ML19224C198
| ML19224C198 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 05/17/1979 |
| From: | Williams F Office of Nuclear Reactor Regulation |
| To: | Eisenhut P Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 7906290348 | |
| Download: ML19224C198 (11) | |
Text
Distribution a
paen, ri3e c
NRR Reading I\\
N[t F. Williams Reading May 17, 1979 MEMOPANDUM FOR: Darrell G. Eisenhut, Deputy Director, Diviston of Operating Reactors T11RU:
Steven A. Varga, Chief, Light Water Reacters Branch No. 4, Division of Project Management FROM:
F. J. Williams, Jr., Technical CooMinster, Division of Project Management
SUBJECT:
CONCERNS RELATED TO TMI-2 EVENT AND BULLETIN 79-05A -
DONQUICK(IE: REGIGH II)
On May 2 and 3,1979, Don Quick of Region II made a verbal presentation to the S. Ysrga review group and other staff members (attendance list enclosed). Based on 'ais own experience and discussions with other IE personnel he provided his assessment of general problems with operations and maintenance and with his perception of problems with the B&W design.
In many instances he cited what he felt were specific design deficiencies and in other areas he described broader problems (e.g., waste disposal area and control boards) applicable to many or all designs.
. provides my writeuo of the presentation.
It has been reviewed by Don Quick and his corrections and additions have been incorporated.
~
No assessment has been made of the concerns expressed. Some have been presented in other forus. Most of them, in my. opinion, represent long range consideration.
Some of the cited design deficiencies could bear on the current considerations regarding the B&W operating plants.
CEIGl?UL SIU C 3 F. J. Williams, Jr., Technical Coordinator Division of Project Managenent
Enclosures:
As stated cc:
H. Denton R. Tedesco E. Case S. Varca P. Mattson F. Willia s 7 00629 0 3/ E, V. Stello Attendance List n
t' e e -
I s. sanauer m -,
jpfag
....c a.
b Yefis/
NI3 Y
~5lM 9 5/ I lj7i 5
NRC FORM 318 (9 *6) NROI 0240 W w. s. aovsawne amr emuriaea orncas s ere - eawa4
/
/
.)
ENCLOSURE 1 COMMENTS AND CONCERNS EXPRESSED BY DON OUI E General Operators - The current situation with regard to operator capabilities is not good. Coerators had a keener sense of awareness 15-20 years ago.
Parameters were closely monitored to keep things on track - there was more concern as to potential problems when things were not on track and actual plant status was better known. The basic philosophy developed over the years is that plants are designed so that they will return into some stable condition following a transient.
Operators have been convinced of this philosophy through training approach and have become overconfident.
Many operators cannot explain effects of jumpers, clearances etc. on system response. We have contributed to the operator problem by our approach of extensive QA, che:k lists. At a result they are further con-vinced that nothing can go wrong if procedures are followed.
In reclity many emergency procedu es are inadequate.
In addition, we provide edicts without bases and without participation on the part of the operators. All of the above have resulted in a complacent attitude which applies to maintenance as well as operators.
In addition - the training has been deficient.
It follows along the classic accident lines and does not prepare the operator for the unexpected.
S'aintenance - The attitude problems discussed above for operators als opply to maintenance.
There are many instances of equipment being out of service too long -
botn safety related ard ron-safety related ccmcenents - both in violation of Tech.
Soecs. and because Tecn. Specs. are too lenient.
Examoles of croblems - (a)
ISI referencesSection XI which gives 96 neurs to interpret test resuits - this can be adced to the 72 neurs permissible cut of service to give a to'.ai of a week -
263 254
. (b) the Tech. Specs. give 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> and some utilities utilize the full 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> rather than going full out to return the eqJipment to service as soon as possible (overtime vs. dayshift consideration).
Imorovement - No easy solution is seen for the operator problem. The following items were presented as tending to alleviate the problem.
Improve training (depth and scope for operators and maintenance personnel).
- Attempt to obtain more operational input for control board and system design.
