ML19224A794
| ML19224A794 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 04/08/1979 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | Eisenhut D Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 7905300378 | |
| Download: ML19224A794 (20) | |
Text
.' )
~
s p
_Eisenhut/Mi 9 Job T 4/8/79 ACRS CCNCERNS TMI -2 INCIDENT 1.
What assurance do we have that the TMI event will not happen at anot.ter BAW reactor temorrow?
The initiating event (loss of condensate and feedwater pumps) i an anticipated transient, i.e., it is expected to occur and may occur tomorrow.
However, the severity of the consequences in the TMI-2 incident was caused by multiple circumstances and actions which are addressed by the April 5, 1979 IE Bulletin 79-05A.
The purpose of that Bulletin is to prevent recurrence of the contributing circumstances and actions thereby preventing recurrence of the incident.
The Bulletin requires licensees to:
review their procedures and operator actions and determine that they are adcquate to prevent a similar incident particularly with regard to termination of HPI flow and tripping of RCS pumps and with regard to reliance placed on pressuri:er level indicators in determining operator actions; review containment isolation singals to determine that proper isolation will be provided; and assure that adequate auxiliary feedwater flow will be provided by observing specific requirements provided in the bulletin regarding auxiliary feedwater systems operability and availability when the plant is at power.
2.
There has been much discussion of this accident as a S&W orcblem.
What makes this accident unigue to 8&W PWRs?
168 059 79053 00 37c,P P
7 U
,---.-e---:,.
~
~
The " loop seal" design of the connection betseen the pressurizer and the hot leg makes the plant susceptible to the so-called "monometer effect." When an extended blowdown through the pressurizer occurs, there is the possibility that the pressurizer will cease to be the highest temperature volume in the primary system.
Higher temperature and higher steam pressurizer in the core then generates a steam volume which could fill part of the core and the hot leg and prevent water in the pressurizer from flowing downward into the core.
Since the operator would likely assume that adequate water level in the pressurizer assures that the core is covered with water, this " loop seal" design could lead to incorrect operator action (throttling or securing water flow to the core).
That is, he would believe the core to be covered when it is not. The loop seal design is present on B&W plants, but is not used on CE plants and recent design Westinghouse plants.
Also, the once-through steam generator design used on B&W plants has a smaller secondary water reservoir, and it is a more efficient design for providing heat transfer to the secondary water.
Therefore, for both of these reasons, it is more prone to quick dryout upon loss of secondary water (feedwate).
w~-
Is there anythina unicue about B&W Containment Isolation Features?
TheTMI-2containmentisolatesonlyonhighpressure(ykpsig)inthe X
containment.
Many other plants (B&W and other vendors) also isolate on ECCS initiation signals.
Such a containment isolation signal, if 168 060
c it had been present on THI-2, would probably have greatly reduced releases from containment during early stages of the TMI-2 incident.
Gus Lainas will provide additional information regarding this question.
4.
Are all auxiliary feedwater systems redundant and diverse All auxiliary feedwater systems (fFW) in FWRs are redundant.
None have only electric motor driven AFW pumps.
Most are diverse and have both electric motor and steam turbine driven pumps.
Some have electric motor and diesel driven pumps.JEome do not have diversity within the Y
AFW, but are diverse to the mainfeedwater and condensate system.
These systems rely on steam turbine driven AFW pumps alone.
5.
Is there anything about the PORVs at TMI-2? On B&W plants?
3 B&W plants (including TMI-2) utilize a de@ign where an electrical signal activates a solenoid which opens a pilot valve which in turn allows pressure to be introduced "oder the main valve seat, which then opens (Dresser pilot operated valve).
Although some PWRs also utilize other valve designs many PWRs supplied by other vendors quite likely utilize similar or identical PORVs (review of records is continuing in this area).
6.
What are the cresent reauirements for coerability of auxiliarv feedwater systems?
168 061
The relevant license conditions, or Technical Specifications for TMI-2 are attached.
