ML20197B599
| ML20197B599 | |
| Person / Time | |
|---|---|
| Site: | Maine Yankee |
| Issue date: | 12/19/1997 |
| From: | Collins S NRC (Affiliation Not Assigned) |
| To: | |
| Shared Package | |
| ML20197B592 | List: |
| References | |
| EA-97-387, NUDOCS 9712240009 | |
| Download: ML20197B599 (18) | |
Text
_ _
UNITED STATES NUCLEAR REGULATORY COMMISSION in the Metter of
)
)
Yankee Atomic Electric Company
)
EA 97 387 and Duke Engineering & Services Co.
DEMAND FOR INFORMATION l
Maine Yankee Atomic Power Company (Lloensee or MYAPCo or Maine Yankee)is the holder of Facility Operating License No. DPR 36, issued by the Nuclear Regulatory Commissio (NRC or Commission) pursuant to 10 CFR Part 50 on September 15,1972. The licenae authorizes the operation of Maine Yankte Atomic Power Station (facility or MYAPS)in accordance with conditions specified therein. The facility is located on the Licensee's she in Lincoln County, Maine.
Yankee Atomic Electric Company (YAEC) provides certain basic components and services to the Licensee. The Y/.EC LOCA Group provided Loss of Coolant Accident (LOCA analyses of Emergency Core Cooling System (ECCS) perfortnance, for use in safety and licensing analyses. It is the NRC's understanding that subsequent to the events described in j
this DGmsnd for information, Duke Engineering & Services (DESS) purchased the YAEC LOCA
- Group, i
11 l
As a result of concerns regarding Small Break Loss of Coolant Accident (SBLOCA) analyses raised by the 197G accident at Three Mlle Island Unit 2, and pursuant to 10 C.F.R. 9712240009 971219 i
PDR ADOCK 05000309 O
.PDR
~
2 5 60.54(f), the NRC required licensees to submit revised, documented Smal M A Loss of-Coolant Accident (SBLOCA) analyses which were to meet the gui of TMI Action Plan Requirements"(NUREG 0737 or TMIl ll.K.3.31. In response to the guidance of item ll.K.3.30, on January 14,1983, submitted a report, YAEC 1300P RELAP5YA: A Computer Program for Light System Thermal-Hydraulk: Analysis"(RELAP5YA) to the NRC. In January 1989, approved RELAP5YA for use by Maine Yankee as a 10 C.F.R. Par 150, Appendix K model, acceptable to demonstrate compliance with the requirements of 10 C.F.R RELAP5YA is a generic, non plant specific LOCA computer code for calculating performance over the Smal! Break portion of the break sp6ctrum.
hem ll.K.3.31 of the TMl Action Plan provides that licensees should submit calculations using the SBLOCA evaluation model approved by the NRC pursuant to II.K.3 30. Although the NRC approved RELAP5YA as an Appendix K model, two pis.nt specific RELAP5YA cvaluation models for Maine Yankee in response to Plan item II.K 3.31: a "Best Estimate" model and an Appendix K model.'
YAEC developed its *Sest Estimate"(BE) RELAP5YA plant specific evaluation June 1990 in order to satisfy a verbal commitment made to the NRC by Maine Y that the item II.K.3.31 plant specific calculations would be completed by June 199 1
5 An Appendix K modelis a LOCA evaluation model which complies with th set forth in Appendix K to 10 C.F.R. I $0.46. A best estimate modelis an evaluation e requirements which provides realistic calculations of LOCA phenomena and uses nomina parameters. The "Best Estimate" SBLOCA model developed by YAEC and referred to combined the two approaches, such that the limiting SBLOCA case was identifed w best estimate calculations using nominal values of input parameters and th temperature of the limiting case was calcula'ed wit K approach.
3 1
because of dlfficulties in developing an Appendix K evaluation model. YAEC prepared a report of the BE RELAP5YA SBLOCA evaluation model analysis in Au0ust 1990. Maine Yankee relied upon that "Best Estimate" RELAP5YA SBLOCA analysis to evaluate the effects of a reduction in steam generator (SG) pressure, as part of a January 1993 wrttien safety evaluation required by 10 C.F.R. $ 50.59, in order to determine whether a reduction in SG pressure involved an i
unreviewed safety question.
