ML20114A542
ML20114A542 | |
Person / Time | |
---|---|
Site: | Perry ![]() |
Issue date: | 01/22/1985 |
From: | Hiatt S OHIO CITIZENS FOR RESPONSIBLE ENERGY |
To: | |
References | |
CON-#185-259 OL, NUDOCS 8501280263 | |
Download: ML20114A542 (30) | |
Text
h . .
RELATED CORRESPONDQCg January 22, 1985 UNITED STATES OF AMERICA pA NUCLEAR REGULATORY COMMISSION l o Before the Atomic Safety and Li ensing Board ES g
' - - % 25
) N0:43 ;
- .~ J In the Matter of
) J c,.I ' F
. M* is .!, .
Docket Nos. 50-440 #-
7, ,: ..q , CLEVELAND ELECTRIC ILLUMINATING ) '
Ac 7' -
"' 4. . ' COMPANY, Et A1. j ) 50-441 4
. , . l , k4.'
- - - - ~
) (Operating License) '
(
.../[, W (Perry Nuclear Power Plant, ) .
'i M. Units 1 and 2) )
- .[ ,
. S h . .' '-
2 ,7 s'.L:y'; y ,
- . .y
- - y r.
3; . n ;
M , f;[ , ,. OCRE UPDATED RESPONSE TO APPLICANTS' ,! ,
'- SECOND SET OF INTERROGATORIES TO OCRE .
Issue #8, Interrogatory #3: OCRE hereby identifies additional ,
documents pertaining to I:: sue #8.
SECY-83-357, Amendments to 10 CFR Part 50 Related to Hydrogen -
Control, August 26, 1983.
Nuclear Safety, Vol. 24, No. 4, July-August 1983, pp. 502-503, on Survivability of Equipment Under Hydrogen Burn January 15, 1982 letter to Harold Denton, NRC, from J.D. Richardson, Chairman of HCOG, eubmitting Hydrogen Control Program
. Document, prepared by-Quadrex Corp.
March 18, 1983 letter to Harold Denton, NRC, from J.D. Richardson, Chairman, HCOG, submitting Westinghouse CLASIX-3 Report, Document No. WCAP-10260 (Non-proprietary version)
January 15, 1982 letter to Harold Denton, NRG,' from J.D.
Richardson, Chairman, HC0G, submitting "CLASIX-3 Contain-
$ a. ment Response Sensitivity Analysis" report >
Ne$ ~
hu eM April 8,1982 letter to Harold Denton, NRC, from J.D. R'ichardson, HCOG, submitting " Report on Hydrogen Control Accident Scenar'ios, Hydrogen Generation Rates and Equipment Sac Requirements"; also letters of October 26, 1982 and Sept. 9,1982, both submitting same report (including report).
- \
EE January 14, 1983 letter to Harold Denton, NRC, from J.D. Richardson, an.o H000, submitting Information on H2 Combustion Test Programs Board Notification BN-83-104, July 26,1983, re BWR Mark III Owners Group Meeting Minutes Concerning Hydrogen Control 13, Memorandum from M.D. of-BWR 1983, Summary Houston,Hydrogen Grand GulfControl ProjectOwners Manager,(Sept.HCOG)
Group
e
~2-Meeting on July 28, 1983 Documents listed in Appendix A to Sept.12, 1983 letter from J.M. Felton, NRC, to S.L. Hiatt, OCRE, re FOIA-83-469 Documents listed in Appendix A to January 30, 1984 letter from J.M. Felton, NRC, to S.,L. Hiatt, OCRE, re F0IA-84-7 Draft Report, "An Assessment of Postulated Degraded Core Accidents in the Grand Gulf Reactor Plant" by R.D. Gasser, Brookhaven National Laboratory, June 1982.
Draft Report, " Analysis of Full Core Meltdown Accidents in the Grand Gulf Reactor Plant" by R.D. Gasser, Brookhaven National Laboratory, August 1982.
Sept. 27, 1982 Memorandum from J.W. Yang to W.T. Pratt, Brookhaven National Laboratory, re MARCH Analysis of Hydrogen Burning During Degraded Oore Accidents for the Clinton Power Station and the Skagit Power Station December 20, 1982 letter from W.T. Pratt, Brookhaven National Laboratory, to Dr. John Long, NRC, submitting a BNL Memo entitled " Degraded Core Accidents in the Perry Power Plant" dated Dec. 13, 1982.
" Review and Evaluation of the GESSAR-II Probabilistic Risk Assessment - Containment Failure Modes and Fission Produce Release", letter report by Accident Analysis Group, Brookhaven National Laboratory, July 27, 1983 (Proprietary portions deleted)
Copies of documents provided by Applicants re Issue #8, as listed in February 25, 1983 letter from Alan P. Jones, CEI, to S.L. Hiatt, OCRE (not attached)
Documents listed in Appendix A to Dec. 7,1982 letter from J.M.
NRC, to S.L. Hiatt, OCHE, re hydrogen explosions Felton,,ff-gas in BWR o systems ( FOIA-82-545 ) <
NSAC Report; items pertaining to hydrogen generation and com-oustion tests: Vol. 2, No. 3 (Aug. 81); Vol. 2, No. 6 (Fall 82); Vol. 2, No. 7 (May 83); Vol. 2, No. 8 (Sept. 83);
Vol. 2, No. 9 ( Jan. 84)
NUREG/CR-1711, MARCH Code Description and Userig Manual, Battelle Columbus Laboratories, Oct. 1980 NUREG/CR-2285, Interim Technical Assessment of the MARCH Code, 'c by J. Rivard, Sandia, Nov. 1981 .
