ML20197D187

Staff Eval of Nedo 21778:Transient Pressure Rises Affecting Fracture Toughness Req for Boiling Water Reactors. Requests Elimination of Criticality Limit Based on Temp for Insvc Hydrotest & Substiting 25 Psi Pressure for 400F Temp

ML20197D187
Person / Time
Issue date:	01/31/1978
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20197D157	List: ML20197D153 ML20197D187
References
NUDOCS 7811210411
Download: ML20197D187 (7)

Text

_ . _

Qs e EilCLOSURE TOPICAL REPORT EVALVATI0t1 Report tio: f1ED0-21778 Report

Title:

Transient Pressure Rises Affecting Fracture Toughness Require-ments for Boiling Water Reactors Report Date: January, 1978 g

Originating Organization: General Electric Company Reviewed by: Engineering Branch and Reactor Safety Branch, Division of Operating Reactors; Reactor Systems Branch, Core Performance Branch, and Materials Engineering Branch of the Division of Systems Safety.

SUMMARY

OF REPORT The report contains justification for a request from the General Electric Company for relief from a provision of Appendix G, 10 CFR Part 50, which i sets minimum temperature limits for reactor core criticality. Paragraph IV.A.2.c.

reads:

Whenever the core is critical, the metal temperature of the reactor vessel shall be high enough to provide an adequtte margin of protection against fracture, taking into account such factors as the potential for overstress and thermal shock during anticipated operational occurrences in the control of reactivity. In no case when the core is critical (other than for the purpose of low-level physics tests) shall the tempera-ture of the reactor vessel be less than the minimum permissible temperature for the inservice system hydrostatic pressure test nor less than 40 F above that temperature required by section IV.A.2.b.

The primary change requested by GE is elimination of the criticality limit based on the temperature required for an inservice hydrotest. A secondary change requested by GE is the substitution of a 25 psi pressure margin for the '40 F temperature margin in paragraph IV. A.2.c. The effect of the two changes requested by GE on the pressure-temperature limits for a BWR with a radiation-sensitive beltline material near end of lit o is shown by the shaded region in Figure 1 (taken from Figure 2-1 of the Ttpical 78112104I l

)

4 Report). At present, the shaded region must be traversed by heating with non-nuclear heat, such as heat from the recirculation pumps, whereas the ,

requested changes would permit tcking the reactor critical to obtain power for heatup in this region. The Report points out that substantial savings could be realized in startup time and in power replacement costs, in the discussion of NEDO-21778 at meetings with GE people, it was explained that BWRs cannot use pump heat during startup as effectively as PWRs can, because the elevation head of water in the reactor alone is insufficient to meet the NPSH requirements of the pumps at all but the lowest speeds. Hence, pump heat is low until there is steam pressure in the reactor.

The report addresses the concern expressed in paragraph IV.A.2.c. of Appen-dix G to 10 CFR 50 that there is a potential for overstressing the reactor vessel as a result of some transient during reactor startup or when , critical due to a subsequent malfunction in the control of reactivity. The control rod drop accident (CRDA) was selected as the limiting case for study, despite its low probability of occurrence, because CRDA conditions bound those of the anticipated transients that would have to meet Appendix G.

The CRDA increases in total energy output and peak energy density directly with the assumed reactivity worth'of the dropped rod. Rod worths are limited such that the peak energy density is less than 280 cal /gm, the staff limit criterion for the event. For this energy density limit there is no prompt dispersal of the V02 fuel into the coolant and thus no prompt pressure pulse and heat transfer from fuel to water, and thus pressure increase, may be determined by conventional heat transfer methods. For a 280 cal /gm bounding event, giving maximum total energy output, the G.E. analysis shows that the resulting pressure transient should not exceed 12.5 psi.

With regard to the potential for a water solid pressure transient that would violate the revised Appendix G limits recommended in the Report, the only events discussed by GE were a rod drop accident during a hydrotest or concurrent with an inadvertent filling of the vessel to water-solid conditions.

No detailed analysis was provided for either event, because GE believes that each is of sufficiently low probability to not be considered.

SVKMARY OF THE REGULATORY EVALUATION B.achground During the evaluation of the Topical Report, consideration was given to additional information obtained from meetings and correspondence as given in the attached chronological listing.

