ML20141H702

Provides NSP Response to Two Questions Concerning Containment Overpressure.Commitment 6 of Noted Submittal Should Be Modified to Include Listed Info

ML20141H702
Person / Time
Site:	Monticello
Issue date:	07/21/1997
From:	Hill W NORTHERN STATES POWER CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
TAC-L97781, NUDOCS 9708010095
Download: ML20141H702 (12)

Text

.

I*

l Northem States Power Company Monticello Nuclear Generating Plant 2807 West Hwy 75 Monticello, Minnesota 55362-9637 July 21,1997 US Nuclear Regulatory Commission Attn: Document Control Desk Washington, DC 20555 MONTICELLO NUCLEAR GENERATING PLANT Docket No. 50-263 License No. DPR-22 Response to Staff Questions Regarding NSP Letter of July 16,1997(TAC No. 97781)

The information included in this letter is intended to address NRR questions which were communicated to NSP via a July 18,1997 telephone conversation by T.J. Kim, Monticello Project Manager, to J. Beres, NSP Licensing Engineer. The questions concerned NSP's submittal of July 16,1997, " Response to Request for AdditionalInformation Regarding Revision 2 to MNGP License Amendment Dated January 23,1997 (TAC No. 97781)."

This letter provides the NSP response to two questions concerning containment overpressure. In addition, this letter contains a supplement to the previous Question 5 response in the submittal cited above and a modification of a previous commitment as described below.

NSP requests that commitment 6 of the above submittal be modified to include the following.

NSP commits to submit a change to the BWROG EOP Committee for evaluation and resolution of the EOP spray limits in regard to adequate NPSH for ECCS pumps.

Please contact Joel Beres, Licensing Engineer, at (612) 295-1436 if you require further information.

/ M A L.

/b2 William J. Hill Plant Manager Monticello Nuclear Generating Plant c:

Regional Administrator-lil, NRC NRR Project Manager, NRC

([) l l

Sr. Resident inspector, NRC

_I State of Minnesota, Attn: Kris Sanda J. Silberg, Esq.

Attachments:

- Affidavit to the US Nuclear Regulatory Commission lil.lllll.llki.l!Il.lIl.llllIl.lllWill c

4

"'r 9708010095 970721 VI' PDR ADOCK 05000263 p

PDR

)

T UNITED STATES NUCLEAR REGULATORY COMMISSION NORTHERN STATES POWER COMPANY MONTICELLO NUCLEAR GENERATING PLANT DOCKET NO. 50-263 s

Response to Staff Questions Regarding NSP Letter of July 16,1997(TAC No. 97781) 1 Northern States Power Company, a Minnesota corporation provides its response for the Monticello Nuclear Generating Plant to a July 18,1997 telephone conversation between the NRC Staff and NSP concerning the subject license amendment. This letter contains no restricted or other defense information.

NORTHERN STATES POVER COMPANY By

1. 'h 4

Wiliiam J. he Plant Manager Monticello Nuclear Generating Plant i

On this M day of7tdy MD before me a notary public in and for said County, personally appeared William J. Hill, Planf Manager, Monticello Nuclear Generating Plant, and being first duly sworn acknowledged that he is authorized to execute this document on behalf of Northern States Power Company, and that to the best of his knowledge, information, and belief, the statements made in it are true.

na.ch

^ ^

' Sarnuel 1. Shirey

['"'V ~ S AMUEL 1. SHIREY

=

I Notary Public - Minnesota

/ notAny eustic.uimA Sherburne County My Commission Expires January 31,2000 L ;;,_

y comm. Exp. Jan. 31,2OOO u

^^

2

t

1. Please clarify the conservatisms that exist in the worst-case peak containment overpressure requested for use in the evaluation of ECCS pump NPSH.

i A. Introduction i

This response provides information on conservatisms that exist with respect to containment overpressure and with respect to NPSH requirements for the ECCS pumps. NSP's license amendment included a request for approval of the use of a maximum containment overpressure of 6.1 psig for the limiting ECCS pump NPSH. Case 3 of Exhibit D of the subject license l

amendment shows the limiting case for ECCS pump NPSH.

