ML20149J801

Forwards Evaluation of Capability of Prairie Island Units to Mitigate Complete Loss of Feedwater Atws,Including Results of Supporting Analyses Performed by W & NSP Nuclear Analysis Group,As Requested by NRC Staff

ML20149J801
Person / Time
Site:	Prairie Island
Issue date:	07/23/1997
From:	Sorensen J NORTHERN STATES POWER CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NUDOCS 9707290103
Download: ML20149J801 (63)

Text

.

Northern States Power Company Prairie Island Nuclear Generating Plant 1717 Wakonade Dr. East Welch, Minnesota 55089 July 23,1997 U S Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 20555 PRAIRIE ISLAND NUCLEAR GENERATING PLANT Docket Nos. 50-282 License Nos. DPR-42 50-306 DPR-60 Evaluation of Capability to Mitiaate a Complete Loss of Feedwater ATWS Attached for the information of the NRC Staff (Attachment 1) is an evaluation of the capability of the Prairie Island units to mitigate a complete loss of feedwater Anticipated Transient Without a Scram (ATWS). The attached evaluation includes the results of supporting analyses performed by Westinghouse and the NSP Nuclear Analysis Group.

The evaluation is provided at the request of the NRC Staff.

When Attachment 1 was completed, the modification to increase the Auxiliary Feedwater Pump discharge pressure switch setpoints had been completed on Unit 2.

At that time the modification had not yet been completed on Unit 1. The evaluation was based on this configuration. Subsequently, the modification has been completed in j Unit 1. Therefore, the Unit 1 response to a complete loss of feedwater ATWS would be l sirnilar to the Unit 2 response discussed in the evaluation.

When Attachment 1 was completed, it was believed that the corrective actions could be I performed during the upcoming refueling outages. These plans have been superseded by the schedule to implement the diverse scram system modification discussed during our meeting with the NRC Staff on July 9,1997.

When Attachment 1 was reviewed by the Plant Operations Committee, the l Westinghouse analysis attached to the evaluation was preliminary. Approval of the evaluation was contingent on receiving the final Westinghouse analysis. The final Westinghouse analysis (Attachment 2) was received later that day with no changes from the preliminary analysis reviewed by the Plant Operations Committee. , (

hVAs l 9707290103 970723

!i i

USNRC ig July 23,1997 In this letter we have made no new Nuclear Regulatory Commission commitments.

Please contact Gene Eckholt (612-388-1121) if you have any questions related to the attached evaluation ,

///

Joel P Sorensen Plant Manager Prairie Island Nuclear Generating Plant c: Regional Administrator- Region Ill, NRC Senior Resident Inspector, NRC NRR Project Manager, NRC J E Silberg Attachments: 1. Evaluation of Capability to Mitigate a Complete Loss of Feedwater ATWS

2. Final Westinghouse Evaluation of ATWS for Prairie Island with Reduced AFW Flow 4

4 1

is i

OPEREVAI.. DOC

USNRC

! 4 July 23,1997 i

! Attachment 1 1

Evaluation of Capability to Mitlaate a Complete Loss of Feedwater ATWS i

I 4

4 F

i 1

i s

t i

0 4

t J

i a

i i

i i

a

, -3. w-r ,,y,re ,.,2- --

i.-- --- . ,.,-, . _, -,,7 . - , .----,--r -..-. , ,- , -

~

COMPLETE LOSS OF FEEDWATER ATWS t

1.0 PURPOSE ,

The purpose of this evaluation is to provide reasonable assurance that the plant is capable of mitigating a complete loss of feedwater ATWS. With the current configuration for the control circuitry, during a complete loss of feedwater ATWS, the motor driven auxiliary feedwater pump would trip after approximately 140 4 seconds into the transient, leaving the turbine driven auxiliary feedwater pump to mitigate the event. This evaluation is only for an interim time period until permanent corrective actions can be implemented.

Corrective actions are scheduled for implementation in Unit 1 prior to restart from the upcoming Unit 1 refueling outage (Fall 1997). Corrective actions are

. scheduled for implementation in Unit 2 prior to restart from the next Unit 2 refueling outage (1998). These actions may not be the final permanent corrective actions.

l

2.0 BACKGROUND

in May,1996, a deficiency in the Auxiliary Feedwater System was discovered where the pumps were not adequately protected from potential damage due to runout for some postulated scenarios. The immediate action was to assign a Control Room operator whose response to the event is to throttle the AFW Pump discharge valves to maintain pump discharge pressure above a pre-set value to preclude runout (cavitation) from occurring.

Subsequent to that time frame, investigation has been performed to develop a means which relies on equipment operation in lieu of operator intervention (as much as possible) to provide this protection. Initially it was thought that flow restricting orifices could be installed which would sufficiently limit the flow to prevent the auxikary feedwater pumps from operating in the runout regimes of the pump curve. However, hydraulic modeling of the AFW System has determined that, with the additional resistance due to the orifice, other design assumptions will not be satisfied. Specifically, the system would not be capable of providing 200 gpm with the Steam Generators at the Safety Valve lift pressure.

Further investigation into available short term equipment related solutions indicates that increasing the low discharge pressure switch setpoint will improve the protection currently provided. For specific scenarios, the increased low discharge pressure setpoints will result in the auxiliary feedwater pumps tripping 4/197 NSP ATWSEVALDOC 1 1

- - - y n ,

~

e to protect them from runout. Per the Safety Analysis Report, the operator would then restart the pump (s). Thus, the increased setpoint for the low discharge pressure switch will provide protection for the pumps and is consistent with the description in the USAR and the NRC SER for installing the low discharge pressure switches. For the turbine driven auxiliary feedwater pump an additional ~

modification has been implemented to ensure that it remains in operation for~a period of time longer than the motor driven auxiliary feedwater pump during the complete loss of feedwater ATWS.

. In Unit 2, the discharge pressure switch setpoints have been changed and the modification to the control circuitry for the turbine driven auxiliary feedwater pump has been implemented.

. in Unit 1, these changes have not been implemented yet The discharge [

pressure switch for the motor driven pump is still set at 500 psig. The  :

discharge pressure switch setpoint for the turbine driven pump is still set at 200 psig. During a complete loss of feedwater ATWS, the motor driven pump would trip at approximately 140 seconds.  :

This evaluation will focus on the loss of feedwater ATWS event. The discussion below refers to the Westinghouse generic analysis and sensitivity studies performed to determine the effect on RCS peak pressure.

3.0 EVALUATION An ATWS event is a postulated operational transient which is accompanied by a failure in the reactor protection system to shutdown the reactor. The analytical basis for the ATWS event assumes that at no time is automatic reactor scram or -

control rod insertion initiated. All other components, equipment, and systems are assumed to operate normally during the event provided that (as discussed in Reference 3):

a. Failure of the equipment, component, or system is not the cause of the transient being analyzed;

b. The function of the equipment, component, or system is not disabled as a consequence of the transient being analyzed;

c. The probability of failure of the component, equipment, or system is reasonably small during the interval of the transient being analyzed.

j This is consistent with Reference 4, Volume 3, which does not require application of the single failure criterion to the systems employed in mitigation of ATWS events.

W e

om NSP ADVSEVALDoc 2

~ .

4 Regarding ATWS events,10CFR 50.62 requires:

"Each pressurized water reactor must have equipment from sensor output to final actuation device, that is diverse from the reactor trip system, to automatically initiate the au'xiliary (or emergency) feedwater system and initiate a turbine trip under conditions indicative of an ATWS."

NUREG 0460 contains criteria for systems required to mitigate the consequences of an ATWS event. Section 7.1.6 states, in part:

"These systems shall be automatically initiated when the conditions monitored reach predetermined level and continue to perform their function without operator action unless it can be demonstrated that an operator would reasonably be expected to take correct and timely action."

NUREG 0460, Section 7.1.6, further indicates that operator action should not be relied upon for the first 10 minutes of the accident. Operator action is acceptable if it is shown that information on the conditions in the reactor and of the mitigating systems is available to the operator, sufficient time is available to correctly assess the situation and take appropriate action, and that the operator is trained in the proper actions.

ATWS events are not considered design bases events. Since these were not considered design bases events, plant specific analyses were not required. The information pertinent to ATWS presented in the USAR is from the ATWS submittal from Westinghouse to the NRC in 1979. This is a generic submittal intended to bound Westinghouse plants.

Five different ATWS events are included in the generic Westinghouse analysis (Reference 3):

l . . Loss of External Load

. Complete Loss of Normal Feedwater

. Loss of Off Site Power

. Accidental Depressurization of the RCS

. Uncontrolled Rod Withdrawl at Power i Auxiliary Feedwater is not required to mitigate the Accidental Depressurization of the RCS ATWS or the Uncontrolled Rod Withdrawl at Power ATWS. During a Loss of External Load ATWS, the main feedwater pumps are still available as the PINGP has motor driven main feedwater pumps. For a loss of off site power ATWS scenarios, the turbine is assumed to trip at time zero; which is bounded by the complete loss of feedwater ATWS where the turbine trips at 30 seconds into the event.