Paragraph IV. A.2.c. was written into Appendix G during the final review by the ACRS in 1973, to satisfy a concern that the reactor vessel material be .

warm enough to be in a ductile condition before undertaking any startup 2

I

• e s

.r operations that could overstress the vessel if some malfunction occurred.

Core criticality was chosen as the specific startup operation to make the requirement definite. The required temperature for an inservice hydrotest was used for the criticality limit, because that would represent the lowest temperature at which vessel integrity at pressures near the relief valve z

settings had been demonstrated on the vessel in its irradiated condition.

Control Rod Drop Accident The staff has accepted the use of the CRDA as a limiting case with regard to overstress of the reactor vessel resulting from a malfunction during the control of reactivity. We also accept the reported estimate of 12.5 psi pressure rise, for the following reasons: First, by having previously accepted an estimate of 280 cal /gm peak fuel enthalphy, we have agreed that a certain physical model of fuel damage is reasonable.

The time constant for the process of heat transfer to the water is estimated at greater than 6 seconds, much too long for there to be significant nuclear-to-mechanical energy conversion. Second, even if the calculated pressure rise (12.5 psi) were off by a factor of 10, the effect on the vessel would not be significant. In this calculation, water level was assumed to be normal, i.e., the large vapor space in the vessel and in the steam lines cushions the pressure rise.

If the criticality limit is modified as requested, it is possible that the reactor could be taken critical to warm up the vessel for a hydrotest. To further reduce the possibility of a CRDA while the vessel is water solid for a hydrotest, it will be necessary to add a requirement to Appendix G that all control rods must be fully inserted during hydrotest.

Despite general agreement that the 12.5 psi pressure rise estimate is reasonable, the criticality margin of 40 F required by Appendix G was retained. The GE recomnendation to substitute a 25 psi pressure margin, a minor part of the requested change, was not accepted because: (a) 25 psi is a auch smaller margin than 40 F at the temperatures of interest, and (b) it is a tradition of the transition temperature approach to state this sort of margin in terms of temperature.

Water Solid Pressure Transients Despite the generally good operating history of BWRs with regard to water solid pressure transients, the staff explored typical existing startup procedures and systems aspects of potential vessel overpressure, goino well beyond the coverage in the Topical Report. As can be seen in Figure 1, startup operations that normally occur af ter achieving criticality now take place at temperatures above the shaded area, whereas the requested rule change would permit these operations to occur at lower temperatures where the pressure limits are reduced and more easily exceeded. Violation of

' pressure limits in a BWR during startup could only occur if water level was not adequately controlled.

3 1

as In thei. at the BWR Simulator, in Morris, Illinois, the staff received the follow 1,3 additional information on water level control.

1. ,For the vessel to go water solid, the level would have to rise about 300 inches above normal (a volume of 2700 to 5600 cu. ft, depending on the plant).

2. There are three high-level alarms - at 10,18 and 25 inches above normal.

3. During startup, there are several pumps that could supply water to the vessel:

a. Condensate booster pumps (shutoff head 400 - 600 psig, typically) would be running - could not be tripped by high level conditions.

b. Control rod drive pumps (shutoff head 1500 psig, typical flow 40-80 gpm) would be running.

c. Feedwater pumps (shutoff head s 1500 psig) are not normally started until steam pressure reaches 300 to 600 psig, according to NED0-21778. Steam driven pumps could not start at the temperatures of interest. The feedwater flow control valve closes on a high-water signal if in automatic level control mode. The high level trip would trip the feedwater pumps.

d. The low pressure coolant injection (ECCS) system (375 psig shutoff head) could fill the vessel in 2 or 3 minutes if inadvertently actuated. (No high level trip.)

e. The high pressure coolant injection system is steam driven and thus is not of concern Dere,

f. The high pressure core spray system trips automatically on high water level.

. 4. Various let-down paths are available during startup as follows:

a. Drain flow to the reactor water cleanup system is manually con-trolled as part of level control during startup.

b. The head vent 'is open to the dry well sump until a temperature of about 190 is reached, at which time it is vented to the steam lines.

c. The main steam line isusation valves and drain valves are open during startup.

4

d In summary, we conclude that the probability of a water solid pressure transient in BWRs is sufficiently low and that the proposed lowering of the criticality limits will not increase the probability significantly.

