ML20246F733

Summary of ACRS Advanced Reactors Designs Subcommittee 881005 Meeting in Bethesda,Md Re NRC Draft SER for Power Reactor Inherently Safe Module Conceptual Design

ML20246F733
Person / Time
Issue date:	11/08/1988
From:	Advisory Committee on Reactor Safeguards
To:	Advisory Committee on Reactor Safeguards
References
ACRS-2605, NUDOCS 8905150073
Download: ML20246F733 (17)

Text

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _.

06RS-Ms05 l;

Qi)fpm, ppe sp/n' DATE ISSUED:

11/8/88 wer ACRS MEETING MINUTES /

SUMMARY

OF THE ADVANCED REACTORS DESIGNS SUBCOMMITTEE OCTOBER 5, 1988 BETHESDA, MD Purpose The purpose of this Subcommittee meeting was to review the NRC staff's draft safety evaluation' report (SER) for the Power Reactor Inherently Safe Module (PRISM) conceptual design.

Attendees ACES NRC D. Ward, Chairman J. Flack, RES J. Carroll, Member R. Landry, RES C. Michelson, Member T. King, RES F. Remick, Member Z. Rosztoczy, RES C. Siess, Member G. Vissing, NRR C. Wylie, Member P. Williams, RES R. Avery, Consultant J. Wilson, RES M. El-Zeftawy, Staff Others G. Van Tuyle, BNL G. Sherwood, DOE C. Allen, SAIC J. Humphrey, DOE M. Subudhi, BNL L. Strawbridge, W-G. Slovik, BNL R. Amar, RI S. Poltorak, SERCH R. Lancet, RI C. Lewe, NUS F. Tippet, GE M. Waterman, INEL G. Gyorey, GE D. Wade, ANL R. Berglund, GE D. Pedersen, ANL R. Ketchel, GE J. Cahalan, ANL K. El-Sheikh, GE l

B. Seidel, ANL P. Macgee, GE H. Alter, DOE T. Geinsterg, BNL Rsc+

' \\

8905150073 881108 PDC Certified B7 1

N r

Advanced Reactors' Designs Meeting. Minutes October 5, 1988 Meeting Highlights,' Agreements, and Requests 1.

Mr.. Ward, Subconsnittee Chairman, stated. the purpose of the subcom-mittee meeting and introduced the other ACRS members and consul-

. tant.

Mr.. Ward indicated that the~ACRS is expected to write a c-letter on this subject at the November _1988 full Committee and not'

~

the October 1988 meeting.

2.

Mr. J. Griffith, ' Associate Deputy Assistant Secretary /D0E, de-scribed the advanced re; tor program approach. Mr. Griffith outlined the' challenges that faces the program as follows:

• Regulatory criteria Public attitudes

• Complexity of the design

• Plant life (operation and maintenance)

• Uncertain load growth l

• Financing The DOE-sponsored approach has developed the following technical responses:

Passive safety through metal fuel and IFR technology j..

Modularity (factory fabrication and advanced instrumentation andcontrols)

• Standardization (Design certification by NRC, component and system reliability)

• Improved waste management (IFR fuel cycle, actinide burning).

Mr. Griffith emphasized the benefits from passive and inherent safety as follows:

)

j Advanced Reactors Designs Meeting Minutes October 5, 1988

• Improves safety margin

• Requires no operator action or external power

• Lead to simpler plants and simpler operations

• Risk of core melt accidents is insignificant Aids in plants capital investment protection deployment flexibility and plant cost reduction

' Aids in public perception of safety 4

'Other advanced design innovative feature example is the potential bottom support cost savings versus the top support for the reactor vessel (M58% or 30.4 million dollars savings).

The DOE expected LMR program schedule is:

• Select reference concept design by beginning of FY 1989

• Decide whether to further develop reference concept by end of FY 1991

• Define future needs for advanced reactor development and demonstration Mr. Griffith commented that there is a high international interest j

and summarized the international cooperation as follows:

• International cooperation is cost effective even though the U.S. is developing modular reactor with metal fuel cycle and l

l

g 1

Advanced Reactors Designs i

Meeting Minutes October 5, 1988 ll

-the Japanese and European' Consortium are developing mono-

-l lithic reactor with oxide fuel cycle

• U.S. completing oxide fuel R&D program through overseas b

cooperation (e.g., fuel transient testing with UK, long-life fuel tests with Japan and remote systems development with Japan,UK, France,FRG) 3.