NUREG/CR-2672, SBLOCA Outside Containment at Browns Ferry Unit One -
Accident Sequence Analysis, by Condon et al, Oak Ridge National Laboratory, Vol. 1, Nov. 1982
NUREG/CR-2726, Light Water Reactor Hydrogen Manual, by Camp et al, Sandia National Lab., Aug. 1983 NUREG/CH-2530, Review of the Grand Gulf Hydrogen Igniter System, by Cummings et al, Sandia National Lab., March 1983 NUREG/CR-2442, Reliability Analysis of Steel Containment Strength, by Greimann et al, Ames Lab., Iowa State Univ., June 1982 NUREG/CR-1967, Failure Evaluation of a Reinforced Concrete Mark III Containment Structure Uncer Uniform Pressure, by Sharma et al, Brookhaven National Lab., Sept. 82.
NUREG/CP-0033, Proceedings of the Workshop on Containment Integrity, W.A. Sobrell, ed. , Oct. 1982, Vols. 1 and 2 NUREG/CR-2897, Calculations of Hydrogen Detonations in Nuclear Containments by the handom Choice Method, Delichatslos, et al, MIT, Sept. 1982 NUREG/CR-2836, Buckling of Steel Containment Shells, Vols 1-4, Lockheed Palo Alto Research Lab., Dec. 1982 t
NUREG/CR-2864, Identification of Safety-Related Equipment for Analysis and Testing in the Hydrogen Burn Survival Program Sandia National Lab., Nov. 1982 HUREG/CR-2898, Reinforced Concrete Containment Safety Under Hydrogen Explosion Loading, Fardis et al, MIT, Sept. 1982 NUREG/CR-3135, Buckling Investigation of Ring-Stiffened Cylindrical Shells with Reinforced openings under Unsymmetrical Axial Loads, Baker et al, Los Alamos National Labs, Feb.1983 NUREG/CR-3225, Data Analysis of the LLNL Hydrogen Igniter Experiments, Altenbach, et al, Lawrence Livermore National Lab. May 1983 NUREG/CR-2847,COGAP: Nuclear Power Plant Containment Hydrogen Control System Evaluation Code, R.G. Gido, Los Alamos National Lab., Jan. 1983 NUREG/CR-2865, Hydrogen Combustion in Aqueous Foams, Baer, et al, Sandia National Lab., Nov. 1982 NUREG/CR-1831, Hydrogen Distribution Af ter a Loss of Coolant
! Accident in the Subdivided Containment of Light Water Reactors, H.L. Jahn, Sandia National Lab. , Nov. 1980 NUREG/CR-2481, Light Water. Reactor Safety Research Program '4 Semiannual Report, April-Sept. 1981, M. Berman, Sandia National Lab., Feb. 1982 .
1 NUREG/CR-2486, Final Results of the Hydrogen Igniter Experimental Program, Lowry, et al, Lawrence Livermore dational Labs, Feb. 1982
4_
Preliminary Thermal NUREG/CR-2730, Hydrogen Burn Survival:en d Test Results, abs,McCulloch, et a Aug. 1982 CT-1380, Schott Letter 10/5/81 to Okrent re Comments on the Hydrogen Control Mechanisms in the Air Filled Mark III Containment Systems Sept. 8,1980 letter to NRC Chairman Ahehrne from ACRS Chairman Milton Plesset, re Additional ACHS Comments on Hydrogen Control and Improvement of Containment Capability NUREG-0831, Supp. 3, Grand Gulf SER, July 1982 The following are submittals and licensing correspondence on the Grand Gulf docket related to hydrogen control:
Date NRC PDR Acc. No. Subject Dec. 19, 1980 8101120257 NRC request to MP&L to perform ultimate contain-ment capacity analysis Jan. 21, 1981 '
8101270843 MP&L response to above Feb. 3, 1981 8102190768 NRC request to MP&L for description of hydrogen mitigation system Mar. 27, 1981 8104030405 MP&L response to above Apr. 9, 1981 8104140410 MP&L selection of igniter system for hydrogen control June 19, 1981 8106240238 MP&L further details on igniter system and contain-ment response Aug. 18, 1981, 8108210239 &L status report on igniter evaluations July 1, 1981 8107080396 MP&L forwards additional information on igniter system Aug. 31, 1981 8109030063 MP&L expanded report on hydrogen control measures Sept. 11, 1981 8109160069 MP&L responses to hydrogen action items Oct. 28, 1981 8111120814 NRC RAI from Chemical Engineering Branch on hydrogen control Nov. 6, 1981 8112030705 NRC RAI re hydrogen -
ignition system .
n-
-5
, Date NRC PDR Acc. No. Subject Nov. 19, 1981 8112090577 NRC revision to RAI of Nov. 6, 1981 Dec. 21, 1981 8112290102 MP&L response to NRC questions Jan. 12, 1982 8202220166 NRC request for hydrogen-
_ control information as soon as possible Jan. l'$, 1982 8201250330 MP&L endorses HCOG CLASIX-3 sensitivity study
'Jan. 19, 1982 8201210192 MP&L forwards report on equipment survivability
?
