation is, any status that would create such a condition were an accident to occur. Using this, interpretation, our status be reportable. 'ould Even if the above described. provisions of 10CFR50.72 were not applicable, another provision i~.- Paragraph (b)(2)(iii, requires that reports shall be made within four hours of any the safety function of structures or systems needed to (a) shut down the reactor and maintain safe shutdown, (b) remove residual heat, (c) control radioactive release, or (d) mitigate the accident. The concerns described in the EDR potentially could have prevented the fulfillment of all four of these.

Again, legalistic arguments have been raised with the fact that, at the time, it was not known for certain that any of these conditions would result. We maintain that the intent of this provision is clear; if it is reasonable to believe, based on our knowledge and experience with our plant and its analyses, that a condition could have prevented the fulfillment of a safety function, then it is reportable.

~~There has been sufficient knowledge to have this reasonable for several months. Zt isn't necessary to cross every "T" and dot every "I" to have that reasonable belief. ln~~

~~'elief addition, other independent formal reports have confirmed our concerns and reinforced this reasonable belief (Reference Mr.~~

~~K. W. Brinkman's report, PLI-72288 of 1/9/92 and Mr. J. R.~~

~~Miltenberger's report, PLI-72367 of 9/9/92).~~

~~The following paragraphs address specific comments made in Mr. Myers'emo:~~

~~First paragraph, we agree with Mr. Myers'onclusion that these concerns <<...should be ~formall brought to the NRC's attention...", and we would like to cite several very important reasons.~~

~~First, it is the law.~~

~~Second, informal reports are not well documented and may be incomplete.~~

~~Third, informal reports don't tend to get the requisite level of attention, either internally or externally.~~

Fourth, formal reports set into motion certain actions, commensurate with the safety significance of the concerns, both internally and externally, that are not necessarily set into motion by informal reports.

~~2 Page 3, Item 1, this section purports to discuss Items 9G~~

~

~~and 9H from the EDR. It actually addresses neither.~~

Item 9G discusses the radiological release from a boiling fuel pool with respect to the increased heat load and the resultant shorter time to boil and increased boiling rate, both of which increase the radiological release.

~~Mr. Myers'iscussion does not, touch on this point.~~

Item 9H discusses the increases in radiological release due to the increase in the number of bundles offloaded (the most significant contributor to the increase in releases) from 184 to 230 (a 25% increase). Mr. Myers defers discussion of this item to another section of his report.

~~He concludes that the existing analysis results are still bounding. Without considering these two factors, it to understand how this conclusion can be is'ifficult reached.~~

3 ~ Page 3, Item 2, this section discusses Items 9E, 9F, 9G, and 9H from the EDR. In this discussion, Mr. Myers concludes that in the pool configuration currently used for refueling, "at the appropriate time post shutdown",

the time to boil would not be significantly different from the 25 hours2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months from the FSAR (This is the same time

currently relied upon by the operators from procedure ON-135-001.). Reference 8 cited by Mr. Myers shows that "the appropriate time post shutdown" does not occur for 14 days. However, until 14 days have past, the time to boil is less than 25 hours2.893519e-4 days 0.00694 hours 4.133598e-5 weeks 9.5125e-6 months (as low as approximately 8.8 hours9.259259e-5 days 0.00222 hours 1.322751e-5 weeks 3.044e-6 months at the beginning of this time) . This is not reflected in the plant procedures.

4 ~ Page 5, in Mr. Myers'ddressing the "Effect of Spent Fuel Pool Boiling on the SGTS", he states that, "The SGTS is apparently designed for the environment (using preheating to reduce humidity)." In fact, SGTS requires this preheating to withstand the environment of a LOCA without fuel pool boiling (inlet conditions of 125 F, 100% relative humidity). It is specifically not designed to accommodate fuel pool boiling conditions (inlet of, 180 F, 1004 relative humidity) as described in Bechtel Calculation 175-17, Rev 4.

Additionally, Mr. Myers'eview only addresses one of the concerns with the SGTS design, the structural integrity of the ductwork. Several other concerns have been communicated in documents and conversations subsequent to the original EDR. These include moisture carryover and/or condensation in the charcoal beds, fusing of the fire dampers in the ductwork (rated at 165 F), exceeding the EQ conditions in the SGTS room, and pressurizing the reactor building, among others, any one of which could incapacitate or degrade the system.