Mr. T. King, RES, summarized the SER status as follows:

Review of the PRISM by RES is complete Some SER sections need updating; Chapter 3 - criteria, Chapter 6 - containment, Chapter 15 - safety analysis Completien of internal reviews by NRR is expected-by November 1988. ACRS and CRGR also by November 1988. The RES staff is planning to submit the completed SER to the Commission by late November 1988 Mr. King indicated that the staff's review is considered a pre-application review for the purpose of providing guidance early in the design process on the acceptability of the PRISM design. The SER does not constitute an approval of the PRISM design but rather documents the Commission's preliminary guidance regarding licensing requirements, including the acceptability of the DOE proposed supporting research and development programs.

Mr. King stated that, based upon the results of the review, the staff overall conclusion is that the PRISM design has the potential to achieve a level of safety at least equivalent to current l

)

5 Advanced Reactors Designs l

Meeting Minutes-October 5, 1988 generation LWRs, provided that a number of safety issues to be resolved (e.g., response to certain bounding events that leads to fuel melt and/or sodium boiling and the potential for positive reactivity feedback accident). Mr. King commented that the final determination of the PRISM acceptability is contingent upon the; following:

Satisfactory resolution of the issues identified in the SER j

Completion of final design and licensing review by NRC Successful design, construction, testing and operation of a prototype reactor prior to design certification f

l Adequacy of containment 4.

Mr. R. Landry, RES, stated that the NRC staff reviewed a prelimi-nary safety information document-(PSID) that was provided by D0E.

]

The PRISM conceptual design has been developed for DOE by the Generic Electric Co. (GE) in conjunction with Bechtel Power Corpo-ration, Borg Warner, Foster Wheeler and United Engineers and Constructors. The conceptual design submitted is for a small, modular, pool type liquid metal reactor.

Each reactor module has:

• Core power of 425 MWt

• Primary Sodium inlet temperature of 610*F

• Primary Sodium outlet temperature of 825'F

• Primary Sodium Flow rate of 40,800 Gpm The overall plant characteristics consist of nine reactor modules with a total thermal output of 3825 MWt (1245 MWe) and one control room. The facility is to be designed to permit siting at 90% of i

O______._____-._'_..____.____

l Advanced Reactors Designs Meeting Minutes October 5, 1988

.!l existing U.S. sites. The design life of the facility.is proposed to be 60 years.

1 The PRISM core is designed to use metal. fuel rather than oxide fuel. The core is designed.to have a small reactivity swing during the fuel cycle.

Inherent feedback mechanisms are expected to result in a. negative reactivity coefficient for all anticipated transients.

Reactivity and power are controlled by six, indepen-dently controlled, control assemblies. Each control assembly is capable of shutting down and maintaining the core in a cold shut-down-cendition.. The primary coolant is forced through the core by four electro-magnetic pumps. The shutdown heat is removed in three ways via:

(a) the main condenser, (b) an auxiliary steam generator air cooling system (ACS),'and (c) the safety-related reactor vessel auxiliary cooling system (RVACS). The LMR experience to date is based on operation of seven facilities (EBR-II, Phenix, PFR, FFTF, SNR300,Monju,andSuperPhenix). Only Super Phenix is a large i

commercial power plant.

Mr. Landry stated that the NRC staff developed definitions for various event categories (ECs) with correspondence to conventional LWR event categories, as:

EC I

- Abnormal Operational Occurrences EC II - Design Basis Accidents l

EC III - Severe Accident EC IV - Emergency Planning Basis Events The selection of a spectrum of accidents which must be considered l

in the design, beyond the traditional LWR design basis accident (DBA) envelope,isconsiderednecessaryforadvancedreactors.

For j

1 I

example, EC III would include:

l 1

w__

Advanced Reactors Designs Meeting Minutes October'5, 1988

Internal events (less.likely. initiating events plus multiple failure events) down to a frequency of ^/10-7/yr.