The analysis for the design basis accident for the limiting ECCS pump NPSH includes assumptions and methodologies that are designed to reduce the amount of containment pressure and maximize the temperature response of the suppression pool. These assumptions and methodologies are extremely conservative to the extent that certain conditions assumed or calculated to exist inside containment do not actually reflect any reasonable operating or post-accident condition. Use of realistic thermodynamic processes that would also serve to increase containment pressure and reduce suppression pool temperature are not credited in the licensing bases analysis. Eliminating these conservatisms would increase the amount of containment pressure available and consequently reduce the theoretical and conservative amount of containment pressure requested for ECCS pump NPSH.

P. Realistic Assumptions that increase Pressure Containment pressure is determined by the mass of non-condensable gases inside primary containment and by the vapor pressure of containment spray water inside containment. Pressure from the non-condensable gas that is used to inert the containment under normal operating conditions, nitrogen, is evaluated using the Ideal Gas Law. Containment spray vapor pressure is simply added to the partial pressure provided by the non-condensable gas to determine containment pressure.

Non-condensable gas mass inside containment is minimized using methods described in the response to Question No. 5 as provided in Reference No.1. Increasing the amount of non-condensable gas inside containment will increase the amount of pressure available in containment. The analytical non-condensable gas mass is influenced by the factors described below.

1) The use of standard atmospheric pressure of 14.7 psia instead of the conservatively low value of 14.23 psia would change initial non-condensable gas mass by about 3%.

Because the containment is inerted with nitrogen, the normal operating pressure of the containment is maintaineo slightly positive at approximately 0.5 psig. This would increase containment pressure at the time of the suppression pool temperature peak by approximately 0.4 psi.

2) The use of 100% relative humid ty is conservative with respect to actual expected humidity levels. The use of a realistic value of 60% for relative humidity for initial drywell conditions would inemase non-condensable gas mass by 4%. This would increase containment pressure at the time of the suppression pool temperature peak by approximately 0.5 psi.

3) Containment ler' age is assumed for the entire 5 day pericJ analyzed. After 5 days the containment approaches atmospheric pressure. The containment leakage rate assumed was the Monticello Technical Specification 3.7.A.2.b.1 limit of 1.2 weight per cent per day. The Technical Specification assumes this leakage occurs at the peak 3

1 containment air pressure of 42 psig. For conservatism, this leakage rate was assumed constant regardless of containment air pressure. The total mass of air assumed to leak over the 5 days of interest was therefore 5 days x 1.2 weight percent per day or 6% by weight. This is a conservative assumption since actualleakage rates would be substantially less than this due to the low containment pressures of concem here, i.e. <7 psig. Use of reakstic containment leakage rates would increase the amount of containment overpressure available by approximately 0.7 psi at the end of 5 days.

Therefore if realistic assumptions are applied to the calculation of containment pressure, the amount of pressure available for the limiting ECCS NPSH pump would increase by approximately 1.6 psi.

1 C. Thermodynamic Processes that Would Decrease Pool Temperature A decrease in the suppression pool temperature has a direct impact on the torus water vapor pressure, A decrease in torus water vapor pressure significantly reduces the required amount of containment everpressure necessary to assure adequate NPSH for the ECCS pumps.

\\

At a suppression pool temperatura of acout 162*F, an atmospheric pressure of 14.26 psia is required to support the limiting ECCS pump NPSH requirement st design flow rates. Since peak suppression pool temperature is 194'F for the limiting case, any parameter that reduces this temperature willimprove NPSH margin. The items described below are conservative assumptions that were used in the evaluation of containment response.

1) The assumed decsy heat was conservatively based on the use of nominal ANS-5.1 decay power curves for 102% of an initial reactor thermal power of 1880 MWt (1917 MWt). The 1880 MWt value is 12.6% above the current licensed thermal power of 1670 MWt and is approximately twice the incremental power level increase required to achieve a 95% confidence interval.

The use of an ANS 5.1 nominal decay heat curve at 102% of 1670 MWt (1703 MWt) would reduce suppression pool temperature from 194.2*F to 184*F. This reduction in suppression pool temperature reduces required containment overpressure for the limiting ECCS pump NPSH by approximately 2 psig.