Even with the single concern addressed by Mr. Myers, he concludes that the condition is unanalyzed, and "it is not clear" if the SGTS would function. This alone, per 10CFR50.72(b)(ii)(A), makes the condition reportable.

5. Page 6, Mr. Myers states that "...the reportability question is whether or not the matter is within our licensing design basis." Although this is certainly one of the criteria from 10CFR50.72 that must be considered, there are others that do not appear to be considered in his report. There is also the question of the correctness of the original design criteria. If it is not correct, and as a result, unanalyzed conditions exist which have the po'tential to compromise the integrity or functionality of a safety feature or system in the plant, then that is also certainly reportable.
Mr. Myers also makes the point that historically the NRC did not consider all of the subsequent consequences of design basis events. This is an interesting historical footnote, but it is not pertinent. The consequences should have been considered, and the fact that they were not does not diminish their safety significance or the obligation to report them and correct them.
Mr. Myers cites three points which he concedes "...do raise some questions:"
The first is that MNP-2 (similar in design to SSES) did consider the effects the boiling spent fuel pool, came to the conclusion that the condition was unacceptable, and upgraded their fuel pool cooling system to preclude this condition. This would seem to lend credibility to the contention that the current design for SSES is inadequately analyzed and probably unacceptable.
The second point made is that our latest COTTAP analysis of the reactor building temperatures is conservative because it takes into account the sensible heat from the fuel pool at boiling conditions. Mr. Myers does not appear to understand the problem with the current analysis. The problem is that this analysis considers the sensible heat ~onl ; it does not consider the latent heat released dur'ing boiling which is many time greater than the sensible heat. Considering the sensible heat only, yields a total building heat load of approximately 5.2 million BTU/hr. Considering the latent heat adds approximately 20 million BTU/hr. Clearly, the results of the latest analysis are very non-conservative.
The third point made is that, he understands that: earlier calculations did include considerations of a boiling spent fuel pool. Indeed, this is true; the previously cited Bechtel Calculation 175-17, Rev 4 is an example, and based on the results of these calculations, concluded that the boiling spent fuel pool was not it was acceptable. However, this conclusion was somehow lost and was not. integrated into the original design and licensing of the plant.
Page 7, first paragraph, Mr. Myers states that "... the preponderance of the evidence says that the effect of the boiling environment on both normal and post accident environments was not considered to be a requirement. for the licensing of SSES."
On a
this point we strongly disagree. First of all, requirement; 10CFR50.49 specifically requires that it was electrical equipment be environmentally qualified "for the most severe design basis accidents." In this case, the design basis accident of LOCA/LOOP, or even LOCA without LOOP under the current operating procedures, mechanistically results in a boiling spent fuel pool which produces environmental conditions which are not analyzed and are likely to exceed the current EQ limits for the safety-related systems in the building.
Secondly, the preponderance of evidence is not that was not considered to be a requirement (We have an EQ it program that's evidence of our understanding of the requirement.), but simply that for the SGTS, and in that case, it was overlooked except the effects were found to be unacceptable.
Page 7, in the second paragraph, Mr. Myers begins a discussion of the loss of fuel pool cooling during normal operation. In this and the ensuing two paragraphs he describes analyses that need to be performed'and conditions that would be required to'e maintained for this event which are not addressed by the current design or operating procedures. He concludes that "...given the time involved [until boiling begins after loss of cooling), it appears feasible to take damage control steps to accommodate the impacts."
We strongly maintain that this is not a valid approach to plant design or operations. Design shortcomings are required to be corrected when they are found, not when they are manifested in actual failures, and procedures to address anticipated accident conditions are also requiied to be developed ahead of time, not while the accident is in progress. It is not valid to say that we will develop these at the time of the event as a part of damage control.