)
YAEC completed its development of an Appendix K RELAP5YA SBLOCA evaluation model, described in its June 1993 report, YAEC 1868: " Maine Yankee Small Break LOCA Analysis"(YAEC 1868). YAEC 1868 states that the Appendix K, RELAP5YA SBLOCA evaluation model was " performed in conformance with the requirements of 10 CFR 50.46 and post TMI Action items ll.K.3.30 and ll.K.3.31....The 10 CFR 50.46 acceptance criteria were met, thus closing out item II.K.3.31 for Malne Yankee." The Appendix K analysis described in YAEC 1868 was used to prepare Core Performance Analysis Reports (CPARs), which were submitted to the NRC as part of Maine Yankee's reload analyses for Cycle 14 and Cycle 15 operations. The RELAP5YA analysis described in YAEC 1868 was the SBLOCA analysis of record throughout Cycle 14 operations, which occurred between October 14,1993, and January 25,1995, but not during Cycle ir 3perations because of the intervening January 3,1996,
" Confirmatory Order Suspending *.uthority for and Limitirg Power Operation and Containment Pressure (Effective immediately), and Demand for information"(Ordorf 61 Fed, Reg. 735 (Janusy 10,1996).
8 Among other things, the Order lim}ted operation of MYAPS to 2440 MWt. pending NRC review and approval of a plant specific SBLOCA which conforms to lMI Action Plan items ll.K.3.30 and li.K.3.31 and which meets the requirements of 10 C.F.R. $ 30.46.
4 The LOCA Group at YAEC prepared the RELAP5YA SBLOCA code and evaluation models for Maine Yankee. The Manager of the LOCA Group ( hereinafter,'the Mansper*
Arst employed in the LOCA Group as an engineer in 1980 and has been the Manager o LOCA Group in the YAEC Nucisar Engineering Department ainos February 1989 participated in the development of the RELAP5YA code. The Manager supervised the i
i development of both the BE and Appendix K plant specific RELAP5YA evaluation models for i
Maine Yankee, and participated in the preparation of YAEC 1868. The Manager supervise LOCA work, wnich includes preparation of CPARs, LOCA codes for other NRC licensee including Vermont Yankee and Yankee Rowe, and safety analyses. The Lead Engineer in LOCA Group (hereinafter 1he Lead Engineer") has been an engineer in the LOCA grou 1978 and has been the lead engineer for MYAPS LOCA analyses since mid 1990. The Lead Engineer performed the quality assurance review of both the 1990 BE RELAP5YA SBLOCA analysis and the 1993 Appendix K RELAP5YA SBLOCA analysis described in YAEC 1861 On December 4,1995, the NRC received allegations that, among other things, YAEC, acting as agent for the Licensee, knowingly performed inadequate analysss of the Em Core Cooling System (ECCS) to support two license amendment applications to increase the rated thermst power at which MYAPS operates, Arst to 2630 MWI, and then to 2700 MWL lt wa further alleged that YAEC management knew that the ECCS for Maine Yankee, if evaluate accordance wth 10 C.F.R. $ 50.46, using the RELAP5YA SBLOCA evaluation model, did not meet licensing requirements.
In response to the allegations, NRC dispatched an Assessment Team to YAEC
. headquariers between December 11 and 14,1995, to examine, among other things analyses, especially the SBLOCA analysis which supported the Licensee's operating c (Cycle 15) retc,ad analyses. Based on the Assessment Team's review, and a meetin
- u.. --
5 the Licensee on December 18,1995, the NRC staffissued its January 3,1996. Order. T concluded in the Order, hier alia, that the Lloonsee had not demonstrated that compute RELAP5YA would reliably calculate the peak cladding temperature for all break aires in smalLbreak LOCA spectrum for Maine Yankee and that, for a variety of reasons, the plant speelfic application of RELAPSYA did not conform to the requirements of 10 C.F.R.