Improve selection of parameters presented to operators and manner of display-ing the information.
- Encourage utility management to attempt to improve such things as; morale, working conditions, attitude toward both operations and maintenance, communi-cation, and plant status awareness.
Details of Concerns (3&W Plants)
Heat Sink - The OTSG causes rapid effects on parameters during transients or trios e.g., pressurizer level changes are mag'iified vs. W or CE design. There is concern that this plant was licensed with this size pressurizer since many transients have given us either OTSG or pressurizer in a nearly dry condition.
It is difficult to return to a stable condition following a transient - partic-ularly tFJ loss of feedwater transient. Accident analysis should be reevaluated in light of knowledge gained from the TMI event.
There is a lack of OTSG level trips - the design was aoparently based on fast recovery fcilcwing seccndary transients, and the primary system design lacks tne carabil,ty c handle :nese t-ansients under certain conditions.
263 255 It is recomnended that OTSG level trips, a larger pressurizer and upgraded auxiliary feedwater (AFW) systems in light of the icw mass and unreliable heat sink (OTSG) be provided.
AFW Systems As stated under Heat Sink, abo
, he recorrends upgrading the existing AFW systems to meet current requirements.
There are general problems with lack of redundant flow paths, and pumps and poor indications of system operability.
Some specific problems were listed based on familiarity with the Crystal River (CR) plant.
CR has one auto steam driven pump and one manual electric driven pump. Surveillance procedures on the electric pump negate auto start of steam pump - the electric pump cannot be carried on the diesels unless other safety related loads are shed - if the motor driven pump has replaced operation of the steam driven pump the steam driven pump will no longer start on auto if we lose the motor driven pump.
At Oconee there is only one AFW systen per unit - manual valving is required to align cross ties.
At ANO-i the design includes a recirculation valve which must open to prevent everpressurization of AFW system under no-f*cw conditions.
At C3 a break anywhere in the 6" startup feed lino causes loss of all emergency feedwater.
i 263 256
4_
At CR the steam break isolation matrix closes all valver (including emergency) to isolate affected OTSG on low pressure.
However, without appropriate check valves it appears that you could isolate both OTSG's thereby requiring manual operations to staply emergency feedwater af ter determining which OTSG had the break.
There is no annunciation on unavailability of pumps or valves not in position.
There is no eme gency flow indication.
It is recommended that AFW systems receive same treatment with regard to status indications as ECCS.
Natural Ciruclation Natural circulation capabilities as well as procedures and the reed for actual testing should be reevaluated.
We should not' be relying on isolated tests.
The capability of going into natural circulation with voids in the system should be evaluated. He thinks that there is a high probability of void existence as a result of the low pressure for HPl actuation (close to saturation).
Review interface with ICS which programs a higher level with no RCP's to enhance natural circulation.
ICS Interface with FW The ICS in not designed as a safety related system but appears to be controlling a safety related system.
In 50me designs it has cnly a single power supply - in ethers it recuires a marual changeover to an alternate oower sucoly.
Westinghouse inclu:es trics on many of the # unctions covered by the ICS therefore they can get away witn a ncn-safety grace design.
The B&W ICS shculd be ucgraded.
L:DJ
-) r) 7
,z t.
PORV The Bulletin requirenents in the area of lower scram settings, increased FORV lif t pressures and reactor trip or, turbine trip alleviate some of the concerns with the B&W PORV.
The B&W PORV type provides paor position indication (PI) to operator - only indication is energization of solenoid which is indirect at best.
Several years ago it was determined that no PI was available and that loss of power caused the valve to fail open. This led to current PI but some PORV block valves do not have Class IE power supply. There is a history of valve failures and he is of the opinion that the valve should be radesigned. The B&W valve was never qualified to pass water and function properly.
The Westinghouse PORV is air operated with direct PI limit switches on valve travel.