Additionally, it should be pointed out that since our letter of October 9, 1992, between ten and twenty engineers have been working in the Allentown office and others at the site, continuously, late nights, and weekends for two weeks to revise the designs and the procedures to justify interim operation. The significance of this is twofold: First, if a JIO is required, this would appear to concede that the existing designs and procedures are inadequate. Second, if such massive and concentrated effort is required to explore
all of the ramifications of making these changes, then it is difficult to imagine how in an accident we could take the right actions or even know what the right actions are, with less people, in a shorter time frame, and under extreme pressure, as a part of "damage control".
The fact that we do not currently have these design features and procedures in place satisfies the conditions for reportability per 10CFR50.72, paragraphs (b)(ii)(A) and (b)(ii)(c) respectively.
Page 8, Conclusion 1, there appears to be no reason given why Item 9I from the EDR is not valid.
Page 8, Conclusion 4, Mr. Myers states that because the effects of fuel pool boiling were not considered as part of the original design basis or licensing basis, 0he concerns are not reportable. This appears to be convoluted logic. Their not being considered in the original design and licensing is, in and of itself, a reportable condition. The effects not being considered is the prime focus of the EDR the reason for concern in the first place. If the original design and licensing bases were not adequate, the fact that it. was not recognized until today does not make them adequate today.
Conclusion 5 appears to agree with this in principle, but not to the point of saying these concerns are reportable.
It is at least gratifying to see that Mr. Myers shares our concern for the potential difficulties of the operators and the need for prompt attention.
Mr. Myers makes the point in the last sentence of Conclusion 4 that these concerns are typical of those uncovered in DBD efforts and implies that DBD concerns are exempt from being reportable.
On this point the regulatory guidance is very clear; NUREG-1397, 2/91, An Assessment of Design Control Practices and Design Reconstitution Programs in the Nuclear Power Industry, states in Section 3.8, Operability and Reportability, "Once the determination has been made that the facility has been or is operating outside its design bases or that systems, structures, and components may be incapable of performing their specified function(s), the requirements for reportability as specified in 10CFR50.72 and 10CFR50.73 become operative and the time clock starts for any affected action
statements as defined in the facility's technical specifications." Additionally, in Section 4.3.2 Reportability, the NUREG states, "The reporting requirements specified in 10CFR50.9, 50.72, and 50.73 apply equally to discrepancies discovered during DDR
[DBD] programs. Therefore, there is no regulatox'y basis to treat discrepancies discovered during the conduct of a DDR program differently than any other reportable item."
10. Conclusion 6 states that the impact of heat and moisture on SGTS should be evaluated immediately, and that, if the conditions are unanalyzed, that is reportable. The fact is that, as described in Item 4 above, one of the concerns was analyzed by Bechtel and found to be unacceptable for SGTS. This should be reportable as Mr.
~~Myers says. The other conditions of concern, as described in Item 4 above, are unanalyzed, and therefore they also should be reportable as Mr. Myers says.~~
~~11. Conclusion 7, we agree that an LER should be produced.~~
12. Mr. Myers'omments in Conclusion 10 are appreciated, and we believe they are sincexe. Unfortunately, they will not be read by the engineers in the tren..hes of the Nuclear Department, of if they are, they will not be believed. To them, this is a test case. Their approach to the EDR System in the future will be governed in large measure by how this issue has been handled, and up until our meeting October 9, 1992, they have seen this issue being brushed off. They will believe what they see, not what they are told.
Again, we appreciate the attention of yourself, Mr. Myers, and all of the others who have been engaged in addressing these concerns. We are gratified that the approach of Mr.
Myers appears to have taken has been more common sensical and less legalistic than others have taken. However, our basic concerns and positions reflected in our letter of October 9, 1992 and subsequent conversations still remain.
As always, we remain at your service.
David A. Loch aum Donald C. Prevatte