2) The suppression pool water volume is conservatively assumed to be at the minimum operating value of 68,000 cubic feet. Suppression pool volume is controlled by operating j

procedures which provide guidance to maintain operating leve!s significantly above the l

lower limit. Technical Specification 3.7.A.1.e requires suppression pool volume to be i

i raaintained between 68,000 and 72,910 cubic feet. Use of the average volume as recommended by procedure would increase the mass of water by 3.6%. Reference 4 determined that, assuming a normal water volame, the suppression pool temperature would be reduced by 2'F to a value of 182*F. This reduction in suppression pool i

temperature reduces required containment overpressure for the limiting ECCS pump NPSH by approximately 0.4 psig.

3) The river temperature, or ultimate heat sink temperature, used in the analysis was 90*F. River temperature has never exceeded 86*F in over 25 years of plant operation.

The mean of the daily average river temperature for calendar years 1995 and 1996 was 49.8'F. Typical daily average river temperatures in the summer range between 70 F and 82*F. Use of the typical summer high aver?ge daily temperature would reduce the ultimate heat sink temperature to 82*F. Peak suppression pool temperature will drop approximately 0.4'F for every 1*F reduction in ultimate heat sink temperature based on 4

i

A sensitivity studies provided in Reference 3. Based on this assumption, suppression pool temperature will drop 3*F to 179*F with the postulated 8*F reduction in river temperature.

This reduction in suppression pool temperature reduces required containment overpressure for the limiting ECCS pump NPSH by approximately 0.5 psig.

4) RHR heat exchanger performance is evaluated by test annually. Nominal test results typically exceed the design heat transfer rate by approximately 10%. Assuming a 5%

increase in RHR heat exchanger design heat transfer coefficient (K), the suppression pool temperature would be expected to be reduced by about 2*F to a value of 177*F.

This reduction in suppression pool temperature reduces required containment overpressure for the 12miting ECCS pump NPSH by approximately 0.3 psig.

Therefore if realistic assumptions are applied to the modeling of thermodynamic processes, the amount of NPSH for the limiting ECCS pump would be reduced by approximately 3.2 psig.

D. Conservatism in NPSH Calculations A significant amount of conservatism was assumed in the evaluation of NPSH. These further reduce the amount of containment overpressure for adequate NPSH as described below.

1) The suppression pool water volume was conservatively assumed to be at the minimum operating value of 68,000 cubic feet. Use of the average volume would reduce the NPSH required by 0.33 feet due to the higher head of water. This conservatism reduces required containment overpressure for the limiting ECCS pump by approximately 0.1 psig.

2) Suppression pool water level is assumed to be drawn down due to holdup of water inside the drywell during the DBA LOCA scenario. This drawdown was conserva+ively determined to be 10.7 inches and included holdup of water in the reactor vessel. A revised calculation of the amount of suppression pool drawdown is in progress that uses more realistic assumptions. Because of the nature of the DBA LOCA, this amount of i

drawdown is unrealistic. The revised calculation reduces the amount of drawdown to 5.5 inches which increase the elevation head. This conservatism reduces required containment overpressure for the limiting ECCS pump by approximately 0.2 psig.

3) The original NPSH required curve for the Core Spray pumps was based on a conservative 1% head degradation versus the Hydraulic Institute Standard of 3% head degradation for determination of NPSH required. Information provided in Reference 5 described additionalinformation on Core Spray pump based on the use of the standard 3% head degradation for the determination of NPSH required. Interpolation of this data based on reasonable engineering judgment would reduce the NPSH required by 4 feet for Case 3. This conservatism reduces required containment overpressure for the limiting ECCS pump by approximately 1.7 psig.

Therefore if realistic assumptions are appli6d to the NPSH calculation, the amount of NPSH for the limiting ECCS pump would be reduced by approximately 2.0 psig.

E. Other Factors One foot of head (0.4 psig) has conservatively been added to account for debris loading for future considerations. This results in a 0.4 psig increase above ECCS pump NPSH requirements.

5

Additional conservatisms, which have not been quantified are incorporated into the suppression pool temperature determination. These include: ambient losses from the containment to the reactor building are not credited, all of the pump heat from the ECCS pumps is assumed to all be transferred to the containment at rated hp, and the entire inventory of feedwater that would add heat is assumed to be transferred to the containment.