I 50.46 rnd thus was not acceptable for use by ti,e Lloonsee. RELAP5YA had be Maine Yankee for Cycle 14 SPLOCA analyses, and had been proposed for use by M Yankee for Cycle 15 SBLOCA analyses, to demonstrate compliance with 10 C.F.R.
requirements for calculating ECCS performance. The Order required the Licensee to subm SBLOCA analysis specific to Maine Yankee for operation at power levels up to 2700 M must meet the requirements of 10 C.F.R. I 50.46, " Acceptance criteria for emergency co I
cooling systems for light water nuclear power reactors," and which must conform to th of NUREG 0737, " Clarification of TMI Action Plan Requirements", items ll.K.3.30 a
'SBLOCA Methods" and " Plant specific Analysis," respectively, and item II.K.3.5, " Autom Trip of Reactor Coolant Pumps During LOCA* The Order suspended authority to ope Maine Yankee at 2700 MWI maximum power and limited power to 2440 MWt, pendi review and approval of the required SBLOCA analysis.
The NRC also in}tiated an investigation by the NRC Office of investigations (01) examine possibh wrongdoing. Of issued its Report No.195-050 on Septemier 6,1996.
Based on the Assessment Team review and the 01 investigation, tne NRC atsff concluded that, as a result of Maine Yankee's use of the Appendix K and BE RELAP5YA evaluation models, apparent violations of NRC requirements occurred as set forth in l
l m
6 oorrespondence to MYAPCo' (hereinafter
- December 19,1997 letter to MYAPCo')
concurrentty with this Demand for information. Preparation of LOCA analyses by th LOCA Group caused Mabe Yankee to apparently violate NRC requirement below.
lll e..
A.
Durina Cvele 14 onorations and in the Cvele 14 and Cycle 15 reload analvaan Maine Yankee used anoarently unaccaotable avajgglisi modela which cou calculate or reliabfv calculate ECCS nerformanr As set forth in the December 19 letter to MYAPCo, Enclosure 4 Apparent Vio and D, Maine Yankee used unacceptable evaluation models to calculate ECCS during Cycle 14 operations, and in the Cycle 14 and Cycle 15 CPARs used to analyses, in apparent violation of 10 C.F.R. $ 50.46(a)(1), because there was a 1
smali break spectrum between break sizes of 0.35 ff and at least 0.6 ft' for wh evaluation model could either calculate or reliably casculate ECCS performance.
To calculate core cooling performance, MYAPCo used the Appendix K RELAP SBLOCA evaluation model described in YAEC 1868 and the Lar9e Br Loss o: Coolant Accident (LBLOCA) analysis described in YAEC-1160, " Appl WREM BASED Generic PWR ECCS Evaluation Model to YAEC 1868 and YAEC 1160 were prepared by the YAEC LOCA Group for MY CPARs for Cycle 14 and Cycle 15 were also prepared for MYAPCo by YAEC, participation of the YAEC LOCA Group.
- See letter dated December 19,1997, from H. Miller, Regional Administra President, MYAPCo.
l l
l 7
i i
in order for LOCA codes to be acceptable, they must not only be capable of calc any point on the brank spectrum, but must be capable of producing reliable calculations.
i RELAP5YA was not capable of calculating at break skes of and greater than 0.35 ft, While th WREM LDLOCA code was capable of calculating ECCS performance at break slees of 0.6ft' and above*, those calculations would have been unreliable et break slees of 0.6 # and some
{
4 portion of the break spectrum above 0.6 W. Because LDLOCA codes conrot be assumed to produce reliable or acceptable results in the Small Break region, in order for the WREM LBLOCA code to be acceptable for calculating cooling performance in the Smol Break re and greater than 0.6 ft', M would have been necessary to vertfy WREM LBLOCA results in that region of the break spectrum against test data in order to estimate those uncertainties. The WREM LBLOCA evaluation model, however, was not used to calculate ECCS performance in the Small Break region of and greater than 0.6# for Cycle 14 or Cycle 15, and likewise was not verifed against applicable test date in that region of the break spectrum? Because MYAPCo's ECCS analyses, singly or combined, were not capable of acceptably calculating any point on break spectrum, M was not possible to analyze any point on the break spectrum and thus to confirm that the limiting break had been identifed.
- The WREM LBLOCA code analyzed break sizes of and greater than 3.6 R in the Cyc and Cycle 15 reload analyses. At the request of NRC staffin early 1996, Maine Yankee analyzed break stres further down the spectrum with the WREM LBLOCA code, and ultim i
did so down to.6ft'. Since no changes were made to the WREM LBLOCA code in so doing, I
apparent that the WREM LBLOCA code was capable of performing calculations in the Cycle 14 and Cycle 15 reload analyses for break sizes down to 0.6 ff.