cc: C. A. Myers A2-4 G. D. Miller A6-3 R. R. Sgarra A2-4 J. M. Kenny A2-4 J. E. Agnew A6-3 D. F. McGann SSES SGA-4 G. J. Kuczynski SSES H. G. Stanley SSES J. R. Miltenberger A6-1 H. W. Keiser TW-16 R. G. Byram A6-1 W. R. Corcoran 21 Broadleak Circle Windson, CT 06095 J. S. Kemper 115 Polecat Road Glen Mills, PA 19342 R. L. Doty A9-3 A. F. Zorfida SSES A. R. Sabol A2-5 W. R. Licht A6-1 J. S. Stefanko A9-3 F.'. Butler A6-3 A6-3 J. A. Zola M. R. Mjaatvedt A6-3 C. A. Boschetti SSES T J 0 Sweeney
~ SSES G. D. Gogates SSES M. J. Manski Enercon J. D. Richardson Enercon
Attachaent 28 PP8L Memo from David A. Lochbaum.and Donald C. Pr evatte to George T. Jones, "Response to Evaluation of EDR G20020", October 28, 1992 (PLI-72751)
Note: In this memo, the authors responded to PPEL's evaluation of the individual concerns in EDR G20020 (Attachment 26). The authors agreed with the technical justification prepared by PP&L showing that two (2) of the nine concerns were not valid discrepancies. The authors also agreed with PP8L that the remaining seven concerns were valid discrepancies, but strongly disagreed that these valid discrepancies did not affect the, operability of the plant and were not reportable.
October 28, 1992 George T. Zones A6-2 SUSQUEHANNA STEAM ELECTRIC STATION RESPONSE TO EVALUATION OF EDR G20020 PLI-72751 FILE A45-1A We have reviewed Mr. Glenn Miller's most recent evaluation of EDR G20020 dated October 21, 1992 (PLI-72711). While we concur with the technical justification provided for two of the concerns expressed in EDR G20020, we continue to disagree with the operability and reportability assessments for the majority of the concerns. A detailed discussion in response to Mr. Miller's evaluation is attached.
In his memo, Mr. Miller suggested that Nuclear Regulatory Affairs reconsider the reportability aspects of EDR G20020 in light of the clarified information. Since the initial evaluation dated October 6, 1992 for EDR G20020 concluded that none of the concerns was valid and Mr. Miller's evaluation indicates that two of the concerns are valid, we suggest that EDR G20020 be formally revised to reflect the latest PP&L position on these nine concerns.
Mr. Miller also indicated that a justification for interim operation would soon be completed. In Mr. Miller's evaluation, none of the nine concerns affect the operability of the plant and a justification for interim operation would not be required. We agree that the justification for interim operation should be completed, but because current operation of SSES is adversely affected by the concerns identified in EDR G20020.
While we are pleased to have the concerns we raised in EDR G20020 addressed individually, our basic concerns and positions reflected in our letter of October 9, 1992 and subsequent discussions still remain.
As always, we remain at your service.
Ceu I'Reysrrg David Lochbaum Donal C. Prevatte Attachment Response to Evaluation of EDR G20020 (PLI-72711) cc: C. A. Myers A2-4 M. W. Simpson A1-2 H. G. Stanley SSES J. f S. Ste anko A9-3 G. D. Miller A6-3 J. R. Miltenberger A6-1 J. E. Agnew A6-3 D. F. Roth SSES M. R. Mjaatvedt A6-3 Z. M. Kenny A2-4 Nuclear Records A6-2 F. G. Butler A6-3
Response to Evaluation of EDR G20020 (PLZ-72711) 1~ The nine concerns identified in EDR G20020 were listed in order of decreasing nuclear safety significance. Mr. Miller's evaluation rearranges the nine concerns and addresses them in essentially order of increasing safety significance. We have ordered our comments to match the order in EDR G20020.
~~2. We concede that EDR G20020 incorrectly stated that the RHR fuel pool cooling assist mode was not seismically designed.~~
The fact that the fuel pool cooling mode of RHR is seismically designed does not materially change any of the concerns expressed in EDR G20020.
3 ~ Mr. Miller's response to EDR G20020 Item A on reactor building heat loads is inadequate for zany reasons. Mr. Miller states that the only scenario in the SSES design basis for loss of fuel pool cooling is a seismic event. However, the SSES design basis also includes loss of offsite power, LOCA, and failures of non-safety related components which "an each result in loss of fuel pool cooling. Additionally, Mr. Kevin Brinckman in his report dated September 1, 1992 (PLI-72288) indicated that the hydrodynamic loads resulting from a LOCA may result in loss of fuel pool cooling. And finally, Reg Guide 1.13 states that the spent fuel pool shall be designed to maintain adequate cooling of the fuel under all normal operating and postulated accident conditions. Therefore, the SSES design basis implicitly covers failure modes for fuel pool cooling other than the seismic event.
Mr. Miller contends that the loss of fuel pool cooling event coupled with a LOCA or LOOP is outside the SSES design basis.
