ML19290A143
| ML19290A143 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 09/12/1979 |
| From: | Milstead W NRC - TMI-2 LESSONS LEARNED TASK FORCE |
| To: | Mattson R NRC - TMI-2 LESSONS LEARNED TASK FORCE |
| Shared Package | |
| ML19290A141 | List: |
| References | |
| NUDOCS 7910170381 | |
| Download: ML19290A143 (2) | |
Text
.r 4
/
UNITED STATES y
g NUCLEAR REGULATORY COMMISSION g
E WASHINGTON, D. C. 20555 o
?
Y
/
SEP 121979 MEMORANDUM FOR:y R Mattson, Director, Lessons Learned Task Force, NRR FROM:
W. C. Milstead, Lessons Learned Task Force, NRR
SUBJECT:
KEMMENY COMMISSION QUESTION REGARDING DIFFERENCE BETWEEN TMI CONTAINMENT RESPONSE AND PEBBLE SPRINGS CALCULATED CONTAINMENT RESPONSE In responding to ACRS question number twenty-six regar. 'ng the effects of TMI-2 type accidents the Portland General Electric Company calculated the Pebble Springs reactor and containment response to an accident resulting in complete loss of main feedwater and auxiliary feedwater. The calculated response indicated that containment pressure rise would be sufficient to initiate an ESFAS signal (containment isolation and high pressure injection) at about 10 minutes into the event. The trip setpoint for Pebble Springs is 4 psig.
The ESFAS setpoint for TMI-2 also was 4 psig. However, in the accident at TMI-2 containment pressure did not reach the ESFAS set point until more than four hours after initiation of the accident. The Kemmeny Comission has questioned the reason for the apparent discrepancy between the actual containment response at TMI-2 and the calculated containment response for Pebble Springs used in evaluation of the reactor system response (i.e., high pressure ECCS injection).
Per your request, I have examined the assumptions used in the analysis of the Pebble Springs containment response. The Pebble Springs analysis was performed for an assumed stuck open pressurizer code safety valve which has a flow capacity of about 5 times that of the TMI-2 operated relief valve (PORV). The Pebble Springs analysis assumed direct release of pressurizer flow to the containment rather than throgh the pressurizer relief tank, which would offer a small delay in the containment pressure rise. The Pebble Springs analysis also neglected the effect of containment passive heat sinks (heat capacity of structures, equipment supports etc) as well as normal containment ventilation and cooling systems which might continue in service following an accident. The ANS decay heat curye multiplied by a factor of 1.2 was also used in calculations of mass and energy flow to the containment. These assumptions would clearly be conservative if used to calculate a maximized containment pressure as is nonnally done for the purpose of evaluating the containment design pressure.
Such assumptions are nonconservative when used for the purpose of determining the expected containment pressure rise for the purpose of evaluating the adequacy of safety systems initiation set points and the subsequent reactor cooling performance.
7 9102 703R(
2225 034
~
e-SEP 121979 R. J. Mattson At Three Mile Island, Unit 2, four containment fan coolers were in operation in the normal cooling mode of operation at the start of the accident and remained in operation until the ESFAS 4 psig containment pressure set point was reached more than four hours after the accident started. At that point the units were not shut down but were switched to the accident mode of cooling. ' Reactor coolant system fluid was not relieved into the containment until about 14 minutes into the event when the pressurizer relief tank diaphragm ruptured.
At TMI-2 eacg Btu /hr for air with a cooling coil temperaturc rise of 2.5 F.
containment fan cooler has a normal cooling rated capacity of about 1 x 10 The same fan cooler in the accident mode (i.e., air and steam environment) has a rated capacity of 50 x 106 Btu /hr with an 84'F cooling coil temperature rise. Back of the envelope type calculations indicate that four fan coolers coupled with the passive heat sinks inherent in the containment should have removed a major fraction of the steam in the containment atmosphere thereby reducing the containment pressure rise and maintaining it at the low pressure (1-2 psig) actually experienced at TMI-2.
This illustrates the need to perform conservative minimum containment pressure analyses for the purpose of evaluating the adequacy of containment pressure ESF set points and possibly best estimates analyses for the purpose of operator training. Analyses of the first type are supposedly done by applicants but, to the best of my knowledge, are not submitted for staff review and are not performed by the staff as inupendent analyses.
It appears that calculations which are used as the basis for establishing containment pressure set points for safety systems should be reported in SAR's and audited by tne staff.
I will bring this to the attention of the Lessons Lea ned Task For idiamC.'Myistea c
Lessons Learned Task Fo e Office of Nuclear Reactor Regulation cc:
R. Tedesco W. Butler G. Lainas J. Kudrick R. Denise J. Shapaker -
LLTF Members 2225 035