ML20128B222

Informs That Addl Info Will Be Required in Order to Continue Evaluation of ECCS Conformance to New AEC Adopted Interim Acceptance Criteria

ML20128B222
Person / Time
Site:	Monticello
Issue date:	05/03/1972
From:	James Shea US ATOMIC ENERGY COMMISSION (AEC)
To:	Ziemann D US ATOMIC ENERGY COMMISSION (AEC)
References
NUDOCS 9212030598
Download: ML20128B222 (6)

Text

- - - -

v. -. . . . . . c. . . . . . .

4 l e  %

e ** UNITED STATES

/l .j., )'  ?. ATOMIC ENERGY COMMISSION

,g ". H !

wasmucros o c. som g 3 1972 Files (Docket No. 50-263)

THRU: Z mann, Chief, ORB #2, DRL Me MONTICELLO NUCLE POWER PLANT - ECCS CONFORMANCE TO AEC INTERIM DESIGN CRITERIA Summarl To continue our evaluation of the Monticello "ECCS Conformance to New AEC Adopted Interim Acceptance Criteria" (NSP report dated September 21, 1971), we require further explanation:

1. of the re@c.4 time f rom 8 minutes to 3 minutes to cover the core by the accumulation of core spray water.

2. of the reduced flood time for equivalent core spray conditions following the DBA.

3. of the peak clad temperature discrepancies within the report for the 0.1 ft2 break.

4. of the maximum recirculation line break described in the FSAR as 5.6 ft2 and in the September 1971 report prepared by GE as 4.2 ft2

5. of the automatic relief system reliability with regard to operation of two or all three of the Automatic Pressure Relief (APR) valves following small primary system breaks.

Discussion By letter dated September 21, 1971, Northern States Power Company (NSP) submitted an analysis by General Electric Company confirming that the performance of the Monticello ECCS is in compliance with the AEC Interim Design Criteria.

According to the FSAR, the core is protected from excessive temperatures following loss of coolant by two independent full capacity core spray systems that draw water f rom the suppression pool and spray water directly v4 9212030598 DR 720503 ADOCK 05000263 PDR

s'. ,

Files MAY 3 1972 on the core. Each system, rated at 3250 gpm with 165 psig pump dis-charge pressure, starts to deliver when the injection valve opens at 450 psig reactor vessel pressure. The core spray pumps are electrically driven. Either of the core spray systems when operated in conjunction with either the automatic pressure relief or High Pressure Coolant Injection (HPCI) system will, according to the GE reanalysis, limit core clad temperatures to less than 2300*F for various postulated loss-of-coolant accidents f rom the smallest detectable leak to the largest pipe break, the complete double-ended severance of a recirculation loop pipe.

The HPCI system is described as a system designed to pump 3000 gpm into the reactor vessel through the feedwater inlet within a reactor pressure range of about 1125 psig to 150 psig under loss of coolant conditions which do not result in rapid depressurization of the pressure vessel.

The single HPCI pump is driven by a steam turbine with steam f rom the reactor v asel. Initial core cooling water is supplied from the condensate storage tank. When this source of water is exhausted, the pump suction transfers automatically to the suppression pool. Operation of the HPCI system is dependent upon reactor water level signals. Either

" low" reactor water level or "high" drywell pressure starts the system and "high" reactor water level will stop it. Operation of the HPCI system is completely independent of ac power and requires only de power f rom the plant de batteries systems for operation. GE has assumed in reanalysis of the ECCS, as they have for other similar power plants, that the HPCI is not availabic af ter the LOCA.

The automatic pressure relief system accomplishes reactor vessel depres-surization by blowdown through automatic opening of the relief valves which vent steam to the suppression pool. For small breaks, the vessel is depressurized in sufficient time to allow either the core spray subsystem or Low Pressure Coolant Injection (LPCI) system to provide adequate cooling to prevent any clad melting

• according to the FSAR.

Two of three relief valves, it is reported in the FSAR, discharge more 3

• The original design basis for the auto relief system was, in conjunction with core spray or low pressure coolant injection, to prevent clad melting for all break sizes (FSAR 6-2.39 and 6-2.3). Two of three j relief valves will reduce the pressure in time to permit rated core

spray flow before clad melting occurs, t

l

Files MAY 3 1972 than 1.6 x 106 pe.inds of steam per hour at 1125 psig, thereby providing a satisfactory backup for the HPCI system. Automatic actuation requires coincident indication of reactor water " low-low" level and drywell "high" pressure. An additional interlock is provided to assure that at least one of four Residual Heat Removal (RHR) (LPCI) or one of two core spray pumps is delivering output pressure. A time delay provides time for the HPCI or feedwater system to restore the proper reactor vessel coolant level before blowdown activation. The GE reanalysis of the ECCS, in accordance with AEC Interim Policy Statement, assumes automatic depressurization for the small breaks less than 0.1 f t2, The LPCI subsystem, an integral part of the RHR, is designed to deliver

?

water to the recirculation loops at the rate of 4000 gpm for each of four pumps when the reactor vessel prescure is 20 psi above the suppression chamber pressure or 2000 gpm when reactor vessel pressure l 1s 262 psi above the suppression chamber pressure (FSAR 6-2.11). The objective of the LPCI system is to restore and maintain the coolant inventory in the reactor vessel af ter a loss-of-coolant accident by pumping water from the suppression pool to the reactor vessel via the coolant recirculation .aops. The LPCI system is designed to provide reactor core cooling for a large spectrum of loss-of-coolant accidents completely independent of the reactor core spray cooling systems.

