ML20003G680
| ML20003G680 | |
| Person / Time | |
|---|---|
| Site: | Three Mile Island  |
| Issue date: | 04/28/1981 |
| From: | Sholly S AFFILIATION NOT ASSIGNED |
| To: | Atomic Safety and Licensing Board Panel |
| References | |
| RTR-NUREG-0667, RTR-NUREG-667 NUDOCS 8104300515 | |
| Download: ML20003G680 (59) | |
Text
'
,/f g 1 SHOLLY, 4/28/81
~1
- 6 g ; y'8 %
&9ffg UNITED STATES OF AMERICA
'e m= 9 NUCLEAR REGULATORY COMMISSION
\ 4.;
ig &
yt
@ , FORE THE ATOMIC SAFETY AND LICENSING BOARD In the Matter of )
)
METROPOLITAN EDISON COMPANY, ET AL. ) Docket No. 50-289 g
) (RESTART) y (Three Mile Island Nuclear ) g Station, Unit No. 1) )
9 y . . , -
p INTERVENOR STEVEN C. SHOLLY MOTION [ N*
TO THE ATOMIC SAFETY AMD LICENSING BOARD TO TAKE OFFICIAL NOTICE OF
(*- g
.< e g/j CERTAIN PORTIONS OF NUREG-0667 4 " ~ '
w Pursuant to the provisions of 10 CFR S2.743 (i) , Intervenor Steven C. Sholly hereby requests the Atomic Safety and Licensing Board to take Official Notice of specified portions of NUREG-0667, TRANSIENT RESPONSE OF BABCOCK & WILCOX-DESIGNED REACTORS, May 1980.
This document, especially recommendations made therein, were addressed by NRC Staff witnesses Denwood F. Ross and Robert A.
Capra during oral testimony on March 19, 1981 (Tr. 15,762 through 15,796) in response to Board Question 7 which dealt with NUREG-0667.
Upon reviewing the state of the record on the Integrated Control System and the participation of the ICS in feedwater transients and feedwater oscillations, and upon reviewing NUREG-0667 (a copy of which was obtained from the NRC's Public Document Room),
it became clear that NUREG-0667 contains much information and certain facts which appear to be directly relevant to Sholly Contention 6-a, the contention dealing with the FMEA of the ICS.
l &6104 300$lg 9 ft
2-l There are also a few pages which are relevant to Sholly, Contention 1, which dealt with containment isolation. ;
1 The pages which are proposed herein for Official Notice by the Board are listed in Appendix A to this motion,~along with-l l a brief explanation of the relevance of the particular pages to l
Contentions 6-a or 1.
It should be noted that if the Staff disagreed with any i
factual materials or recommendations within NUREG-0667, there has been ample opportunity for the Staff to bring such disagreements to the attention of the Board and the parties. According to 4
page 1-6 and 1-7 of NUREG-0667, the following individuals 1 participated in the B&W Reactor Transient Task Force which authored NUREG-0667:
- a. J. Anderson, ORNL (presumably John L. Anderson, who led the Oak Ridge National Laboratory review ;
of the B&W report on the ICS (BAW-1564), in a consultant role to the Task Force; the ORNL ;
l review of BAW-1564 is Sholly Exhibit #2 in this '
proceeding.
i
- b. D. Thatcher, NRC (presumably Dale F. Thatcher, l
l NRC Staff witness on the Failure Modes and Effects Analysis of the ICS in response to ;
i Sholly contention 6-a; also Technical Monitor I I for the contract with ORNL to review the l I B&W report on the ICS).
I
- c. R. Capra (presumably Robert A. Capra, who was the NRC Staff witness on Board Question !
Number 7, which was concerned with NUREG-0667).
- d. T. Novak (presumably Thomas Novak, a senior NRR l Staff member who appeared before the Board to ,
address the matter of Staff allocation of resources to the Restart hearing).
et ww+- e e see+---=eem m-- -.- w se- - + -**mweeer +e-w.em-wm----wwww,m. e- -,+e r+ww--i-----wm-y _ e wo--rw-e- gw w w + ,sr v e m-e-t-a--i-ywm----yereste--wms--**ve-- n +-- -Wa+-ym+wWu---=g-"pa**w* -
- Witnesses Thatcher and Capra appeared in this hearing to testify on contentions and Mr. Novak appeared before the Board to address the allocation of Staff resources to the Restart hearing. Certainly, if the Staff had any disagreement with the facts and recommendations set forth in NUREG-0667, there was ample opportunity for such disagreements to be brought to the attention of the parties and the Board.
On the basis of the facts set forth herein and in the j
attached Appendices A (setting forth the specific pages and reasons for Noticing each item) and B (copies of each relevant
! page for Board reference), Intervenor Steven C. Sholly requests l
l that the Board grant this motion.
DATED: 28 April 1981 RESPECTFULLY SUBMITTED, l l
Steven C. Shollyy Union of Concerned Scientists
- f 1725 I Street, N.W. i Suite 601 ]
Washington, D.C. 20006 l
- Affiliation for mailing purposes only; Intervenor Sholly represents only himself in this proceeding. UCS is represented by Ellyn Weiss, Esquire.
)
l
i SHOLLY, 4/28/81 APPENDIX A Specification of portions of NUREG-0667 proposed for Official Notice by the Atomic Safety and Licensing Board l
l ITEM # PAGE # DESCRIPTION OF SIGNIFICANCE OF ITEM l
l
- 1. 2-2 The second full paragraph on this page addresses the significance of the OTSG in the B&W design, particularly the first sentence of the paragraph which states:
"It is clear that the OTSG is unique in terms of its capability to affect either rapid cooldown or heatup of the reactor coolant system."
l This is significant since the OTSG, as explained I in the Commission's August 9th Order, is a i primary cause for the sensitivity of the B&W design to secondary system perturbations, particularly feedwater transients in which the ICS has been demonstrated to take part.
- 2. 2-2 to 2-3 The general findings of the B&W Reactor Transient Response Task Force are described as follows:
" (1) Confirmation that B&W-designed plants are more responsive to secondary side perturbations than other pressurized-water reactors.
(2) The once-through steam generator design is technically sound; however, it requires a highly interactive and responsive control system (i.e., the integrated control system).
(3) A high degree of overall plant interaction is inherent in the integrated control system and the once-through steam generator.
(4) Based on the design features and the l
faster response of B&W plants during transients and upset conditions, the operators may be required to take more rapid action and have.
a better understanding of instrument response than operators on plants having other designs."
These findings are very important to the consid-eration of failure modes of the ICS and to the degree to which operator familiarity with such failure modes may affect operator response to ICS and ICS-related failures.
_ _ - _ . . . . ~ . _ , , _ . _ . _ _ _ . . _ _ . _ _ _ _ _ , _ . . _ _ . - _ _ _ _ _ . . _ _ _ _ _ _ .-.__-- _ _... _ ____ ___ _
- 3. 2-5 thru NUREG-0667 discusses specific recommendations 2-7 related to instrumentatiort and control, in ,
particular recommendation 5 which is directly related to the ICS and i':s power supplies.
- 4. 2-9 Recommendation 8 under Instrumentation and.
Control directly relates to Sholly contention 1.
- 5. 2-9 thru Recommendations on Design and Operational 2-10 Matters are related to the ICS, particularly recommendations 9, 10, 11, 12, and 14.
- 6. 4-6 Last sentence on the page states:
" Loss of instrumentation and failure-induced control system response (e.g., failure of control systems or their input signals) have contributed to transient severity or hampered operator attempts to stabilize plant conditions."
This is an important conclusion regarding the consequences of power-failure induced instru-mentation loss and the response of the ICS to ICS or input failures, a conclusion which was not included in the Staff testimony on ICS, in the B&W FMEA (BAW-1564) , nor was it in the ORNL review of BAW-1564. This conclusion is important in assessing the importance of .
improvements in ICS response.
- 7. 4-7 The analysis of NNI/ICS power failure incidents contained in the first paragraph of Section 4.4.1 of NUREG-0667 is significant since it shows quite clearly the results of such power failures.
It also demonstrates the these events occurred more often with the reactor at power than in shutdown conditions.
- 8. 4-8 The information in the first paragraph of this page supplements that referred to in Item 7 above and is-similarly important in showing how NNI/ICS power failures are linked to feedwater transients.
7
,w.- .-g--- , w,+-m - , , , --w,,- --
, ,-,-,---r--.--,-,- -----wam---nnw..,r~,-ww-enw--,w-w,r--.-m-,~w-.r,--.-wmn*,-w.-------, --we-e~-.- --------~s,v ---e' -
~
T
- 9. 5-18 thru Conclusion (1) under Section 5.2.2.4 regarding 5-19 the disadvantages of the OTSG design is significant since it' serves to explain how the OTSG responds to secondary system perturbations (such as those caused by ICS failures'and ICS-related failures). .
I i
- 10. 5-19 thru These pages contain a detailed discus'sion of 5-24 undercooling and overcooling events, in which, based on the FMEA performed by B&W, the ICS can become involved. This subject was also discussed in response to questions raised by Administrative Judge Jordan with NRC Staff witnesses Ross and Capra (Tr. 15,773 thru 15, 781), and is thus supplementary to that cross-examination. It is also, in my view, more coherent than the oral response to the cross-examination by Administrative Judge Jordan referred to above.
- 11. 5-50 thru These sections (5.3.2 and 5.3.3) explain the 5-52 NNI and its interaction with the ICS, a subject which, although important in the context of ICS power supply failures, is not fully addressed in the Staff's testimony or the B&W report on the ICS (BAW-1564). This information is therefore supplemental and illustrative of a main concern raised by B&W and the Staff in their recommendations to ~
B&W plant owners, including the Licensee.
Page 5-52, second paragraph, also addresses the~ topic of mid-scale failures, which is highly relevant to the ICS FMEA and which was previously addressed in cross-examination.
- 12. 5-55 thru This section, beginning at the bottom of page 5-56 5-55, discusses the genesis of I&E Bulletin 79-27 which is referenced by the Staff in numerous filings and reports.
l
__. _ _ - - _ . _ _ _ - . _ _ - - _ _ . . . . . _ _ .,_m.. -_ _ . , , _ - . - . _ _ - _ . . _ _ _ . . _ . _ _ . . _ .
_4_
- 13. 5-57 The paragraph at mid-page (first full paragraph) axplains succinctly why the ICS,is related to the TMI-2 accident, thus providing the nexus which the Board has previously expressed difficulty in perceiving. This is essential to the record and is something which the Staff has not thus far provided.
