ML20198D075

From kanterella
Jump to navigation Jump to search
Safety Evaluation Supporting NSP Proposed Change to EOPs to Use 2/3 Core Height as Potential Entry Condition Into Containment Flooding
ML20198D075
Person / Time
Site: Monticello Xcel Energy icon.png
Issue date: 12/10/1998
From:
NRC (Affiliation Not Assigned)
To:
Shared Package
ML20198D069 List:
References
NUDOCS 9812220253
Download: ML20198D075 (6)


Text

. . _ ... _ - ..... _ . _ . _ _ __ ___ _ _ _ _ _ _ _ ___ _ _

,da carg p k UNITED STATES

! NUCLEAR REGULATORY COMMISSION wasHWOTON, D.C. 30e06 4001 SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION RELATlNG TO EMERGENCY OPERATING PROCEDURES NORTHERN STATES POWER COMPANY MONTICELLO NUCLEAR GENERATING PLANT DOCKET NO. 50-263 l

l

1.0 INTRODUCTION

l In a letter dated November 18,1997, Northern States Power Company (NSP), the licensee for l Monticello, requested approval of a deviation from the Boiling Water Reactor (BWR) Owners' l Group Emergency Procedure Guidelines (EPGs), Revision 4. The licensee also submitted additional information in letters dated October 1,1998, and November 10,1998. This deviation will permit the revision of the Monticello Emergency Operating Procedure (EOPs) to change the reactor pressure vessel (RPV) water level necessary for adequate core cooling from the top of l the active fuel (TAF) to 2/3 core height. The EPGs require containment flooding if the RPV l level cannot be maintained at TAF irrespective of the emergency core cooling system (ECCS) pump operability. NSP requests to change the RPV level at which containment flooding is initiated from TAF to 2/3 core height. Containment flooding actions will not be initiated provided the RPV level is maintained at 2/3 with two or more ECCS pumps, including at least one core 1 spray pump, injecting to the vessel. If fewer than two ECCS pumps, or no core spray pumps, l l

are operating, containment flooding will be initiated at a higher RPV level, i.e., at minimum l steam cooling reactor pressure vessel water level (MSCRWL).

l l

2.0 BACKGROUND

l t

in January 1997, the staff completed a System Operational Performance inspection (SOPI) of the Monticello residual heat removal (RHR) system. During this inspection, unreviewed safety questions associated with the containment long-term cooling analysis and the net positive suction head requirements of the RHR and core spray pumps were identified. The unreviewed

i. safety questions were resolved by the NRC with the issuance of Amendment No. 98 to Facility Operating License No. DPR-22, dated July 25,1997. One of the conditions imposed by.the NRC with the license amendment was a requirement to process a 10 CFR 50.59 evaluation to change the EOP definition of adequate core cooling to 2/3 core height. The licensee request for the approval of the deviation is a follow-up to the license condition imposed by the staff.

, in Revision 4 of the EPGs, the BWR Owners' Group replaced the spray cooling contingency l with a containment flooding contingency. This was done in recognition that (1) flooding the p RPV and containment to TAF places the reactor in a stable condition for long-term cooling and

minimizes reliance on active systems for core cooling, and (2) the largest break sizes will result in RPV levels below TAF and will require the containment to be flooded eventually as part of l 9812220253 991210 PDR ADOCK 05000263 P PDR ENCLOSURE

y", ', g l .

4 E

\

accident recovery. Once core submergence is reestablished, any residual concerns regarding the adequacy of core spray cooling are also alleviated. However,8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> or more may be

- required to complete the containment flooding evolution and reestablish core submergence.

Any deficiencies regarding the effectiveness of core spray will be manifested within this interim period regardless of the decision to proceed with or defer containment flooding, and are, therefore, not impacted by the NSP request.

3.0 EVALUATION l

l

' During a loss-of-coolant accident (LOCA), operators will control the RPV level using the level control section of the EOPs. Operators will attempt to keep the RPV level above TAF using the l ECCS and other pumps. Revision 4 of the EPGs requires that reactor water level following a l LOCA be restored to above TAF before adequate core cooling is assumed to be achieved, if l RPV level cannot be maintained at TAF, immediate steps will be taken by operators to flood the l containment using all available sources of water.

(-

In the new Emergency Procedures and Severe Accident Guidelines (EPG/ SAGS), Revision 1, containment flooding is initiated at a lower RPV level, i.e., at MSCRWL. The BWR Owners' l

Group EPG/ SAGS define MSCRWL as the lowest RPV water level at which the covered portion l of the reactor core will generate sufficient steam to prevent any cladding temperature in the uncovered part of the core from exceeding 1500 degrees F. This water level is utilized to preclude fuel damage when RPV water level is below the TAF.