&5.97 NSP ATWSEVR. Doc 3

_ . ~ . _ _ _ _ -

For the complete loss of feedwater ATWS, SG pressure decreases as secondary side inventory is reduced, and the motor driven auxiliary feedwater pump would trip. The SG pressure profile (USAR, Figure 14.8-19) shows that SG pressure initially increases to the Safety Relief valve setpoint. After the secondary side mass decreases, the SG pressure decreases. USAR, Figure 14.8-19, indicates that SG pressure will decrease below 700 psia at 126 seconds of the event. Approximately 15 seconds after this point the motor driven auxiliary feedwater pump will trip (the 15 seccnds account for the time delay associated with the low discharge pressure switch). The turbine driven pump would trip or lose driving force at secondary side pressures below approximately 200 psig. This is conservative as testing has shown that the turbine driven pump can continue to provide flow at steam pressures below 100 psi.

3.1 Probabilisitic Evaluation From a probabilistic perspective, the loss of main feedwater ATWS initiator is a low probability event. The probability at Prairie Island of a transient causing a complete loss of main feedwater followed by failure of the reactor protection system to trip the control rods into the reactor is approximately 4.2E-6/yr. This number is low because of the high reliability of the reactor protection system and  ;

the low transient initiating event frequency at Prairie Island. Evaluation of ATWS l events were also included in the IPE submittal (required by Generic Letter 88-

20) for their contribution to the overall core damage frequency. The ATWS evaluation was based on the methodology of WCAP 11933 (Joint Westinghouse Owners Group / Westinghouse Program: Assessment of Compliance With ATWS l Rule Basis for Westinghouse PWRs).

NUREG 0460 forms the basis for the ATWS rule making (10 CFR 50.62). In NUREG 0460, the NRC states:

"The staff has maintained since 1973 and reaffirms today that the present likelihood of severe consequences arising from an ATWS event is acceptably small and presently there is no undue risk to the public from ATWS. This conclusion is based on engineering judgment in view of: (a) the estimated arrival rate of anticipated transients with potentially severe consequence in the event of scram failure; (b) the favorable operating ,

experience with current scram systems; and (c) the limited number of operating reactors."

As further stated in NUREG 0460 as the population of reactors was expanding the NRC considered the rulemaking riecessary to overcome the aggregate risk increase. However, this does not reduce the significance of the NRC 4/5 97 NSP ATWSEVAUXX' 4

~

• ~ -

_ _ 1 I

conclusions regarding the low probability of the anticipated transient nor the high probability of proper functioning of the scram system: -

During the ATWS rule making, the NRC considered imposing shorter term actions on utilities to attain compliance. However, in lieu of trying to attain the shorter term corrective action, the staff elected to proceed with the full rule making procedure. As justification for using the time consuming rule making process to attain corrective action, the staff relied on the low probability of the events as the basis for not issuing a generic letter or bulletin requiring quicker response from licensees (Reference 4).

3.2 Reactor Trip Reliability A complete loss of normal feedwater event will produce a reactor trip signal from one or more of the following (Reactor Protection Design Bases Document):

. OTAT

• High Pressurizer Pressure

. High Pressurizer Level

. Lo-Lo Steam Generator Level Thus, there are several diverse signals available to generate the reactor trip.

The reactor trip system is considered highly reliable. Following an actual ATWS event at Salem in the early 1980s, the NRC issued Generic Letter 83-28. The purpose of Generic Letter 83-28 was to improve the reliability of the reactor trip i systems at nuclear power plants. Generic letter 83-28 requested licensees to take actions in four general areas:

a. Post-Trip Review The purpose of this action was to ensure that the causes for unscheduled

! shutdowns were fully understood prior to plant restart.

b. Equipment Classification and Vendor interface The purpose of this action was to ensure that components necessary for accomplishing required safety related. functions were properly identified and controlled. In addition, assurance was provided that vendor information for safety related components is complete. i

c. Post-Maintenance Testing  !

1 1

4.'197 NSP ATWSEVALDoc 5

l l

The purpose of this action was to ensure that post maintenance testing of safety related components were adequately identified and performed.

d. Reactor Trip System Reliability Improvements The purpose of this action was to assure that vendor recommended reactor trip breaker modifications and associated reactor protection system changes were completed (this required shunt trips for Prairie Island). In addition to the modifications, this action was aimed at assuring that a comprehensive program of preventive maintenance and surveillance testing was implemented for the reactor trip breakers.

USAR, Table 7.4-6, summarizes all the actions completed in response to these actions and the associated NRC Safety Evaluation Report which reviewed and approved the NSP response.

The addition of the reactor trip breaker shunt trips (safety related) provided an additional diverse means for tripping the reactor trip breakers. These enhancements to system design, operation, maintenance and testing further increased the reliability of the reactor trip system.

The reliability of the reactor trip system provides a high level of confidence that the system would perform as designed when called upon, and that during a complete loss of feedwater event, the reactor would automatically trip.

As discussed in Section 3.1, the NRC denotes in NUREG 0460 that scram systems are very reliable. This statement was made well before the actions were initiated in response to Generic Letter 83-28. These actions have further improved this reliability.

Based cn the reliability of the reactor trip breakers (RTB), the NRC issued Supplement No.1 to Generic Letter 83-28 (Reference 21) which canceled two of the actions initially considered necessary (life testing of trip breakers and program for preventative replacement). In this letter the NRC states:

"In light of this RTB operating experience, the staff has concluded that the actions already completed pursuant to GL 83-28 have been effective in improving RTB reliability to open and that further actions to address the end of life degradation in breaker relia.bility are not justified."

3.3 Conservatisms in the Generic Analysis:

The complete loss of feedwater ATWS analysis is based on a generic ,

Westinghouse ATWS analysis. As such there are several conservatisms in the om Nsr wrWSEVEDOC 6

analysis which were used to bound all Westinghouse plants. Some of these conservatisms and the expected impact en a plant specific basis for the PINGP are discussed below.

t

a. Loss of Feedwater event Loss of normal feedwater could result from a malfunction in the feedwater or its control system, a loss of Instrument Air, etc. The majority of the plant transients which effect feedwater would only cause a partial loss of feedwater (e.g., one main feedwater pump trips). With a partial loss of feedwater, the remaining operating portions of the feedwater system would be available to mitigate the event. That is for a partial loss of feedwater, the auxiliary feedwater system is not required to mitigate the event.

b. Decay Heat vs. Operating History The ATWS analysis assumes beginning of core life considerations for conservatism. The decay heat is also conservatively assumed to be due to a core that has been operating for 8000 hours0.0926 days <br />2.222 hours <br />0.0132 weeks <br />0.00304 months <br />. Although used in the 1 analysis, these two assumptions are mutually exclusive. By combining l these two assumptions in the same analysis, the reactor heat is increased which worsens the transient.

c. Turbine Trip For the loss of feedwater ATWS analysis, the Turbine is assumed to be l 4

tripped at 30 seconds into the event by the signal from ATWS Mitigation System Actuating Circuitry (AMSAC). However, there are other trips (loss of main feed pumps) which could cause the turbine trip much sooner in the event. The continuation of turbine operation, assumed in the analysis, results in a loss of SG mass much sooner. An earlier turbine trip would leave more mass in the SG, and increase the time before the secondary side inventory is decreased.

d. Auxiliary Feedwater Initiation For the loss of feedwater ATWS analysis, Auxiliary Feedwater is assumed to be initiated at 60 seconds into the e. vent by the signal from AMSAC.

The AMSAC design at the PINGP willinitiate auxiliary feedwater at 27.5 seconds into the event (at full power). In addition, there are other signals l (loss of main feed pumps, low-low SG level) which could initiate AFW much sooner. The earlier initiation of AFW would result in an increase in available SG secondary side inventory; which would increase the time before the secondary side inventory would be decreased.

4497 NSP ATWSEVALDOC 7 l

'N _

.y _

- ~

e. Auxiliary Feedwater Purge Volume

. The loss of feedwater ATWS analysis assumes that auxiliary feedwater is injected into the feedwater pipe at a temperature of 130F, 500 cubic feet upstream of the steam generators (from discussion in Section 5.2 in Reference 3), such that the cooler water enters the steam generator after i this volume is purged. Reference 3, Table 3-1-a, indicates that for a two loop plant a purge volume of 261 cubic feet was used. A review of piping i isometrics show that the actual volume at the PINGP is less than 2 cubic feet upstream of the steam generators. This means that the cooler

- auxiliary feedwater is injected into the SG significantly earlier than for the generic analysis. This would remove more heat from the reactor coolant system and reduce the transient. This has been factored into the analysis (Reference 23).

i

f. Auxiliary Feedwater Injection Temperature The loss of feedwater ATWS analysis assumes that auxiliary feedwater is l injected at an inlet temperature of 130F (enthalpy of 100 Btullbm). The i maximum design AFW injection temperature at the PINGP is 100F (enthalpy of 68 Btullbm). This would remove more heat from the reactor

coolant system and reduce the transient. This has been factored into the analysis (Reference 23).

g. Steam Generator Initial Mass The loss of fers ater ATWS analysis assumes an initial SG fluid mass of l

101,600 lb. At normal full power operation (which is limiting) the nominal  !

fluid mass is 107,000 lbm (Reference 26). This provides increased i available SG secondary side inventory; which would increase the time before the secondary side inventory would be decreased.

i

h. Automatic Rod Control 1

The ATWS analysis assumes that automatic rod controlis not available.  !

l Reference 22 states:

! "A benefit would be realized from control rod motion, if the rod 1 control system is in the automalic mode, no credit is taken for this benefit because the rod control system may not be in the automatic mode when ATWS occurs."