Jan. 21, 1982 8201260450 MP&L report on detonations Mar. 2, 1982 8203100181 MP&L forwards report on hydrogen control Mar. 10, 1982 8203160318 MP&L re'sponds to NRC letter of Jan. 12, 1982 Mar. 31, 1982 8204020429 MP&L addresses hydrogen burn effects on drywell Apr. 1, 1982 8204060086 MP&L submits vendor analysis of containment airlock seals Apr. 1, 1982 8204060035 MP&L responds to NRC questions on power supplies for igniters Apr. 5, 1982 8204070304 MP&L response re con-tainment atmosphere mixing Apr. 6, 1982 8204120356 MP&L discusses pool swell loads from drywell hydrogen burns Apr. 8, 198C 8204120182 MP&L submits information on igniter qualification
- testing Apr. 20, 1982 8204230267 MP&L responds to NRC.
questions of Mar. 30, 1982 Apr. 30, 1982 8205040539 MP&L forwards revision to emergency procedure and tech. spec. rc ignition system
Date NRC PLR Acc. No. Subject
.May 10, 1982 8205140434 MP&L advises of 2 revisions to. emergency procedures re ignition
- system
. May 17, 1982 8205270043 NRC letter. advising MP&L of incorporation of icaues raised re.Sequoyah June 6,.1982 .8206090051 MP&L submits additional information on contain-ment response re heat sink nodalization June 11, 1982 8206140151 MP&L forwards infor-mation re effects of pressure on equipment survivability June 11, 1982 8206160167 MP&L submits information .
re drywell head negative pressure capacity June 14, 1982 8206160178 MP&L evaluates emergency procedures re ignition system June 15,.1982 8206160159 MP&L submits revision to emergency procedure re reactor pressure venting
^
June 23, 1982- 820'7010023 NRC submits list of e outstanding information-
. required to complete r . evaluation of hydrogen issue June 25, 1982 8206290066 MP&L submits information
- re equipment survivability June 25, 1982 8206290538 MP&L forwards evaluation of response of personnel 4
airlocks to hydrogen detonations
- June 25, 1982 8206290310 MP&L response to NRC i RAI re pool impac,tjloads l July 1, 1982 8207020102 MP&L submits information re local detonations July 6, 1982 8207080358 MP&L provides information re hydrogen control equipment
(.
-7 Date NRC PDR Acc. No. Subjec.
Jan. 28, 1983 8302010528 MP&L informs NRC that detailed review re hydrogen control system procedures in progress Feb. 14, 1983 8302160295 MP&L's environmental qualification report on igniters Mav 11, 1983 8305130216 MP&L forwards response to questions in Feb. 3, 1983 letter re HCOG program June 16, 1983 8306230167 MP&L forwards " Grand Gulf Drywell Break Sensitivity Summary" June 23, 1983 8306270285 MP&L clarifies capabilities of igniter system under all circumstances July 22, 1983 8308050594 NRC comments re HC00 presentation of June 29, 1983 Aug. la, 1983 8308170018 MP&L response to above Aug. 23, 1983 8308250092 MP&L additional response to NRC's July 22 letter Construction Appraisal Team Inspection Report 50-440/83-31,50-441/
83-30, Nov. 7, 1983, pp. V-4 to V-7 NED0-10977, GE "Drywell Integrity Study: Investigation of Potential Cracking for BWR/6 Mark III Containment" Aug. 1973 Dec. 23, 1983 letter to J. Keppler, NRC, from M. Edelman, CEI, re response to CAT report, with pp 1-2 of response to Appendix A Nuclear Safety, Vol. 25, No. 3, May-June 1984, pp. 350-372, " Hydrogen Comoustion and Comtrol in Nuclear Reactor Containment Buildings",
L. Thompson et al.
Nuclear Safety, Vol. 25, No. 1, Jan.-Feb. 1984, pp. 53-74, " Hydrogen Behavior in Light-Water Reactors", Berman and Cummings. j NUREG-0011, Sequoyah SER, Supplements 3 and 4 NSAC Report, Vol. 2, No. 10
^
Letters dated August 3 and 23, 1984 from J.M. Felton to OCRE (FOIA-84-509 )
NUREG/CH-2549, " Background Study and Preliminary Plans for a Program on the Safety Margins of Containments" Blejwas et al.,
Sandia, May 1982 NUREG/CR-3234, "The Potential for Containment Leak Paths Through Electrical Penetration Assemblies Under Severe Accident Conditions", W. Sebrell, Sandia, July 1983 NUREG/CR-2475, " Hydrogen Combustion Characteristics Related to Reactor Accidents", Berlad et al, Brookhaven, July 1983 NUREG/CR-3131 Vol.1, " Containment Integrity Program FY82 Annual Report", Blejwas et al, Sandia, March 1983 Letters dated August 21 and 29 and November 2,1984 from J.M. Felton to OCHE, FOIA-84-577 NUREG/CP-0038, " Proceedings of the Second International Conference on the Impact of Hydrogen on Water Reactor Safety" NUREG/CR-2080, "A Review of Hydrogen Detection in Light Water heactor Contutnments", Neidel et al, Sandia, Feb. 1982 NUREG/CR-3463, "An Evaluation of HECTR Predictions of Hydrogen Transport", Wester and Camp, Sandia, Sept. 1983
" Discussion Report on Mark III Containment lnterface Issues", by John M. Humphrey, June 30, 1983 Documents made available by Applicants re Issue #8, identified in letters from Bradley Ferrell, CEI, to OCRE dated August 14, May 1, and August 8, 1984.
Oct. 17, 1983 letter from MP&L to NRC: second supplementary response to NRC 7-22-83 letter Oct. 25, 198351etter from MP&L to NRC: clarification of Aug. 23, 1983 letter Feb. 9, 1984 letter from MP&L to NRC. on containment negative pressure capability Feb. 9, 1984 letter from MP&L to NRC: Quarterly Status Report on hydrogen control program Feb. 13, 1984 letter from NRC to MP&L: RAI on negLtive pressure capability of containment A Feb. 22, 1984 MP&L letter to NRC responding to Feb.13 RAI April 4, 1984 letter from MP&L to NRC: response to Staff RAI's of Dec. 8, 1983
n.