F. Conclusions Given the above, a significant amount of conservatism exists in the derivation of the requested amount of containment overpressure. The cumulative effect of these conservati!;ms, when applied to the limiting ECCS pump, provide reasonable assurance of successful pump operation. The realistic analyses above demonstrate a NPSH deficit of up to 0.9 psig for the worst case period of 16 hours1.851852e-4 days <br />0.00444 hours <br />2.645503e-5 weeks <br />6.088e-6 months <br /> compared to the recuested analytical value of 6.1 psig.

Therefore, the requested amount of containment pressure can be considered as a prudent additional reserve of available pressure such that NPSH considerations would not affect pump operation for the duration of the design basis LOCA.

6

a II. Can the reactor core be adequately cooled following a LOCA without taking credit for containment overpressure?

A. Introduction NSP is requesting a predetermined amount of containment pressure to operate the ECCS pumps in a design basis LOCA. This amount of pressure was conservatively chosen to be below the calculated amount of pressure which would be available in the containment. Even if a design basis LOCA event did occur and if the thermodynamic processes that serve to increase containment pressure from the energy transfer were assumed not to be present such that the containment pressure were to remain at approximately atmospheric pressure, the Monticello plant i

is designed to provide a defense in depth to assure that the necessary core cooling still occurs and that the plant can be safely shutdown.

Plant specific EOPs have been developed for Monticello and are based on BWR emergency procedure guidelines. These procedures are designed to address a wide range of accident events including those considered outside the design bases of the plant. This range of events includes a LOCA event with a subsequent loss of containment pressure. Control room operators are well trained on these procedures and frequently exercise these procedures during simulator training.

B. LOCA With Torus at Atmospheric Pressure The following is an evaluation of a LOCA with the torus assumed at atmospheric pressure throughout the accident.

1) First Ten Minutes Following the Event For the first ten minutes of the event, no operator action is assumed. For a large break LOCA the reactor would quickly depressurize and the low pressure ECCS pumps would automatically start injecting into the reactor vessel at runout flow conditions. A limiting case with a LPCI loop select error was analyzed and submitted as Figure E.1 in Reference 2. Starting at about 85 seconds into the event, there would be an NPSH deficit for the ECCS pumps that would increase-to 3.16 ft.

at 10 minutes. At 189 seconds the two core spray pumpo will have reflooded the core (Reference NEDC-31786P, DBA Suction Break with a LPCI injection valve failure to open, water level versus time is shown on Page A-36). By reflooding the reactor core, adequate core cooling is assured.

The maximum core spray flow used in the current SAFER-GESTR analysis is 89% of the maximum core spray flow assumed in current NPSH calculations (Reference USAR Figure 14.7-8 and Exhibit E of Ref. 2). Use of flows assumed for the SAFER-GESTR analysis will provide adequate core cooling for the full range of break sizes. It is expected that sufficient core spray flow will be available to meet the core cooling requirements with the NPSH deficit based on testing performed by the pump vendor-Sulzer Bingham Pump Company. Bingham Pump Company Report S. O. 280685 documents that similar pumps could run with NPSH deficits for several hours with no observable damage. The testing also showed that a similar pump with an NPSH deficit of 10 feet provided 90% flow for 30 minutes without damage.

2) Entry into the Drywell Flooding Procedure Following the first ten minutes into the event, operator response is assumed. EOP Procedure C.5-1100 instructs the operators to attempt to restore reactor level to the normal operating level.

Due to the reactor internal design, this would not be achievable for the large break LOCA event.

The procedures would then instruct the operators to maintain reactor level above the top of active 7

)

fuel. i his level would also not be achievable. If the operators cannot restore and maintain level above the top of the active fuel, Drywell Flooding, C.5-2004, is entered.

Another possible path for the operator to arrive at the Drywell Flooding EOP may include the determination by the operating crew that RPV levelis unknown. This determination may result 1

from the high temperatures that would be seen by the RPV water level instrument sensing lines in the drywell. If these temperature limits are exceeded, the EOPs (C.5-1100) direct the operator to enter C.5-2006, RPV Flooding. RPV flooding will result in an entry into drywell flooding since solid conditions in the reactor with 3 SRVs open cannot be achieved with a large break. Based on the above two situations, drywell flooding is the resultant end point even if RPV flooding is briefly entered.