- Neither the 1996 performance of the WREM LBLOCA code down to break stros of 0.6 ft nor any vertfication against test data in 1996, even if it had been done, would have made the WREM LBLOCA code acceptable for use in calculating break stres down to 0.6 ff during 14 operations or the Cycle 14 and Cycle 15 reload analyses. Section 50.46(a)(1) requires an ECCS code be acceptable before being used, and does not permit the substitution of engineering judgement for the ability to analyze any point on the break spectrum to confirm that the limhing breaks were identifed.
.-.-._--.,,.-.m.m.
,-.--.,_-..-.--.,---..---m
- _ -m,_.--
..r
.--.,---,...,or m
i a
The Manager of the LOCA Group had the primary responsibility at YAECl that LOCA codes developed by YAEC complied with 10 C.F.R. $ 50.46 and 10 C
{
Appendix K requirements, and the responsibility for bringing compliance problems to th l
sttention of YAEC management. It is necessary for engineers in the LOCA Group to b with Section 50.46 requirements 5 order to adequately perform their duties. The told 01 investigators that she understood that herjob description required her to be 10 C.F.R. l 50.46 and Appendix K, and that she was expected to report compliance p the Manager.
The Manager and the Lead Engineer were identified by the YAEC 1868 report as
" major contributors"to the RELAP5YA analysis described in YAEC 1868. The Manag participated in the development of the RELAP5YA code, supervised the development of the RELAP5YA analysis described in YAEC 1868, and was one of the preparers of YAEC-1 The Manager of the LOCA Group supervised all LOCA work, which includes the L8LOC analyses for Maine Yankee. The Lead Engineer was employed in the LOCA Group as a engineer in 1978, has been the principal engineer for Maine Yankee LOCA analyses s niid 1 ppd, and performed the quality assurance review of the RELAP5YA S8LOC described in YAEC 1868. As a resuh of their duties, the Manager and the Lead Eng also familiar with the WREM L8LOCA evaluation model and analyses developed by YAEC for use by Maine Yankee. Additionally, the CPARs for Cycle 14 and Cyci prepared with the participation of the YAEC LOCA Group, including the Lead Enginee approved by the Manager, among others.
The apparent failure of the RELAP5YA S8LOCA analysis described in YAEC 1868 to comply with Section 50.46(a)(1) is of heightened concem because the oscillations a in the analysis became more severe at larger break sizes, increasing the risk that r
v-
-.---,.-..,-----,-..----..,---,.-.---,---.n,--
- - - -~,--- -.---, -
9 l
breaks had not been identl6ed. The Lead Engineer told Olinvestigators that ath RELAP5YA had failed et 0.35 ft', she nonetheless believed, during her revi that the worst case had been bounded (the limting transient had been iden
}
told Ol investigators that, despite the fed that RELAP5YA had folied at 0.35 ft, he was nonetheless con 6 dent, during the development of the RELAPSYA SBLOCA analy YAEC 1868, that the limting break had been identitled because Peak Cladding T
?
calculations had trended downward with incrossing break size. This rationale is neither acceptable nor credible because, as explained above, RELAP5YA experienced inc t
oscillations as break size increased and because there was a portion of the break s j
i which there was no acceptable evaluation model capable of Salculating coolihg perfor In view of the above, N is reasonable to conclude that the Manager and the Lead Engineer knew that there was a portion of the break spectrum between 0.35 ft and ft' for which no NRC. approved, scooptable LOCA evaluation model was capable of j
i calculating or reliably calculating cooling performan:a.
In view of the above, N is also reasonable to conclude that as a resuN ofits pre and review of the RELAP5YA SBLOCA analysis described in YAEC 1868, and Rs and approval of the CPAR's used to support MYAPCo's Cycle 14 and Cycle 15 r YAEC caused MYAPCo to be in apparent violation of 10 C.F.R. I 50.46(a)(1).
B.