-g_
Sept. 9,1982 letter from HCOG to NRC (HGN-006)
Oct. 26, 1982 letter from HCOG to NRC ( HGN-007)
May ll,1983 letter from HCOG to NRC (HGN-011)
Handouts and transcript of HCOG-NRC meeting of March 2, 1984 (see FOIA-84-170)
April 2,1984 letter from HCOG to NRC (HGN-016)
Handouts from HCOG-NRC meeting of August 28-29, 1984 June 7, 1984 letter from HCCG to NRC (HGN-017)
May 23, 1984 letter from MP&L to NRC, third quarterly status report on hydrogen control program
- Letter dated November 28, 1984 from J.M. Felton, NRC, to OCRE (F01A-84-797)
NUREG/CR-3742, " Buckling of Steel Containment Shells Under Time-Dependent Loading", Babcock et al, Los Alamos National Lab.,
May 1984 NUhEG/CR-3273, " Combustion of Hydrogen: Air Mixtures in the VGES Cylindrical Tank", Benedick, et al, Sandia National Lab.,
May 1984 Letter dated January 4, 1985, from J.M. Felton, NRC, to OCRE (FOIA-84-892) 1 NUREG/CP-0041, Proceedings of NRC Tenth Water Reactor Safety hesearch Information Meeting, January 1983.
4 9
6
i Issue #8, Interrogdtory N10 ;
(a) The documents related,to the use of igniters os a hydrogen ;
control system (including containment integrity and equipment ,
i survivability) are identified above.
1 (c) OCRE does not bel'i eve that igniters will safely control ,
hydrogen generate $ from o 75% metal-Water reaction (the ,
appropriate ' scenario
- now that'the new hydrogen rule is finalized) for the reosons delineoted below. The criterio for
" sorely' controlling hydrogen are those set forth in the new hydrogen rule.
The reasons for OCRE's belief that Applicants cannot meet the new hydrogen control rule include, but are not limited to, the <
following,
- 1. the manual initiation of the distributed igniter system does not suorontee its prompt actuation in those occident scenarios in which the operators fail to recognize that a degraded core occident is in progress or commit 0,ther errors resulting i'n the system's unavailchility, 2, powering the dittributed igniter system from diesel ,
generator-backed AC Esc buses will make the distributed igniter system unavailable during a station blockout occident; in addition, powering the hydrogen analyzers (which may be used with the distributed igniter system) from this some system would result in their unavailability for on extended period of time
- ,. . n, -- , - . . - c- , _ , - . ,
e following a station blockout occident, as they require a 5 hour-Worm-up period, (If the hydrogen concentration in the contoinment is unkncwn at i
' the time power is restored in o station blackout occident, the igniters could be octuated when the conditions favoring o detonation are present.)
- 3. the design of the i9 niter assembly, particularly the Spray a
shield, and the placement of these assemblies in ,I-beams and near ceilings makes it unlikely that they will ignite hydrogen below the downward propogotion limit of opprox. 9-10% (NUREG/CR-530 ot 195): An ACRS consultant (Gorry Schott) hos also suggested that the spray shield would och as o 'condie snuffer'.
Applicants' analyses assume that the igniters ignite hydrogen ok o concentration of'8 volume-%; calculated pressure and f temperature rises are greater at 10% than at 8%,
- 4. Applicants are using the CLASIX-3 computer code to predict temperatures and pressures from hydrogen deflagrations in the PNPP containment. CLASIX-3 is on inadequate and non-Conservative code, so much so that the NRC Stoff recently .
instructed Applicants not to use it. See the August 30 1984 letter from Youngblood to Edelman, Question 480.55.
The CLASIX-3 code predicts very low temperatures and Pressures resulting from hydrogen deflagrations. The state-or-the-ort HECTR code, developed by Sondio National Loboratories ,
e
-LA-specifico11y for the onolysis of Mark III hydrogen combustion, predicts pressures 2 to 3 times higher than does CLASIX-3.
NUREG/CR-2530 discusses some of the possible causes of CLASIX-3
~
non-conservatism (low flame speed,-treatment of vacuum breaker flow and reverse flow through the suppression pool, and spray heat transfer (CLASIX-3 predicts higher pressures with sprays than without)). .
- 5. The Sondia National Laboratories review of the Grand Gulf distributed igniter system (virtually identical to Perry's) concluded that the system was " marginally adequate", meaning that the uncertainty bonds ossociated with the calculated containment failure pressure overlap the uncertainty bonds ossociated with the c'olculations of burn pressures, but if midronge values are chosen for both factors, the calculated pressures will be below the failure pressure.
These uncertainties are susbstantial. The Sandio reviewers suggested
~
that other hydrogen control methods might be more desirable for.
Mark III plants. NUREG/CR-2530, pp. 9-11.
6.-Both HECTR and CLASIX-3 use hydrogen source terms calculated by the MARCH code. Earlier versions of MARCH do not properly model the BWR core. Applicones do not know which version of MARCH was used for the CLASIX-3 input. MARCH also models o full core meltdown occident, so the MARCH hydrogen source term was orbitrarily modified to yield a 75% metal-water reaction without
- breach of. the reactor vessel. Furthermore, only one MARCH run was used for all possible scenarios. (See Applicants' Answer to L
w
- /5-OCRE Interrogatory 13-65.)
There is no evidence that the version of MARCH used,.and the modifications thereto, either are conservative or occurately describe the behavior.of the PNPP reactor system during and i
ofter o degraded core occident. The limitorions and errors in-the MARCH code are discussed in~NUREG/CR-2285; fond include (but
~
are not limited to) improper decoy heat model, time step sensitivity, numerical instabilities'cousing oscillations in Zr oxidation rate, and sensitivity to user-chosen values of input variables. .
It was concluded that the review of MARCH did not support the sometimes-voiced opinion that MARCH is conservatives indeed, MARCH con be non-conservative with respect to small LOCA sequences. NUREG/CR-2285, Pp. 1-2, A4.1-14.