1 With the rod control system in automatic during a complete loss of feedwater ATWS, the control rods would move in adding negative 4/5/97 NSP ATWSEVALDOC O

~

~

reactivity and decreasing the severity of the event. With the reactor at full power, the rod control system is in automatic the majority of the time.

Manual rod control is only used for specific scenarios (e.g., failure in the rod control system, repositioning. etc.). Thus, the time period that the rod control system is in manual is very small. If the time period on a yearly average that the rod control system was in manual is coupled with the probability of a complete loss of feedwater ATWS occurring (see Section 3.1), the probability of the event occurring with this assumption would be significantly smaller.

i. Peak RCS Pressure Acceptance Criteria The complete loss of feedwater ATWS is limited to an acceptance criteria for reactor coolant system internal pressure of 3200 psig maximum. 3200 psig corresponds to the Emergency Condition (Level C) maximum allowable pressure per ASME, Section Ill, based on the reactor vessel nozzle safe ends and studs. This is very similar to the limits in the USAR, Table 12.2-12. During the late 1960s, Westinghouse performed work in determining the ultimate strength criteria for stainless steel clad vessels and piping. For typical reactor vessel materials the design failure stress (corresponding to 50% total strain) is well above the yield stress for the material. This indicates that there is significant margin between the acceptance criteria and the failure point for the reactor vessel nozzle safe ends and studs.

Therefore, there are several conservatisms in the generic analysis vs. the PINGP design. Use of the plant specific values in lieu of these conservatisms will reduce the severity of the transient by (1) Increasing the time available before SG pressure decreases sufficiently to trip the motor driven AFW Pump, and i

(2) Resulting in less stored heat in the RCS at the time of the motor driven AFW pump being tripped, and (3) Provide assurance that additional margin exists in the allowable RCS internal pressure above the acceptance criteria.

3.4 Expected Plant Response The information pertinent to ATWS presented in the USAR is from the generic ATWS submittal from Westinghouse to the NRC in December 1979 (Reference 3). The analyses in Reference 3 were performeo for a four loop plant, intended to bound Westinghouse plants. ATWS are not design bases events. Since (5s7 sse ATwstv.u.coc 9

~

l .

ATWS are rot design bases events, plant specific analyses were not required.

Without a p. ant specific analysis, it is not possible to definitively determine the

! plant response. However, the plant's response can be evaluated within the I . bounds of the Reference 3 analysis. '

l In resporse to a complete loss of feedwater' ATWS event, both the motor driven and turb'ne driven auxiliary feedwater pumps start. The ATWS analysis 4

assumes this occurs at 60 seconds into the event. As previously discussed, this will ocr.ur much earlier (Section 3.3.d, above) in this scenario at the PINGP. At i approximately 130 seconds into the event, the Reactor Coolant System (RCS) '

' reaches the peak pressure condition. The RCS pressure then decreases quickly. Approximately the same time into the event, the Steam Generator pressure decreases to approximately 700 psia. In Unit 2, approximately 15 seconds after this point the motor driven Auxiliary Feedwater Pump would trip on low discharge pressure (the 15 seconds accounts for the time delay associated with the low discharge pressure switch). In Unit 1, the motor driven pump low discharge pressure trip occurs at 500 psig with a time delay. Thus, the trip would occur a few seconds later in Unit 1 than in Unit 2. This trip protects the ,

pump from possible damage due to runout, and allows the pump to be available for the Operator to restart later in the event. Tripping the Auxiliary Feedwater Pump is consistent with the Safety Analysis Report for the function of the low discharge pressure switches. The turbine driven Auxiliary Feedwater Pump continues to operate until Steam Generator pressure further decreases to less than approximately 200 psig. In Unit 1, the turbine driven auxiliary feedwater l pump remains operating due to the lower discharge pressure setpoint. In Unit 2, i the turbine driven auxiliary feedwater pump remains operating due to the Reactor Trip Breaker closed signal bypass of the discharge pressure switch trip signal (Modification 96AF01, Part 2, implemented Unit 2 Refueling Outage in  :

1997).

The Westinghouse analysis for a complete loss of feedwater ATWS (Reference

3) contains several sensitivity studies showing the effect on the analysis of varying the number of AFW Pumps, the time of initiation of the AFW System, and degrading the AFW flow. The AFW System at the PINGP is rated at 50% of the capacity assumed in the Westinghouse analysis (Reference 3). Thus, q starting both pumps is_ equivalent to the sensitivity study performed assuming the largest AFW Pump fails to start. In Reference 3, the peak RCS pressure in the reference case is 2753 psia (2 loop,100% AFW flow). According to Westinghouse, this peak pressure was incorrectly reported; i.e., this is the peak pressure for the loss of load ATWS. Peak RCS pressure for the complete loss of feedwater ATWS ( 2 loop,100% AFW flow) should have been reported as 2711 psia.

The attached Westinghouse evaluation (Reference 23, attached) summarizes i the results of analysis performed to determine the effect of tripping the motor l i

l l

4197 NSP ATW$EVALDoc 10

q- -.

~

, .y driven auxiliary feedwater pump after steam generator pressure decreases

. below 700 psia then flow ceases from the turbine driven pump when steam generator pressure decreases to less than 200 psig. This evaluation was performed using the PINGP configuration of 400 gpm auxiliary feedwater flow initially, then reducing the auxiliary feedwater flow to 140 gpm after the motor driven pump trips, then reducing the auxiliary feedwater flow to zero when the steam generator pressure decreases below 200 psi. Analysis were performed with this auxiliary feedwater system configuration at moderator temperature coefficients (MTC) of both -8 pcm/F and -4 pcm/F (Reference 23, attached). The results in Reference 23 conclude that RCS pressure remains below the limit of 3200 psig for both cases. The 95% MTC at the PINGP is more negative than -7 pcm/F (References 24 and 25, attached).

Although the motor driven auxiliary feedwater pump will trip, it is not required to operate to maintain RCS pressure less than the acceptance criteria. As discussed above, the configuration of both pumps starting and operating through the RCS peak pressure transient is acceptable per the attached evaluation (Reference 23). Restarting of the motor driven pump is at the discretion of the Operator. This procedure is simple and is performed from the Control Room.

3.5 Operator Response to a Complete Loss of Feedwater ATWS As discussed in the plant response with the current configuration, this evaluation does not rely on operator actions to demonstrate that the auxiliary feedwater system can complete its function (s) in a complete loss of feedwater ATWS.

However, Operators will take actions during a complete loss of feedwater ATWS which will reduce the event severity and enhance the overall response.

ANSI /ANS 58.8 (Reference 11) contains guidance for determining the appropriateness of operator actions. A difficulty in using this document for review of response to an ATWS is that Reference 11 is for design basis avents, and as previously discussed, per Reference 4, ATWS is not a design basis event. An appendix to ANSI /ANS 58.8 discusses the bases for the timing requirements for operator actions. Part of these bases were developed from experimental programs (General Physics, NRC, Westinghouse, and EPRI). The accident prevention group (APG), through the NRC, developed models for evaluating the timing of operator actions. The conclusions of these analyses I state:  !

"The APG analyses, with a few exceptions, validate the conservatism of the present timing requirements. These exceptions are generally of two classifications:

4197 NSP ATWSEVALDOC 11 l

l

l - .

L "1. Anticipated Transients Without Scram (which are not design bases events). Operators have extensively trained for these events, and i

they can successfully complete safety-related actions in mere seconds."

The second classification is not relevant to this justification. Thus, ANS 58.8 indicates that significantly shorter time periods for operator action are justified during an ATWS. The above statement in ANSI 58.8 that operators are extensively trained for ATWS events is also valid for operators at the PINGP based on classroom and simulator training and testing.

For an ATWS, the Operators would respond per EOP FR-S.1 (Response to Nuclear Generation /ATWS). The indications to the Operators would allow them to unambiguously diagnose the ATWS event. If the Operators did enter E-0 (Reactor Trip). step 1 would transition them directly to FR S.1. The first three ,

steps in FR S.1 are to (1) verify reactor trip, (2) verify turbine trip, and (3) verify auxiliary feedwater flow. The first two steps are memorized steps. FR-S.1 is meant to be used for any type of ATWS event. That is, the steps are the same .

regardless of which ATWS related event is in progress This makes the actions relatively simple to carry out successfully.

The " Response Not Obtained" step for the first action in FR-S.1 (verify reactor trip) directs the operator to manually trip the reactor. If the reactor will not trip, the operator is directed to verify automatic control rod insertion or manually insert the control rods. These actions will ensure that the reactor is shutdown and mitigate the ATWS event. Although not credited in the ATWS analysis, insertion of the control rods (either automatic or manually) is available and will mitigate the event.

The " Response Not Obtained" step for the second action in FR-S.1 (verify turbine trip) directs the operator to manually trip the turbine. Tripping the turbine will decrease the severity of the event.

Following verification of turbine trip and auxiliary feedwater initiation, the operator will borate the reactor coolant system. This will shutdown the reactor and mitigate the event. ,

When the actions in FR S.1 are complete and in the event that all secondary side inventory were lost, the operator would . respond per FR H.1 (Loss of Secondary Heat Sink). With no secondary heat sink available, the operator is directed to initiate feed and bleed of the reactor coolant system to maintain core cooling.