However, SSES FSAR Chapter 6.2 reports that the LOCA scenario used for containment functional design is postulated to occur simultaneously with a LOOP and a safe shutdown earthquake.
The calculated reactor building heat loads are inputs to the EQ program for safety related components located in the reactor building. Since operation of these safety related components is assumed for core and containment cooling in the containment functional design analyses, it is necessary that the reactor building heat load calculations consider a loss of fuel pool cooling. The latent heat load from a single spent fuel pool is approximately four (4) times greater than the current total calculated reactor building heat load and could result in calculated reactor building room temperatures exceeding E{} values for safety related components.
PP&L's implementation in 1988 of procedures to manually initiate shedding of all non-Class lE loads in the reactor building 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months after a LOCA to control reactor building room temperatures should have been an opportunity to properly
Response to Evaluati.on of EDR G20020 (PLI-72711) address the consequences of a boiling spent fuel pool. The 10CFR50.59 safety evaluation for this activity should have covered the effects of loss of fuel pool cooling, particularly considering the change was made to prevent excessive reactor building room temperatures and loss of fuel pool cooling will adversely affect these same room temperatures.
Mr. Miller states that one SSES response to a loss of fuel pool cooling is the RHR system. The RHR fuel pool cooling assist mode has adequate decay heat removal capacity to handle the load, but it is a non-safety related system which has never been used at SSES and may never have even been successfully pre-operationally tested. Furthermore, use of the RHR fuel pool cooling assist mode is described in the SSES FSAR and SER only to supplement fuel pool cooling for the emergency heat load case (full core offload). And finally, as Mr. Brinckman states in his report dated September 1, 1992 (PLI-72288), use of RHR in the fuel pool cooling assist mode following a LOCA is an unanalyzed condition which may compromise core and containment cooling.
Mr. Miller states that the second SSES response to a loss of fuel pool cooling is to "allow the fuel pool to boil with makeup supplied by ESW with c'onsideration of either SGTS operating on Zone III or providing a vent path from Zone III."
Per Mr. Dave Pai, the SGTS will not operate i.f the fuel pool boils because the fire dampers isolate at inlet temperatures above 165 F and the calculated inlet temperature resulting from a boiling spent fuel pool is =180'F. The normal reactor building ventilation relied upon by Mr. Miller to cool and vent Zone III is a non-safety related function which cannot be relied upon in this manner. Additionally, the current design of the reactor building HVAC system and the standby gas treatment system for the LOCA scenario do not permit the alignment proposed by Mr. Miller. To utilize such an alignment would require extensive analyses to determine its feasibility and the design modifications necessary to accomplish thi.s operation.
Mr. Miller reports that the SSES IPE determined a very low probability of LOOPs lasting over 24 hours2.777778e-4 days 0.00667 hours 3.968254e-5 weeks 9.132e-6 months . Such information would support a justification for interim operation, but cannot be used to eliminate a design requirement. In addition, Mr. Miller's contention that fuel pool cooling would be restored prior to the pool boiling after a LOCA is inconsistent with the assumed LOOP duration specified throughout the SSES FSAR. For example, the design basis for the ultimate heat sink water inventory provides makeup to both boiling spent fuel pools for the 30 days period following the
Response to Evaluation of EDR 020020 (PLI-72711)
SSES design basis LOCA/LOOP, not just 24 hours.
Mr. Miller states that "although the fuel pool cooling system and other non-safety related systems are not specifically analyzed for the effects of hydrodynamic loads it that they will be able to perform their normal functions is expected following a broad spectrum of design basis events" and that "credit for these systems is not needed to meet the design basis". The design response to any postulated design basis accident must not rely on non-safety related equipment.
Additionally, "it is expected" as Mr. Miller states is insufficient rigor for design analyses of nuclear safety functions.
Mr. Miller states that a loss of fuel pool cooling could be handled by allowing the pool to boil, providing makeup from the ESW system, and operating Zone adverse consequences.
III ventilation to avoid We assume that Mr. Miller is conceding that SSES could not have endured a boiling spent fuel pool with the design features and procedures currently in place without significant adverse consequences.