Several MARCH input variables con, by'the value cnosen for them by the user, drastically ofrect hydrogen production from the Zr-Water reaction. These include FDCR (fraction of Zr to be oxidized before RPV failure), IHWA (determines whether molten Zr 1
con react with. steam). DPAR'T (particle diameter, determines s'urface area of debris during core slumping), and FZMCR (distribution of Ze throughout the Particle). The values or these variables must' be chosen consistent with experimenhol dato on core behovior during decroded cor.e occident'. s There is no .
evidence that this hos been done. _
- 7. Applicants' onolysis of the ultimate copocity of the PNPP contoinment vessel is deficient, os it has not been demonstr,oted*
.x
{_
-/$ -
that the analytical techniques used occurately or conservatively represent actual containment behavior under the loads imposed by hydrogen. combustion.
(Applicants' onalysis is contained in a document entitled " Perry Nuclear Power Plant Units 1 and 2, Ultimate Structural capacity
'of Nork.III containments.')
~
Significant uncertainties exist in the analysis of ultimate contoinment capacity. The computer codes available for onolyzing nonlinear inelastic behavior of steel structures have not been qualified by comparison to test-to-foilure of containment-like structures. No ultimate failhre mechanism for bioxiol and trioxial stress is universolly accepted. NUREG/CR-2549, p. 9. Desi9n codes such as ASME are not oppropriate for determining contoinment behavior beyond design conditions.
NUREG/CR-3131, Vol. 1, p. 9. (For these reasons experiments are being conducted on scale models of containments at Sandia National Laborotories.)
Since the function of the containment is to serve os o leak tight botrier, the oppropriate standard for containment failure is one of containment leakoge. Applicants have.provided no v
correlation of their calculated ultimate containment capacity with the degree of' containment leakoge. .
Applicants have used stress 05 the failure criterion in their
.)
onalysis of the PNPP containment. When considering containment leokoge, displacement and deformation are more important factors. Non-uniform' deformation of the vessel wall and on ,
e
f[-
equipment or personnel hatch could lead to leakage of gaskets or seals. Differential motion between penetrations ("hard points' or fixed points) could cause 19akage. While these effects have not been analy ed or quantified by Applicants, they do admit <-
'that *1orge deflections" will occur in the containment vessel before reaching the calculated ultimate pressures are reached.
Although Applicants did consider the ability of some ,
penetrations to withstand pressures greater than design, they were apparently analy:ed sep'orately, oport from the analysis of the vessel. The effect of penetrations on the ability of the vessel to withstond hydrogen combustion loads must be considered. The presence of pen'etrations (even if reinforced in accordance with ASME rules) will lower the buckling resistance of a shell. NUREG/CR-3135. Buckling is considered a controlling vessel failure mode for PNPP.
Applicants' onolysis is also deficient because it has not considered the simultaneous effects of hydrogen combustion loads and loads from SRV discharge or from steam discharge through the dryWell LOCA Vents on the integrity of the PNPP containment, as is required by General Design Criterion 4.
- 8. Appliconts have foiled to consider the effects of defects in the fabrication, construction, and repair of the containment vessel and associated structures and components on containment integrity, Newport News Industrial Corp. was responsible for the fabrication and erection of the containment vessel. This ,
s.
-/f- .
company has o history of quality assurance violations. It is known that welds with rejectable indications were left unrepaired in the PNPP containments: their use was justified by a fracture mechanics analysis which did not consider the loads resulting from hydrogen combustion.
These defective welds were accepted by Newport News and Applicants the first time, and were only rejected by a re-review of weld radiographs.
Applicants recently filed a report pursuant to 10 CFR 50.55(e) regarding drywell equipment hatch covers, fabricated and supplied by Newport News, which had not received proper inspections.
The fracture mechanics analysis report on the rejectable welds contains a curious reference to earlier work done by the onolyst (APTECH Engineering Services) for Applicants concerning a fracture mechanics analysis of containment stiffener flange welds at Perry. The onl,y reason such on analysis would be performed is if there is some question about the quality of these welds.
It is quite probable that additional, undiscovered (or undisclosed) defects exist in the Perry containments which would compromise their strength. Any analysis of containment ultimate copocity must address the as-built condition of the containment (including weld defects, material defects or domoge, and geometric imperfections or asymmetries) to be reliable.
- 9. Applicants have failed to consider containment failure ,
4 I
m.
-/f~
, e through leakage through penetrations using temperature-sensitive elastomeric materials os seals, seats, gaskets, or seolonts.
Such penetrations include electrical penetration ossemblies, equipment / personnel hatches, and purge-vent butterfly volves.
'It has been estimated that leakoge through these structures could be o predominate containment failure mode. 'NUREG/CR-2182,-
Vol. 1, p. 149; drafts of NUREG-1037.
- 10. Applicants have foiled to consider steam bypass of the suppression pool, resulting from hydrogen burn induced domoge to the dryWell Wall, os o mechanism for containment failure. The containment would fail from steam overpressure if the pool is bypossed. Domoge to the dryWell Wall could result from local detonations or from domoge to penetrations (especially those penetrations utilizing elastomeric seols or gaskets, e.g.,
equipment hatches) from the high temperatures of hydrogen combustion.
Applicants have not onoly ed the ability of the drywell Wall or its Penetrations (or components such as vacuum breakers or the dryWell purge system) to successfully Withstond the pressures and temperatures associated With hydrogen combustion, including locol detonations.
It has been estimated that on opening in the drywell wall as small as 4 inches in diameter would eliminate steam flow through the suppression pool. See FOIA-83-460, Letter Report, July 27, 1983, by Brookenhoven National Laboratory, re review of GESSAR-
'n PRA, containment Failure Modes and Fission Product Release, p."