Although not required to demonstrate that the auxiliary feedwater system can complete its function (s) of ensuring that RCS peak pressure does not exceed the 4:sm sse awstvat.coc 12

_ . . - , . . - - - - - - ~. - . . . -. . . - . . . -

i ,

1 l acceptance criteria in a complete loss of feedwater ATWS, these actions provide I

assurance that the operators will mitigate the event much sooner tha.n assumed i in the analysis.

i 3.6 Conclusions l

Based on the above evaluation the auxiliary feedwater system (as currently ]

i configured) is capable of mitigating a complete loss of feedwater ATWS. To i

summanzel this conclusion is based on:

. The extremely small probability of a total loss of feedwater ATWS. The NRC used this very low probability as justification for proceeding through

' the rule making process for ATWS.

j . The high reliability of the reactor trip system. Actions taken in response to NRC Generic Letter 83-28 have further enhanced this reliability.

l  :

l

. There are several conservatisms in the total loss of feedwater ATWS )

} generic analysis which taken in the aggregate provide reasonable '

assurance that the RCS would not be over pressurized during a complete loss of feedwater ATWS even if the motor driven AFW Pump were to trip and needed to be restarted, in the event the motor driven auxiliary l feedwater pump trips, restarting the motor driven pump is a relatively simple evolution which is performed from the Control Room.

c . The AFW System at the PINGP is rated at 50% of the capacity assumed 3

in the Westinghouse analysis. Then, to protect the pumps from runout, j the auxiliary feedwater pump will trip later in the event. Westinghouse 1 has evaluated the effect on RCS peak pressure for the current configuration at the PINGP and concluded that the results are acceptable J

' (Reference 23, attached).

l

. Operator response to a complete loss of feedwater ATWS is not credited I in the analysis of the expected plant response. However, there is reasonable assurance that these actions will be taken and further mitigate this event.

l l

l 4

i 4'597 NSP ATWSEVALDOc 13

REFERENCES

1. USAR, Section 11. " Steam and Power Conversion."

2. USAR, Section 14, " Safety Analysis."

3. Westinghouse Letter NSP-80-3, "ATWS Report, December Submittal to NRC,"

dated January 14,1980. [NS-TMA-2182]

4. NUREG 0460, Anticipated Transients Without Scrams for Light Water Reactors.

5. Licensee Event Report 96-10-00, " Determination that the Auxiliary Feedwater Pumps are not Protected Against Runcut for all Accident Conditions," dated June 19,1996.

6. Westinghouse Letter PlW-P-540, " Auxiliary Feedwater System Sizing," dated October 1,1969.

7. Westinghouse letter PlW-P-519, " Auxiliary Feedwater System Criteria," dated September 5,1969.

8. Pioneer Service & Engineering Co. Letter to Northern States Power, PIP-N-353,

" Auxiliary Feedwater System - Feedwater Line Rupture Accident," dated October 6,1969.

9. FR-S.1, " Response to Nuclear Generation /ATWS." l i

10. C28.1 AOP4," Restarting an AFWP after low suction / discharge pressure trip." l

11. ANSl/ANS 58.8-1994, "American National Standard, Time Response Design Criteria for Safety Related Operator Actions."

12. Safety Evaluation No. 470, " Auxiliary Feedwater Pump Runout Protection."

13. NRC Safety Evaluation Report for NUREG 0737, ll.E.I.1 and ll.E.I.2, dated March 22,1982.

14. NUREG 3214-CR, " Summary of NRCs LOFT Program Experiments," July 1983.

15. NRC Generic Letter 83-28," Required Actions Based on Generic implications of Salem Events."

16. USAR, Section 7, " Plant instrumentation and control Systems."

45 97 NSP ATWSEVALV.>C 14

- - ~*- .,

4

17. DBD-SYS-08, " Reactor Protecticn System."

18. Isometric Piping Diagram XH-106-130, Feedwater Piping Inside Containment. j l

19. WCAP'5890," Ultimate Strength Criteria to Ensure no Loss of Function of Piping and Vessels under Earthquake Loading."

20. Modification 96AF01, Auxiliary Feedwater Pump Runout Protection.

21. Supplement No.1 to NRC Generic Letter 83-28," Required Actions Based on Generic Implications of Salem Events."

22. WCAP 10858-A, Revision 1,"AMSAC Generic Design Package."

23. Westinghouse Letter SAE-TA-126, " Evaluation of ATWS for Prairie Island (NSP/NRP) with Reduced AFW Flow." .

. 24. NSP Internal Correspondence, "P218 MTC at Hot Full Power," dated March 31, 1997. ,

25. NSP Internal Correspondence, "MTC for P118 at 13.791 GWD/MTU, dated April f 1,1997.

26. Calculation ENG-ME-312, " Calculation of Steam Generator Volume at Power."

l l

I l

l i

s e

4597 NSP ATWSEVALDoc 15 l l' I i l

,,. ~ . - - _ _- _ _ . , . _ , . - _ -_ , . _ __

p: 04 '97 16 CE E: no NLL!cEMPG 4:2 27a act: 70 2-1612rc?994 F.02 2a SAE-TA-97-126 Systems Analysis Engineering From  : Transient Analysis WIN  : 284 4897 Date  : April 5,1997 l f 8 Subject : Evaluation of ATWS for Prairie Island I f .." 1 (NSP/NRP) With Reduced AFW Flow l 6 h$ $$

KEYWORDS: NSP/NRP

REFERENCES:

1) CN-TA 97-045, Rev.1

2) NS-TMA-2182

3) SAE-TA-97-121 To  : N. S. Kury EC E 4-12 cc  : J. S. Ivey EC E 4-12 S. D. Rupprecht EC E 4-14 l

Attached for transmittal to Nordern States Power are the results of an evaluation of the Loss of Normal Feedwater ATWS event. This evaluation was performed to support operation of Prairie Island Units 1 and 2 with a reduced AFW flow configuration resulting from a trip of the motor-driven AFW pump when the secondary-side pressure falls below that corresponding to a steam generator pressure of 700 psia and a trip of the turbine-driven AFW pump when pressure falls below 200 psia.

l The evaluations performed (Reference 1) are based on the Westinghouse generic 2-Loop PWR ATWS analysis model with Model 51 steam generators and consider various plant specific changes as identified in the attachment. The use of the generic ATWS model as a basis is consistent with the ATWS analyses reported in NS TMA-2182 (Referecce 2) which supports the basis of the Final ATWS Rule, 10CFR50.62.

Please transmit the attached information to Nordern States Power for their use in assessing plant operability. Note that the Information contained herein supersedes that recently provided via Reference 3.

If you have any questions regarding the information contained in the attachment or on the scope of this evaluation, please contact me.

4 / , .

Gary G. Ament Reviewed by: D. H. Risher Transient Analysis Transient Analysis Attachment

_ _AJ: c4 '97 16:29 FF CU) CFL L CEF$:rG 4122744011TC9-16123b5954 P.02/24 _ . ,

Attachment to SAE-TA-97-126 l

. 1 1

EVALUATION OF ATWS FOR PRAIRIE IS1AND l WIT 11 REDUCED AFW FLOW i

l

m -- _ . _ .. .. . . . . , _ _ _ . . _ _ _ _ _ _ ._ ,_

FF: 04 '97 is: 9 R GD CFL LIces:NG __412 374 4011 TC 3-1612 205934 F.04/24. . _ _ l

, ~'. .

1 LYTRODUCTION

. i I

The purpose of this lecer is to document an ATWS evduation performed for Prairie Island Units 1 & 2 (NSP!NRP) with reduced AFW flow conditions.

The tctal AFW flow capacity for Prairie Island Units 1 & 2 is 400 gpm. This total AFW f4ow capacity is supplied to the steam generators from one meter-driven and one turbine-driven AFW pump. This  !

j APW flow rate corresponds to the 50% AFW flow sensitivity analysis (i.e.,50% of the 800 gpm AFW flow capacity) contained in the '79 ATWS subminal, NS-TMA-2182 (Reference 1) for the generic 2-Loop i

PWR configuration. In addition, to prevent AFW pump runout conditions, Prairie Island Units 1 & 2 are equipped with a low pressure AFW pump trip logic that results in tripping the motor-driven AFW pump when pressure conditions at the pump outlet fall below a condition corresponding to a SG pressure of 700 psia. This results in only the turbine-driven AFW pump being available for conditions when the steam generator pressure is 2 200 psia and < 700 psia. Per the request of Northern States Power, l

-Westinghouse has been contracted to evaluate these AFW flow conditions following a Loss of Normal l Feedwater (LONF) ATWS event where steam generator pressure drops below 700 psia.

In addition, Prairie Island Units 1 & 2 are also licensed to operate with a positive moderator temperature

! coef6cient (h1TC) at part-power conditions. As a result, the evaluations performed herein also consider the effects of a less negative full power hfTC at 6e request of Northern States Power.

]

l EVA1.UATION In the generic Westinghouse ATWS analyses presented in NS-TMA-2182, analyses were performed for a 2-Loop PWR configura:lon assuming a tctal AFW flow capacity of 800 gpm and a moderator i temperature coefficient of -8 pcm/*F. For the LONF ATWS analysis of this 2-Loop plant configuration with Model 51 steam generators, the maximum RCS pressure reached is 2711 psia. This is well below a pressure of 3200 psig corresponding to the ASME Service Level C stress limit as identified for Westinghouse PWRs.

For the purposes of this ev'a luation, the generic 2-Loop PWR LONF ATWS model with Model 51 steam generators was reanalyzed assuming the following plant specific conditions:

Hot full power nominal vessel Tavg of 560'F.