- 14. 5-58 First full paragraph (mid-page) explains the genesis of the ICS FMEA requirement, noting that it was a B&W-proposed requirement which was confirmed by Commission orders and that the FMEA requirement is related to TMI-2 related concerns about B&W plants.
- 15. 5-66 Conclusion (4) and the associated recommendation are directly related to Sholly Contention 1.
This conclusions and recommendation are refe renced I in cross-examination of NRC Staff and Licensee witnesses on Contention 1 (Tr. 7352 and 7384).
- 16. 5-69 thru This discussion, beginning with the second 5-70 paragraph on page 5-69, is directly relevant to Sholly Contention 15 (Human Factors Engineering Review of Control Room Design) as well as to th~e ICS. It discusses the advantages and disadvantages of the B&W simulator facility in providing training to B&W reactor operators. .
! 17. 5-78 The first full paragraph identifies the importance l
of having a qualified I&C Technician on-duty
! on all shifts, and suggests that the Crystal River-3 accident could have been longer in duration (perhaps more serious as well) had there not been a qualified I&C Technician on site at the time of the accident.
- 18. 5-83 This conclusion and recommendation reinforce the discussion on page 5-78 identified by Item 17 above.
.n_
- 19. 6-2 Mid-page,-niere is a discussion which reveals that the Crystal River-3-related IREP program does not include NNI and the ICS within the scope of the IREP program in a quantitative manner.
- 20. 6-3 thru Mid-page, there is a further discussion of 6-4 the short-comings of the IREP program regarding
{
the ICS and additional information on common cause failures related to the ICS, a topic which was addressed by the ORNL review of BAW-1564 and discussed briefly in cross-examination on Sholly Contention 6-a. At the top of page 6-4, there is additional discussion of common
' cause failure modes.
- 21. 7-8 The first full paragraph discusses operator action related to feedwater transients and how operating practice at B&W plants may increase the chance for such operator errors.
- 22. 7-12 thru Beginning with the first full paragraph .
i 7-13 on page 7-12 is a discussion of the impact of water in the main steam lines, which was the subject of limited discussion on the record under cross-examination of NRC Staff witnesses by Administrative Judge Jordan '
(Tr. 15,781 thru 15,783). This was also discussed briefly in the ORNL review of l BAW-1564 (pages 11-12).
- 23. 7-36 thru Item 8 beginning at mid sge on page 7-36 7-38 is directly related to Sholly Contention 1 and explains quite well the significance of the recommendation, something which NRC Staff and Licensee witnesses were unable to address (Tr. 7352 and 7384).
- 24. 7-43 thru Discussion of the benefits of specific procedures 7-44 for ICS/NNI failures, and the value of these procedures and related training.
- 25. Table B.1 This is a historical summary of NNI/ICS B-2 thru power failures and supplements the data in B-8 BAW-1564 very well.
I
\
l
)
SHOLLY, 4/28/81 APPENDIX B Copies of pages of NUREG-0667, with title page, for which Official Notice is requested in this motion.
t a
t i
l n
\
1 i
i I
l i
+
l J
l 4
l r
D
NUREG N l
l l
Transient Response of '
Babcock & Wilcox-Designed Reactors 1
Manuscript Completed: April 1980 Dsta Published: May 1980 BErW Reactor Transient Response Task Force Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, D.C. 20555
/% ,
.. . . . N' i
I -
I h I
l
The Task Force h o conducted an essment of the effe ive. css of the short- and lo tern lessons lear d actions resultin rom the efforts of e Lessons rned Task Force d the Bulletins and rders Task Force. t e. assess-men as based on the erating history of e B&W plants durin he put-TMI-2 accident period. he results of this eview did not revea any major deficiency in the requ ements being applie to B&W facilities. nstead, implementation of the requirements appe to have led to an erall improvement in e ponse of B&W-desi d plaats to various ransient events. The sk Force supports the ex itious implementati of these requirement , since it is believed t they will contrib e to improving to the afety of all ope ing plan .
h It is clear that the OTSG is unique in terms of its capability to affect either rapid cooldown or heatup of the reactor coolant system. However, replacement of the OTSG does not appear to be practical or a necessary action for operating plants, especially when weighed against certain other safety advantages of the OTSG. Furthermore, this Task Force does not believe that complete plant shut-down of the B&W plants is either necessary or desired with regard to public health and safety. -
The general findings of the Task Force may be stated as follows:
1 j 1 Confirmation that B&W-designed plants are more responsive to secondary 4
bl(1) side perturbations than other pressurized-water reactors.
i i
l 2-2 y c , . . , _s .. __ __.p- . . _ . , - . . . .,,m ,,,.y..y_ , - _ , _.,,y_._,,.c,, , ,-m.
l I
l (2) The once-through steam generator design is technically sound; however, it requires a highly interactive and responsive control system (i.e., the integrated. control system).
(3) A high degree of overall plant interaction is inherent in the integrated control system and the once-through steam generator.
I (4) Based on the design features and the faster response of B&W plants during l
i transients and upset conditions, the operators may be required to take more rapid action and have a better understanding of instrument response than operators on plants having other designs.
I The specific rec endations of the Ta Force ~ focus on mini zing the conse-quences of se ndary side perturba ons (e.g., providing ore reliable instrumen tion and control sys ens, assuring availab' ity of heat sink, d improv'ng plant recovery ac ons). As such, thes recommendations f into fo main action areas: (1) auxiliary (or em gency) feedwater, ) instru-mentation.and contr , (3) design and oper ional matters, an (4) general areas of improve nt to enhance the s ty of the B&W-des ned reactors. Th does not mea that efforts to redu the frequency of ransients should continue A discussed in Sectio of th'c; report, o determine the ef ectiveness of each of the recomme dations ara to as ss scheduling prio ties, an evaluatio i of the risk red tion potential a ociated with each these recommenda 'ons has been pe ormed by the Pro ilistic Analysis aff. In addition l
prelimi ry resource esti tes have been prov" ed for planning p oses.
1 2-3
, ~ ,
l 3 _
!I initiate A ' system flow shoul e reevaluated to pe t automatic I
initia on of AFV in a no timely manner to pre ude steam generator d ut (i.e., AFW sy em automatic start an nticipatory loss of f d.<a te r).
I In addition, th evel of secondaiy co ant in the steam genee ors should be automat ally controlled by t AFV system in a manne to prevent over-coolic of the ' reactor cool system during recov y from feed.<ater transfer d that an appropriat ignal be orovided to erminate feedwater flow the steam generate before overfilling t es place.
(3) Installt on of a diverse-dri AFW pump should be expe ted at the Davis-Be 1 facility. -
~
l -
l l (4) The' steam lin reak detection and miti tion system should be a 4 fled as neces ry to eliminate adverse ~ teractions between it the AFW .
, sy .
In addition, to fur + er assure -heat sink ava' ability, the steam l
l line break detection an mitigation system shou be reevaluated and modfried in such anner that it is capab of differentiating tween an actual st line break and under oling or overcoolin vents caused by fee er transients. ,
/
Instrumentation and Control (5) B&W plants should improve the reliability of the plant control system, l particularly with regard to undesirable failure modes of power source, l
l signal source, and the integrated control system itself. Specific recom-cendations for improvecent in the plant control system include the following:
1 2-5 De
l l
l l
(a) The power buses and signal paths for non-nuclear instrumentation and l associated control systems shculd be separated and channelized to reduce the impact of failure of one bus. .
(b) The power supply (including protective circuitry) logic arrangement I should be reconsidered to eliminate "mid-scale" failures as a pre-ferred failure mode for instrumentation. " Full-scale" or "down-scale" failures may be preferred in that they give the operator more posi-i tive indication of instrumentation malfunction.
(c) Multiple instrument failures, typically caused by power loss, should l be unambiguously indicated to guide operator selection of alternate .
instrumentation that is unaffected by the failure.
l (d) If control system failures or response to failed input signals can l
cause substantial plant upsets (e.g., required action by engineered safety features or safety valves), the control system should have provisions for detecting gross' failures and taking appropriate defensive action automatically, such as reverting to manual control or some safe state.
I I
l (e) The NNI power buses should be reviewad and rearranged, as necessary, l
to provide redundancy of indication of each reactor coolant and secondary system loop. That is, where indicators for one loop are provided, one channel should be powered from NNI "X" and the other from NNI "Y," instead of loop "A" being powered from NNI "X" and loop "B" from NNI "Y."
2-6
-, -.n- ,. -- - - . , . . - . - - -
k (f) Prompt followup actions should be taken on the recommendations contained in BAW-1564 (Integrated Control System Reliability
~
Analysis).
(g) NRC has reviewed the recommendations contained in NSAC-3/ IMP 0-1 (Analysis and Evaluation of Crystal River Unit 3 Incident). The staff agrees l
that licensees should evaluate the effectiveness of these recommendations l for their plants, especially with regard to the NNI/ICS aspects.
l l
b (h) Prompt followup actions should be taken on IE Bulletin 79-27.
(6) A mini set of parameter should be establishe enable the opera r
~
to ssess, plant status. The set recommended the Task Force foi ows:
i (a) Wide range eactor coolant sys ressure, I
(b) Wid range pressurizer le ,
I
( Wide range reactor oolant system temp atures: hot leg ch loop),
cold leg (each oop), and core out t (two or selectab ),
O l
(d) Makeup nk level, l (e) actor building pr su re , ,
I 2-7 i-l._ _ . _ _ _
a continuo or trending display of i e thermocupies. This splay
,. not be indicated in the ntrol room at all ti ut say be called up on demand from computer.
(8) B&W plants should provide a safety grade containment high radiation signal to initiate containment vent and purge isolation in addition to the pre-sently required signals (i.e., containment h!gh pressure and low pressure ESFAS actuation).
Design and Ooerational Matters (9) Following a reactor trip, pressurizer level should remain on scale, and system pressure should remain above the HPI actuation setpoint. The systes response (e.g. , secondary pressure) should be modified to meet the above two objectives. Meeting these objectives should be independent of all manual operator actions (e.g. , control of feedwater, letdcwn isolation, and startup of a sakeup pump).