For Monticello, the reactor water levels are:

1 Level inches

, MSCRWL -150 2/3 Core -174 l l

  • Vessel Zero is 477.5 from the bottom of the vessel l

Therefore, both the EPGs, Revision 4, and the EPG/ SAGS, Revision 1, would direct the operators to flood containment early in a design-basis accident (DBA) LOCA event, since the  !

corresponding RPV level (2/3 core height) is lower than both the TAF and the MSCRWL.

Flooding the containment early in a DBA-LOCA to minimize the consequences of potential .

additional failures may not provide the best total plant response. For example, requiring l l containment flooding per the EPGs, Revision 4, instructions would require primary containment i venting to maintain the containment pressure below the containment design pressure. Venting j

! is not expected to occur until after the hard pipe vent is submerged, however. Therefore, I venting would have to be through the standby gas treatment (SBGT) system. At the l containment pressures at which venting would occur, damage to the SBGT duct would probably I occur. This could affect the ability of the plant to use a filtered, elevated release pathway. The Monticello containment pressure response to a DBA-LOCA shows the pressure would be less  !

l than 5 psig within 6 days. Deferring containment flooding and the time of containment venting

( could reduce the potential for damage to the SBGT ductwork and other consequential failures that could result from flooding. Accordingly, the licensee prefers that this decision be made by '

I I

l

e *8 l

I 3 l Technical Support Center (TSC) personnel following a thorough analysis of the particular plant conditions and consideration of all alternatives.

The licensee states that the current EOPs are too conservative'in not crediting core spray or low pressure coolant injection (LPCI) for providing adequate core cooling with reactor water level at 2/3 core height. Normally, there are two core spray pumps, and the probability of simultaneous failure of both pumps is low. Moreover, even if the pumps fait during post-LOCA conditions, they may be recoverable. The licensee has determined that one core spray pump is sufficient for adequate core cooling.

The licensee further believes that following the EPGs, Revision 4, approach of immediately flooding primary containment creates the following conflicts with the licensing basis plant response described in the Monticello Updated Safety Analysis Report (USAR):

RHR and RHR service water pumps would not be aligned for long-term suppression l pool cooling.

USAR environmental quali6 tion, shielding, and radiological analyses may no longer be applicable because conditions are different from those originally assumed.

RPV venting to the condenser, and ultimately to the environment, is required to flood the reactor vessel. This creates a vent path not previously considered in the USAR radiological analysis.

Flooding early in the accident may result in damage to the SBGT system if it is used for l venting. This could result in vent paths not considered in the USAR radiological analysis l and may render vital areas of the plant inaccessible.

l The licensee referenced Topical Report NEDO-20566A, " General Electric Company Analytical Model for Loss-of Coolant Analysis in Accordance with 10 CFR 50, Appendix K - Volume 2,"

September 1986, which describes conformance to the five acceptance criteria of 10 CFR 50.46, including long-term cooling. NEDO-20566A indicates that the uncovered portion of the core would be adequately cooled by steam cooling if any ECCS pump is running and RPV level was maintained at 2/3 core height. However, there may be some situations in which the core is operated outside the limits assumed in the NEDO document, and in which core damage may occur.

l In order to assess the applicability of NEDO-20566A to Monticello, the staff performed a series of audit calculations of a postulated long-term cooling scenario in which only LPCI was available. The staff used a simplified single channel staff-generated TRAC BF1 model. The analyses were performed to assess the cooling of exposed fuel over a range of axial power distributions. The NEDO assumed a mid-plane power peak, which was bounding at the time of the report. Plants have since gone to more efficient fuel management, which is no longer bounded by the NEDO. The new fuel management results in a peak above 2/3 core height for the end of the fuel cycle. The staff evaluated the impact of using a limiting top-peaked axial j power distribution as well as the mid-plane peaked distribution used in NEDO 20566A. The staff performed a parametric study and generated 1500 degrees F peak centerline temperature

, (PCT) isotherms as a function of in-channel collapsed level and bundle power. The results i

1 .

T 4

4 were compared to results from the MSCRWL analysis (performed by the BWR Owners' Group) and the results presented in NEDO-20566A. The results are in qualitative agreement for the same power profile, and show that the hot node needs to be covered to ensure adequate core cooling. The results also show that there is very little sensitivity to bundle power. It is important to mention, however, that there is uncertainty in the results due to the method used and the assumptions in the analysis. No attempt was made to assess the magnitude of this uncertainty.

The staff could not conclude that there is adequate core cooling at all times during the fuel cycle when the RPV level is maintained at 2/3 level without core spray.

Following the completion of the staff analysis, the staff had telephone conferences with the licensee on November 3 and 6,1998, to discuss the implications of the model results. It was i

agreed that Monticello could not reasonably assure adequate core cooling in the exposed I j portion of the fuel following a LOCA leading to a LPCI-only injection scenario with core

, uncovery for approximately 20 percent of an average cycle. By letter dated November 10, i 1998, the licensee committed to revise its deviation request by modifying the procedural step l used to decide when to flood containment to specify that one of the two ECCS pumps available j must be a core spray pump.