400 gpm AFW from 60 seconds into the event until 15 seconds af:er the steam pressure falls below a pressure corresponding to a steam generator pressure of 700 psia.

}

140 gpm AFW at a steam generator pressure from the condition above until the steam pressure falls below a pressure corresponding to a steam generator pressure of 200 psia.

No AFW after the above condition until operator action is taken at 10 minutes into the event.

p ca 's 1s:a0 p on c:L LM2ns:rc

~

at: ra 201: C 2-is:2reE39a P.05 24

.n AFW purge volume of 2 ft'. 1 Maximum AFW temperature of 100*F.

50% of the total AFW flow is delivered to each steam generator.

The full power moderator temperamre coefficient of-8 pcm/*F and -4 pcm/*F, corresponding to i values for which a more negative moderator temperature coefficiem at hot full power conditions should exist for k 95% of the cycle.

4 The results of these LONF ATWS analyses fer the generic 2 Loop PWR with the Frairie Island plant specine conditions described above show that the maximum RCS pressure reached is 2943 psia which occurs for the case modelling a -4 pcm/'F moderator temperature coefficient. For the case with a

-8 pcm/'F moderator temperature coef6cient, the maximum RCS pressure reached is 2772 psia. This

is still well below the 3200 psig pressure limit associated with ATWS events. In all cases analyzed, the 4 peak RCS pressure occurs before the steam generator pressure falls below 700 psia and, hence, occurs during the ATWS mitigation phase when full AFW flow capacity is available.

s As was the case with the generic ATWS analysis presented in NS-TMA-2182, some magnitude of voiding is predicted to occur. Of all the Westinghouse ATWS analyses performed and summarized in NS TMA-2182, the most limiting RCS pressure condition that occurs is that following a Loss of Load ATWS event for the 4-Loop PWR configuration with Model 51 steam generators. In this case, main 1 I

feedwater is assumed to be lost due to the consequential loss of condenser vacuum which supplies steam to the mrbine-driven main feedwater pumps. In 1981, the LOFTRAN analysis results for this limiting ATWS case were bench-marked to results of analyses performed using a 2-phase flow NOTRUMP computer code model. A comparison of the LOFTRAN and NOTRUMP results showed good agreement and, hence, support validation of the results presented in NS TMA-2182 for the limiting RCS pressure Case.

As part of the evaluations performed herein, the magnitude of voiding predicted to occur was also examined and found to be significantly more than that predicted in the limiting 4-Loop PWR Loss of Load ATWS analysis bench marked against the 2 phase flow model as described above. However, evaluations were perfcrmed to demonstrate dat under these conditions, the RCS pressure transient will not exceed the maximum RCS pressures reported above and that e liquid inventory remaining in the RCS at 10 minutes into the event is suincient to ensure that the core remains covered. At 10 minutes, it is assumed that appropriate operator action can then be taken to reestablish AFW and initiate long term recovery procedures.

The transient analysis results for the two Less of Normal Feedwater ATWS cases considered in this l evaluation are illustrated in Figures 1 through 18. Figures 1 through 9 reflect the transient conditions for the case with a -8 pcm/*F moderator te=perature coenicient at full power conditions. Figures 10 through 18 are those for the case with a -4 pcm!'F moderator temperature coeffelent.

~

/AF: 24 '97 16:41 FR CW) ChL LICEGING 412 274 4011 TO 3-1612 r C5?E4 ~ P. 2G. 24 l d

CONCLt'SIONS The results of the evaluation performed herein demonstrate the acceptability of tripping the motor-driven AFW pump during a Loss of Normal Feedwater ATWS event when secondary-side pressure conditions falls below those corresponding to a steam generator pressure of 700 psia and with a subsequent loss of the turbine-driven AFW pump when steam generator pressure falls below 200 psia.

His acceptability is based on demonstrating that the peak RCS pressure does not exceed a value of 3200 psig, consistent with RCS pressure limit imposed on Westinghouse PWRs for ATWS events and, although a significant amount of voiding is predicted to occar in the RCS due to the loss of primuy-to-secondary heat removal capability until 10 minutes after event initiation, the loss of primary-side inventory through the pressurizer relief and safety valves is less than that which would lead  !

to core uncovery. i REFERENCES l

1. NS-TMA-2182, December 30,1979, 'ATWS Submittal," T. M. Anderson (Westinghouse) to

. Dr. S. H. Hanauer (NRC).

l l

l 9

l 4

I

,p.p-_ _

-y- ,n_._,- , - - - - - ,, -

-,_ pp:: 24 '97 16:42 FP ClJ) CFL LICEN3 NG di~r 2"4 4011 TC 3-1612205924 P.07/24 - '

. . .. j 1

i Figure 1 - '

Prairie Island Loss of Nonnal Feedwater ATWS Event with a ~

- 3 pcm/*F Moderator Temperature Coemdent at Full Power J

J 1

Nuclear Power versus Time s

1 .' 2 l

2 1

5 a

z 8r o

L v -

6.

es -

N c 6-c_ -

o

.c e

, =

Z . 4 -[

j

~ l 1

2-1 i

0 d Id0 2d0 3 d,0 4d0 5d0 680 s

lime (Sec) 1

• f,Ft 24 '97 16:42 F: GO CFL LICEN3DG ~2*2 374 aill IU 3#1612332335 F* '

J Figure 2 Prawie Island Las of Normal Feedwater ATWS Event M&a

- 8 pcm/*F Moderator Temperature Coemeient at Full Power i

Core Heat Mux versus Time j , p _. ..

1-4 z 8IW O -

L__

v .

X

~'

6-

\

C .

N u -

1 C

O ,4--

i M

2-M b

I l 0 0 0 0 1d0 2d0 3d0 460 5d0 600 Time (sec)

e r,p: aa '97'1s: 42 F: GD CFL LIGG;hG 412 Na C11 IG .3-1612E3D 'O' Figure 3 Prairie Island Loss of Normal Feedwster ATWS Event

~

with a

- 8 pan /*F Moderator Temperature Coemeient at Full Power Core Mass Flow versus Time 1

1.2 I

n 8- -

u 5 -  %

v

!c

_o _

g cc O7 I~

=

E .

u u _

c v .4-2-

N-0 3do /

0 Id0 260 460 500 60'0 Time (SOC) i

p.p: c4 'c; is;43 FF (U) CP'~~LICESI?G #U ##D Figure 4 Prairie Island Loss of Normal Fee 6 vater ATWS Event with a i - 8 pcm/*F Moderator Temperature Coefficient at Full Power i

RCS Average Coolant Temperature versus Time d

IAVG

- - - - :sn  !

700[

1

,s is, s i

~

650 -

r-i e,

C -

w

= .

C u

o 600 -

c w

Q C ~

' C H

E j 550f l

' 1 500  :  :

1 l l 0 180 200 300 400 500 i

600

)

1 Iime (Sec)

E I

P. ki24 -

7 AP: e.: '97 is:a: ::, (W) CPL LICEN9 X 412.274 4011 TO 3-1612 M 904 Figure 5 -

Prairie Island Loss of Normal Feedwater ATWS Event with a '

- 8 pcm/*F Moderator Temperature Coemcient at Full Power Pressurizer Pressure versus Time 2700 >

i i

_ {

2600 -

[

n o

'v; r Q

v h e

=

2500 L m _

tn S

ct _

m o -

E

=

2400 -

m -

en

^

y -

f 2300 --

l-2200  : l l . .'

d 100 200 300' 400 560 6dC Iime $6C

- - -. . . _ . =

, *;,pq aa '97 is:aa FR dte CFL LICEN5ING ' 212 E4 4011 E 6-1$12330"U * '

Figure 6 Prairie roand Ems of Normal Feedwater ATWS Event with a ~

- 8 pcm/'F Moderator Temperature Coetncient at Full Power Pressurizer Water Volume versus Time i100

~.

, 1000 ,- W"

~

n m .

=:

v c f E F

.2 900 '-

O EU

%w .

ilt: .

R 500 I

= .

3 m

m -

e u ~

c_ .

700 '-

600 ! ' ' ' '

! 0 ISO 20 360 480 560 600 l

Time (S6C) t

__- . ;. . .. _ _ . . . . - - . ~

~ '

~ f o C4 'N 16:44 PP (W) CFL LICESING 412 374 4011 TC'G-141f!205994 P.13/24 Figure 7 Prairie Island Loss of Normal Feedwater ATWS Event with a

- 8 pcm/*F Moderator Temperature Coemcient at Full Power RCP Outlet Pressure versus Thue 2300 1 l 2700 - ( l

.9 en w

c_ -

~

c)

~

N m

o 2500 *l

, i r

~

'Ei

= -

a o -

0 2500 --

o I

m -

c_ L C i C L

_J 2400 1 2300 - , ,

d Id0 2d0 3dO' 4d0 5d0 600

,. Iime (Sec) t

. i,pf a g is:ai :: (W) GPL LICEN5 DG 412 3~4 4011 TO 3-1612332395d ~ , E

• 1

~

Figure 8 ~

Prairie Island Loss of Normal Feedwater ATWS Event with a 8 pcm/*F Moderator Temperature Coemcient at Full Power 4 4/d6 Steam Generator Pressure Pjwe'r versus Time 1200 I_

1000 -

o -

'Q 800-[

C _

E _

m e _

to S 600 --

c_

o en .

@ 400 -

O a -

200 T_

0 l . .