MYAPCo maintained informatlan and anhmittad to the NRC Core Analvals Rancr'A in an=.it of Cycle 14 and Cvcis is reland a-armee wheh anaarentiv ;;re not camalata nnd ve:arata in all material reae-^=
As set forth in the December 19,1997 letter to MYAPCo, Enclosure 4, Apparent' Violation E, Maine Yankee maintained information and submitted to the NRC Core P r
Analysis Reports (CPARS), in support of Cycle 14 and Cycle 15 reload analyses,
_,ay,+y,-wr-"tES'W'
~
10 not complete and accurate in all material respede, in apparent violation of 10 C.F.
MYAPCo used the RELAP5YA SBLOCA analysis described in YAEC 18685 to prepar CPARs in order to demonstrate compliance wth 10 C.F.R. $ 50.46. YAEC 1868 conesels the lack ok an scooptable evaluation model to calculate ECCS performance for a portion of the break spectrum between 0.35 ff and at least 0.6 ff (see Section Ill.A.
making statements which imply.at the complete break spectrum was analyzed or ospa analysis, s.nd by stating that a sufficient number of break s!*es was analyzed such that the limiting LOCA conditions were identified in conformance with the requirements of 10 C.F.R.
I 50.46(a)(1).
Through its submission of YAf C 1868 to MYAPCO, YAEC provided to MYAPCo information required to be maintained by the NRC which apparently was not complete and accurate all matenal respects. Based upon the Assessment Team review and the 01 I
investigation, it appears that no Maine Yankee personnel realized that the RELAP5YA code 1
failed at 0.35 ff or that there might be a portion of the break spectrum for which there was no acceptable evaluation model to calculate ECCS performance, and that no one at YAEC informed MYAPCo personnel that REl.AP5YA had failed at 0.35 ff. The Manager to.W Ol investigators that he did not think that the failure of RELAP5YA at 0.35 ff was significan does not recall discussing this failure of RELAP5YA with the manager in charge of the Engineering Section of the Licensing and Engineering Group of MYAPCo )Ao was the i
MYAPCo manager kept directly appnsed by YAEC and by the Manager of the develop the plant specific REl.N'5YA evaluation models on a continuing basis.
i 4
' YAEC-1668 is a record required to be maintained by the NRC because it is part o
- documentation underlying the CPARs, which are records required to be maintained by
$ 50.71.
-.. ~ -
11 Statements in YAEC 1868 that RELAP5YA ' terminate #' after SIT a ft' break site case (see YAEC 1868, at 22) do not reveal the non40mpliance wRh t
10 C.F.R. $ 50.46(a)(1). That language would not signify to an indiviouel wthout expert LOCA codes that RELAP5YA had failed and was not capable of calculating ECCS pe at break skes of and prester than 0.35 ff. N YAEC 1868 had explichly stated that RELAP5 failed at 0.35 ft' and was not capable of calculating ECCS performance at break stres of an greater than 0.35 ft', and that there was no NRC approved, acceptable evaluation model which could analyze a portion of the break spectrum, the description of the RELAP5YA SBLOCA evaluation model would have been complete and accu / Ate.
The Manager participated in the development of the RELAP5YA code, supervised the development of the RELAP5YA evaluation model described in YAEC 1868, and assisted in th t
preparation of YAEC 1868. The Lead Engineer performed the quality assurance review of RELAP5YA SBLOCA analysis described in YAEC 1868. Both the Manspor and the Lead Engineer were familiar with the WREM LBLOCA analysis developed by YAEC for Maine Yankee. It is reasonable to concluds that the Manager and the Lead Engineer knew that there was a portion of the break spectrum, between break sizes of 0.35 ft and at least 0.6 ft', for which no NRC approved, acceptable LOCA evaluation model was capable of either ca or reliably calculating coohng performance. (See Sedion Ill.A appra).
Based on the above, R is reasonable to conclude that, as a retut of YAEC's preparati and review of YAEC 1868, YAEC provided MYAPCo with information that was not comple securate in all material respects, and thus caused MYAPCO to be in apparent violation 10 C.F.R. I 50.9(a).
l
l.
l i
12 C.
Durina Cvels 14 aaaratians and in the evc!s 14 and cvels is CPi
"-d an annarantiv unmeaantable SBLOCA avaluation m
"=f which over artMad -m -- p.