-/ d -
t 3-5. -
- 11. Applicants have foiled to consider domoge to structures, systems, and components within the dryWell (including the RPV and recirculation pumps and piping) from " reverse' pool swell loods (and thermal shock) caused by pressur.i otion of the
- Wetwell from hydrogen combustion. Such loods could domoge the reactor vessel and associated piping, perhaps leading to o full core melt accidents the containment will then foil due to pressurization from non-condensible gases generated from corium-Co nc r e t e interaction. ,
Officiols of Mississippi Power & Light, the Grand Gulf' licensee, have admitted that " violent overflow of the suppression pool into the drywell' con result.from hydrogen combustion.
NUREG/CP-0038, p. 291. There is no evidence that the effects of these loads have ever been onoly ed by Applicants.
II. In some of the more benign scenarios, the pressures and temperatures resulting from hydrogen combustion may not be sufficient to foil the containment directly. However, the resulting pressures will be high enough to keep both troi'ns of containment spray .operatin9. The PNPP containment spray is one of several functions of the Residual Heat Removal System. Other RHR modes include coolont injection (LPCI, port of the ECCS),
shutdown cooling, steam condensing, and suppression pool cooling. ,
Containment spray is automatically initiated on high containment pressure (9 psig) and takes precedence over,other RHR functions.'
e
-g y .
-There are no procedures available for turning the containment spray off. If both trains of contoinment spray are operating, suppression pool cooling is lost. The pool will overheot, and couse loss of those reactor coolant makeup systems taking suction from the pool. This would cause o full core meltdown occident with resultant containment f o i,1 u r e .
13..The point above assumes that containment sprays will.be avoilable. Applicants, in their analyses of hydrogen -
combustion, assume sproys will be operating and will serve'os a moJor heat sink, thereby lessening the pressures and temperatures of hydrogen combustion. Because of the mutual dependence of sproys and LPCI on the availability of the RHR system, it is unreasonable to assume that sprays are available in a degraded core occident involving o 75% metal-water reaction, which only occurred due to the loss of coolant makeup systems, including LPCI.
Thus, Applicants should not model or take credit for containment sprcys in any of their analyses or experiments justifying their hydrogen control system, except where the sProys result in non-onservatism, as in item 12, above.
(Another non-conservative effect of containment sprays is that of increasing turbulence, thereby resulting in complete combustion and increased flame speed, cousing higher pressures and temperatures.)
- 14. Applicants have not demonstrated that detonations connot occur in the drywell. In o drywell break occident, hydrogen,and' 4.
6 .
so -
steam will be released from the primary system directly into the
~
drywell. The drywell air mass will be expelled to the wetwell
'through the vents'. The drywell atomsphere will-then consist Lonly of hydrogen and steam and is non-flammable, However,
. oxygen will'be introduced into the drywell by the vacuum .
'. breakers ond the purge compressers. Radiolytic.decompos,ition of ..
s' team and Water in the vessel and drywell will also yiel'd ~'
f oxygen, '
Applicants assume that the oxygen entering the drywell from the purge compressers and vocuum breakers will cause " inverted
' diffusion flames *, which supposedly result in benign, pressure and temperature excursions.
There is no evidence that such " inverted diffusion flames" will occur,;porticularly in o steom-rich otmosphere, (Applicants neglect the effects of radiolysis.) A more likely outcome is that combustion Will be suppressed by the high steam concentration, and the' oxygen will gradually mix unirormly with the hydrogen and steam. At some point in the occident the steam will tutdenly condense, either from outflow of subcooled water from the break after core recovery, or pool, overflow from hydrogen burn pressure in the wetwell or from the normal action of-the suppression pool makeup system, automatically actuated 30 minutes ofter o LOCA si9nol, After steam condensation, a detonable mixture will exist in the drywell, which would be ignited by the glow plug igniters.
Applicants have not onalyzed the ability of-the drywell or any
e
~LI.
structure within (including the RPV) to withstand such o
-dekonotion.
- 15. Applicants have not. demonstrated that detonations, quasi-detonations, and detonceion-to-derlogroeion eronsition ('00T')
would not occur in the PNPP contoinments. These types or hydrogen combustion
- yield pressures in excess or those from normal derlogrations, and may result in impulsive (dynamic) loading.
DDT and quasi-detonations con result from flame front accelerot' ion due to obstacles. De,tonations con be'directly initiated by a strong gas jet. Although the 'classicol' detonation limits for hydrogen were long thought to 18 and 56%,
recent research has greatly. expanded these limits', especially on the low end.
Detonations have been initiated in hydrogen concentrations as low as 13.8%. Flame speeds as high as 150 m/see in 10% hydrogen has been observed. Researchers' now believe that there is no limiting hydrogen concentration below whic.h detonations connot occur. Instead, the hydrogen concentration needed ror o detonation is related to the geometric scale or the enclosure.
9 The larger the scale, the lower the hydrogen concentration needed. NUREG/CR-2726, P. 2-51s NUREG/CP-0038, P. 9618 NUREG/CR-2530, P. 194.
The configuration or the Mark III containment is likely to cause flame C celerotion. NUREG/CR-2530, P. 194. Local detonations in s the Wetwell could occur from temporary inerting due to steam,
e-
- 4,4. -
concentrations (caused by drywell bypass leakoge and spray ineffectiveness below the operating floor) or oxygen depletion.
I NUREG/CR-2530, p. 110. Applicants have failed to consider the effects of detonations (local or global). DDT, or accelerated t
flames on the integrity of the containment or the components within.
- 16. Detonations are made more likely 'ond the effects of deflagrations more severe, by the action of ioni ing radiction, which will be present in the drywell and containment during o degraded core occident. None of the research programs proposed 1
or completed either by Appliconts and their oget.ts or by the NRC and its contractors has addressed this effect. This factor must be considered, as any sources of free radicals or chain corriers will ofrect hydrogen combustion phenomeno. NUREG/CR-2475, p. 4.