We agree with Mr. Miller that EDR G20020 Item A is a valid deficiency. We strongly disagree with Mr. Miller's contention that this valid deficiency has no safety significance and is not reportable. If the fuel pool boils, the existing reactor building heat load calculations do not account for the latent heat load.
4 ~ Mr. Miller's response to EDR 620020 Item B on the impact of ESW makeup water is inadequate for the reasons as given in Comment 3 and because his response does not address all of the potential impacts.
The ESW flow supplied to the fuel pool is controlled by manually positioning a throttle valve. If the ESW flow rate to the fuel pool exactly matches the boil-off rate, then the level in the fuel pool will be maintained constant.
ESW flow rate is lower or higher, then the fuel pool level If the will drop or rise accordingly. The most probable outcome will be for more ESW flow than is required to be'supplied to the fuel pool. Under this scenario, both the moist air from fuel pool boil-off and the water from fuel pool run-off must be considered. The adverse nuclear safety consequences include pressurization of the refueling bay and/or secondary containment, flooding, component failure due to humidity and condensation, and HVAC ductwork failures due to either flow blockage from condensed vapor or collapse from the added water weight. As stated in Comment 3, the SSES ultimate heat sink
Response to Evaluation of EDR G20020 (PLI-72711) design analysis covers ESW makeup to both spent fuel pools throughout the 30 day period following the SSES design basis LOCA/LOOP. However, this analysis is incomplete and invalid because the consequences of the 5 million gallons of water delivered to the spent fuel pools on systems and components in the reactor building is not taken into account.
We agree with Mr. Miller that EDR G20020 Item B is a valid deficiency. We strongly disagree with Mr. Miller's contention that this valid deficiency has no safety significance and is not reportable. If ESW is supplied to a boiling fuel pool, the consequences are virtually unanalyzed. For example, the boil-off might result in pressurizing the refueling bay and challenging secondary containment integrity. As Mr. Miller states, "additional analyses are warranted to further quantify the effects of evaporation and boiling conditions on the Zone III atmosphere and the potential transport of moist air to other locations in the reactor building." However, these analyses are needed to support SSES operation within its existing design basis.
Mr. Miller's response to EDR G20020 Item C on manual ESW valve actions is inadequate because it is based on predictions of no more than 14 fuel failures. Mr. Miller's logic that recent evaluations of actual fuel thermal response during design basis accidents indicate no fuel failures occur would support tearing down secondary containment and removing the standby gas treatment system if it were justified.
Mr. Miller states that the radiation levels determined at SSES in response to NUREG-0737 requirements are "clearly based on degraded core conditions which are beyond the design basis LOCA." Nevertheless, the requirements in NUREG-0737 were imposed by the NRC following the TMI accident and are clearly within the SSES licensing basis.
Mr. Miller claims that "airborne radioactivity sources from containment leakage are required to be analyzed for environmental qualification of equipment but not for personnel access." This claim is preposterous, and we believe that the SSES Operations staff does not support this position. If airborne radiation sources are considered, and they would be y g ~ ~
building could be entered, severe core damage is not
g'eactor required for the reactor building to be rendered inaccessible.
Therefore EDR G20020 Item C is a valid deficiency because the makeup supply to a boiling spent fuel pool may not be available post-LOCA due to inaccessibility of the ESW manual
Response to Evaluation of EDR G20020 (PLI-72711) valves.
Mr. Miller's response to EDR G20020 Item D on fuel pool instrumentation is inadequate because that it instrumentation is incomplete. Reg required for Guide 1.97 requires initiating and monitoring safety functions be qualified. The fuel pool instrumentation may be powered from Class 1E sources, but instrumentation it will has not function been established in the environment that this in which it will be exposed post-LOCA. In addition, Mr. Brinckman in his letter dated September 1, 1992 (PLI-72288) reported that "the fuel pool trouble alarm in the control room cannot be counted on for reliable indication." The readouts of fuel pool level and temperature are at local panels in the reactor building and would also be inaccessible post-LOCA. Mr. James Miltenberger in his letter dated September 9, 1992 (PLI-72367) indicated that the fuel pool level instrumentation needs to be upgraded to provide reliable control room indication.