(10) The B&W licensees should perform sensitivity studies of possible modifi-cations which would reduce the response of the OTSG to secondary coolant flow perturbations. Both passive and active seasures should be investi-gated to mitigate overcooling and undercooling events.
i (11) Modifications should be made to the plant, to the extent feasible, to reduce or eliminate manual immediate actions for e ergency procedures.
e 2-9
(12) A qualified Instrumentation and Control Technician (I&CT) should be pro-vided on a round-the-clock basis at all operating B&W reactors.
(13) Lectu should be dev ped and given pr y to all lice ed personnel neerning the C al River 3 event well as their ant-specific 1 s of NNI/ICS lysis. A means to valuate the tr ning (e.g., qu es) should e included. This aining should audited by th ffice of pection and Enfor ent.
(14) Licensees should develop and implement promptly plant-specific procedures concerning the loss of NNI/ICS power. These procedures should enable the operator to bring the plant to a safe shutdown condition. These procedures shall be audited by the Office of Inspection and Enforcement. Furthermore,
( the Task Force endorses the effort by B&W to develop abnormal transient operational guidelines and recommends full utility support be given to this program.
l (15) Ma tory one-week simula raining should be require or all licensed B&W operators. The aining should be oriented to or include under-l cooling and ov ooling events, solid system eration, and natural circ -
tion cool g. Upgrading of simulator p ormance in accordance wit he rec endations of the TMI-2 Actio lan (NUREG-0660) should b expedited.
(16) The NRC should review t criteria for RCP restart ring recovery from non-LOCA transient as provided in B&W small-b ak guidelines. Restar ing the RCPs ovides the operator with ssurizer spray and thu greatly improve lant pressure control.
2-10
, . - - . .- ,----,-m- -
sr-~+mw. a w. --
of heat sink) has ven RCS temperature and ssure high enough to cause a reactor on high pressure. Init ion of auxiliary feedwater fl o gain the heat sink, combi with steam bypass system ope ion, has on occasion caused ov oling and rapid depressuriza of the RCS to a pressure low enoug initiate high pressure inje n. This depressurization has~
o rred even though a second mak pump had been started and let n isolated (both manual operator ac ns immediately after the react rip). Per emergency procedures after e TMI-2 accid 5nt, reactor coo pumps have been tripped ,
l followi reactor trip and HPI initiati n low pressure. Excessive re.ac essel and pressurizer cooldow tes have occurred during overcoo
- g conditions.
Feeding steam generato per procedure to the 95 percent el on the operate l
l range for nat circulation cooling of the RCS s on occasion caused another loss partial loss of heat sink condi . This has occurred becaus duced sten pressure has been sensed the steam line break detect and mitigation system, where install ," which has automatically iso d steam and feedwater -
connections e steam generator.
During events in which the V opened and stuck open, act on the plant has ,
been aggravated b aster depressurization, 1 minimum RCS pressure, sic l
recovery pressurizer level by the pressure injection syste , and a ger time to establish st plant conditions. -
Loss of instrumentation and failure-induced control system response (e.g, failure of control systems or their input signals) have contributed to tran-sient sev'rity or hampered operator attempts to stabilize plant conditions.
1 4-6 4
l l
4.4 Summary of Ev i
Events of terest to the Task F ce invol've power pply failures to non-nuc ar f instru ntation (NNI) or th ntegrated contro system (ICS), reactor tr s, P0 actuations, and fe ater transients. An historical summary o NI/ICS power failures in plants is provi in Appendix. B, Table . Appendix B,
, o Table B.2 con ins (1) details of eactor trips with PORV ctuation in B&W plants th occurred before t TMI-2 accident, and the documented num r of a itional automatic r actor trips in B&W pl s during which PORY pening s not reported. T le B.3 contains deta of all unplanned r ctor trips in B&W plants si e the TMI-2 acciden Licensee estimates the effects of the two dif rent setpoints for h pressure reactor p and PORV actuati
~
are inc ded in Tables B.2 B.3 (i.e., what wo the effect have b n dur g these events if esent setpoints had en used before the I-2 accident and if pre-TMI-2 s points were in effec instead of presenp etpoints).
4.4.1 Analysis of Data in Table B.1 (Appendix B) l lSince December 1974, a total of 29 NNI/ICS power failures have been identified.
l l
( Twenty-one of these events caused reactor trips,17 caused PORV actuation, and 4 resulted in engineered saf'eguards (high pressure injection) actuations. A steam dump valve stuck open in one event, and feedwater transients occurred in 19 of these events. A pressurizer safety valve lifted in one event and a PORV. ,
stuck or failed open in three events. Three ICS power failures that occurred while the reactor was at power did not cause reactor trips, and in the remaining five power failures the reactor was in a shutdown condition when the events l l occurred.
l l l
4-7 i s
-m -
-w- g. --w- .-
,g y- - - , , , . , -% ,..,,..y-,g,,e,.,,.,e,-,s. -+r-- , - - , . - - , . - + - - , . . - - - - , , , - , . , - - - . . . . . .,-,,,,-,m---,-,w, - , - - - . , - ,.
Based on these data, NNI/ICS power failure perturbations have been severe enough, considering the B&W-integrated plant response characteristics, to cause reactor trip in almost all events (Note: in most of these instances, the reactor trip was tne result of a consequential feedwater transient).
I
( Approximately 18 percent of all observed feedwater transients have been caused by NNI/ICS power failures. Approximately 10 percent of all reactor trips have b en associated with NNI/ICS power failures. These percentages and the data j l
in Tables B.2 and B.3 (discussed later) indicate that many other initiators of ;
feedwater transients and reactor trips exist. The data in Table B.1 appear to show that, given an NNI/ICS power failure, it is very likely to result in a ;
szvere feedwater transient that will trip the reactor on high pressure (even at the pre-TMI-2 reactor trip setpoint of 2355 psig).
/ \
4.4.2 Ana sis of Data n Table B.2 (App dix B)
The ata precedi g the THI-2 acci nt suggest a pre nderance of rea r trips n response o secondary plant ransient or ups conditions tha were reflected across t steam generator and caused reac r coolant syste (RCS) pressure excu ions. Prior to e TMI-2 accident no ani.icipato reactor trip- o casion of turbin rip or loss of edwater existe Reactor- trip th PORY cctuation occur d approximately 49 times. Othe automatic reac+ r trips, during whic PORY opening wa not reported, o urred 83 times .n B&W plants prior t the TMI-2 accid t. Notwithstan ng inaccuranci which may e 'st in the ta, it is signi cant to note t t (1) RCS press re excursions ave been f
1 vere enough to use not only P0 actuations (2 5 psig old s point) but also reactor ip (2355 psig o setpoint or e , ual trip bas on operato 4-8
+
l _
g-
- - - - , - - - , - , -n -- - g --9--a em +--,%----m-- -- ei,-+9-w-- s77-
' s 5.2.2.3 erational Advantages Ass iated with OTSG Design ;
efore discussing in d il the performance chara ristics of an NSSS with an
/
OTSG design, it i of benefit to understand e operational advantages wh make the OT a desirable design. One . the most desirable featur of an OTSG the capability to produc uperheated steam which res s in lower oisture in the turbine an onger turbine life, an obv s economic advantage.
Moreover, this superb t produces a slight increa in plant efficiency, al a desirable eco ic advantage. Operationa experience has also indi ed favorable e integrity in the OTSG sign compared to inverte -tube 1
desi , and can, in part, be at ibuted to the low second side water invento l -
l and associated lower cont inant concentration. Th e benefits, however, obtained at the cos of a system highly res ve to secondary side rturbations.
5.2.2.4 Conclusions and Recommendations 1
1 (1)
Conclusion:
The Task Forces recognizes that the OTSG has certain operational advantages !
with respect to tube integrity and steam properties that-make it an attractive design. However, we conclude that other characteristics of the design result in a system that is highly responsive to secondary side l i
flow perturbations. Specifica*ly, the relatively small volume of the secondary coolant, together with the rapid change in heat transfer area with variations in coolant level in the OTSG, result in conditions that .
produce significant mismatches between tne heat generated in the nuclear core and the heat removed by the OTSG during anticipated transients.
5-18
~
i i
h These mismatches are reflected into the RCS, and cause primary coolant
- volume variations and pessure change that result in unnecessary challenges topressurereliefdevicesortheengineeredsafetyfeatures. The Task Force understands that these are efforts under way in the industry to
- investigate means to improve the response of the OTSG to secondary coolant
- perturbation. We endorse and encourage these efforts.
Recommendation:
l The Task Force recommends that licensees be required to perform sensitivity studies of possible modifications to reduce the response of the OTSG to t
- secondary coolant flow perturbation. Specifically, we recommend that passive i
l and active measures be investigated to mitigate overcooling and undercooling events.
i -
- l i -
i 5.2.3 Performance aracteristics wi OTSGs -
l
) \
l
/ There are t basic types of ondary side perturba 'ons that affect the prim system. These events that overco the primary system move ,
more heat than is eing generated in th core) or undercool t primary s em ,
(cannot re e all of the heat b ~ g generated in the e). , 1 s
5.2.3.1 Undercooling Events l
Undercooling events initiated from the secondary side usually involve either a l 1
reduction in or loss of feedwater flow to the OTSGs. For loss of feedwater l I
(LOFW) events, all of the PWR vendors, prior to the THI-2 accident, calculated l 5-19
,~
- l
.-_...,_______..J
on a conservative basis that power-operated relief valves (PORVs) would open due to high reactor coolant system pressure during the early stage of the transient.
Because a B&W steam generator holds about 27 to 30 full power seconds (FPS) worth of inventory compared to Westinghouse or Combustion Engineering steam generators that hold approximately 90 FPS worth of inventory, the primary system of B&W-designed plants pressurize faster and reach the PORV setpoint sooner during a LOFW event. B&W has calculated (Ref. 24) that it will take about 8 seconds after a LOFW to reach the high pressure reactor trip setpoint.