0 100 200 300 400 5d0 S00 Time (SEC) y w

~

, AF: 24 '97 16:4? FF Ga CFL CICENEING d'2 274 4Ci1 70 3-16122 05394 - P* N 2d ~

i .

Figure 9 Prairie Island Loss of Normal Feedwater ATWS Event with a

~

- 8 pon/'F Moderstar Temperature Coeffident at Full Power Total Reactivity versus Time 5000 r

f f

0 L

, L L

F

^E -5000 L o -

b -

N> I L

1

=oeo -10000 L c:::

L

~

e d -15000 I

-20000 -

-25000 l l  :  :  :

0 100 200 300

• 400 500 600 IIfile(S6C)

1

~ P: 24 '97 16:46 FF Cu) CFL LICEN5ING 4:2 374 4011_TC 5-1612TICE554 F.16/24 ,l

~

• l l

l Figure 10 Prairie Island less of Normal Feedwater ATWS Event I with a

- 4 pcm/'F Moderstor Temperature Coemdent at Full Power 1

Nudear Power versus Time l

l i

12 1-m 8 --

z O ~

L.

v _

b e -

a e 6--

c_

C -

_S2 u .

=

Z 4 2- -

> 1 1 1 0 l .'  :  ; . -l 0 100 200 300 .

400 500 600 Tima (S6C) r- *

- ffe 04.'97 16:46- FR ' CW) CFL LICE 53NG 412 U4 4011 IIM -151232EI # P'i7'2 Figure 11 Prairie Island Loss o'r Normal Feedwater ATWS Event with a 4 pcm/'F Moderator Temperature Coemcient at Full Power Core Heat Flux versus Time 1-

.a-Z -

O b 6 --

x

" ~

_3

' . l

%d D ~

~

e b*

w w

O

- 4 2-I

=

_ l 0 ' '

i 0 1d0 2d0 3d0 4d0 560 Ed0 Time Sec 1

l l

l l

l l

~ '- P 1G'24

- g,p ca '97 1 stas FR G.D CPL LICENSING 412 Tra 4011 70 5-16122 05394 Figure 12 Prairie Island Loss of Normal Feedwater ATWS Event with a 4 pcm/'F Moderator Temperature Coefficient at Full Power-Core Mass Flow versus Time 3

f -

o L -

v c

c F w

C e . 6L {

25 - f c

i y

• O d-

,e N

a 0

d 150 20 300

• 4$0 5d'0 680 Time (sec) m 4

p: 24 '97 1s:47 gp (W) CFL LICEh3DO "4 "

4 - "' a go',4 in :-1612E053y4

' ~~ P.19/24 --

Figure 13 Prairie Island Ims of Nonnal Feedwater ATWS Event with a 4 pcmPF Moderator Temperature Coemeient at Full Power '

RCS Average Coolant Temperature versus Time Avc

. 5A7 700 s

\' t ,

. I

/

650 -

u .

I C'

c) .

C -

w -  !

8 600 -

Y -

E3 4

c. -

E -

i c>

550 --

\

b w

l Time (sec)

- AF9 04 '97 16:49 FP (U) CFL LICE 6:NG - 412 274 4011 70 3-1612 ROT?S2 F.20/24 _. -

y , l i

1

. Figure 14  ;

Prairie Island Loss of Normal Feedwater ATWS Event with a

- 4 pen /*F Moderator Temperature Coefficient at Full Power Pressunzer Pressure versus Time l

1 1

2900

\

2800 -  !

l a 2700 L m

CL.

W ,".

C)

• w E 2600 L

~

M I

c.  ;-

w l

.Rc 2500 h I

L m -

E -

c- 24,00 a _

M P

2300 - -

M M

2200  :  :

0 160 200 300 400 500 600 Time (sec) 4

.- ff? 04 '97 16:45 :: CW CFL i.! CENSING 4'2 74 4CM TO 3-1612ECT984 P Il d

l l

l Figure is i

Prairie Island Loss of Normal Feedwater ATWS Event '

with 2

- 4 pcm/*F Moderator Temperature Coefficient at Full Power Pressurizer Water Volume versus T~une i100 l

i I

l 1000 -

R N .

l v

= .

i e ~

E

_a 900 --

B w .

e a -

E= .

u

'i

.5 o 800 -

m m .

en D

m -

c. -

700 --

600 l l l 0 100 200 300' 400 500 600 Time (Sec)

- . go24,'s?,y:JS FR UJ) CFL L!C&lSING '412'374 4011 TO G-16123205?S4 F .'?.2 /24 ,

4 Figure 16 -

Prairie Island Loss of Normal Feedwater ATWS Event with a

- 4 pcm/*F Moderator Temperature Coemeient at Full Power RCP Outlet Pressure versus Time 3000 ;

{

2900 --

n

.9 .

[c.2800-y 2

- i

\ l' I

U -

~p 2700 - -

\

"5 .

C b C- 2600b

& C a2500F c

] _

2400 - _

EW 2300 ,

0 t$0 2b0 360 N0 d0 600 Time (sec)

. 'AFR 2e '97.16:49 FR CLD CPL LICENSING 412 374 4011 T0-G-16122205964 P.23/24

.- ~

j

. Figure 17 Pndrie Island Iass of Normal Feedwater ATWS Event with a d Pcm/*F Moderator Temperature Coefficient at Full Power c 4{sINT Steam Generator Pressure Pow'er versus Time

/

1200 1000 -l I

n .

e ,

'G i

v

c. 800 I e 1 h - 1 54

' m -

l 600 - )\

C i A

S 400 -

o ,

a _

l P

200 I .-

g ,

7 0 1dC 20 3h0- 480 Sd0 '

500 Time (SCC)

. _ - _412 374 4011 TO 8-16123205994_ _ _ .P.24/24

~~

A: 04 '97 16:49 FA CUTTPL LICENSING .___..._....._.

' ~

- ~

l.

t l

~

i

• Mgure 18 Prairie Island Loss of Normal Feedwater ATWS Event with a

-4P em/*F Moderator Temperature Coefficient at Full Power Total Reactivity versus Time 5000 4

o

, l i

F

^

E -5000 F o -

e -

y *

.g .

• g -10000 -

e -

e 1 g .

o C

+ -!5000 -- j

-20000 r- >

-25000  ? - - - ,

0 100 250 3d0 460 5$0 600 Time (SCC)

~

. Internal Correspondelice' - - - - - '

D '- -

)

oat. April 1,1997 From Paul T. Kelly tocanon RS 10 -

l To Tom Breene Locauen PI l

subject MTC for P118 at 13.791 GWD/MTU I

l Tom, The calculated MTC for P118 at 13.791 GWD/MTU is -13.96 pcm/ F. This does include the +1.19 pcm/ F bias.

If you have any questions contact me at 337-2059.

l (LO -

Paul Kelly ,

Engineer Associate l Fuels Department RS-10 l

4 4

b

~ ~ ~

~

Intema[ Correspondence -

cate March 31,1997 From Tim Tasto tocation RSQ 10 To Tom Breene tocation Prairie Island subject P218 MTC at Hot Full Power NAD has analyzed the MTC for P218 at hot-full-power operating conditions. Two exposure points were chosen for this analysis - 0.0 Gwd/Mtu and 1.0 Gwd/Mtu.

These calculations yield best-estimate MTC's of -6.4 pcm/ F at 0.0 Gwd/Mtu, and , ,

-7.1 pcm/*F at 1.0 Gwd/Mtu of cycle expcsure. The values include a bias of +1.19 pcm/ F.

The predicted cycle length is 21.530 Gwd/Mtu (from the P218 SOR, NSPNAD-96006, Rev 0). Based upon this cycle length, it can be stated that NAD calculations show that the predicted hot-full-power MTC for P218 remains more negative than -7.1 pcm/ F for 95% of the total cycle length.

. i If you have any comments or questions, please contact me at extension 2069.

fW l Tim Tasto Fuel Resources - Nuclear Analysis and Design (NAD) t 4

e Y

e t

4

USNRC July 23,1997 Attachment 2 Final Westinahouse Evaluation of ATWS l

for Prairie Island with Reduced AFW Flow 1

4 l

1 l

4 OPEREVAL, DOC

~ ~

.- AFP 05 '97 22:10 FP ClO CPL LICE!!5 !t G 4122744011TO'E-161222'0594 P.02/24 NSP-97-501 NSD-SAE-TPM 97 002 Netlear Se:wces hsiCn Westirighouse Energy Systems Electric Corporation g, e Pittstwp PevsyNea 15'30 0355 April 5,1997 Mr. Joel Sorensen Northern States Power Company Prairie Island Nuclear Generating Station 1717 Wakonade Drive East Welch, MN 55069

REFERENCES:

1)CN-TA 97 045

2) NS-TMA-2182

Dear Mr. Sorensen:

Attached are the results of an evaluation of the Loss of Normal Feedwater ATWS event. This evaluation was perforrned to support operation of Prairie Island Units 1 and 2 with a reduced AFW flow configuration resulting from a trip of the motor-driven AFW pump when the secondary-side pressure falls below that corresponding to a steam generator pressure of 700 psia and a trip of the turbine-driven AFW pump when pressure falls below 200 psia.

The evaluations performed (Reference 1) are based on the We stinghouse generic 2-loop PWR ATWS analysis rnodel with Model 51 steam generators and consider various plant specific changes as identified in the attachment. The use of generic ATWS model as a buis i:. consistent with the ATWS analyses reported in NS-TMA-2182 (Reference 2) which . support the basis of the Final ATWS Rule.10CFR50 62.