For a water solid oressure transient to occur, (a) there would have to be an error or malfunction that caused an unanticipated inflow from some pump (b) the operator would have to ignore three high-water alarms (in some cases, there would have to be failure of automatic protection

• feature), and (c) the normal let-down paths would have to be reduced "

through error or malfunction.

One positive aspect of the proposed lowering of the criticality limits is the reduction in time spent at temperatures where fracture is of concern.

Such a reduction of time spent in this region would imply less chance of occurrence of a maintenance or testing operation that might inadvertently flood up and pressurize the vessel. 5

,R,egu_latory Position The Regulatory Staff agrees that it is desirable to amend Appendix G, 10 CFR 50 to concur in the requested deletion of the criticality limit based on the minimum permissible temperature for the inservice hydrotest. In its

- place, the criticality limit at very low pressure would be based on considera-tion of fracture prevention in the flange regions that are highly stressed by the bolt preload. The Regulatery Staff does not concur in the requested change of the criticality margin from 40 F to 25 psi. The staff-recommended pressure-temperature limits are illustrated by the dashed lines in Figure 1.

A proposed package of amendments revising Appendices G and H is now being ,

reviewed within NRC. Paragraph IV. A.2.c. has become paragraph IV. A.l.f. in i the current draft. It reads as follows:

When the core is critical (other than for the purpose of low-level ,

physics tests) the temperature of the reactor vessel shall not be less than 40 F above (22 C above) the minimum permissible temperature of paragraph e. of this section nor less than 60 F above (33 C above) the reference temperature of the closure flange regions that are highly stressed by the bolt preload. All control rods shall be fully inserted .

during hydrotests. .

r 5 .

f 4

e

9 p .

ATTACHMEi4T I CHRON0 LOGICAL LISTING OF f1EETINGS AND CORRESPONDENCE REGARDING NED0-21778

July 13,1977 Le uer (MFN-270-77) to the Secretary of the Commission, U.S. NRC, from Glenn S. Sherwood, flanager, Safety and Licensing, Nuclear Energy Systems Divison, General Electric Co.,

Subject:

. Comments on Proposed Change to 10 CFR 50, Appendix G, " Fracture Toughness Requirements",

noted in Federal Register of liarch 31, 1977, Volume 42 -

No. 62. This letter contained the first sugge:, tion from GE regarding the change to Appendix G.

December 6, 1977 Meeting at San Jose, CA, at which an early draft of NED0-21778 was discussed.

January 24, 1978 Letter to the Commission, Attn: Olan D. Parr from J. F. Quirk, Manager BWR Standardization, Transmitting the Licensing Topical Report, " Transient Pressure Rises Af fecting Fracture Toughness Requirements for Boiling Water Reactors," NED0-21778, F. E. Cooke, L. A. Kelley, C. J. Peone, R. L Shingleton and H. T Watanabe.

March 16,1978 Informal meeting in Bethesda with some of the authors I of the report.

April 6,1978 Letter (HTW-78-016) to' V.S. NRC, Attn: Mr. W. S. Hazelton from H. T. Watanabe, Principal Licensing Engineer, Operating Plant Licensing, General Electric Co.,

Subject:

f1EDO-21778.

May 26,1978 Letter to H. T. Watanabe from W. S. Hazelton,

Subject:

NED0-21778, which transmitted informal questions for use as agenda for upcoming meeting.

. June 7, 1978 Meeting in Morris, Illinois at location of BWR Simulator.

e

I,,

.l. l l l l

. ~E-4 12 b " =

> :e

'y2x' -

8

[isQ.$

es d-ll-

s

!s E E g

'/ N ilMO

/ f - 40 F -

/

1

.o 800 -

! N

.g -

1-CP.lTICALITY LIMITS  !

-[i RE QU ESTE D IN c: N E DO-21778 / \\

' 600 - -

/

. / \ ,

! PRESENT

/ CRITICALITY 400 -

f E

I i

200 -

s CRITICALITY LIMITS- '

PROPOSEDIN APP G & H REVISION O  ! '  ! I  !

O 100' 200 300 400 500 600 TEMPE R ATU R E, "F

~

FIGU RE 1.

PRESSURE-TEMPERATURE LIMITS AT END OF LIFE FOR SOME BWR's (FROM FIGURE 2-1 OF NEDO-21778)

I