( Westinghouse (W) (Ref. 25) and Combustion Engineering (CE) (Ref. 26) predict trip setpoints on low steam generator secondary side level will be reached for LOFW at about 20 seconds and 17 seconds, respectively. For Westinghouse and Combustion Engineering the primary system pressure rise prior to reaching these setpoints is negligible. These results are indicative of the more responsive nature of B&W plants to undercooling events. Since the TMI-2 accident and inversion of the PORV and reactor trip setpoints on B&W plants, the responsiveness to undercooling events leads to a lower challenge rate.to ,
overpressure relief devices (both PORVs and safety valves); however, it no'w reflects a high challenge rate to the plant protection system. .
Undercooling events in B&W reactors, prior to TMI-2, usually resulted in lifing the PORV and discharging primary coolant to the pressurizer quench tank. Reactor trip on high pressure was usually precluded because the trip setpoint was purposely set higher than the PORV actuation setpoint to allow the plant to " ride through" loss of load events without reactor trip. Although most of the recorded challenges to PORVs are from B&W plants, these types of 5-20 1
1
, - - - , - - - w ,-- - - -__, , , -,-- .,,-- ,.----,,,,,--..--% ,. .,
, -,c --s.-w --..%,----,-- -,.-.-,-u-,-.w+ -, , + + , - -- ,,<-*r.--y--.e% ry-w---r.,w.-...y-,-
events have also been known to challenge the PORVs on both Westinghouse and Combustion Engineering-designed plants. Because PORV actuation is not in itself a reportable occu.rence, the data base regarding the number of PORV challenges, to date is incomplete (Ref. 27). Since TMI-2, the setpoints for the PORV actuation and reactor trip have.been inverted on plants with B&W l reactors. This action was taken to reduce the number of challenges to the PORV and hence reduce the probability of a PORV failure leading to a loss-of-coolant accident. This action however, consequently increases the number of challenges to the reactor protection system. This was recognized by the ACRS (Ref. 28), and they recommended a continued evaluation of this action on plant safety. In addition to the above action, two anticipatory reactor trips were added that will trip the reactor on turbine trip or loss of feedwater. This was also done to reduce the numbte of challengas to the PORVs.
Despite the actions just mentioned, there still remains an underlying problem resulting from failure related to the integrated control syst'em (ICS)~.- These failures involve the severe degradation of the feedwater (e.g., control valve closure, main feedwater pump runback) without producing an anticipatory reactor trip. The resultant dryout of the steam generator and loss of heat l sink would produce a reactor trip on high system pressure. The startup of the auxiliary feedwater pumps and the rapid introduction of relatively cold feedwater into the upper elevations of the steam generator overcools the primary system and can produce a rapid depressurization transient which may result in actuation !
I cf ESFAS on low primary system pressure. Overcooling and its effect on primary l
system behavior is discussed below. j l
\
I l
5-21
l 5.2.3.2 Overcooling Events .
l i
Depressurization of the primary system results from secondary side overcooling:
This overcooling usually occurs because of overfeeding a steam generator, .
demanding too much steam from the steam generators, or introducing excessive l
amounts of relatively cold auxiliary feedwater into the steam generator.
i l
l The depressurization of the primary system is caused by primary coolant shrinkage i
due to cooldown. During normal reactor trips, this depressurization is limited by the secondary side response designed to control secondary steam pressure and maintain the core average temperature at a minimum of 547*F. For B&W-designed plants, the primary system has typically been observed to depressurize to between 1700 and 1800 psig during reacb r trips that are not compounded by feedwater upsets. For events that overcool the primary system in excess of the normal cooldown experienced during a trip, th h minimum pressure will decrease further and could reach the ESFAS actuation setpoint during some events. Primary system behavior for a trip at the Davis-Besse plant is shown
, in Figure 5.7.
Typical reactor trip transients in Westinghouse-designed plants result in a lesser primary system depressurization (about 2000 psig,.or slightly below).
Although newer Westinghouse plants typically operate with a higher differential temperature across the core than G&W plants (64*F versus 50*F) and at a higher l
core outlet temperature (~618*F versus $605*F), the minimum pressure reached after reactor trip is typically higher.
5-22 4 , _.
Very little pre-TMI-2 information for plants designed either by Westinghouse -
- or Combustion Engineering is readily available concerning plant response to events that overcooled the primary system in excess of the normal cooling l
l expected following a reactor. trip. It should be noted, however, that since TMI-2, three events that depressurized the primary system'to the HPI actuation 0 setpoint have occurred in plants with reactors designed by Westinghouse and Combustion Engineering. Two of these events-involved stuck-open turbine' bypass
! valves and one was the result of a steam generator tube rupture.
Overcooling events analyzed in safety analysis reports are only concerned with demonstrating acceptability with caspect to the minimum departure from nucleate boiling ratio (MDNBR) and the overpressure limit. 'Because of this limited concern, th?se events are never carried out more than about 100 seconds into the event, and it cannot be determined from the overcooling event analyses if the HPI actuation setpoint would be reached. However, any sustained over-cooling event.would be expected to depressurize the primary system to the .9PI actuation setpoint for both Westinghouse and combustion Engineering designed plants. -
If an overcooling event depressurizes the primary system excessively, then the HPI will be actuated and emergency core cooling water will be injected into
! the primary system. For all B&W operating plants, except Davis-Besse, the HPI pumps (which are the charging pumps in the injection mode alignment) are capa-ble of injecting water at pressures above the PORV and safety valve setpoints.
Thus, operator action is required to throttle the HPI flow in order to prevent the primary system from going water solid. The need for the operator to throttle the HPI pumps does not pose a direct safety ccncern, since the 5-23
? _ _ _
consequence of operator failure to throttle back the HPI flow does not directly ,
challenge core integrity. However, the rapid depressurization of the primary system as a result of reactor trip, as well as steas. generator overcooling events, has evolved a set of questionable operator responses having the poten-tial to compound minor events into more serious ones. For any non-LOCA event in which HPI is actuated, the injected HPI water contains a high concentration of boron that must be removed from the primary coolant prior to. restarting the plant. This is a time-consuming p acess, so it is highly desirable to either prevent or terminate HPI actuation as soon as possible. Thus, for reactor trips and other non-LOCA overcooling transients, the operators at B&W plants have historically secured letdown and started a second makeup pump in an attempt to mimimize the depressurization. There is also evidence that feedwater may be throttled to minimize the depressurization. Prior to the TMI-2 accident, HPI actuations were routinely terminated rapidly because the operators assumed the actuation was not due to a LOCA but rather to system response to over feed events or even spurious actuation. Some operators would even manually actuate HPI to arrest the pressure decay, which would subject the HPI nozzles to an accountable stress cycle. Thus, although HPI actuation in itself does not result in a safety concern, the various operator actions precipitated by the -
actuation could ultimately place the plant in an unsafe condition. .
5.2.3.3 C clusions and commendations (1
Conclusion:
Auxi ary feedwater in owered-loop p nts is delive d to the ste nerators through n auxiliary f dwater spray ng in the up- r elevations 5-24
/
- f generator fe system, the reac r, or the turbine ge rator. The control -
system pr vides limiting a ons to ensure proper elationships between the gener ted power, turbi header pressure, fee ater flow, and reactor po r.
- The ICS was d igned to be able to p vent a reactor trip for ny anticipated plant up s ranging from minor- sets, such as small loa hanges or small feed ter heating upsets, t ajor upsets, such as 1 of one reactor cool t mp, loss of one main edwater pump, or turbi trip from 100 percen power.
Following a re or trip, the ICS cont s steam generator le 1 at a minimum level setp 'nt with the startup f dwater valves to prov e decay heat removal.
Upon 1 s of both main feedw er pumps, this minimu evel control is accom-
,pl hed with the auxili feedwater valves. S uld loss of all four r ctor coolaat pumps occur the level is controll at a higher level in se steam generator (i.e , 50 percent on the op ating range indication to help promote ,
natural c culation. Following eactor trip, the ICS so provides contr.ol of t steam pressure with t turbine bypass valves r the atmosphere d p alves (depending on th availability of the co enser and circulat' g water). ;
d 5.3.2 Non-Nuclear Instrumentat' ion Design B&W plants utilize ~a set of instrumentation classified as the non-nuclear instrumentation (NNI) to provide a significant amount of the input information to various plant control systems, including the ICS. Additional input informa-tion for plant control is obtained from the reactor protection system (RPS) :
and nuclear instrumentation (NI) system. In addition to providing information to the plant control systems, the NNI supplies control room information for 5-50 !
. . - _ _ _ _ A
tha main control board, plant computer, alarm annunicator, and display informa-tien from the RPS and the engineered safety features (ESF) systems. Therefore, tha NNI is the major source of information from which the operator can determine cenditions in the primary and secondary systems.
Th2 NNI also plays a large part in reactor coolant system pressure control by i providing control signals for the pressurizer spray valve, heaters,f and PORV.
I NNI Pressurizer level is controlled by the makeup and purification system.
provides the pressurizer level signal for automatic control of the makeup and purification system, low pressurirer level heater cutoff, and control room I
indication. i Typical B&W NNI systems (and integrated control systems) consist of two subsystems, "X" and "Y." This would imply a certain degree of channelization and/or redundancy. More channelization occurs in some plants than'others.
For certain parameters such as steam generator level, pressurizer level I and steam pressure, redundant sensing instrumentation exist with switches on the osin control board that are used to select one of the redundant sensors for .
both control and indication. However, the full advantage of redundancy is often compromised by the use of one indicator, and, furthermore, operating experience indicates that the sensors and signal conditioning are not channel'-
ized in a balanced manner. For example, it appears that one subsystem ("X" or In addition, one "Y") contains a proportionately larger amount of equipment.
subsystem may contain signal conditioning (e.g., temperature compensation nstworks) for sensors in both subsystems. Therefore, there is no true redundancy for the NNI.
5-51
. , . . , , _ . , _ , _ _ . _ . _ _ . . , . . , .__,m , , , , , ._
ICS and NNI Power Supply Design l 5.3.3
- l Th2 power for the NNI/ICS systems has not been required to be on Class IE (safety grade) power sources because the systems were not designated as part of the plant protection systems (RPS or ESF). However, in order to obtain the most reliable source of power for these systems (to prevent plant unavail-cbility) most of the plants provide the source of power from the Class IE
~
vital buses. In some plants, the vital buses utilized are not c1tssified as l
tha Class IE buses; however, they do have similar reliability because they can receive power from a battery source in case of loss of offsite ac power.