As set forth in the December ig,1997 letter to MYAPCo, Enuosure 4, Apparent Volations F and G, the RELAP5YA BBLOCA evaluation model described in Y to calculate core cooling performance in Cycle 14 operations and in the Cycle CPARs, apparently did not satisfy the requirements of 10 C.F.R. 3 50.46(a)(1), be result of incorrect calc 0tations of the penetration factors, which arose from misappli Alb Chambre penetration correlatiorf, the anWy6's provides no basis to assume ful of the emergency core cooling system ir#ction and provides no basis to derive the loss coefficient of 600 used for the split downcomer nodalization. These deficiencies resu prediction of core cooling and overstatement of the conservatism of the model.
1 Junction 057' was the only communication between the intact loop and the b I
downcomers. A very large loss coefficient of 600, compared to 0.6 in the original m for Junction 057 in YAEC 1666. The loss coefficier$ were calculated based on pene factors ranging from 0.96 to 1.0. If the Alb Chambre correlation had been appl penetration factors would have been calculated in the range of 0.6657 to -0J767, which is 4
' The Alb Chambre Correlation is an empirical correlation developed by Gar Paul L Chambre (" Correlations for the Penetration of ECC Water i Downcomer Annulus," Nuclear 2ngineering and Design,53, PP 237 248 penetration factor of the ir@cted ECCS water penetrating the downcome)r ann for calculation of the
- plenum,
' A junction is a connection that provides a flow path or communication betw or volumes, in a plant nodslization model. Junction 057 is the flow junction between tw (nodes) of the downcomer in the reactor vessel downcomer nodalizat
k 13 meaningless result because the calculat! arm would have been less than aero? Such j
omiculations also indicate other possible errors in application of the Alb Chambre correlation.
j The Manager supervised the preparation of RELAPSYA SBLOCA analysis described'in i
YAEC-1868, and the Manager was one of the preparers of the YAEC.1968 report. The Lead i
j Engineer performed the Quahty Assurance (QA) review of the analysis described in YAEC 1he Manager and the Lead Engineer should have realized during their work associated with the RELAP5YA analysis described in YAEC 1868 that the Alb Chambre correlation had been Incorrectly applied. In particular, an adequate QA review would have revealed the errrits and the unacceptability of the RELAP5YA SBLOCA analysis described in YAEC.1868.
In view of the above, it is reasonable to conclude that YAEC cauesd MYAPCo to rely o an unacceptable SBLOCA evaluation modelin apparent violation of 10 C.F.R. I 50.46(a)(1).
D.
MYAPCo used an annarantiv unacemotable Best Estimate RELAPSYA B evaluation model to calculate ECCS nerformgg;g, L
As set forth in the December 19,1997 letter to MYAPCo, Enclosure 4, Apparent Violation H MYAPCo performed a safety analysis in order to determine if a decrease in steam gene pressure involvec an unreviewed safety question, pursuant to the requirements of 10 C.F.R. i 50.69, arid in so doing used an unacceptable Best Estimate RELAP5YA evaluation model to calculate SBLOCA ECCS performance, in apparent violation of 10 C.F.R. I 50,46(a)(1). The
~
Section 50.59 analysis was prompted by equipment degradation which had caused a reduction
'A Mnetration factor is the fractional or partial penetration of ECC injection water into the i
downcomer, defined as the ratio of the volumetric flow rate of the ECC water penetrating int lower plenum to the volumetric inlet flow rate. The penetration factor ranges from 0 for no penetration (total bypass) to 1 for full penetration. Penetration f9ctors either greater than 1 or less than 0 are meaningless because iney are physically imppelbie.
g ll t
,.m.
L.~.
.._.,.._...-..._L.
.-..~..._-.-..__..-..._.,~.....~--,,..---.m..-
-,, _ _ ~. -. - - _., _ *..,.,, -...
14 in SG pressure, wh!ch was possit,1y a non conforming condition, as the opera had fallen below the pressure assumed in the SSLOCA analysis of ricord."