If you have any questions regarding the mformation contained in the attachment or o - the scope of this evaluation. please contact me on (412) 374-4481.

Qf .c 4 l

l Nancy ' 'ury Technica Projects 5 anager

/ Attachment i

ec: Kenneth Albrecht Paras Shah I

'The musuun ofNSD is toprovide our customere mth pwek, cqunpment and servan that set the standards ofescenew sn the nuclear inds.stry *

%PR O'S '97 s3:10 FF .CLQ CPL LICEt61HG 4123*44011h08-16123305984 P.03/24 1

EVALUATION OF ATWS FOR PRAIRIE ISLAND Wmi REDUCED ARY FLOW l

l l

1

AFP 05 '97' 23:11 FP 04 OPL LIGN5 dig J12 74 4011 TO 9-16123305M4 P 04/24 -

n INTRODUCTION The purpose of this letter is to document an ATWS evaluation performed for Prairie Island Units 1 & 2 (NSP NRP) with reduced AFW flow conditions.

The total AFW tiow capacity for Prairie Island Units 1 & 2 is 400 gpm. This total AFW flow capacity is supplied to the steam generators from one motor-driven and one turbine-driven AFW pump. This AFW flow rate corresponds to the 50% AFW flow sensitivity analysis (i.e., 50% of the 800 gpm AFW flow capacity) contained in the '79 ATWS submittal, NS-TMA-2182 (Reference 1) for the generic 2-Loop PWR configuration. In addition, to prevent AFW pump runout conditions, Prairie Island Units 1 & 2 are equipped with a low pressure AFW pump trip logic that results in tripping the motor-driven AFW pump when pressure conditions at the pump outlet fall below a condition corresponding to a SG pressure of 700 psia. This results in only the turbine-driven AFW pump being available for conditions when the steam generator pressure is 2 200 psia and < 700 psia. Per the request of Northern States Power, Westinghouse has been contracted to evaluate these AFW flow conditions following a Loss of Normal Feedwater (LONF) ATWS event where steam generator pressure drops below 700 psia.

In addition, Prairic Island Units 1 & 2 are also licensed to operate with a positive moderator temperature coefficient (MTC) at part-power cond.'tions. As a result, the evaluations performed herein also consider the effects of a less negative full power MTC at the request of Northern States Power.

EVALUATION l

in the generic Westinghouse AT"is analyses presented in NS TMA-2182, analyses were performed for a 2-Loop PWR configuration assuming a total AFW flow capacity of 800 gpm and a moderator temperature coefficient of-8 pcm/*F. For the LONF ATWS analysis of this 2-Loop plant configuration with Model 51 steam generators, the maximum RCS pressure reached is 2711 psia. This is well below a pressure of 3200 psig corresponding to the ASME Service Level C stress limit as identified for Westinghouse PWRs.

For the purposes of this evaluation, the gerwric 2-Loop PWR LONF ATWS model with Model 51 steam generators was reanalyzed assuming the following plant specific conditions:

Hot full power nominal vessel Tavg of 560'F.

J

- 400 gpm AFW from 60 seconds into the event until 15 seconds after the steam pressure falls below a pressure corresponding to a steam generator pressure of 700 psia.

140 gpm AFW at a steam generator pressure from the condition above until the steam pressure falls below a pressure corresponding to a steam generator pressure of 200 psia.

- No AFW after the above condition until operator action is taken at 10 minutes into the ew.t.

'AFP 0'~'97 23:11 FP (td) OPL-LICEN?!NG - 412 3~4 4011 TO S-16123305M4 P'.09 24- ,

- AFW purge volume of 2 ft'.

! - Maximum AFW temperature of 100*F.

- 50% of the total AFW flow is deliyered to each steam generator.

- The full power moderator temperature coefficient of -8 pcm/*F and -4 pcm/*F, corresponding to values for which a more negative moderator temperature coefficient at hot full power conditions should exist for a: 95% of the cycle.

The results of these LONF ATWS analyses for the generic 2-Loop PWR with the Prairie Island plant specific conditions described above show that the maximum RCS pressure reached is 2943 psia which 1 occurs for the case modelling a -4 pcm/*F moderator temperature coefficient. For Ge case with a

-8 pcm/*F moderator temperature coefficient, the maximum RCS pressure reached is 27% psia. This is still well below the 3200 psig pressure limit associated with ATWS events. In all cases analyzed, the peak RCS pressure occurs before the steam generator pressure falls below 700 psia and, hence, occurs

< during the ATWS mitigation phase when full AFW flow capacity is available.

As was the case with the generic ATWS analysis presented in NS-TMA 2182, some magnitude of voiding is predicted to occur. Of all the Westinghouse ATWS analyses performed and summarized in NS-TMA-2182, the most limiting RCS pressure condition that occurs is that following a Loss of Load

. ATWS event for the 4-Loop PWP configuration with Model 51 steam generators. ?n this case, main feedwater is assumed to be lost due to the consequential loss of condenser vacuum which supplies steam

^

to the turbine-driven main feedwater pumps. In 1981, the LOFTRAN analysis results for this limiting l ATWS case were bench-marked to results of analyses performed using a 2-phase flow NOTRUMP computer code model. A comparison of the LOFTRAN and NOTRUMP results showed good agreement and, hence, support validation of the resula presented in NS-TMA-2182 for the limiting RCS pressure case.

As part of the evaluations performed herein, the magnitude of voiding predicted to occur was also examined and found to be significantly more than that predicted in the limiting 4-Loop PWR Loss of Load ATWS analysis bench-marked against the 2-phase flow model as described above. However, ev!!uations were performd to demonstrate that under these conditions, the RCS press 're transient will not exceed the maximum RCS pressures reported above and that the liquid inventory remaining in the RCS at 10 minutes imo the event is sufficient to ensure that the core remains covered. At 10 minutes, it is assumed that appropriate operator action can then be taken to reestablish AFW and initiate long term recovery procedures.

The transient analysis results for the two Loss of Normal Feedwater ATWS cases considered in this evaluation are illustrated in Figures 1 through 18. Figures 1 through 9 reDect the trarsient conditions for the case with a -8 pcm/*F moderator temperature coefficient at full power conditions. Figures 10 threagh 18 are those for the case with a -4 pcm/*F moderator temperature coefficient.

~

. .] APR 05 '97 23:12 FP ClJ) OPL *-L ICEN9 I NG , 41,2 37.4 4011 TO 8-16123305984 P.06/24 .,

[

i CONCI.USIONS ,

i I The results of the evaluation performed herein demonstrate the acceptability of tripping the motor-driven AFW pump during a Loss of Normal Feedwater ATWS event when secondary-side pressure conditions falls below those corresponding to a steam generator pressure of 700 psia and with a subsequent loss of the turbine-driven AFW pump when steam generator pressure falls below 200 psia.

This acceptability is based on demonstrating that the peak RCS pressure does not exceed a value of 3200 psig, consistent with RCS pressure limit imposed on Westinghouse PWRs for ATWS events and, although a significant amount of voiding is predicted to occur in the RCS due to the loss of primary-to secondary heat removal capability until 10 minutes after event initiation, the loss of primary-side inventory through the pressurizer relief and safety valves is less than that which would lead to core uncovery.

REFERENCES

1. NS TMA-2182, December 30,1979, " ATWS Submittal," T. M. Anderson (Westinghouse) to Dr. S. H. Hanauer (NRC).

1 I

, APF 05 '97 ~23:13 FR (tJ) OFL LICEtGING 412 ~T4 4011 TO G-16123305984 P'.07/24

~

Figure l' Prairie Island Loss of Nonnal Feedwater ATWS Event with a

- 8 pcm/'F Moderator Temperature Coemeient at Full Power Nudear Power versus Time 1.2 l

1-m 8-2 o -

6 _

E3 -

4 m c 6--

c- '

E5 I

$= $

Z .4-2- -

0 -

l  : l l  :

0 100 200 300 400 500 500 Iime (Sec)

= - -- -" - - -

-^ - -

~

. AFP 05 '97*23:13 FF;(LD OPL LICBGING 4123744011TO8-16123305h34 P.09/24

- Figure 2 Prairic Island Loss of Normal Feedwater ATWS Event with a 8 pcm/'F Moderator Temperature Coemcient at Full Power Core Heat Mux versus Time 1-2 --

. I t-p n

z 8-C (%

?-1 sy x *

~3 C ' 6--

= _  :

c.)

B C.) .4 f l 1

M 2--

4 0 l l l l l 0 100 200 300 400 500 600 Time (SOC) l

~

~

. APP 05 '97 23:14 FP (IJ) OPL LICEN9IiG - 4123744011708-16123Ob84 P.09724 '  ?

Figure 3 Prairie Island Loss of Normal Feedwater ATWS Event with a

- 8 pcm/*F Moderator Temperature Coefficient at Full Power Core Mass Flow versus Time i.2 .