I Regardless of the " quality or reliability" of sources, power supplies do fail.
Most notably these have been caused by inverter failures or transfer switch
- failures. Therefore, events such as reactor trip can be initiated by a sing.e channel power supply failure.
I Durinp such events, the operator may also lose a significant amount of control room inforration on which he would normally rely. This is clearly an unaccep-table situation. The above emphasis is added to point out that arguments can b2 made that the operator does have sufficient information to safely shut down l tha plant; however, this may require special procedures and unfamiliar operating modes. In addition, the operator must be able to recognize the need to revert to these procedures, which can be complicated if the operator does not readily recognize the failure mode (e.g., because of "mid-scale" failures) and, therefore, does not utilize the proper procedures immediately.
5-52 g I
t
3 taken direc from protect n system instrumentatio . In a few plants, is -
informa on from the RP and ESF is not availabl in the control room 1
t l e amount of inf rmation available to t control room operat is very much l
plant depende .
The NSSS vendors re ommend utilization o safety system '
instrument ion to interface with ontrol systems and th the operators in l lvariet of ways. It is impor nt to understand th in the past the on {
\ >
l l ! re atory restriction re rding the use of the afety system instru entation, j I s an input to contro systems, was that th safety function not e comptomised l by using the prot fan signal for ches other purposes. Th efore, the l
l extent to whic the operator has saf y system instrumen tion available a
)
the statio where routine operat n takes place varie widely. The qua ity I
)
- of saf y grade and nonsafet grade instrumentati displayed on o ating j l' . pa 1s, evailable in the omputer, in the ann ciator system, a in the l sequence-of-events re rder is.different f r each operating ant'. In j recognition of thi situation, the NRC as accelerated ' e review and pla ed '
'implementatio, of Regulatory Guide .97, Revision 2, Ir.strumentation r Light-Wate Cooled Nuclear Pow Plants To Asses Plant and Enviro Conditions During nd Following an Ac ' dent" (Ref. 34). his regulatory ide addresses #
t.S need to provide a m imum amount of fety-related dis ays for the ope tor not only to mitigat the consequences of transients an accidents but a o to' monitor the perf rmance of safet/ ystems.
/
I n response to an event which occurred at the Oconee Nuclear Station, Unit 3, en November 10, 1979, NRC issued IE Bulletin 79-27 (Ref. 8). This bulletin rsquested licensee.s to review buses supplying power to safety and nonsafety-related instrumentation and contro' gystems that could affect the ability to l
l 5-55
achieve a cold shutdown condition. The bulletin also requested that emergency procedures be revised or prepared to include identification of alternate -
indication and control circuits that would, be available in the event of loss of each power bus.
Other events before and after this Oconee event (i.e., Rancho Seco and Crystal River 3) indicate that response to the IE Bulletin alone may not adequately resolve the problem of inadequate operator information. The Task Force is proposing a requirement for a set of safety-related indicators to be located in the control room to provide the operator with adequate information to assess plant conditions during and following anticipated operational transients.
5.3.5 Instr mentation and Con al Operational Consi rations Prio to the THI-2 acci ent, B&W plants total about 35 years of mbined erating data. I hat time period, as ported in the B&W " tegrated Control System liability Analysis" AW-1564, Ref. 35), ere were 310' reactor tri , about one-third of ich are attributed o the ICS or ICS-rel ed equipme . Comparative inform ion submitted by , in a letter from J. .
Tay r (B&W) to D.F. Ross RC) dated October , 1979 (Ref. 36), wo d indicate at B&W plants have a ower trip rate tha other PWR vendors.
l I
Prior to TMI-2, here had been at .ast one very signifi nt event at a B&W plant involv ng almost all of e control sytstem as cts discussed above l This was he March 20, 197 ancho Seco " light b incident." The ent was init ted by a dropped ight bulb that cause a short, which in rn caused l
ss of power to t NNI-Y subsystem. T 's caused the contr systems to 5-56 I
i
/
receive erroneous i ormation and caused large amount of the fau y information .
-to be presente on the main control card. The event resulted n a reactor trip, hig ressure injection uation, and an excessive oldown rate for th2 p nt. It could be c ectured that the event ma ave been significantly fferent were it no or the fact that the PORY ock valve was manually cicsed and that condary cooling (to at leas one OTSG) was availabl at all times.
he TMI-2 accident did not involve the failure of any of the previously discussed control systems. It did, however, involve a loss of main feedwater which i
l could have been initiated by an ICS failure. It involved the temporary loss
~
of auxiliary feedwater, which could also have been caused by the ICS. It also involved problems related to instrumentation, although not attributable to any instrument failure but rather to the lack of direct information, misinterpreta-tion of information, and too much emphasis on one plant parameter (i.e. ,
pressurizer level). It also irvolved a stuck-open PORY that could have been cru.ed by its control system failure.
After the THI-2 cident, a series of fur r events occurred at plants.
Most notabl .fere the increase in rea or trips (as reported an August 23, I 1979 me ing between the staff an he B&W licensees, Ref 37) and the Oconee 3
~
l and rystal River 3 loss o~ N /ICS power supply eve s of November 10, 1979 nd February 26, 1980, re ectively.
l l
, Subsequent to the I-2 accident, the N and the licensees took number of 1 -
cctions on th W plants in the ar of control systems. I an attempt to avoid cha enges to the PORV, t si.aff issued IE Bulle 79-05B requiring
/
/
i 5-57 e
, , _ _ , , --.c-- -
w----w-g-- ---
the licensees to r se the setpoint of contro ystem actuation of the P0 from 2255 'g to 2450 psig. The reac protection system trip po for hig ressure was simultaneous 1 owered from 2355 to 2300 ps . In conjunction 4
- with these changes, it w also proposed by the licens (and confirmed by Commission Order) at hard-wired, anticipatory r ctor trips would be i alled for turbin rip and loss of feedwater (as nsed by the loss of b pumps).
Thes rips were implemented in the art term as part of t plant control system (hence, the term contr grade) and are to be lemented in the long term as part of the rea . r protection system .e., safety grade).
/ /
The THI-2-related concerns with the ICS were addressed in a number of ways.
j j The short-term portion of the Orders required that the plants have procedures to initiate and control the auxiliary feedwater flow independent of the ICS.
To determine the (potential) contributions of the ICS in plant upsets, B&W proposed to perform a reliability analysis including a failure mode and effects analysis (FMEA) of the ICS. This was confirmed by the long-term portion of the Commission Orders.
Simul neously, the Lesson earned Task Forc ade related reco dat' ions in .
he area of auxiliar feedwater automat initiation. Requi ment 2.1.7.a of NUREG-057P,(Re . 13) required con ol grade automatic ~ itiation and in i long ter required that thi actuation system b pgraded to a saf system.
T implementation o his requirement on plants would ef ctively remove this function om the ICS, since t CS is not a saf system.
T staff's review of th CS Reliability lysis (BAW-1564) ef. 35) was -
initiated with the id of consultant rom Oak Ridge Nat nal Laboratory. e
/
5-58
-.-e , -- - - + ,--
- ,-- , ---,.y- , , .,,-.-.m m.- . - . - .
- - -y . _ _ - . --m -------,.w.- - * - - - -- - - - .
1 i
be indi ed in the con ol room at all ti s but could be c led up on de d from the co ter.
l (4)
Conclusion:
'S' The operating history of B&W plants has shown that as a result of opera-tional transients, such as loss of feedwater events or overcooling transients ,
I that lead to hign pressure injection actuation, a high probability exists i l
that reactor coolant inventory may be released from the system through the pressurizer PORV or safety valves. During the course of this event, l
radioactive gases, which collect in the top of the pressurizer, would be l ._
j expelled. If during this transient a containment vent a'nd purge operation were in progress, it would be important to isolate containmsnt as quickly as possible to minimize the release of radioactivity to the environment.
Although all plants provide a safety grade high containment pressure isolation signal to perform this function, this signal may prove inadequate',
during small or intermediate releases of coolant to the containment building, because a buildup of containment pressure to the actuation setpoint may not occur with the valves open. It appears that the most .
effective way of providing containment atmospheric isolation, as required, would be through the use of containment high radiation signals.
Recommendation:
Provide safety grade containment high radiation signals to initiate contain-
! ment vent and purge isolation in addition to the presently required signals (i.e., containment high pressure and low pressure ESFAS actuation).
5-66
_ _ _ _ . _ _ ,- - - . - m_---.,-w- --
. , . - - - - - - , _ - , , - . . - . . - , . . , , ,_-4, ,.-.6-- .~,.w ,,. mw , ,. ,.-,g , . ,,,--,-- ,w . - . . - -
/ _
At the r sent time the is only one rational simul or for each o the B&W d CE vendor de igns. The B&W imulator is loc ed in Lynchbur , Virginia d i, represent ive of the Ran o Seco control r om. The CE si ulator is located at Wi dsor Locks, Con cticut, and is r presentative the Calver Cliffs co rol room. Ther are two boiling ater reactor 4R) simulat s (Dresd n 2 and Brown's erry) and five W tinghouse simu ators (Zion Indian Po t, Surry, McGui , and Sequoyah).
The simulator training that B&W operators receive is unique in that there is much more uniformity in plant systems, instrumentation, and operation than any other vendor design. The single exception is Davis-Besse 1 which has the raised-loop design and low-head high pressure injection pumps. Oderwise, the primary reactor coolant system (RCS), once-through steam generator (GTSG), ICS, emergency core cooling systems (ECCS), and plant systems including the makeup and purification system are very much the same from plant to plant. The trainees therefore do not encounter the degree of unfamiliarity that sometimes occurs in training conducted on other vendor-type simulators. The staff believes that this is a distinct advantage in the training of B&W operators. Also, the startup certifications of operators at B&W plants must be conducted at the B&W simulator ,
(requirements for GE are similar). However, allowance has been made in the past for some Westinghouse facility operators to obtain startup certifications on the CE simulator and vice versa. ,
I-The disadvantages of the B&W simulator training are (1) age and fidelity of the simulator, and (2) counter productive simulation for Davis-Besse operators.