HPI Initiation point is too close to saturation. The capacity of HPI to remove decay heat (without heat sink) is questioned. The staff should relook at cooling capaci*y and reevalua procedures which call for the operator to throttle ficw. All procedures calling for ocerator to throttle a core cooling ficw should be reviewed.
o r' O
,s/ b)
(. J O Level-Pressurizer & OTSG Should look at B&W plants for trips based on OTSG level. Westinghouse has several trips - level - mismatch etc.
B&W has no trip on pressurizer level.
Westinghouse has a high level trip.
There appears to be an inconsistency in the review of these designs.
Recommends tne addition of trips for the B&W plants in addition to upgrading the level instrumentation to safety grade on both pressurizer and OTSG.
The FW trip is not always effective - in some cases the FW pump is set back to minimum speed. This results in no real feed flow but does not give reactor or turbine trip.
Reactor Coolant System Instrumentation Existirg instrumentation is not properly placed on RCS to give operator the information needed to assess his problem.
When bubble formed (TMI-2) pressure and temperature indicated subcooling, therefore, TH is not providing proper indication of the upper plenum temperatJre or there is a lag time.
Should have instrumentation in highest system points to evaluate natural circu-lation performance.
No olants - E!,W, Westingbouse or Ccmbustion Engineering (except for scme older clants' have uccer head temceratures.
Tre tyces c# installed instruments must te reliable under voided canditicns.
aestinghcuse Tave maj be very inaccurate if correct cypass ficw is not available.
Q
.} b >z
- ) T:
uJi
7-Control Board Not enough human engineering or operations input has gone into tt:e design and layout of the control board.
Control boards have not been designed to permit effective operation under a transient condition.
Key parameters should be identified for voiding or LOCA situaticns.
These parameters should be grouped in a display configuration so that they are readily apperent to the operator.
H Recombiners 2
CR does not have a reccmbiner - thay rely on purge which is totally inadequate.
As a result of iMI-2 we need to take another look at containments from the standpoint of dead ended volumes etc. in order to prevent local buildup of explosive mixtures.
Gaseous venting from the RCS has net been a concern in the past but obviously needs thought now.
One possible design approach is the installation of vents from RCS high points to the pressurizer gas space with controlled bleed from that point.
Contairment Isolation All vende s have unique oroblems.
Most B&W plants use only a containment oressure sicnal wnich is not satisfactory.
2b3 20
. Bulletin calls for isolation of lines not essential to core cooling on Safety injections. Thinks this is a vague requirement and doesn't recommend that we isolate RCP cooling water, cooling to rod drive motors or cooling to ventil-ation system coolers. We should keep them available until a containment pressure signal isolated them - don't isolate or safety injection - keep cooling water available to the containment.
Another potential isolation problem is aoplicable to St. Lucie 2 where HPI pumps don't inject at normal system pressure.
Isolating normal charging path could then result in no-injection at higher pressures.
Waste Discosal The tankage and waste processing has been inadequate for years. Not enough attention has been paid to the waste disposal systems - in many cases these systems were field designed and installed.
Have always given a lot of problems with regard to operator exposure and accessability. Most licensees have run borderline on waste disposal cor bility.
TMI-2 has only highlighted this existing problem.
One sumo cump contaminated a whole building. Why does the design include a line from the containment building sump to the auxiliary building sump? This exists on several plants.
Controls re generally located in a cart of the building with difficult access
'/~ts have resulted in sericus inaccessability.
2 h 3 t_ b
-9 Sulletin Comments Based on Plant Visits What is interpretation of running RCP? Does it mean 'un to destruction?
Is vendor sure the pump can operate in a steam environment? Obviously need to consider requirement to have pumps available during recovery phase of any accident.
Why run HPI for 20 minutes? The subcooling criterion appears to be the important Concern.
General concern - unless the operators are aware of our bases for some of these edicts we can't expect to get their cooperation.
O e
~
ENCLOSURE 2 ATTENDEES - CON OUICK PRESENTATION PAY ?, 1979 NRC R. McDermott A. Oxfurth B. Clayton B. Boger N. Wagner T. Wambach M. Williams H. Silver Frank Orr R. W. Woodruff Harry Rood Don Quick
__