External events beyond those included in EC II

• Using engineering judgement, additional bounding events to account for plant specific uncertainties, e.g.:

(a)

Inadvertent withdrawal of all control rods, without scram-for 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> (b) Station blackout for 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> (c) Loss of forced cooling plus loss of RACS/RVACS, with 75%

inlet blockage after 36 hours4.166667e-4 days <br />0.01 hours <br />5.952381e-5 weeks <br />1.3698e-5 months <br /> (d)

Instantaneous loss of flow from one primary pump with coast down flow for other pump (s)

(e) Steam generator tube rupture (f) Large Na Leaks (g)

Flow blockage of one fuel assembly The NRC staff believes that the identification and selection of accidents is consistent with the Commission's severe accident policy statement, particularly those proposing the use of a mecha-nistic calculation of source terms and a shift in emphasis from accident mitigation to plant protection.

Events will be selected

'using engineering judgement, complemented by PRA.

Mr. Landry, described the inherent safety characteristics of the PRISH design. The inherent safety refers to the reactor design 1

Advanced Reactors Designs Meeting Minutes October 5, 1988 features that cause it to transition to a lower power level during overheated conditions, for example:

Doppler feedback - for metal fuel it is a smaller factor than it is for oxide fuel, but it still adds negative reactivity on a power increase

• Sodium density / void feedback - this is a positive feedback.

As long as the sodium is subcooled, this contribution is modest Axial expansion - negative coefficient and it is a function of burnur

• Radial expansion - negative coefficient, slow acting L

l

• Bowing - limited free bow to keep coefficient negative Control rod driveline expansion - results in inserting control rods further into the reactor, adding negative reactivity

• Vessel expansion - results in withdrawal of control rods and a positive feedback With respect to the PRISM safety objectives regarding these feed-backs, GE is designing to achieve the reactor power runback (shut-down) in response to reductions in both heat removal and reactor sodium flow rate.

l l

l Mr. Landry described the containment system for the PRISM.

It is l

not a conventional containment.

It will be designed however, to l

provide a leak tight boundary that tends to contain the accidental release of core fission products and primary coolant.

It will also L_____._________.____

f

)

Advanced Reactors Designs Meeting Minutes October 5, 1988 l

be designed to' withstand the static and dynamic loads resulting

'l from'a primary sodium leak accident. The containment is to'be a second, or guard, vessel surrounding the reactor vessel. The staff believes-that the acceptance of the design without conventional containment will require extensive review, testing and demon-stration, in addition to a satisfactory compliance to the criteria-l specified by the staff, e.g.:

l~

10 CFR 50, Appendix I and 40 CFR 190 limits.should be' met' The 10% of'10 CFR 100 dose guidelines should be met The fission product retention capability must be demonstrated

• Different emphasis and types of 0/A, surveillance, in-service inspection, etc., should be provided.

Mr. Landry summarized the favorable factors in the PRISM design as follows:

• Passive means of decay heat removal and reactivity insertion resulting in power level reduction are available i

• Potential for only minor core damage and fission product release for many events

• Reduced dependence on human actions and reduced vulnerability i

to human error Long response time under many accident conditions Capability to demonstrate by test significant safety features and plant performance

i Advanced Reactors Designs Meeting Minutes October 5, 1988 The above favorable factors are contingent on the following:

• Satisfactory resolution of all issues identified

• Completion of R&D i

Completion of final design and licensing review Successful design, construction, testing and operation of prototype reactor 5.

Mr. G. Van Tuyle, BNL, summarized the safety analysis study that was' performed by BNL for the PRISM design. Three beyond the design basis events (BDBE) have been analyzed. These events have a low probability of occurrence, and it is assumad that there is no intervention from either the reactor control or protection systems.

Loss of Flow (LOF) - the unprotected loss of prim ny flow transient was analyzed for B0EC conditions. The transient analyzed was a simultaneous trip of both the primary und intermediate pumps without any control rod action. The results show that the trip of the primary and the intermediate pumps causes a rapid flow coastdown and a rapid coolant temperature increase. The increased temperatures activate the inherent feedbacks of the core and leads to a reduction in the power level. There is sufficient negative feedbacks from the grid plate, Doppler, bowing, and radial expansion to override the positive feedbacks from the sodium, axial expansion, and control rod drive line withdrawal.