Because the B&W simulator was one of the lirst of 1.ts kind, there is a distinct lack of fidelity in some areas. For example, modifications had to be made to 5-69
simulate both the TMI-2 and Crystal River 3 events. Two phase conditions in areas of the RCS other than the pressurizer and multiple failures of instrumen-tation were not part of the computational model.
i i The " feed and eed" method of core ooling for spme smal break LOCAs and loss of feedwate events is not appli able for the Davis-B se facility. Also, t auxilia feedwater and stea generator level con o1 system are distinc y diff ent from the simul or. These differenc s are mentioned becau the ystal River 3 oper ors credit the simul or training on solid ystem opera-
- tion and natural irculation as being v ry beneficial in the response to the loss of NNI e nt on February 26, 1 0.
5.4.2 Assessment of Licens Operators at B&W F 111 ties i
- 5.4.2.1 Discussion
/
The logical m had of ass'essing t role o'f B&W opera rs with regard'to t Ir ability t cope with transien events, particular secondary side upse , is to ma a comparison with perators of other ndo'-type' r reactors. Unfortunately, no olid data base ex ts for such a comp ison. The only app cable data that
, exist for compari n are the Licensee vent Reports (LERs) ttributed to licensed pers nel error sorted by endor types. One st be careful in taching a great d 1 of significance t these data. These ERs have enly bee categorized by li nsed personnel erro since January 1978 nd different crit ia have been ed in differentia ng such events. F example, a Westi ouse four-loop plant has raported e most (24) licen d personnel errors from 1978 to the present with a ther Westinghouse ant and a BWR clos behind (21 each).
/ 5-70
these pr cedures addres d the loss of NNI/IC power. At the ime of the I l Crys al River event, rily Rancho Seco had rocedures that ddressed NNI/ICS '
wer failures a the resulting impa on the plant. The Task Force b ieves !
that each B& facility should imp ment procedure concerning the rtial or
)
l total los of NNI/ICS power.
Twenty-one minutes after the event began, power was restored to the NNI "X" l bus and plant recovery began in accordance with the applicable procedures. I The delay in restoring power was attributed to the I&C technician first being l directed to the wrong power supply due to an ambiguous annunciator. It was not until power was restored that the operators were able to throttle HPI and stop the coolant release from the pressurizer safety valve. Had this event
! occurred on a back shift without I&C coverage, it is unknown how long the event would have continued before power to the lost instrumentation would have been l regained and the event terminated. Since B&W facilities are uniquely susceptible j to this type of failure, the Task Force is recommending that round-the-clock coverage by qualified I&C technicians be provided at all operating B&W reactors.
t .
Throu out the first alf-hour of th event, the operati crew was very aw e of ubcooling an the importance f maintaining an a quate margin to s rated l -
1 conditions. ey followed t ir training and pr edural guidance fo HPI termi-nation cr' eria and RCP ip requirements. en NNI power was r stored, they
) were erating under olid system pressur control. In this condition, the l
R is completely ull of water with o steam bubble in e pressurizer nd pressure is c trolled with make and letdown. Th rystal River erators ,
)
l l
l l
! l 5-78 l
,, . m. .. . . ., .
e,- , , , . , - -----,, ,- -
,-n., . - - - - - - ,- -- .
l i
A l (4)
Conclusion:
l There have oeen 24 ins'tances of loss of NNI/ICS. power supplies at B&W
)
facilities while the reactor was critical or at power (see Appendix B, i Table 8.1). Twenty-two of those resulted in a reactor trip and four events in HPI actuation. The longest period of time before power was restored is believed to be 21 minutes. Had a qualified I&C technician not been j
l l
available, it is unknown how long this condition would have continued.
Recommendation:
l l
The Task Force recommends that qualified I&C coverage be provided on a round-the-clock basis at all operating B&W reactors.
1 l .
I e
5-83
' - / .. .,
(2) Th identificatio of common cause f lure mechanisms important to ndividual pla systems or to gr ps of systems. he potential aflure j mechar. isms nsidered include, or example, hu n interactions' and common support s stems such as se ce water, comp ent cooling wa er, end elect c power.
/ /
Because of the need for timely results in the IREP study, some particular aspects of the plant either have not been included in the Crystal River study or have not been thoroughly investigated. These include:
(1) He cen:fderatf- h2: beer g?fer te the pctenti:1 effect: f extern:1 c;;nt:
sudi a= caiuiyuc.kos, fleeds, r.earby hazardeue c.eterial related 3 :fdent: ,
et:., :nd (2) The details of control systems and plant instrumentation (such'as non-nuclear instrumentation and the integrated control system) and fault propagation through these systems has been treated in anly a qualitative manner.
s /
6.2 Th IREP Crystal fver 3 Study ..
6 .1 Discuss n As noted bove, the initial plant selected f evaluation in he IREP study has b n the Crystal Riv r Unit 3 Nuclear lant. Complet on of this work s pr sently expected in ay 1980. Based n the prelimin ry results of t s study, it has been shown at:
/
4 6-2 [
(1) The c e melt probabili and public risk sociated with the rystal River I ant are dominated y transient-initi ed accidents; (2) The total ss of all feedwate (main and auxili feedwater) is a ghly signif ant accident; and
/
(3) The plant's auxili cooling water stems couple many the plant's i
/ engineered saf y features syste l
l
, indicating a pot tial highly sign -
icant comm cause failure e hanism.
i
/
l Because of the Ifmitations in IREP discussed in Section 6.7, certafn issues I
have not been resolved or failure mechanisms identified. These include:
(1) The common cause failure potential resulting from ICS failures and inter-actions has not been quantitatively determined; and 1
(2) The specific common cause failure mechanism exhibited in the February 26, t
1980 event at Crystal River (loss of the 24 Volt de power supply to NNI
~
"X") was not identified prior to the incident. '
6.2.2 Conclusions and Recommendations (1)
Conclusion:
l The IREP Crystal River study has tentatively concluded that transient-l induced accidents are highly significant contributors to the likelihood l
l of core meltdown in the Crystal River plant. Implicit in this conclusion 6-3
- +
is the determination that support systems to vital equipment (e.g., cooling .
water, ac and dc power) couple this equipment to a degree that common cause failure potentials are significant. ,
Common cause failure! modes resulting from NNI/ICS faults have been qualita-tively identi,ied, f but the probabilities of such common-cause failures have not been calcul ted. We agree with the IREP staff judgement that a rigorous quantification of these probabilities would require a large-scale effort, exceeding the constraints placed upon the IREP study.
f Recommendation:
l l
- The Task For e recommends that e IREP Crystal Ri' r study be complet i
! and thoro hly documented in n expeditious man r, with the results provided j to appr priate parts of t agency in a clear manner. At that ti , we reco end the action 1 ted below:
(a) The Probabi stic Analysis Staff consider the need or additional
/ Crystal ver work to examine particular unresol ed questions w ch ,
may b evident, and reexam e the scope, met ds, and format f the 1 ,
fi st IREP study so tha modifications may e made prior' the initiation of further IREP work; and (b) Appropriate sta within the Offi e of Nuclear Re ctor Regulatio (in coordina on with the Prob ilistic Analys s Staff) make empt I
i determina ons with respec to the need for modifications the Crysta River p ant.
6-4
t ,
transient- uced LOCA. 'The dif' rence between B&W and o r designs is r confin to the case of del ed auxiliary feedwater rts. Prompt A starts do not caus undercooling transient . Outright (sustaine failure to start i equally serious with o without responsive s am generators. s, B&W plants place a remiun upon the reli lity with which the uxiliary feedwater st ts are properly time The penalty for te starts is an incr ed likelihood of tra ent-induced LOC .
s -
,The most prominant common-cause failure mechanism we can identify that I
v) causes both delayed auxiliary feedwater starts and sustained ECCS failures lies in operator ermr. A practice of trying to avoid over-cooling incidents tends to make such errors more likely. On the other hand, the experience of having had a THI accident, the operator retraining it spawned, and the other changes made since the accident have gone a long way to reduce the likelihood that such scenarios muld start or would progress to core damage once started. Nevertheless, our event tree-fault tree studies suggest that transient induced LOCA which cannot be ,
isolated and which occurs in conjunction with ECCS failure may be among -
the dominant mutes to core damage, i.e., to an accident, although we think it very unlikely that such a scenario would also entail the l
i failure of containment fan coolers as well as sprays. Thus, transient-induced LOCAs should not be prominant causes of severe accidents.
, s It is own that B&W plants ave somewhat more requent trips than er PWRs, particular since the THI-i f red alterations the trip ,
setpoints. Thes excess trips seem be originating f minor secondpr 7-8 1
.. .. ~ 4. c ,$
_o.. .. _ . . . .
i work to ustain core cool g. However, ECCS wil have higher-than-no reliability un r these conditions ause its successfu start aused the LOCA i the first place. re is no reason to lieve that e or with the such incident are likely to be co led with ECCS fail failure o containment fan co ers or sprays.
It has been suggested that a reactor trip together with a failure to Q throttle main feedwater in a B&W plant muld rapidly fill the OTSG's and result in water in the main steam lines. No such instances have occurred but comparable upsets in the Integrated Control System have been observed.
The main steam lines and valves may not be qualified for t'he weight or the water-hammer potential associated with this scenario; they might rupture. The characteristic range of times to (111 the steam generato. i and main steam lines is a very few minutes, perhaps too rapid to give much confidence that the operators would consistently trip the feedwater: ,
pumps or stop valves in time to avoid main steam ifne brea'ks.
Such scenarios wuld affect the risk of' severe accidents only if the break produced flooding that defeats support systems for essentially all of the active engineered safety features, i.e., essential DC power, AC l power, or possibly essential auxiliary cooling water systems, and do so with a probability that rivals station blackout or Event V. Such scenarios would have a significant effect on the likelihood of core damage only if the flooding defeats energency feedwater and HPI (feed and bleed cooling) and does so with a probability that rivals other commn-cause or multi-fault scenarios such as loss of all feedwater and HPI failure.
7-12
. s
, - , , . . - , - - _..----%.- - - - - ---,.-,----,y--, . . - ,-y, , , _ . , - . _ . , , , . , _ - -
9 - ,-7,-- - . w--. > p ,.y,, , ,-
E In either the case of accidents or severe accidents, the significance of the water-solid main steam line break scenarios seems to rest upon the potential for massive flood damage in essential compartments of the auxiliary building. If such flooding does not take place, there appears to be little direct threat to ultimate core cooling or containment integrity.