As such, containment flooding will not be initiated provided two or more ECCS pumps, including at least one core spray pump, are running. This is the minimum number of ECCS pumps available in the spectrum of DBA-LOCAs. If fewer than two ECCS pumps are running, or if no core spray pumps are running, operators will start to flood the containment if the level cannot be maintained above the MSCRWL. The situations involving these failures are beyond the DBA. These two situations are discussed below.

With two ECCS pumps, including one core spray pump, operating, uncovered portions of the core are cooled by the core spray pump. In the event the core spray fails during the event, the uncovered portion of the core is cooled by steam cooling. Steam cooling is the mechanism of core cooling whereby steam updraft through the uncovered portion of the reactor core is sufficient to prevent the temperature of the hottest fuel rod from exceeding the PCT of 1500 degrees F. Water in the covered portion of the reactor core and the lower plenum is the source of steam, if fewer than two ECCS pumps, or no core spray pumps, are available, either initially or as a result of subsequent equipment failures, operators will flood the primary containment if the reactor water level cannot be restored and maintained above the MSCRWL. This is consistent with the BWR Owners' Group EPG/ SAGS.

If all pumps fail in the first 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> (the approximate time to flood the containment to TAF), both the NSP and the BWR Owners' group guidance would lead to similar results, i.e., core damage will occur in both cases. However, the consequences may be worse using the EPG/ SAGS, since it will require earlier venting that will occur closer to the time of core damage. If pump failures occur after 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />, the BWR Owners' Group strategy would be better because containment would have been flooded and core damage prevented. However, pump failures after 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> without recovery of at least one ECCS pump are not likely.

.. e..g s

5 The proposed approach is preferred by the licensee over the BWR Owners' Group approach, since it would enable Monticello to respond to a DBA-LOCA while remaining within the design-basis capabilities of the plant as described in the USAR. The proposed approach would also minimize the potential for damage to systems that could be used to recover from the event by deferring venting until primary containment pressure is reduced. As previously stated, if fewer than two ECCS pumps, or no core spray pumps, are available, Monticello would flood l primary containment if reactor water level could not be restored and maintained above the MSCRWL. When the primary containment flooding is initiated, the EPG/ SAGS would be entered and decision-making responsibility would be transferred to personnel in the TSC. The l two ECCS pumps (with at least one core spray pump) criterion was chosen because this is the

! minimum number of ECCS pumps available for injection into the reactor following a LOCA.

In the event the LOCA occurs late in the fuel cycle and core spray is subsequently lost (a beyond DBA scenario) or not effective in cooling all fuel channels, cladding oxidation and limited fuel damage may result. As mentioned previously, this would occur during the period l required to complete the containment flooding evolution and reestablish core submergence, i

and would occur regardless of whether containment flooding is initiated promptly or deferred.

Prompt initiation of containment flooding offers the benefit of reestablishing core submergence earlier in the event. However, containment flooding will necessitate containment and RPV l venting prior to completion of the flooding evolution and possibly coincident with fuel damage.

This would likely lead to consequential equipment failures (e.g., SBGT ductwork) and larger I

fission product releases than the alternative strategy of delaying containment flooding until the associated releases from containment could be better managed. Therefore, the NSP proposal to defer the decision to flood containment offers a potential for reducing risk in this situation.

4.0 CONCLUSION

At the time the EPGs, Revision 4, were developed, the TSC staff did not have structured guidance for responding to events beyond the scope of the EPGs. Accordingly, the guidance

regarding initiating containment flooding was placed within the EPGs. Since that time,

! however, the BWR Owners' Group has developed additional guidance in EPG/ SAGS for use by the TSC, and the role and capabilities of the TSC have been more clearly established.

1 '

i The staff concludes that the licensee's proposed change to EOPs to use 2/3 core height as the potential entry condition into containment flooding is acceptable. For the spectrum of accidents within the design basis of the plant, the EOPs will not direct the operators to flood the containment. However, the TSC will be expected to evaluate plant conditions, adequacy of the core spray for long-term cooling, and consideration of alternative strategies. Reliance on the TSC for decisions regarding containment flooding is consistent with the way that the decision to flood would realistically be made. In the event of additional fa,ilures that exceed the plant ,

design basis, the EOPs direct the operators to flood the containment. The staff agrees with the licensee's evaluation of the conflicts between the plant design basis and the EPGs, Revision 4, I

l

Pi . . 7 3

l. .

! .4 L

6 i

an'dwith the potential benefits of delaying the decision to vent. Therefore, the licensee may deviate from the EPGs, Revision 4, and use 2/3 core height for establishing the basis for long-term heat removal with at least one core spray pump running.

Principal Contributors: G. Thomas R. Palla R. Frahm, Sr.

A. Ulses

' Date: December 10, 1998 t

.. _- .- - -