MYAPCo's Jenuary 1993 Section 50.59 ar.elysis relied upon an analysis cf th 1
i the reduchon in SG pressure on ECCS performance prepared by YAEC, wh other analyses, the Best Estimate (SE) RELAP5YA SSLOCA analysis. The Y prepared " impact of Low Steam Generator Pressure on LOCA Analysis"(LOCA 0104),
January 25,1991, which used the June 1990 SE RELAP5YA analysis for the SS of the study. LOCA 9104 was approved by the Manager of the YAEC Nuclear Department for the Manager A YAEC Memorandum, " Impact of Lower Steam Generato Pressure on the Safety Analysis"(NED 91 18), dated January 28,1991, relied on LOC to evaluate the impact of reduced steam ponerator pressure on the MYAPCo saf The Manager was provided a copy of NED 91 18, which was also transmitted to the Vice 4
President of MYAPCo Licensing and Engineering, on January 28,1991. Finally,in a Y memorandum dated May 29,1992, and titled " Steam Generator Pressure and Heat Transfer Coemeient Monttoring", YAEC concluded that "the lower initial SG pressure did not affe results of the licensin0 analysis," based in part on NED 91 18. The May 29,1992, memo states: 'This memo is safety related". The Manager approved the May 29,1992, m t
which was transmitted to the Vice President of MYAPCo Lloonting and Engineer 4
" The SG pressure assumed *n the SSLOCA analysis of record (the CE SSL between Cycle 4 and mid-Cycle 13, and thus at the time of the January 1991 SG reduction analysis performed pursuant to Section 50.59, was approximately 477 p
-intomal memorandum frum F. John Card dated April 7 1989 (His 14 87 210 April 8,1989, at full power, SG pressure was 827 psig, and that on October 30,1954
), states that on MWt SG pressure was 825 ps;g.
I r
m.m
-..,_.. -. -..~. -.
._..-.... ~.. _ -.... _.,
m.-..
mo
._..,.m.m.
,,-.m..
.m-
t 15 t #92. The MYAPCo Section 50.59 analysis of reduced steam 9enerator pressum referenced and relied upon NED 91 18" and the May 29,1992 memorandum.'8 A Section 50.59 analysis cannot confirm that ECCS performance will be adequate unless the Section 50.59 analysis uses LOCA evaluation models acceptable to demonstrate compliance Wh 10 C.F.R. 5 50.46. Maine Yankee's reliance upon the BE RELAP5YA model was in apparMt vio*ation of 10 C.F.R. l 50.46(a)(1) because the January 1989 NRC SER approved RELAP5YA for application to Maine Yankee as a full Appendix K evaluation model, not ts a BE evaluaten oMel, to demor. strate compliance with 10 C.F.R. l 50.46. Additionally, the BE 82LAhWA evaluation model apparently did not comply with 10 C.F.R. Part 50, Appendix X, because: (1)'ie BE model calculated decay heat with the 1979 ANS Standard l
t j
rather than the required 1971 ANS Standard plus 20 percent; and (2) the BE model calculated the two phase criticel flow with the RELAP5YA mechanistic model rather than the required Moody critical flow model.
l H is reasonable to conclude that the Mana9er of the YAEC LOCA Group was aware, I
before the studies of the impact of reduced steam generator pressure on LOCA analyses were parfom.td, that the BE version of RELAP5YA did not conform to the SER and that hs use without NRC approval would be une coptable. A memorandum dated January 2,1990, prepared by the Manaper, states that the Best Estimate RELAP5YA SBLOCA code, which the Mana9er had proposed to develop to resolve the Three Mile Isl6nd Action Plan Nem II.K.3.31,
" Maina Yankee's " Technical Evaluation" associated with the Section 50.59 analysis referenced NED 91 18 as MYP 910098, the identification number Maine Yankee assi ned to 9
NED 91 18 upon Ms recolpt.
'8 Maine Yankss's "10 CFR C0.59 Determination" referenced the May 29,1992 memorandum as MYP 92 0605, the ident$ cation number Maine Yankee assigrwd to the May 29,1992 YAt a memorandum upon ha receipt.