• Reactivity Insertion (TOP) - this event was assumed to occur at full power. The severity of this event is determined by j

the amount of reactivity associated with the inserted rods

l Advanced Reactors Designs Meeting Minutes -

October 5, 1988 during operation. The most severe case was analyzed with all 6' rods are withdrawn (maximum worth of 35 cents including

. uncertainties). The initial jump in power occurred due to the positive reactivity insertion and. sodium density changes, but the heat up of the fuel support structures then' activated the inherent negative feedbacks' The major negative reactivity mechanisms'were the Doppler, axial expansion, and bowing.

l.

• Loss of Heat Sink (LOHS) - this event assumes that the flow'in the intermediate loop quickly goes to zero. The heat. gene--

rated in the core stays within the primary system. The primary system begins to heat up, activating the inherent reactivity feedbacks. The power level drops to less than.5%

and the net reactivity remains strongly negative during the transient.

• T0P/LOF - BNL analyzed this' event that includes 35 cent reactivity insertion (TOP) in addition to tripping the pumps.

BNL concluded that some safety margin still exist, however serious challenge (if unscrammed) would be of a concern.

This event is considered to be of a very low probability to occur.

BNL used the SSC and MINET computer codes to simulate the reactor system and independently assess the postulated transients.

The final conclusions by BNL regarding the PRISM inherent shutdown are:

}

• Analysis contains significant uncertainties

• Margins for TOP, LOHS are substantial

• Smaller margins for LOF events

7-._ __

Advanced Peactors Designs-Meeting Minutes October 5, 1988 Vulnerability of EM pumps coastdown via synchronous machines and cable links is a concern

• Safety. tests are needed to confirm margins EM pump coastdown must be assured and. conclusions for passive-cooling are:

i

• RYACS performance was confirmed and shown to be fault tolerant -

• Long time for heat-up

• Heat removal is a strength for the PRISM design 6.

Mr. J. Flack, RES, presented the PRA study for the PRISM design.

The PRISM PRA is a level three conceptual PRA that includes the systems, containment, and consequence analysis. The assumed location for population distribution is the GESSAR-II site. This PRA does not consider:

Sabotage External events other than (limited) seismic

• Startup accidents

• Multi-module interactions

• Accidents due to radioactive sources outside the reactor vessel Normal plant effluent releases l

6

~

Advanced Reactors Designs Meeting Minutes October 5, 1988 1

Mr. Flack stated that best estimate values (no uncertainty dis-tribution) were used. There are three general classes of initia-tors:

(a) reactivity insertions excluding seismic, (b) seismic, and (c) heat removal faults. The study considered these initiators to bound many of the concerns. The impact of DC power, instrument air, service water, and interactions among support systems have not been addressed because of lack of detail.

l Mr. Flack commented that there are some specific issues that the staff is concerned about, e.o.:

• The frequency of station blackout was estimated to be 3x10-5/yr. (comparable with LWRs).

PRISM however does not have Class IE energency diesel generators and must run back reactor pcwer to pick up house loads during loss of offsite power.

' GE assumed the ability of plant control system and BOP to runback power in nine out of ten transients and that has to be demonstrated.

Tube ruptures due to seismic effects that lead to steam generators failures and common cause failures that could result from fatigue, thermal shock, and aging have not been modeled in the PRA analysis.

Excluding large earthquakes, no common cause failures of the synchronous machines were postulated.

l l

EM pumps raise a special concern, as their instantaneous loss f

without scram could result in a destructive power excursion, the consequences of which are difficult to predict.

1 l

l f

l E____---_-_--------

i

w:

'0 r-j;

^

o z

g

.c

~

Advance'd' Reactors Designs

. ~ Meeting Minutes October 5, 1988

' Source term estimates may be low as a result of extrapolating oxide fuel to metal fuel.

y i

• The role of the operator is not apparent from the PRA.

The NRC staff concluded that the PRISM PRA has provided'a prelimi-nary overview of the' plant's vulnerabilities (e.g.; seismic events and primary pump failures drive the LOF sequences',.which are believedtobeLthedominentriskcontributors). However, due to the large uncertainties in the phenomenological. treatment of the core response and consequence analysis, it is not clear how useful the PRA-will be in the decision process, o

7.