The susceptibility of B&W plants to loss of all essential dC or DC power or loss of all HPI and EFW due to water-solid main steam line breaks and subsequent flooding should be reviewed. If a deterministic analysis suggests a real possibility of such a scenario, then a probabilistic evaluation should be perfonmed.
l These consid ations of B&W p nt characterist s are summarized i Table 7. . We conclude t t B&W plants ar not significantly fferent from her PWRs in th r vulnerability r susceptibility severe a idents - thos hat dominate th nuclear risk. '
B&W plants ave a different file of suscept lity to core d . age accide s than do other P s. They are mo likely to inc transient-in ced LOCA but the nes with high he HPI pumps may less likely incur core damag from a loss of a feedwater. B plants are e likely than er PWRs to ha'v over- or underc ling incident , transient- ,
i induced CA, etc.
e 7-13 y ..
I is coupl with Task Forc recommendation 6 (Saf y-Grade Vital Instr'..entPanel). B ause the latter recor:nendation calls for th p ision of sever indications of RCS s s, we believe that t overshadows the tential benefit res ing from the improv use and display the themocouple in cation. Thus, while feel that bett r use and display of e themocouple indi tion muld be a desi able capability, we lieve that the ins ation of the
's ety panel" is distin ly more important, n this context, is
, recommendation appea to be of low import ce for incident nd accident mitigati and of negligible ..portance for sev accidents. ,
- 8. Safety-Grade Vent /Purae Isolation on a High Radiation Sicnal This Task Force recommendation calls for the installation of safety-grade isolation equipment on the reactor building vent / purge systes which muld be actuated on high radiation levels in the reactor building. This is of concern because, for some events, isolation of the vent / purge systeri on high butiding pressure or low RCS pressure might not occur until after the release of some radio 3ctive ,
ma terial . For example, for a total loss of feedwater accident (i.e., both main and auxiliary feedwater fail), RCS pressures muld climb rather than drop sufficiently to cause the building isolation on low RCS pressure. Further, the operation of the purge might prevent building pressures from reaching the other isolation setpoint; thus, automatic isolation might not occur. Under such circumstances, operator actions to isolate the vent / purge system might not occur until some material (e.g., radioactive gases released from the 7-36
, . - - ~. ~ w - - , - - - - - - - p- ,- - - , , - -,,-
I expelled coolant) has escaped through the system. To cope with such a situation, a vent / purge system isolation on high radiation level in the reactor butiding has been recomended.
In essence, the intent of this recommendation is to rubstitute automatic isolations (on high radiation) for operator-initiated isolations for that class of accidents where the " normal" isolation-initiating signals would not be received. The consequences of not providing such an isolation can be thought of as the differenc? in the magnitude of release if an automatic isolation vere to occur and if the isolation were dependent on operator action. Since the concentration of radioactive material in coolant is relatively low, we believe that the increased time required for human actuation of the vent / purge system isolation muld result in only a small difference in the radioactive ' release. For this reason we believe that this recommendation is of negligible value with respect to severe acci-dents, and low value for accidents. We also believe, however, that it could be important (in the severe accident category) to assure .
that these valves fail closed on loss of power, so that isolation occurs in the event of such potentially severe accidents as station blackout. .
We note that the above conclusions on the relative merit of this recommendation are based on the conclusion that small releases of radioactive material during an incident will result in negligible health effects within the surruunding public. If, however, the 7-37
-th -
E objective is to prevent any release of radioactive material, this recommendation clearly is more desirable; for this reason we believe .
it is of moderate value with respect to coping with incidents. We also note that an anticipatory trip of the containment purge isolation valves could also be triggered on high pressure in the reactor coolant drain tank.
- 9. System ksponse Modifidtions to Prevent P/essurizer Level Loss /nd ECCS tuation F owing a reac r trip in a B&W p nt, the reactor coo nt undergoes significant <ntraction as it co s; as a result, the ressurize~r ,
level and CS pressure drop bstantially. To co with this, opera are trained to ickly isolate letdo flow and start an add fonal make-up (HP pump, so that shri age is accounted for additional coola t injection into th RCS. Even with such operator intery tion, however, the plants have a history f occasional s ondary side malfun ions leading to react t'ips, r
losses of ressurizer level, nd ECCS/HPI actuations on low RCS 1 .
pressu e). This Task Fo recommendation calls r the examinati of eans to reduce the severity of the post-tr p RCS transient, hat the frequency f level loss and HPI a uation is reduc 1
A reduction i the frequency with whi pressurizer le 1 is lost and/or EC is actuated in overcoo in' g accidents i useful in
- severa ways. Frequent ECCS a uations due to ercooling transi ts may condition operators to pect all ECCS tuations to be s rious nd encourage them to di able the autost t of emergency f dwater Y f l 7-38
. . . ~ .-
I
- 13. Doerator Trainino on the Crystal River Incident
- 14. Development of Plant-Specific procedures for Loss of ICS/NNI l
We have chosen to consolidate Task Force recornendations 13 and 14 l4 into one for the purposes of this risk evaluation because of their similarity in intent. Reco:=endation 13 of the Task Force calls for specific operator training on the events of the February 26, 1980 incident at Crystal River. Recorrendation 14 addresses the need for plant-specific precedures to assist operating crews when ICS/NNI failures occur in the future.
We believe that the reduction in the likelihood of operator errors dtuing ICS/NNI-caused transients requires operator training involving both retrospective and forward-thinking views. The Task Force's recorcendation on Crystal River training provides one aspect of the retrospective training; however, this specific training alone does pose questions regarding the need for training on other similar .
events, e.g., the Rancho Seco " light bulb" incident or others identified from LERs as having the potentiai to be accident pre-curso rs. We believe that this type of training could be highly valuable in " preparing" the operators for possible future accidents..
The Task Force recocnendation on plant-specific procedure development addresses the need for forward-thinking training. Since it is a virtual certainty that operators will be faced with ICS/NNI failures l in the future (which may be similar to or different from past '
7-43
.i events), we believe it important that more general training on coping with such even'ts be provided. -
We believe that this combination of training for ICS/NNI faults can
'be of relatively high effectiveness for this type of transient.
Other recommendations reduce the significance of these incidents, I
e.g., recommendations 2 and 6. We believe that en an overall basis, these recommendations are of high value for incidents, high value for accidents, and moderate value for severe accidents.
, . 15. Increased Simu r Traininr1 i .
This Task F rce recommendation lls for the requir ent of a one week pe year simulator tra ing course for all perators in B&W pla s (this training i nowoptional).
I We believe that s reew.endation h both positive and egative
/ aspects. On e positive side, s simulator traint can be ;
in.po rtan to the understandin of plant behavior d ing transient even . LCCAs, etc., and us be a useful mean to reduce the .
elihood of operator error during real e nts(e.g.,brys 1
~
River type "incid s" and THI-2 type " ccidents"). We elieve that making s training mandato , rather than op onal, is of moderate v ue for incidents, derate value fo accidents, and l
negli le value for sever accidents. ,
e negative aspects f this recommen tion result fro our concern -
about the limita .ons of the avai ble simulator c ability.
First, the B simulator is m e to resemble t Rancho Seco ntrol
/
7-44 y .
,4 -
,,-,--,-_,._-,,,---,,-,,.-......_-,'O---,-n.
e TA8LE 5.1 HISTORICAL St# MARY OF MI/ICS POWER FAILURES Power Initial Source
- Date Plant Probles Feeduster Reactor PORV ESF Plant of Cause failure Transient Trip Opened Actuation Status Info.
A.
NN1/ICS Power Failures Which Resulted in Reacter Trio
- 1. 06 DEC 74 Arkansas 1 Lost MI-Y power Water leakage into 8 NNI Yes Hi Press Yes supply & condenser Main Chiller load 801 4, 10 vacuum peps controller
- 2. 8 JUL 76 Arkansas 1 Loss of NNI Power, Melntenance Personnel NNI trip on RC HI Yes Hi Press Yes 94% 1, 4, 10 Pressure
?' 3. 02 MR 77 Crystal
" loss of ICS Power Inverter Diode Failure River 3 (RPS
- 4. 26 FEB 80 Crystal Loss of MI X 24 +24 VDC NN1-X bus River 3 PORY opened (2P5 - shorted causing the NNI Yes Hi Press yes E5-HPI 995 1, 7 did trip) stuck power supply to open open 51 and 52 discon- .
necting *24 VDC free VAC feed
- 5. 12 JAN 79 l Davis-Besse 1 Failure of Inverter Accidental grounding of NNI Yes Hi Press yes to vital power containment hydrogen 1005 1, 8, 9, 10 source (reactor l T
- analyzer causing trip and turbine inverter fuse failure, j
Q trip) $FRC5 trip O ,
y " Sources of information: Listed on page 8-8.
O l
I W
3:=
P .
a I
TABLE 8.1 (Continued) .
t
) '
Date Plant Power Feedwater Reactor Initial Source Problem Cause PORV E5F Plant of Failure Translent Trip Opened Actuation Status Info.
j A.
1 NN1/ICS Power Failures Which Resulted in Reactor Trio
- 6. 14 JUL 76 Oconee 1 ICS power was Power receptacle incor- ICS lost (reactor rectly wired into ICS Hi Press yes 1005 1, 5, 10 i- , tripped) Power Source. When High
] Voltage power supply was
- plugged in, line circuit 1
breaker tripped
', 7. 14 DEC 78 Oconee 1 l Lost power to Teve Shorted while trouble- NNI Ves recorder Press-Temp ES-HPI
" shooting alare Indication 985 5, 8 4, 8, 25 DEC78 Oconee 1 j Loss of all Power Inverter Fuse Failure ICS Yes Hi Press yes (ac) (RPS - did trip)
IDE 1, 5, 10 j 9. 11 JUL 74 Oconee II Lost ICS autopower Melntenace Personnel l ICS Low Press yes ES-HPI 805 5 i
- 10. 23 SEP 74 stuck open Oconee 11 Loss of Pover to t
Inverter Fatiture ICS Yes Hi Press yes 1 ICS Power Panel 954 1, 5, 10 1l Board 2K1. (Reactor did trfp) i
.I ICS. Manual VAC non-1E panel Yes Hi Press No No; Hi CDR 995 1, 5, 8
- i T transfer to backup (inverter fuse failure) power required.