Mr. Miller also states that the "instrumentation available to the operator is not required to be qualified since the design basis loss of spent fuel pool cooling is not coincident with the DBA LOCA conditions." However, Reg Guide 1.13 requires adequate cooling of the spent fuel pool to be available for all normal operating and postulated failure accident conditions.
occur due to the Since a fuel pool cooling can hydrodynamic loads or environmental conditions associated with the design of the design basis LOCA, basis. In it should addition, have if been the considered existing as part unqualified instrumentation provides a false indication of increasing fuel pool temperature or dropping fuel pool water level post-LOCA, personnel may be unnecessarily exposed to radiation as they enter the reactor. building and refueling floor area to respond to the perceived threat.
EDR G20020 Item D is a valid deficiency because the fuel pool instrumentation may not be adequate to provide the operator with sufficient information to implement and monitor required safety measures.
Mr. Miller's response to EDR G20020 Item E on the analytical time to boil for the maximum normal heat load case is inadequate because it does not address the prob)em reported.
EDR G20020 Item E did not dippute the 9.79x10 BTU/hr heat load value versus the 12.6x10 BTU/hr value, but rather that the maximum normal heat load upon which the time to boil calculation was based was rendered non-conservative by changes in fuel types and operating cycle lengths. EDR G00005 was written in 1990 on the subject of outdated FSAR Chapter 9
Response to Evaluation of EDR G20020 (PLI-72711) data. Item E was included in EDR G20020 to identify another consequence of the problem originally reported in EDR G00005.
It is our understanding that EDR G00005 has not yet been resolved over two years after it was initiated.
EDR G20020 Item E is a valid deficiency because the basis for the time to boil calculation reported in the FSAR is invalid, but as we clearly stated in our memo dated June 22, 1992 to Joe Zola (ET-0471) it does not affect the present operation of SSES because the existing decay heat loads in the fuel pool are less than 9.79x10 BTU/hr. This deficiency represents a potential safety concern because the maximum normal heat load at SSES may exceed 12.6x10 BTU/hr when the spent fuel pool is filled to capacity.
Mr. Miller's response to EDR G20020 Item F on the time to boil for the emergency heat load case is inadequate for many reasons. Mr. Miller states that "no time-to-boil calculation t
for this configuration [emer e c he lo d case] is required since the RHR system will be in operation or available." The time to boil calculation for the maximum normal heat load case is provided in FSAR Appendix 9A even though the fuel pool cooling system is initially operating because its failure must be considered. An equivalent calculation must be provided for the emergency heat load case because the RHR fuel pool cooling assist mode could fail. As Mr. Miller points out, this calculation should consider the additional water inventory available. But this calculation must also account for operational events such as putting the fuel pool gates in to isolate the reactor cavity from the spent fuel pool volume.
Mr. Miller states that "single failures of the RHR system are not required for this design basis for the emergency heat load" and references SRP 9.1.3. SRP 9.1.3 indeed supports the position that a single active failure need not be considered for the emergency heat load case. However, since the emergency heat load case is defined as a full core offload in SRP 9.1.3 and SSES FSAR Chapter 9 and refueling operations described in the SSES FSAR do not entail full core offloads, then PP&L's routine use of full core offloads conflicts with the FSAR and increases the probability of "new" and "unanalyzed" events with consequences potentially more severe than the analyzed event.
It is our understanding that the RHR fuel pool cooling assist mode has not been used at. SSES and was not even successfully pre-operationally tested.
Response to Evaluation of EDR G20020 (PLI-72711)
EDR G20020 Item F is a valid deficiency because current SSES refueling operations routinely place the "emergency heat load" in the spent fuel pool without a corresponding time to boil analysis.
~~9. Mr. Miller's response to EDR G20020 Item G on the radiological release calculation is inadequate for the reasons stated in Comment 7.~~