Until the. effects of ioniting radiation have been determined, the 'OCCepted" values for proPogation limits, ficme speeds, i
detonation " limits', and temperature / pressure risei are suspect.
- 17. There is no reasonable assurance that equipment within'the containment and drywell needed to maintain containment integrity and to maintain the core in a safe condition will survive the ,
effects of hydrogen combustion. Small-scale tests have indicated thermally-induced failures:
Fettwol tests - melting of solder (even within protective enclosures) and teflon wire insulations Sandia VGES tests -
melting of plastic fon blades ofter o single burn (NUREG/CR-3 273, p. 11: Sandia' FITS tests - severely chorred cable ,
I
-cs- ,
insulation. NUREG/CP-0038, p. 1134. Thermal effects are expected to be worse in large-scale structures (e.g.,
containments).
Large-scale tests conducted by EPRI ot the Nevado Test Site have also resulted in equipment failures. Electrical cable Jockets experienced charring, cracking, floking, culging, and/or s wrinkling. Operational anomalies were observed in limit switches and pressure transmitters. EPRI Quick Look Report dated Feb. 17, 1984.
- 18. Experiments conducted in a 1/20 scale'a model' of the Mark III containment h' ave revealed that continuous diffusion flames will exist in the Wetwell when the hy.drogen release P (e t e exceeds
- 0.4 lbm/sec, full scale. While diffusion flames do not pose on overpressure threat to the containment, they do create a severe thermal environment. The Hydrogen control Owners Group, of ,
which Applicant is o member, has admitted that dato from the 1/20 scale tests, extrapolated to full scale, indicate that equipment survival is in doubt for diffusion flames resulting rrom o 75% metal-water reaction.
/
In fact, it is doubtful that the Mark III con tolerate diffusion 4
flames persisting for more than 15-20 minutes. (FOIA-83-469)
/
An estimate of the thermal environment resulting from 5,tonding diffusion flames above the suppression Pool in a Mar.k III containment performed by Sondia indicates that the Potential exists for generating heat fluxes and temperatures large enough
! to domoge equipment in the wetwell. (FOIA-84-509) ,.
o.
_ _- . - - ,--~_m ,
~24 ~
4
- 19. In order to reduce these, uncertainties, HCOG'is conducting experiments in a 1/4 scale facility. The purpose of these tests is to determine the thermal environment to which equipment in the containment would be exposed. However, no actual equipment is to be used in the facility. Instead, transient heat fluxes V-are or calculoted from measurements ond extrapolofed tojmeasured to full scale. The calculated thermal environment,would the'n be used in o computer-onalysis of equipment response.
HCOG's own consultants, COMBEX, Inc., recommended full scale testing of diffusion flome phenomeno specifically to ovoid the difficulties of scolin9. 1 letter from Mississippi See 3-2 82 Power & Light to NRC.
In addition, the 1/4 Scale test facility is to serve os a model for all 4 of the remaining Mark III plants, Grand Gulf, Perry, Clinton, and River Bend, even though these contoinments are not identical. River Bend hos fan coolers, which the other plants lock. The other plants are of different si es, have different
, spray flow rates, and are constructed of different materials. A
~
facility trying to model 4 different' plants will model none of them correctly.
The difficulties and uncertainties introduced by using a scaled facility (measurements, calculations, and extrooplations of heat flux: effect of hydrogen release points (vents or SRV): and effect of Pool behavior on combustion) could all be elimi,noted j by performing testing at full scale using actuoi items of ~
equipment to determine their survivability.
- 20. In addition to the uncertainties mentioned above, the 1/4
e 4
-26 scale facility will not oppropriately model Mark III combustion phenomeno. Deflagrations will not be properly modeled because-the containment surface creo to volume ratio is not conserved.
The exiscence of diffusion flames connot be confirmed because drywell leakage is not modelled. .
The existence of diffusion flames is predicated on the 1
assumption that all the hydrogen released to the containment is released to the suppression pools the hydrogeD will bubble up through the pool and be released at its surface. As the hydrogen diffuses into the wetwell cir, it is ignited by the -
glow plug igniters above the pool as a diffusion flame.
If some or all of the hydrogen is released to the wetwell not through the pool, but rather by leokoge through the drywell wall, the above assumption is no longer valid. Diffusion flames
+
may not exist at 011: instead, hydrogen. combustion will be by deflagration, which is not properly modelled in the 1/4 scale L- .
facility. Or, diffusion flames could occur at the point of hydrogen rele'ose from the drywell wall, if the lenkoge is concentrated at a point. The thermal environment here might be more severe than that predicted by the 1/4 scale tests.
It is even possible that hydrogen will occumulate in oreos near the drywell wall and later detonate. The octual behavior cannot be predicted without valid experimental dato, y Drywell leakage must be considered as it is excessive in the i Mark III, A former General Electric contoinment engineer has claimed that the leakoge is 50 great that the action of the
i
(
-2C .
dryWell purge system would drive the hydrogen in the dryWell out through the Wall and not through the pool. See June 30, 1983 Discussion Report on Mark III containment Interface Issues by John H. Humphrey, p. 3.4-5. This analysis assumed a tech spec alloWohle of 0.14 sq. ft. Perry's is 0.168 54 ft, and is likely to be larger due to Applicants' practice of drilling' holes in the steel drywell liner for anchor bolt instal 10 tion.
Drywell lookoge has never been properly analyzed by Applicants or by General Electric. PNPP drywell lookoge should be measured ond included in a full scale test facility.
- 21. The hydrogen release rates to be modelled in the 1/4 scale racility ore to be calculated by the 'BWR Heat Up Code'. This is a proprietary computer model which has undergone limited review by outside parties. This review, however, has revealed non-conservatisms in the code.
The non-conservatisms include artificio11y low core upper plenum temperoture, o decoy heat model that underestimates decoy power, uncertain and incomplete ; ore spray and blockage models, and incorrect intropellet heat conduction models.