It requires the reactor to be shutdown in 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> if one train is inoperative and I hour if both trains are inoperative.
The Bulletin 79-05A issued April 5 requires at least this same level of operability at all SkW designed reactors.
A copy of the Tech.
}
Specs. for these plants is attached.
D The Tech. Specs for other PWRs vary somewhat depending on date of initial license and plant design.
A table summarizing the operability requirements is attached as are the individual plant specifications.
m 168 062
SUMMARY
OF OPERABILITY REQUIREMENTS FOR PWR AUXILIARY FEEDWATER SYSTEMS (TIME TO SHUTDOWN)
B&W Designed PWRs (Bulletin 79-05A - 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> & 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br />)
Plant One Train Inoperative Two Trains Incoerative Three Mile Is -1 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> 0 hours Three Mile Is -2 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> 1 hour Davis-Besse-1 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> I hour Crystal River-3 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> I hour Oconee 1, 2 and 3 indefinite Indefinite Arkansas-1 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> 0 hours Rancho Seco 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> 0 hours CE Designed PWRs Palisades indefinite (with fire pump) 0 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> Ft. Calhoun 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> 0 hours Maine Yankee 148 hours0.00171 days <br />0.0411 hours <br />2.44709e-4 weeks <br />5.6314e-5 months <br /> 0 hours s
Calvert Cliff 1 and 2 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> 1 hour Millstone 2 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> 1 hour 168 063
St. Lucie 1 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> 1 hour Arkansas 2 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> I hcur W Desianed PWRs Beaver Valley 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> I hour D.C. Cook 1 and 2 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> I hour Farley 1 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> 1 hour Ginna indefinite 0 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> Haddam Neck ind'efinite 0 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> Indian Point 2 indefinite 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> Indian Point 3 indefinite 0 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> Kewaunee indefinite 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> North Anna 1 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> I hour Point Beach 1 and 2 indefinite 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> Praire Island 1 and 2 indefinite 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> Robinson 2 indefinite 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> Saleu 1 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> I hour San Onofre 1 indefinite 0 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> Surry 1 and 2 indefinite 0 hours0 days <br />0 hours <br />0 weeks <br />0 months <br /> 168 064 l
W Desianed PWRs (continued)
Trojan 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br /> 1 hour Turkey Point 3 and 4 indefinite 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br /> Yankee Rowe
-- 1 hour1.157407e-5 days <br />2.777778e-4 hours <br />1.653439e-6 weeks <br />3.805e-7 months <br /> No second train Zion 1 and 2 indefinite 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> 6
N 168 065
~
h w
a 7.
Is there anything unicue about the TMI containment isolaton features?
Five other operating B&W designed plants have similar designs as TMI, however, the actions described by the Bulletin will preclude a similar occurrence.
The large majority of other operating plants have containment isolation systems that by design would have prevented flooding of the Auxiliary Building (i.e., loss of containment).
Most plants utilize safety injection as a signal to initiate containment isolation in addition to containment pressure.
SI was initiated 2 minutes for these other oparating plants at that time, rather than at 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> as was the case at TMI.
Therefore, little radioactivity would have been released.
8.
What is the single basic difference in the olant designs of other~-
coerating olants that might by itself preclude a similar incidenTas at TMI?
We currently believe that the single most important difference in other PWR designs is related to the location of the pressurizer and routing of its surge line.
SWRs of course do not have a prersurizer.
Other types of PWR desiIs locate the pressurizer and surge line so that core levels are directly reflected in the pressurizer where reactor system level is measured.
The TMI design requires operator interpretation of a number of instruments to properly identify reactor coolant system level.
Other design.; lend themselves to more direct measurement.
168 066
9.
Have se learned anythina about the__ acability of instruments to withstand LOCAs?
To date, we have confirmed that much of the instrumentation will withstand high radiation exposures without failure.
Only one critical instrument has failed to date.
10.
Manacement Organization Q.
What are your concerns?
1.