EDR G20020 Item G is a valid deficiency because the basis for the time to boil calculation reported in the FSAR is invalid, but as we clearly stated in our memo dated June 22, 1992 to Joe Zola (ET-0471) it does not affect the present operation of SSES because the existing decay heat loads in the fuel pool are less than 9.79xl0 BTU/hr.
10. Mr. Miller's response to EDR G20020 Item H on nonconservative activity terms in the radiological release calculation is adequate. Based on the technical justification provided by Mr. Miller, we agree that EDR 620020 Item H is not a valid deficiency.
Mr. Miller's response to EDR G20020 Item I on maximum time prior to makeup is adequate. We agree that EDR G20020 Item I is not a valid deficiency, but we recommend that calculation 175-14 be either revised to clarify its purpose and usage or deleted.
2. Mr. Miller focused much of his evaluation on the ability of SSES to withstand a loss of fuel pool cooling during a refueling outage. EDR 620020 did not emphasize this aspect to the same degree, although many of the problems are just as pertinent under this condition. In fact, the current SSES practice of performing full core offloads each refueling outage places the station in a very vulnerable (and unanalyzed) condition. At the point when the common RHR system outage is entered on the unit in refueling, the operating unit's fuel pool cooling system is handling the entire heat load from the cross-tied fuel pools. A design basis LOCA/LOOP at this time subjects the station to a loss of fuel pool cooling at a time when one unit's RHR system is totally unavailable and the remaining unit's RHR system is dedicated to core and containment cooling functions.
realistic events such as the single failure of one RHR on the If operating (LOCA) unit and/or installation of the fuel pool gates to the reactor cavity on the unit in refueling are considered, the potential consequences can be quite severe.
In any case, this condition is routinely entered by SSES
'ithout the necessary analyses to support it.
Attachwent 29 PP8L Memo from Glenn D. Miller to George T. Jones, "Evaluation of EDR G20020 - Spent Fuel Pool Cooling Issue", October 29, 1992 (PLI-72763)
~ ~ ~
October 29, 1992 George T. Jones A6-2 SUS(UEHANNA STEAN ELECTRIC STATION EVALUATION OF EDR G20020 - SPENT FUEL POOL COOLING ISSUE PLI-72763 FILE A45-lA Attached, please find a copy of NE-092-002, Rev. 0 "Loss of Fuel Pool Cooling Event Evaluation". This evaluation is provided in response to your memo PLI-72640. This report is intended to supplement my previous evaluation of EDR G20020 transmitted in PLI-72711.
This document evaluates the various accident scenarios identified by EDR G20020 for which a loss of fuel pool cooling can be expected. In addition, this document provides pertinent design basis information with regard to the Fuel Pool Cooling system and recovery from a loss of Fuel Pool -Cooling event.
This report concludes that recovery from all design basis accidents is possible without compromising safety related equipment, while assuring that the spent fuel pool will remain sufficiently cooled. Therefore, it can be concluded that the plant is operable and continued safe operation of the plant is assured with regard to the concerns raised in EDR GZ0020.
This report also recommends a number of immediate and short term actions (see Section 5.0) which will enhance plant safety and reduce the risk to the plant environment resulting from the events studied. Implementation of these actions is required to support the conclusions reached in this evaluation.
If you have any questions regarding this evaluation, please contact me at your convenience.
Glenn D. Miller CC: G. J. Kuczynski SSES J. E. Agnew - A6-3 C. A. Hyers - A2-4 H. R. Mjaatvedt - AS-3 H. M. Simpson A1-2 D. F. Roth SSES H. G. Stanley SSES J. M. Kenny - A2-4 J. S. Stefanko - A9-3 F. G. Butler A6-3 J. R. Miltenberger A6-1 D. C. Prevatte - A6-3 Nuclear Records - A6-2 D. A. Lochbaum - Enercon
Attachaent 30 PP8L Engineering Report, "Loss of Fuel Pool Cooling Event Evaluation for EDR 8620020", October 29, 1992 (NE-92-002 Rev. 0)
Note: This report documents the extensive engineering effort undertaken by PP8L to explore the loss of fuel pool cooling event. It recommends substantial modifications and procedure changes in order for the units to handle a loss of fuel pool cooling under all design conditions. This report concludes that the operability of the plant is not affected, but seems to base this conclusion on the conditions ~1~
all the recommended modifications and procedure changes are implemented, not as the plant exists currently.

ML17157C140

Contents

Text

Reference:

SUBJECT:

Dear Mr. Jones:

Subject:

SUBJECT:

Dear Mr. Jones:

Navigation menu

ML17157C140

Text

Reference:

SUBJECT:

Dear Mr. Jones:

Subject:

SUBJECT:

Dear Mr. Jones:

Navigation menu

Search