1 m 8--

6 -

b o ~

3:

.O -

u_

, 6--

m 1

5 _

e -

o rJ 4__

8 2-0 i

0 100 2do 300 460 5d0 660 Time (Sec)

ffF 05 '97 23:14 FP (W's GPL5.1CElGitG

\

415 374 a'011 TO 8-16123305994

~

P.10/24 1

Figure 4 i 1

Prairie Island Losswith a of Normal Feedwater ATWS Event i

- 8 pcm/*F Moderator Temperature Coefficient at Full Power  ;

i RCS Average Coolant Temperature versus Time i s

m se  ; -

i 700 I- '

F

/ Nli g h - 'I 650 I ,  ;

[ '

~

f

@ f ,

C 3 _

r o

g 600 -- i 3 L I

E e -

550 - , .

~ '

i 500 , '

i O ,

160 280

, i 3$0 t

480 560 Time (SEC) Uc t

,~~APF" f '97 23:15 FP (W) OPL LICB GING 415 774 401L TO G-16123305984 P.11/24 4

Figire 5 ~

Prairie Island Loss of Normal Feedwater ATWS Event with a

- 8 pcm/*F Moderator Temperature Coefficient at Full Power Pressurizer Pressure versus Time i

i 2700

-4

\lI

\'

2600 --

~

9.g .

S -

e 2500 --

= _

2 -

c 5~> -

E 2400 --

=

0

~

c _

2300 - -

1 2200 l l l l l .

0 100 200 300 400 500 600 4

Time (SeC)-

. APP C5 '97 23:15 FR GD OPL LfCEtGItG 412 374 4611 TO S-16123305984 P.12/24 i

Figure 6 Prairie Island Loss of Nonnal Feedwater ATWS Event with a

- 8 pcm/'F Moderator Temperature Coemcient'at Full Power Pressurizer Water Volume versus Time i100 1000

~

n O -

3 _

o G

.5 900 -

c b - ,

c - \

i 3: _

u R 800 --

• c 3 i lA S -

i 700 --

600 l ,

O 100 200 300 400 560 600 Time (Sec)  ;

1

AFP e5 '97 23:16 3 FP OJ) OPL LICENSING 41 374 4011 TO 9-16123305984

'F.13/24-. '_

. Figure 7 Prairie Island Ims of Normal Feedwater ATWS Event with a

- 8 pcm/'F Moderstar Temperature Coemdent at Full Power RCP Outlet Prssure versus Time 2800

- i ,

2700 I q . 8 m -  :

c.

~

CJ E -

E 2600 --

[ ~

z .

! c

C -

c_ 2 5 0 0 --

o Z ~

m -

C -

8 2400 -

2300 0 Ido 2do 3do 4d0 500 500 Time (SOC)

.g .

APR 05 '97 23:16 FR O2D OPL LICEN5Ita3 _412 374 4011 TO 6-16123305964 P.14/24 -

Figure 8 Prairie Island Loss of Normal Feedwater ATWS Event i M&a .

8 pcm/*F Moderator Temperature Coefficient at Full Power )

i 1

Steam Generator Pressure Power versus Time 1200 1000 ~

D 9 -

'm{ 800 -

v _

i 2

a M .

8 600 --

g ~

0 c.n _

g" 400 -

a -

l 200 - -

0 l  :  ; l 0 100 200 300 400 50 500 Time (SeC)

z. -

3

.y .

. APR 05 '97 23:16 FR CU) OPL LIGN5!NG 412 374 4011 TO 8-16123305954 P.15/24

_ l i

-l Figure 9 -

Prairie Island Ims of Normal Feedwater ATWS Event with a -

- 8 pcm/*F Moderator Temperature Coefficient at Full Power Total Reactivity versus Time 4

i I

5000 L

I t ';

i O -

l 4

)

h

.i

^ -5000 1 E .

o .

1

a. _

v J

3 -

5 u: -10000 --

, u

! a -

o cc _-

• ,_o -

8 -15000 i-4 1

-20000 --

4 _

-25000 . l 4b0 560 600 0 100 200 3d0 f s i

lime (SeC) l l

-- ~~ .

}

ffP 05 '97 23:17 FP GD OPL LICENSIHG 412 374 4011 TO 3-16123305994 P,16/24

. Figure 10 Prairie Estand IAss of Normal Feedwater ATWS Event with a

- 4 pcm/*F Moderator Temperature Coemeient at Full Power Nuclear Power versus Time t.2 .

1-8- -

z o

h _

l D -

o 6-c o -

_o o .

a Z 4__

2L t i  !

0 I f 0 lb0 2do 30 4d0 560 500 Ti m e Se C 3

- _ _ - - ._. -. . . -. .- - . . . - . . - - . . _ . . . . n,

.. \

- ~. . _ . . . - .

. APR 05 '97 23:22 FP CW) OPL LICEtGING 412 274 4011'TO S-16127305~G4 P.17/24 Figure 11 Prairie Island Ims of Normal Feedw3ter ATWS Event with a pcm/*F Moderator Temperature Coemcient at Full Power Core Heat Flux versus Time 1

m

,8--

mu 5

• O 6 5 ---

x _

E N

~

~

i o 4_

u m.

O 1

2-- j

>G

$ h 0

0 fA0 2do 3$0 4d0 5$0 600 4

Time (sec)

~

~ [ APP C5 '97- 23:23 FP TH) OPL LICEtGING P.,19/24 412 374 4011 TO S-16123305 i4 l

Figure 12 -

Prairie Island Loss of Normal Feedwater ATWS Event with a

- 4 pcm/'F Moderator Temperature Coefficient at Full Power Core Mass Flow versus Time s

2 1-e 8__ -

, o w -

v

_9 -

u_

m 6--

m

• l C i OE _  !

c) 3 -

U .; _ -

. )

2-N_

0 l  !

l O 100 200 300 400 50 500 Time (SeC) e

, APP 05 '97 23r23 FF (U) OPL LICENS110 412 374 4011 TO S-16123305364 P.19/24 Figure 13 Prairie Island Loss of Nonnal Feedwater ATWS Event with a

- 4 pcm/*F Moderator Temperature Coefrident at Full Power RCS Average Coolant Ternperature versaw Time

- w0 sc 700 i

/

,' vi,

/

1. i 550 !-

& I Tcn -

a -

v m l 3 600 yl Eu L 1

c h3 -

W .

550 - -

500 , ' '

0  ?

100 200 300 400 5dc 600 Time (Sec)

I i

1 J

. AFP 05 '97 23:23'FR CW) OPL LIC&GitG- 412 3~4 4011 TO G-16123305984 . P. 20 '24 - ~~

g

- Figure 14 Prairie Island Loss of Normal Feedwater ATWS Event with a l

- 4 pcm/*F Moderator Temperature Coefficient at Full Power l

l 1

Pressurizer Pmsure verms Time 1

2900 - -

- i 2800 - -

)

L 1

I o.n2700 -- y M

O S e -

3 m

2600 - -

Z

c. -

.R 2 5 00 --

a -

8 -

8 - ,

l C- 2400 -- t

\

l 2300 k l 1

2200 ;

0 td0 2d0 300 4d0 500 600 Time (SEC) l l

l

P.21/24 r#P 05 '97 23:24 FR ClJ) CP LICEtG dig- 412 775 4011 TO 8-16123205984 -

l . .

4

^

Figure 15 Prairie Island Loss of Normal Feedwater ATWS Event with a

- 4 pcm/'F Moderator Temperature Coefficient at Full Power Pressurizer Water Volume vens Time

1100 ,

f -

1000-[

~

0 e

es -

E

.2 900 --

o  !

1 L

u R 800 -

'c -

a

'n M

C) '

o_ ~

\

700 --

l l

l 600 l  : l 0 100 200 300 400 500 500 i n.

lime Sec i I

l 1

l l

1

APR O! '97 73:24 FP GJ) OPL LICEtG!tG 412-374 4011 TO 8-16123305934 P/22/24 r ,

Figure 16- .

Prairie Island Loss of Normal Feedwater ATWS Event with a

- 4 pcm/'F Moderator Temperature Coemcient at Full Power RCP Oudet Pressure versus Time 3000 1 u

F

(

2900 - -

T -

c.n

c. 2 8 0 0 --

l E -

ff

- i 0 -

S 2700 -

c. -

~

T

= _

, 2 O 2600 - -

c. -

U c::

N _

c 2500 --

a 8 -

i T

2400 -- .

l l 2300 l l 400 500 600 300 O 100 200 Time (SOC) e

FFP C5 '97 23:25 FP GD GPL LICEttsitG 412 374 dati TO G-161-23305984 , P.23/24 i Figure 17 l

l Prairie Island I.oss of Normal Feedwater ATWS Event with a '

4 pcm/*F Moderator Temperature Coemcient'at Full Power Steam Generator Pressure Power versus Time 1200 ,

I 1000 -

l

\

1 9 ~

1 800 -

a g _

S 600 -

c ~

O (A -

o" 400 --

2 -

1 200 -

>d pp 0 '

!  ; l 0 100 200 300 400 500 600 Time (SeC) l l

1 l

AFR 05 '97 23:25 FR CW) GPL LICEtGItG 412 374 4011 TO.8-16123305?94 - P;24/24 _ , _

l

• i __

l

, Figure 18 Prairie Island Loss of Normal Feedwater ATWS Event with a 4 pcm/*F Moderator Temperature Coemcient at Full Power Total Reactivity versus Time 50001 -

g i

[ l 0

l h_

^

-5 0 0 0 --

E 2-

> i

'g - 10 0 0 0 t-E3 I T l~

s s 2 -15000 '-

M

-20000 m

-25000  !  !  :  :  :

0 100 200 300 400 500 500 lime SEC e

Nm ^ a a nC Il a