It is expected that the time required for the operators to begin the EOP drywell flooding sequence is 30 minutes from initiation of the event. This response time is conservatively estimated based on observed operator performance in simulator scenarios. During this time, core spray and/or LPCI would provide makeup for boiloff, and the NPSH deficit would remain less than 10 ft.

According to USAR Section 10.2.5, the amount of injection necessary to makeup core boiloff is 400 gpm after 15 minutes into the LOCA.

)

3) Drywell Flooding i

C.5-2004, Drywell Flooding, will direct the operators to flood the drywell with all available j

systems. Operators are directed to keep one core spray system aligned to the torus and to align the remaining ECCS pumps to the condensate storage tanks (CSTs). These actions provide the following benefits.

Relatively cool water (~100*F) is now being added to the reactor core from the LPCI System and/or from the Core Spray system which are aligned to the CST.

Torus water level will increase. This will add available NPSH to the operating core spray i

pump.

The cooler water and increased elevation head together with less friction head loss as the number of pumps is reduced would likely allow for continued operation of the Core Spray Pump irrespective of the torus pressure.

i The ECCS pump NPSH concerns related to containment pressure are eliminated while the pumps are aligned and operated from the CST.

The LPCI flow, of itself, would be more than adequate to assure adequate core cooling during this time period. The drywell flooding procedure does not instruct the operator to ignore NPSH and vortex limits on the operating Core Spray and RHR pumps, thus the operator would be controlling injection within the allowable limits for the conditions present. If operator action is assumed to throttle core spray flow to nominal rated flow rates, it is likely that the NPSH available would be sufficient for successful core spray operation. Throttling of pump flow would also extend the time that the pumps would be able to operate from the CST as a suction source.

Assuming the limiting case of an EDG failure, the CST suction source would provide core cooling for approximately 40 minutes assuming 8000 gpm through two RHR pumps. It is now approximately 1.25 hours2.893519e-4 days <br />0.00694 hours <br />4.133598e-5 weeks <br />9.5125e-6 months <br /> into the event.

8

4

4) RHRSWSystem When the drywell flooding procedure is entered, the EOPs direct the use of the safety grade RHRSW System as a means to flood the drywell. Preparation for this evolution would commence when drywell flooding is entered. This is accomplished per EOP support procedure C.5-3203, Use of Alternate injection Systems for RPV Makeup. When the LPCI pumps have exhausted the CST inventory, the LPCI pumps would be secured and the RHRSW pumps would be used :o provide an inexhaustible supply of cold river water to the reactor vessel via the LPCI piping. Each RHRSW pump is capable of supplying 3500 gpm. This water source is totally independent of containment pressure and suction strainer losses to support NPSH.

The Division 1 RHRSW sub-system is the only loop that can be cross-tied to the LPCI system. If the Division i EDG is the assumed diesel failure, the Division i 4kv bus must be supplied from the Division 11 EDG. This evolution is addressed by procedure E.4-11, Restore Bus 15 from Bus 16.

In order to start the RHRSW pumps with an essential bus transfer load shed in effect and an ECCS initiation signal present, the load shed would be bypassed per abnormal procedure C.4-H, 4

Restoration of Plant Loads. The procedures are all used and repeatedly trained on during operator simulator training.

The drywell flooding process would continue until indication is available to the operator that the containment has been flooded to the top of active fuel with the reactor vented to assure that the core is indeed covered. Adequate core cooling is assured from the initiation of the assumed accident up to the time when the containment and reactor level, in equilibrium, both correspond to i

the top of active fuel.

C. Other Available Water Sources for Core Cooling

1) Fire Protection System i

uner available and inexhaustible source that utilizes river water is the fire protection system

,ch consists of an electric fire pump or the diesel fire pump each of which is nominal ly rated at 1500 gpm. This system can also be aligned to inject to the reactor vessel via the LPCI piping per EOP support procedure C.5-3203. The LPCI connection is a hard-piped tie-in that was installed as a result of the MNGP IPE. The fire protection syste7;) capacity is wellin excess of that required for adequate core cooling by providing makeup for boiloff.