The man in charge of the plant - Units 1 and 2 was not stationed at the plant.
The job was a corporate management job with the principal location offsite.
We felt there should be a single individual in charge of the entire plant, at the plant, to avoid any possibi; inter-unit conflicts, resolve prioritie:, handle site items such as security, emergencies, etc.
There was a lack of clarity about shift supervisors.
Hcw were these c erns resolved?
1.
There was to be a plant superintendent, or one of the unit superintendents was to be designated as in charge of the plant.
This is acceptable.
168 067 4
- ~ ' ~ ~
~
4
2.
The organization was clarified.
Shift foreman (SR0s) on each unit reported to a single shift supervisor (SRO) who had overall plant responsibility.
He was to be in charge of the site in off shifts.
This is acceptable.
11.
Secondary System Line Breaks Q.
Is there any connection between this item and the TMI-2 event?
No.
There was no steam or feed line break at TMI-2.
There are aspects of the analyses and the hardware discussed in this issue which entered into the actual event, but none of the future modifications would have had any effect on the event.
The SER discusses isolation of feed and emergency feed lines and steps taken to preclude single failure causing such isolation, but we did not consider total manual lockcut of all emergency feed valves in our review.
s m
168 068
TMI-2 ACRS CONCERNS 12, Hydrogen Charging Line oc What is this hydrogen used for?
As a scavenging gas in the makeup tank to remove oxygen from the reactnr coolant.
6 Where does the coolant water come from and where does it go from the makeup tank?
Comes from the letdown line through various filters and cther treatment.
After deoxygenation it is reinjected into the reactor coolant via the makeup pumps.
C Can hydrogen be absorbed in the water into makeup tank?
The tank is at approximately 100 F and 15 psig - very little hydrogen will be absorbed under these conditions.
ol How large is the MU tank?
600 CF total - 400 CF water, 200 CF hydrogen
- g. How is the H2 admitted to the MU tank?
There are 5 " bottles" of hydrogen (1000 SCF each at 2400 psi) manifolded together.
One bottle at a time is valved through a pressure regulator into the makeup tank.
paccicent?Is there any way to introduce this hydrogen into the reactor in an 168 069
_. vm emem w
The makeup pumps serve as the high pressure injection pumps during an accident.
On an SI signal the valve from the BWST to the HPI pumps opens.
Procedures require that the valve from the MU tank be closed; nowever, even if it is not closed, equipment elevations tank charging procedures and pump alignment are such that a water seal will still be maintained in the MU tank discharge piping precluding H2 entering the pump (Met Ed had committed to close the MU tank valve automatically by the first refueling).
In addition, the SI signal will close redundant valves in the hydrogen charging line to the MU tank, limiting the amount of hydrogen available.
gH What if these equipment features and procedures do not work?
If (1) the MU tank valve is not closed and (2) the hydrogen isolation valves do not close automatica!!y and (3) tank charging procedures are not followed, then a maximum of 1400 SCF of hydrogen (40 CF at 100 psi and 400 F) could be drawn into the RCS.
If in addition all five hydrogen bottles are valved in the pressure regulator, rather than just one, approximately 155 CFR could be drawn in.
(The sze of the bubble in the TMI-2 event was in the order of 1000 CF).
/c Was there any indication that any hydrogen in the TMI-2 bubble came from tnis source?
(Still checking at site).
168 070_, m-y
~ ~
DISPOSITICN OF ITEMS IN ACRS LETTER CN TMI-2 1.
Augmented Startup Program to verify load folicw and power transient core power characteristics.
Disposition An acceptable augmented startup program was performed on an identical core at Rancho Seco.
These results in conjunction with a closely monitored normal startup program at TMI-2 sere felt to be adequate.
See p. 18-1 of SER Supplement No. 1.
2.
Asymmetric Loads on Reactor Vessel Disposition:
Based on the plant design, the probability of occurrence of the break, and the generic review in progress, we concluded that plant operation prior to final reslution woould be acceptable.