However, the most severe effect on hydrogen production is produced by the input variable T0X0FF. Above the core bundle temperature specified by T0X0FF, hydrogen production in the code Will be irreversibly terminated. This behavior is contrary to ,g a
that observed iq' Power Burst Focility and Sandia ACRR -
experiments, Which showed that hydrogen production continues (and is enhanced) through fuel and clod melting and loss of core
r
'o
-J7- .
geometry.
(F0IA-84-797. re Oct. 3-4. 1984 HCOG-NRC Meeting) ,
Thus, the BWR Heat Up code must not be used until these errors (and others, which will only be uncovered by a complete review) are corrected.
20, Analytical methods for determining hydrogen release rates and qualtities, hydrogen' mixing and combustion, and containment and equipment response are all severely limited by the substantial uncertainties ossociated with these phenomeno, o
Smoti-scale combustion experiments, for example, are not even opplicable to hydrogen combustion in a containment, os research hos now demonstrated the dependence or flammobility and detonability limits on ignition type and strength, vessel geometry and si:e, the nature and geometry or obstacles present (and their heat transferproperties), and gas velocity.
NUREG/CP-0038. p. 50. (We would odd the presence or ionizing
. radiation to this list.)
(An example or'the dramatic errects or scale on hydrogen combustion is the results or Sondio experiments on. hydrogen a burning in aqueous rooms. These rooms were errective in limiting pressure rise in smo11 scale tests, but at larger scale there is evidence that rooms con cause accelerotion to detonotson. NUREG/CP-0041, Vol. 5, p.
363.) sv These uncertainties will be reduced to on acceptable level only by full-scale testing in a rocility exhibiting real Hork III behavior and geometry (including equipment within the .
m.. _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ ._ _ _ _ . . _ _ _ _ _ _ . - _ _ _ _ _ _ _ _ _ _ _ . . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
r _.
e 45-containment Which may accelerate flames), using a full spectrum-of accident scenarios and resultant hydrogen and steam release rates that occurately reflect BWR core behavior in a deproded core occident.
Otherwise, reasonable assurance that PNPP con Withstand the burning of hydrogen produced from o 75% metal-Water reaction without additional core domoge or loss of containment integr.ity simply does not exist. .
Respectfully submitted.
We .
Susan L. Hiatt OCRE Representative 8275 Hunson Rd.
Mentor, OH 44040 (216) 255-3150 60 Y
4 4
I, Susan L. Hiatt, duly sworn depose.and day that:
- 1. I am responsible for the' answers in the attached OCRE Updated Response to Applicants' Second Set of Interrogatories to OCPE;-and
- 2. those answers are true and correct to the best of my knowledge and belief.
W Susan L. Hiatt Sworn to and subscribed before me this If day of January, 1985.
~
4
... . .P. .u
. ; . . '.. \
W:
' s
. . <n i. mv U!':~y};~'24.sy f* ,
Notary Public MARIE S. CASSER. Notary Publ6c STATE OF OHIO.(Lake Co y)j
- ) N2 My Commission Expires _,/8 /N
[**fp W7 0
8 Y
. :' , il.. a O
7 -
-7
-. , e. . ,
, ,c . ,
CERTIFICATE OF SERVICE ,
i ,
! it This is to certi~fy that copies of the forego'ing were served ;by deposit in the U.S. Mail, first class, posta'ge prepaid,.this' -
93d day of Om , 1985,to tho'se on the - -
service list b,elow. g 7
- 4
- i. . e. ,
I . '
r.. :'
,g. ,
d .
Susan L. Hiatt
. . g, g .,
>7*
,. . f , .: 'y SERVICE. LIST . gl.('; 3, ' ,
. .....,.a.,- ,,
..i& A'i,. ,. . '
f_ .. _ ,
.i' - .
. . s, ~
- JRMES P. GLERSON, CHRIRMAN
'. Terry Lodge, Esq. ,
i RTOMIC 5AFETY & LICEN5ING BORRD 618 N. Michigan St.
"513 GILHOURE DR. Suite 105
. . 5ILUER SPRING, ND 20901 .,
, Toledo, OH 43624
.., f
. . . .-~ 4 ; . .d -
'Dr. Jerry"R. Kline -
Atomic Safety.&.Licen. sing Board.
. i. ,,
U.S. Nuclear. Regulatory Commission b '!
20555 - ,- 7 .0. :.; ,
WasEington i','D., C . ,
'; j,Ql
' 7 ] [f,f ' .y ,
, .
- c..t Mr..Glenn O. Bright jf .
i, -
. :., Af Ih ' ,' ~
Atomic ~ Safety &. Licensing Board /
U.S. Nuclear Regul'atory Commission E ,
- p . ,, y / Z [ -
e'.,.*..*,
. Washington, D.C. 20555
.' 4 -
- r. Colleen P. Woodhead, Esq. . .
- s
- Office"of'the" Executive Legal Director '.
l 7 ,,
b' U.S. Huclear. Regulatory Commission ,,
Washington,.D.C. 20555 ,
Jay.SilberJ, E'sq. [. ' *
. Shaw, Pittman; Potts, & Trowbridge p;
- 1 1800 M Street, NW i '
y
. Washington, D.C. 20036 - .
p ,
Docketing #& Service Branch f .
.0ffi'ce of'the Secretary , ,
'. #.g U.S.. Nuclear RegulatorysCommission .
cM Washington, D.C. 20555 ..
. . '. s.- .
t Atomic' Safety.&, Licensing, Appeal. Board' Panel y,. ,- ,,
U.S. Nuclear. Regulatory' Commission
- . ., . . ,'f, '.f , '
ic. ,'/
. Washingtoni D.C. 20555 -
,