- . Q I
After a loss of
,! P _
trip) t O
l N
! 25 1 :x=ss l V"~* -
]
- 4 t-
I i
7A8LE d.1 (Continued.'
l Initial Source Power Feedwater iteactor PORV ESF Plant of Data Plant Probles cause Failure Transic.t Trip Opened Actuation status Info.
j A. ' MI/ICS Power Fallures Which Resulted in Reactor Trio
- 12. 22 MOV 74 Rancho Seco Temporary loss of Due to Loss of "J" MI Yes Hi Press Yes 32Y 1, 6, 10 power to M I "Y" inverter (fuses blew) and "Z" busses
! (Reactor did trip)
, Inverter "C" failure.
- Trtpped. Lost 1 Power to ICS
[ i4. 31 DEC 74 Rancho Seco Lost temporarily Inverter Fuse blev. ICS Press-teep yes ' 400 1, 6, 10 power to "J" inverter, and
! ICS/MI power j (Reactor tripped)'
- 15. 28 DEC 7% Rancho Se,co Loss Power to ICS Unknown. ICS Yes Yes Yes 7 405 1, 6, 10 bus "X" (Reactor l tripped).
i ll 16. 16 APR 75 Rancho Seco Lost of power to unknown. . ICS input Yes 7 35% 1, 6 i
Inverter "B" Vital l T Power Bus.
- 17. 20 MAR 78 Rancho Seco Lost power to M I Dropped Light Bulb MI Yes Hi Press yes & H1 CDR 725 1, 6, 8 l C "Y" power supply safety 1
y (Reactor did trip). Ilfted
! lpt:s
! C l 2 r-- -
l
- v. ,
4 :.
i TABLE 8.1 (Continu'ed) 1 Data Power Initial Source Plant Probles Feedwater Reactor PORV ESF Plant of Cause Fa11ure Transient Trip 1
Opened Actuation Status Info.
A.
leII/ICS Power Failures Wich Resulted in Reactor Trio i 18. 2 JAN 79 Rancho Seco Loss of NNI Power Technictan inadvertently ICS input Yes Hi Press yes (RPS - dId trip) shorted NNI power supply. 1005 1, 6, 10 19 05 JAN 79 Rancho Seco Loss of one ICS j Electrical short by ICS Yes Hi Press yes Power Supply Technician Hi COR 1005 1, 6, 8, 10 20 22 APR 79 Rancho Seco Loss of RC flow
$fgnal to ICS A Inverter trip ICS input Yes Hi Press No i 1005 1, 6, 8, 10 8.
IMI/ICS Power Failures Which Old Not Result in a Reactor Trio
} 1. 18 NOV 77 Arkansas 1 Loss of ICS Power Inadvertent shorting ~ ICS No
, Supply by Shorting of ICS power supply No 1, 4 .
l of ICS Power by Technician.
Supply. RPS did not trip - Reactor l ' Level reduced to 605 I and then returned to normal level.
- 2. 21 APR 77 4
! Crystal River X-Power Supply to slew Fuse in vital ICS Yes No ICS was lost, SG Sus 8". No 465 1, 7, 10 level went to SOE Turbine bypass and Atmos. dump valves opened, lost MFW.
3 27 NOV 79 D Oconee III ICS inverter Component failure ICS No failure No No No 995 1, 5 Q (defective inverter Logic cards in 3K1)
C:3
~
- t:::
h r---
w ome -
w t> l"i
l . .,
i I
j
' TAaLE 8.1 (Continued) .
4 Initial Source Feedwater Reactor PGAV ESF Plant of Power Info.
Cause Fatture Transtant Trip Opened Actuation Status
- Date Plant Problea C. NNI/ICS Power Failures Which occuraJ Durina Testina or Refuelina Outanes 1
- 1. 29 0CT 75 Davis Besse 1 Fallure of Power Defective Component Supply Monttor Module BCC0
- 6625070A1 1
- 2. 11 MAR 76 Davis Besse 1 Fallure of NNI Defective Component Power Supply .
Monitor Module
' BCC0 #6625070A1 1
- 3. 24 MAY 76 Davis Besse 1 Failure of MMI Defective Component i
i Power Supply Monitor Module BOC0
- 6625070A1 1
- 4. 11 FEB 76 Davis Besse 1 Failure of ICS Defective Component Power Supply Monttor Module BCC0
- 6625070A1 j 1
- 5. 11 NOV 76 Davis Besse 1 Summary of Power Suemary of Failure Supply Monitor Module Fallures 1, 5 O 6. 3 MAY 79 Oconee III ICS hand power Component failure (pressurtzer WR level ICS NA N/A N/A No Cold shutdown breaker trip. -
due to defective recorder)
% recorder cord and plug assembly
-g
>r- .
1 i I
I 1
i
}
i 4 TABLE B.1 (Continued)
)
j j Date Plant Problea Power Feedwater Reactor PORV
- Initial Source 1 Cause E5F Plant of Failure Transient Trip Opened Actuation Status 4
i C. Info.
{;
NN!/ICS 7.
Power Failures Which Occurred Durina Testino or Refuelino Outanes 29 MAR 73 TMI-2 Loss of vital power Inverter fuse fatture MI i
to MI X bus, PORY Yes Yes Yes 0
) opened (RPS - did stuck 1. 8 trip) open I
i '
i 1
4 I as i
i t
l -
\
1 -
I I
@ TE: For sources of information see next page.
O ll23 C=3 '
i lllC
- i. as '
I i l3>
i u ...,
i j
\ .
t l
i.
' l Table 8.1 (Continued)
Sources of Information
- 1. Letter from J. H. Taylor (84W) to D. G. Eisenhut (NRC),
Subject:
Responses to Questions of March 4,1980 concerning incident at CR-3, dated March 12, i
1980.
- 2. Letter from J. G. Herbeln (Met-Ed) to H. R. Denton (NRC),
Subject:
Responses to NRC letter of March 6, 1980 concerning incident at CR-1, dated March 13,1980.
- 3. Letter frc,e R. P. Crouse (TECO) to R. W. Reid (NRC),
Subject:
Responses to NRC letter of March 6, 1980 concerning incident at CR-3, dated March 13, 1980, i 4. Letter fore C. L. Steel (AP&L) to H. R. Denton (NRC),
Subject:
Responses to NRC letter of March 6,1980 concern 13q incident at CR-3, dated March 12 1980.
J
- 5. Letter from W. O. Parker, Jr. (Duke) to H. R. Denton (NRC),
Subject:
Response to NRC 1etter of March 6,1980 concerning incident at CR-3, dated i March 12, 1980.
T' 6. Letter from W. C. Walbridae ($ MUD) to H. R. Denton (NRC),
Subject:
Response to NRC letter of March 6,1980 concerning incident at CR-3, dated March 12
' " 1980.
- 7. Letter from J. A. Hancock (FPC) to H. R. Denton (NRC), Subject Response to NRC letter of March 6,1980 concerning incident at CR-3, dated March 12, 1980.
- 8. Licensee Event Report File
- 9. U. 5. Nuclear Regulatory Commission, " Operating Units Status Report " NUREG-0020 (Printed Monthly - commonly referred to as the " Gray Book").
I
. 10. Letter from G. C. Moore (FPC) to R. W. Reid (NRC),
Subject:
Information concerning the lift frequency of the PORV and safety v'alves, dated November 15, 1 1979.
l 7 T l C C"3 C C"3
- ::A :::c i C C"3 .
e :::C 2 CT3 .
2 g .
( 2 2:=,
I r r- ...,
I
UN87ED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE ATOMIC SAFETY AND LICENSING BOARD [
r In the Matter of )
)
METROPOLITAN EDISON COMPANY, ET AL.~~
) Docket No. 50-289
) (RESTART)
(Three Mile Island Nuclear )
Station, Unit No. 1) )
CERTIFICATE OF SERVICE
~
" ~'I hereby certify that single copies of INTERVENOR STEVEN C,.
l )
SHOLLY MOTION TO THE ATOMIC SAFETY AND LICENSING BOARD TO TAKE l OFFICIAL NOTICE OF CERTAIN PORTIONS OF NUREG-0667 were served on the following persons by deposit in the U.S. mail, postage paid, first I class, on this 29th of April, 1981.
I Steven C. Sholly G/
l Ivan W. Smith, Esquire George F. Trowbridge, Esquire I
Administrative Judge Shaw, Pittman, Potts & Trowbridge Atomic Safety and Licensing Board 1800 M Street, N.W.
l I U.S. Nuclear Regulatory Commission Washington, D.C. 20036 Washington, D.C. 20555 Robert W. Adler, Esquire Dr. Walter H. Jordan Attorney for the Commonwealth Administrative Judge 505 Executive House I Atomic Safety and Licensing Board P.O. Box 2357 881 West Outer Drive Harrisburg, PA 17120 Oak Ridge, TN 37830 Jordan D. Cunningham, Esquire Dr. Linda W. Little 2320 North Second Street Administrative Judge Harrisburg, PA 17110 l Atomic Safety and Licensing Board l 5000 Hermitage Drive Louise Bradford i Raleigh, NC 27612 TMI Alert '
315 Peffer Street
- Docketing and Service Section Harrisburg, PA 17102 U.S. Nuclear Regulatory Commission ;
Washington, D.C. 20555 Cail Bradford ANGRY James R. Tourtellotte, Esquire 245 West Philadelphia Street i Office of the Executive Legal York, PA 17404
- Director U.S. Nuclear Regulatory Commission Marjorie Aa M t I Washington, D.C. 20555 R.D. 85 Coatesville, PA 19.320 l Marvin I. Lewis <
6504 Bradford Terrace Dr. Judith Johnsrud Philadelphia, PA 19149 ECNP )
. 433 Orlando Avenue , i Robert O. Pollard State College, PA 16801 609 Montpelier Street Baltimore, MD 21218
- Ellyn R. Weiss, Esquire l Harmon and Weiss 1725 I Street, N.W.
- Mr. George Jett Suite 506 i ,
Ceneral Counsel Washington, D.C. 20006 1
- I '"*9'**"
- INDICATES HAND DELIVERY
, y j ATT!H Docket Clerk i
l !
P00R umumis l
t u4 l
~
b' 1
-_ .. --