2) AdditionalWater Sources Othes oystems that could be used during a LOCA to flood the drywell without offsite power available include the CRD System and the SBLC System. The CRD pumps would require the ECCS load shed to be bypassed in a similar manner as the RHRSW pumps. The SBLC is an engineered safety feature. The SBLC system would be initiated as soon as drywell flooding was entered. This would provide approximately 30 gpm from the SBLC tank in the reactor building.

When the existing SBLC tank has been exhausted, demineralized water can be lined up via existing hard-piped connections and injected to the vessel via the SBLC System using EOP support procedure C.5 3203. The demineralized water storage tank contains a volume of approximately 30,000 gallons dunng normal operation.

D. Conclusion Given the above, containment pressure above atmospheric levels to support NPSH requirement j

is not necessary to successfully mitigate a design basis LOCA at Monticello. Although the primary ECCS may be degraded when post-accident increases in torus temperature have reduced NPSH available, sufficient methods are available to maintain adequate core cooling and 9

containment integrity even if containment pressure is artificially held to atmospheric levels. These methods utilize systems that have capacities well in excess of that required to sufficiently remove core decay heat. These methods are implemented using existing procedures that the operators are continually trained on.

1 i

10

Supplement to NSP Response to Question 5 of NSPs letter dated July 16,1997, " Response to Request for AdditionalInformation Regarding Revision 2 to MNGP License Amendment Dated January 23,1997 (TAC No. 97781)."

By Attachment A to a letter dated February 27,1997 (Ref. 6) Commonwealth Edison provided a benchmark analysis for the SHEX code against the UFSAR suppression chamber response for Dresden Units 2 and 3 (Please note that this correspondence also applies to NSP's previous discussion on Question 5 in Ref.1). The UFSAR pressure response was derived from a previous 1

minimum pressure analysis where containment sprays had been assumed activated at 600 seconds. The SHEX code was shown to give conservative results with respect to calculated containment pressure. This benchmarking analysis was subsequently accepted by the Staff.

The same version of the SHEX code (04) that was used for the Dresden analysis was used for the Monticello minimum pressure analysis. Within SHEX, the same spray modeling was used for both Dresden and Monticello. One hundred percent of spray efficiency was assumed for both plant analyses. Although the ECCS configuration at MNGP is somewhat different as described in NSP's response to Question 4 (Ref.1), both plants include a GE Mark I containment, and the containment modeling for both plants is identical with the exception of plant specific configuration inputs (e.g. number of pumps, flow rates). Containment sprays are assumed in both analyses to be activated at 600 seconds. In addition, the nature of the MNGP containment transient response to spray initiation as shown for Cases 3,6, and 7 of Exhibit D to the subject license amendment is very similar to that shown in Figure 6 of Attachment A of the Commonwealth Edison letter.

Given the above it is reasonable to assume that similar results would be obtained for the Monticello plant and the Dresden plant in regard to the minimum pressure case and that it is reasonable to conclude that the benchmarking analysis is also valid for Monticello.

11

__._._m._._

1 1

References:

1.

Response to Request for Additional Information Regarding Revision 2 to MNGP License Amendment Dated January 23,1997 (TAC No. 97781), Manifest Date July 16,1997 2.

Revision 2 to License Amendment Request Dated January 23,1997 Update of Design Basis Accident Containment Temperature and Pressure Response (TAC No. M97781) 3.

March 12,1997 Comed letter to NRC with a Subject of Dresden Nuclear Power Station Units 2 and 3 AdditionalInformation Regarding Application for Amendment to Facility Operating

{

Licenses DPR-19 and DPR-25, Appendix A, Technical Specifications, Section 3/4.7.K,

" Suppression Chamber," and Section 3/4.8.C, " Ultimate Heat Sink.*

4.

NEDO-32418, Monticello Design Basis Accident Containment Pressure and Temperature Response for USAR Update, December 1994

5. July 16,1997 NSP letter from W. J. Hill to NRC titled, Request for Information Regarding MNGP License Amendment Dated June 19,1997 (TAC No. 97781)

6. Letter from J. Stephen Perry, Commonwealth Edison Co., to NRC Document Control Desk, February 27,1992 i

i

)

12 q