Evaluation Peak clad temperatures have been calculated by the General Electric Company for the complete spectrum of break sizes in the pressurized core cooling system up to and including the design basis double-ended recirculation line break. Using the calculational methods and assumptions that have previously been reviewed and accepted by DRL, peak clad temperatures remain below 2300*F for breaks smaller than 0.1 ft2 with automatic depressurization followed by core cooling with both spray systems or one core spray system and low pressure coolant

inj ection by two pumps. For breaks larger than 0.1 ft2 but smaller than the 4.2 f t2 (DBA double-ended recirculation line break), either two core spray systems or one core spray system and low pressure coolant

, injection by two pumps is sufficient to limit peak clad temperatures

' to less than 2300*F. The active fuel cladding metal water reaction is less than or about 0.1% for all break sizes, acceptably below the 1%

value specified in the AEC Interim Design Criteria. Selected examples of assumed primary coolant breaks show that peak clad temperature

i

(~

4 Files May 3 B72 i

i transients endure above the normal operating 600-650'F temperature i for as long as 350-400 seconds and that the temperatures in some cases remain above 1500'F but below 2300'F for as long as 220 seconds, in our opinion, this is insufficient time, temperature, or volume of clad

' to cause core damage as a result of clad embrittlement that could interfere with the emergency core cooling process.

We have observed that the calculated clad temperature increases, following all LOCAs up to and including the double-ended break of the recirculation l

piping, are arrested by core flooding prior to reaching 2300*F at any j location in the core. The following observations require further eval-uation by DRL:

i

1. Figures 11,16,17 and 18 of the subject report show core flooding f rom two core spray systems following the design basis LOCA within approximately 175 seconds, whereas the FSAR states "the accumulation of core spray water in the bottom of the vessel covers the core in about 8 minutes for the maximum design break". There appears to be a time 4

discrepancy.

2. Figures 11,16,17 and 18 of the subject report show core flooding following the design basis LOCA within approximately 175 seconds, a period of time that is noticeably shorter than the 215 seconds previously reported in FSAR Figure 6-2.25

and 185 seconds in FSAR Figure 6-2.49. According to the most recent calculations, core flooding time appears to be 10-40 seconds f aster than for comparable conditions reported in the FSAR.

i

3. Figures 13,14 and 15 of the subject report show peak temper-atures that are 400-450*F higher than the temperature reported i in Figure 1 for 0.1 ft primary 2

coolant system break. There appears to be an inconsistency in reporting the peak clad temperatures calculated by GE in accordance with the AEC Interim Policy Statement on Emergency Core Cooling Systems.

4. The DBA primary system break while not identified numerically in the subject _ report is estimated f rom Figure 1 to be about-4.2 ft2 This value is in agreement with some of the references in the FSAR but not with Figure 6-2.1 of the FSAR which shows that the maximum recirculation line break size is 5.6 ft2 Resolution of this apparent difference in the i

a Z

i

Files MAY 3 1972 size of the DBA is important because extrapolation of the Figure 1 results to the larger break size results in peak clad temperatures in excess of 2300*F, the limit specified by the AEC Interim Acceptance Criteria for ECCS for light water reactors.

5. Table I and Figure 1 of the subject report show that peak clad temperatures occur following a 0.05 f t2 break which is now defined as an intermediate break. When compared with Figure 6-2.34 in the FSAR, it is evident that the reanalysis has resulted in a significant shif t of the peak clad temperature from the approximately 0.15 ft 2 break to

0.05 ft2 break. Intermediate and large breaks, i.e. , breaks

previously defined as a break size greater than 0.1 ft2, required no HPCI or auto depressurization. The peak temper-ature based on the recent GE calculations now occurs in the region formerly described as the small break range of breaks (i.e., 4.0.1 ft2) and HPCI or auto depressurization is a pre-d requisite to satisf actory cooling by LPCl or CSCS. Since it is assumed in the GE analysis that the HPCI is inoperative,

- depressurization in time for effective core cooling by spray and flooding must be accomplished by the Automatic Pressure Relief System. Figure 3 shows the peak clad temperature and Figure 10 shows the reactor vessel water level following a 0.05 ft2 break as a function of time. We cannot determine whether two of the three automatic relief system valves were assumed to open to discharge 1.6 x 106 pounds of steam per hour at 1125 psig as described by the FSAR or if all three valves were assumed to open to depressurize in time for acceptable emergency core cooling; nor can we explain the transient increase in reactor vessel water level at about 180 seconds af ter the break or verify that the two-minute i

time delay in opening the automatic pressure relief valves af ter sensing high drywell pressure and low reactor vessel water has been properly considered by GE.

W" f , 0 1

Files SY 31972 During telecons on April 7, 13 and 14, 1972, NSP was requested to pro-vide responses to the above concerns so that our evaluation of the Monticello "ECCS Conformance to the New AEC Adopted Interim Acceptance Criteria" can be completed.

,_ w James J . Shea Operating Reactors Branch #2 Division of Reactor Licensing cc: DJSkovholt, DRL TJCarter, DRL DLZiemann, DRL JJShea, DRL RMDiggs, DRL t