See p.182 of SER Supplement No. 1.
3.
ATWS Disposition:
We concluded that operating limitations until final resolution of the gener@c ATWS issue were not necescary.
See p. 18-3 of SER Supplement No. 1.
4.
Flood Emergency Plan Disposition; We concluded that additional information furnished by the applicant satisfied the ACRS concern.
See p. 18-3 of SER Supplement Nc. 1.
5.
Fire Protection Disposition ',
We reviewed the applicants' fire hazards analysis and fire protection program reevaluation, including a visit to the plant.
Required plant modifications and an acceptable implementation schedule were identified.
The license was conditioned to require such implementation.
See: p.
18-4 of SER Supplement No. 1; p. 18-1 of SER Supplement No. 2; p. 9-1 of SER Supplement No. 2; Facility Operating License No. OPR-73.
I68 071 e
t 8
6.
Hermetic Seals on Instrumentation in Containment Disposition *,
The applicant identified all applicable instruments and described their construction and related maintenance procedures.
This concern was also added as an ACRS generic concern, and as such its resolution will be considered for this plant when developed.
See p. 18-4 of SER Supplement No. 1; p. 18-1 of SER Supplement No. 2.
7.
Interference with Safety Functions of the DC System by Non-Essential Loads Disposition:
We reviewed the applicants response to this concern, and concluded that the present design is consistent with that of other plants found acceptable.
The ACRS requrested that this matter be considered in our generic review of reliability of power supplies.
We will consider any changes identified in this review for TMI-2.
See p. 18-4 of SER Supplement No. 1; p. 18-1 of SER Supplement No. 2.
8.
Release of.iydrogen From Hydrogen Charging Line Disposition:
The hydrogen charging line, which runs only through a portion of the auxiliary building, was rerouted and redesigned to preclude effects on safety related equipment.
See p. 18-5 of SER Supplement No. l' p.
18-2 of SER Supplement No. 2.
9.
Instrument Line Failure Affecting Plant Controllability Disposition:
Applicant analysis showed that no instrument line breaks presented plant controllability problems of significance to publis safety.
Our.
generic review of the plant interactions will also consider instrument line failing.
See p. 18-5 of SER Supplement No. 1; and p. 18-2 of SER Supplement No. 2.
10.
Manacement Orcanization Disposition:
Additional information on this matter was subsequently submitted by the applicant which satisfied our concerns.
See p. 18-5 of SER Supplement No.1; p.18-3 of SER Supplement No. 2, p.13-1 of SER Supplement No. 2.
168 072
11.
Secondary System Line Breaks Disposition:
This ACRS item refers to our open issue on this subject described on p.15-2 of the SER.
The applicant has not completed its analyses of a spectrum of steam line breaks and we had not yet completed our review of the recently reviewed feed line break analysis.
We had required modifications to assure that single failure could not prevent isolation of the feed and emergency feed systems.
Subsequently, our review of secondary system breaks was completed.
We concluded that modifications to the secondary system were required to assure that consequences of breaks were mitigated by safety grade equipment, and that operation until such modifications were implemented (by the end of the first refueling outage) was acceptable.
License conditions were imposed to assure such implementation.
See p. 15-1 of SER Supplement No. 2.
- 12. Additional Means to Follow the Course of an Accident Disposition:
This issue was considered generic in nature and as such would be dealt with on TMI-2 when a generic solution was developed.
See p. 18-6 of SER Supplement No. 1.
13.
Sabotage Disposition:
We performed additional review on particular structures and concluded that their design provided an acceptable degree of security.
We also required submittal of an amended security plan in compliance with 10 CFR Part 73.55.
See p. 18-6 of SER Supplement No. 1.
The amended security plan was approved and made part of the license on February 23, 1979.
s 14.
Generic Items Disposition:
We prepared Appendix C to the SER which noted the disposition and status of appropriate generic items.
See p. 18-6 of SER Supplement No. 1; p. C-1 of SER Supplement No. 2.