As a result of the Subcommittee discussion, some of the Subcommit-tee's members and consultants expressed some concerns in regard to the following:

• Dr. Kerr questioned the meaning of the statement made by the NRC staff that indicates the Commission's expectation of higher safety enhancement of the PRISM design'than current generation of LWRs but it is not required.

1 Dr. Kerr commented that the staff should not use in its

]'

reviews GDC from the LWR's technology that could be obsolete for the advanced designs.

• Dr. Avery expressed some concern that the bounding events used in the analysis do not lead to and completely cover core l

disruptive accidents.

• Dr. Kerr asked why did not the NRC staff obtain and make use of existing data from Super Phenix commercial power plant?

l i

l 4

t Advanced Reactors Designs Meeting Minutes October 5, 1988

• Mr. Ward questioned the role of reactor operator for this design.

He pointed out that the two major power reactor accidents that have occurred, TMI-2 and Chernobyl, were caused l

by wrong but purposeful actions by operators.

PRA does not do a good job in describing or estimating the likelihood of such events.

Dr. Kerr questioned the decisions and assumptions that were made by the NRC staff to account for the defense-in-depth philosophy.

Mr. Carroll commented that the sabotage issue especially by the informed insider need to be addressed in great details.

• Dr. Siess questioned the scenario for one EM pump failure and not the others and what are the bases for the random selec-tion. He is also concerned regarding the common-mode failure of more than one pump.

Mr. Michelson expressed some concern regarding the exclusion of external events in the PRISM analysis and how much study was performed regarding the fire hazards for the sodium.

• Dr. Remick commented that it is not clear if the NRC staff knows enough about 10 CFR Part 50 reprocessing and application

.for the advanced design.

Mr. Ward expressed some concern in regard to the flow blockage and the lack of analysis and experimental data to support the staff's argument that this event is considered to be remote.

l 4

Advanced Reactors Designs Meeting Minutes October 5, 1988 Dr. Remick commented that the containment issue need to be discussed further, indicating that he is not convinced that reactors should be built at this time _in the. United States without containment.

Dr. Siess commented that based on the fuel geometry the statement made.by the staff that one control rod can shutdown the reactor should not be an absolute statement. Mr. Ward agreed.

Dr. Kerr expressed some concern regarding the PRISM design that does not have Class 1E emergency diesels and AC power.

• Mr. Carroll commented that this design should have a technical support center similar to LWRs design.

Dr. Avery expressed some concern regarding the operation of nine modules with one single' control room. He commented that the installation of control equipment during the shutdown or operation of one module while the rest of the modules are operating could be disruptive and pose a concern. Dr. Remick agreed.

1 Dr. Remick commented that having one reactor operator operat-

)

ing 3 or 4 reactors all at different states could be of a concern.

• Dr. Avery expressed some concern in regard to the containment issue.

He commented that the NRC staff and DOE rely heavily on the argument that it is not possible for significant activity to escape, and that has not been shown and proven yet.

L

~

e Advanced Reactors Designs Meeting Minutes October 5, 1988 q

• Dr. Siess commented that additional information is needed regarding the seismic isolators.

For a seismic initiating event, the reactor itself is only designed to 0.39,.the seismic isolators are designed to 1.0g to decouple the reactor vessel from the earth's horizontal oscillatory motion.

However, not all of the safety components are isolated (e.g.,

primary pump coastdown system, steam generators).

Future Activities The Subcommittee Chairman is planning to brief the full Committee in October 6-7, 1988, regarding the Subcommittee activities.

In addition, the NRC staff, DOE representatives and subcontractors will brief the full Committee in October 1988. The ACRS is not requested to write a letter on this subject at the October 1988 full Committee, however, a letter is expected at the November 1988 meeting.

NOTE:

Additional meeting details can be obtained from a transcript of this meeting available in the NRC Public Document Room, 2120 L Street, N.W., Washington, D.C. 20006,(202)634-3273, or can be purchased from Heritage Reporting Corporation, 1220 L Street, N.W., Suite 600, Washington, D.C. 20005,(202) 628-4888.

1

- - _. - _ _ _ _ _. _ _ _ _ _