I m
_ - _ _,,. - ~ _.
it ddfors from the approach approved by the SER, and suggests that the proposed BE method and results be submitted to the NRC. The Manager to hvostigators that this had not been done, that he believed that MYAPCo perso discussed the possibilfty of a BE approach to close out item II.K.3.31 with 8
l Manager for MYAPCO, that the NRC staff did not issue an approval of that pr Manager was not aware of any NRC approval of that proposal, and that he l
RELAP5YA model would not be acceptable for use in licensing me*sts, t
j tt is also reasonable to conclude that the Manager knew that the analysis whi performed regarding the effects of a reduction in steam generator pressure was a safety analysis which would be used by Maine Yankee in a Section 50.5g a other safety analysis. The Manager stated to 01 investigators that he did riot kno 3
YAEC analyses were in fact used by Maine Yankee to perform a Section 50.
Because of the very nature of their work, however, LOCA group engineers, inc Manager, would have to have known that the YAEC analyses of SG pressure re intended by Maine Yankee to be used for a Section 50.5g or other safety analy 31,19g0, Maine Yankee Servios request (MY A 24-82, Rev. 3), NED gi 18, '1m e
Steam Generator Pressure on the Safety Analysis", and the May 2g, igg 2 me i
"Stosm Generator Pressure and Heat Yrensfer coefficient Monitoring", explici YAEC studies were performed to determine the effect of the SG pressure re MYAPS safety analysis, in view of the intended use of the YAEC analysis, have provid >d Maine Yankee with an enaPysis which met NRC requirements.
In view of the above, it is reasonable to conclude that by approving the u i-providing to Maine Yankee an unacceptable analyses of the effects of reducti on LOCA analyses YAEC caused Maine Yankee to apparently violate
\\
I 17 10 C.F.R. $ 50.46(a)(1) by relying on an unacceptable SSLOCA evolustion mode ECCS cooling performance in propering a section 60.5p analysis.
i j
J IV 1
i The foregoing situation raises serloco questions concoming regard for and adh NRC requirements ans concoming management control and supervision over licen Questions are raised es to whether YAEC and/or DE&S willin the future provide co
' l accurate information to licensees and to the HRC; whether YAEC and/or DEAS are able to otherwise conduct their activities in accordance with the Commission's req I
whether YAEC and/or DESS shou!d be pen..itted to provide LOCA analyses or other safety related analyses to NRC licensees.
i Further information is needed to determine whether the Commission ca reasonable assurance that in the future licensees can rely upon YAEC and/or DE&S complete and accurate information; whether YAEC and/or DEt" gre willing and able to otherwise conduct its activities in accordance with the Commission's requirements; an YAEC and/or DESS should be permitted to provide LOCA analyses or other sa analyses to NRC licensees.
f 4
+
V
- Acwdingly, pursuant to sections 161c,1810,182 and 186 of the Atomic Ene 1954, as amended, and the Commission's regulatJons in 10 C.F.R. $ 2.204, the Com needs the following information to determine whether enforcement action should be ta I
a
. ~. -.
18 against Yankee Atomic Electric Company (YAEC) and/or Duke Engineering a
Company to ensure future comphanoe, on the part of NRC licensees, v.ith NRC req A.
An explanation why, in view of the matters set forth above, the NRC should permh any NRC Licenses to use the services of YAEC LOCA Group and/or DEAS, to the extent that YAEC LOCA Group was
', transferred to CE&S, to perform LOCA analyses or any safety-related analyses to meet NRC requhements.
8.
An orptanation why the NRC should not cons! der the inadequate t
analyses, which apparently caused MYAPCo to be in violation of NRC requirements, to be the resuN of wilfulness, either deliberateness or careless disregard, on the part of YAEC and/or DESS personnel.
VI.
Ya tkee Atomic Electric Company and Duke Engineering & Services Co. are requ submh their responses in writing and under oath or affirmation, to the Director, Of6os of Nucle Reactor Regulation, U.S. Nuclear Regulatory Commission. Washington, D.C. 20555, w days of the date of this Demand for Information. Copies also shall be sent to the Direct Ofnce of Enforcement, the Assistant General Counsel for Hearings and Enforcement at th same address, and to the Reg'onal Administrator, NRC Region I,475 Allendale Roed, K Prussia, PA iM06-1415.
1
-. -~
19 After reviewing the tosponses, the NRC will determit.e whether further action is necessary to ensure compliance wth repuistory requirements.
FOR THE NUCLEAR REGULATORY COMMISSION [
i da or CWee of Nuclear Reactor Regulation Dated at Rockville, Maryland this 4,1 day of December,1997 l
i
_..