168 073
Cuestions and Answers - Exemptions General
- 1. ct What decision criteria are used in decidino whether to issue an Exemption (Confirmation Order) or to issue an order requiring a snutcown.
Public health and safety is paramount and must be assured before any exemption is issued.
Two recent cases can be used as examples.
The shutdown of the five reactors because of the error in the seismic design methods was determined to be required because no analyses were available to assure that design requirements could be met.
In the case of the small break ECCS analysis error, analyses were available and compensating actions could be and were taken to maintain the required level of safety.
O Why is the reliability of oower included as a consideration in the ECCS exemotions?
Our regu!ations (10 CFR 50.12) provide that the public interest be served if an exception is to be issued.
}-
Small Break ECCS Error Exemotions (Orders)
)< ct Why was Exemption for TMI-l issued after the accident at TMI-2?
The Exemption for TMI-1 was issued prior to the accident at TMI-2, March 16, 1979.
It was published in Federal Register on March 30, 1979, two days after the accident.
7.b Does the deficiency covered in the_E.x.emption relate to the TMI-2 accident in any way?
If not, why not?
168 074
Not directly.
Although the Exemption relates to allcwing operator action to adjust high pressure injection system ficw, the event is a small pipe break in a particular location of a particular section of the primary coolant system.
The adjustment is for HPI ficw out of the break.
The high pressure injection system was actuatea in te TMI-2 event and the operator stopped HPI.
However, the action of the type specifically called for in the Exemption was not required.
J'. C Was there an Exemotion on this subiect for TMI-2?
Yes, although it was in the form of an Order for Mcdification of License.
J'.d' Do all operatina B&W tyoe olants have a similar orcblem?
All except Davis Besse which, because of its raised loop confi jura-tion, does not have this problem.
1 4 What is the status of the corrective action for all B&W tyce olants?,,
Arkansas Unit 1 - approved mcdification will be completed prior to startup frcm current refueling outage.
Crystal River 3 - Modification under review - Startup frca refueling is mid-June 1979.
Schedule for completion not firm yet.
168 075
Oconee 1/2/3 - Modification approved - modifications to be completed at next refueling -
Unit 3 - July 1979 Unit 2 - first suitable outage after December 13, 1978, Unit 1 no later than Fall 1979 reload.
Rancho Seco - Modification approved - to be installed next refueling.
Three Mile Island 1 - Modification approved - to be installed no later than next refueling (April 1980) or first outage after September 30, 1979 projected to last at least 30 days.
Three Mile Island 2 - Staff has reviewed modification - modification will be completed prior to resumption at operation.
e 168 0/6
DAVIS BESSE ACRS CONCERNS is 1.
What is te disoodition of the ACRS concerns for Davis Besse l?
See letter from Denton to Ahearne, dated December 20, 1978.
w 168 077
DAVIS BESSE 1 ACRS CONCERNS Kal 2.
Seismic Reevaluations ca.When is the first refueling outage scheduled?
March 1980.
If the seismic reanalysis indicates scme deficiency in strength of some systems, support on equipment what will the NRC require?
Modification of facility systems to meet acceptance criteria.
6 Is the seismic reevaluation at Davis Besse 1 the same problem as the seismic evaluation prooiem or tne tive piants wnica were souumuaa?
No, the problem at Davis-Besse relates to the acceleration value which was used in the analysis and the method of correlating earthquakes with the va aes of applied acceleration.
The problems at the five plants rela:a to an error in a particular computer code which was used to analyze those five plant's systems.
c When will the guidelines be transmitted to the licensee?
They were transmitted January 30, 1979.
ECCS nn +5a crec
,f+ne ice',nca nf 3.
Why does the staff reauire, ore an21vsi<
a licensee? Why di dn' t_the staff wait to issue the license until after it had accepted '.he ECCS analysis in total?
The kind of analysis required was only to quantify the margins of previous evaluations which show that the proposed linear heat generation limits are well in compliance with the regulations.
168 078
.-