ML19347D465
| ML19347D465 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 02/24/1981 |
| From: | Clements W NRC |
| To: | NRC |
| Shared Package | |
| ML19347D463 | List: |
| References | |
| NUDOCS 8103170739 | |
| Download: ML19347D465 (2) | |
Text
gtMiouq'o d
UNITED STATES
' "! V.
- e NUCLEAR REGULATORY COMMISSION k~
i msmucioN. o.c. 20sss
'%() ~
3 e
February 24, 1981 GFFICE OF THE 00f.1 MISSION E R MEi13RANDUM TO T:iE FILES FROM:
William L. Clemer.tsi. @
SUJECT:
~
MEETING WITH I&E ON INVESTIGATION OF INFORMATION FLOW DURING TMI ACCIDENT On February 13, 1981, Commissioner Bradford met with Victor Stello, Norman Moseley and: members of the IE Investigation Team to discuss the investigation and subsequent enforcement action.
The discussion focused on the three questions posed by Commissioner Bradford in his memoranda of January 23 and 28 to tne EDO.
The first question concerned whether the training of the TMI operating personnel would have provided an adequate foundation for Chwastyk to conclude that tne pressure event in the containment building resulted from the combustion of hydrogen generated in a metal / water reaction.
The members of the investigation team stated their opinion that the trainino provided to TMI personnel could not have led to such a conclusion.
The FSAR (which forms a basis for operator training) indicates that in a credible accident hydrogen conc.ntrations in the containment building would not become a problem for some time (on the order of months).
The operators would expect hydrogen to be generated through the radiolysis of primary coolant.
In such a circumstance, hydrogen concentrations would not reach the flammability range for 4-5 months.
The FSAR treat; an incident involving a significant metal / water reaction as an incredible event.
The. investigators stated that they have never directly asked Chwastyk what led him to conclude that significant quantities of hydrogen were present.in the reactor building on March 28.
Their analysis of his training did not go beyond that founded on the FSAR.
Rather, they assumed that in an accident situation operating personnel will mentally run through all accident scenarios in which they had been trained (i.e.,
listed as credible accidents in the FSAR).
The investigators also indicated tnat Chwastyk s testimony about the role of hydrogen in the pressure evant became more detailed in his later interviews.
Chwastyk was first interviewed on May 21 by I&E, well after the hydrogen explosion and metal / vater reaction had been identified as the cause of the pressure spike.
TFe investigators stated that the notoriety surrounding the event and th! resulting inquiries from NRC, Congress, the media, etc. probably cau;ed Chwastyk to become confused about the state of his knowledge on March 28.
In the period shortly 8108170'J Y
after the accident, hydrogen became an issue of great sensitivity.
~ ne icvestigators surmise that this led :hwastyk and perhaps others to misremember the precise time at which they first learned of or surmised the presence cf hydrogen.
Tne second cuestion c oncerned whether Sir, " :;e nas aware of tne 23 PSIG pressure spike on March 28.
Tne investigators incicatec that tne term "p essure spike" has beco.e a gentric cescrictier. cf :ne pressure ever.: encc-cassine all indicators, i.e., tne spike on the pressure recorcer, actuation of the containment sprays and related alarms and the
" thud" heard in the control room.
They believe that when Zewe testified "I found it so hard to believe that anyone who was in the control room observing anything would have missed that (the spike)" (draft House Report, page 130), he actually meant that virtually everyone in the control room was aware of one or more of the indicators.
They concluded that Iewe did not mean to imply that everyone in the control roor,was aware of a real pressure event or even that everyote was aware of the spike on the chart.
The investigators stated that the Ross testimony suggests that M'.ller saw (or was otherwise informed of) the spike on the cressure recorder immediately after it occurred. However, trey believe tnat Miller paid insufficient attention to this new information as he prepared to leave the control room to brief Lt. Governor Scranton.
Coasequently they concluded that, if he was aware of the spike for a moment, he forgot about it in the rush of events. The investigators also note that it would be pointless for Miller to deny awareness of the spike by itself since those who did see it did not (except for the Chwastyk testimony) believe that it was real.
Tnus, knowledge of the spike would not have implied any immediate knowledge of its causes.
that he changed the strategy for cooling the reactor following pressure event as a result of a conversation with Gary Miller.
investigators etated that Chwastyk had assumed a role in the Unit 2 The control room without first being briefed in detail on the events which had transp'ered before his arrival or on the strategy then being employed
.for cool % g the reactor.. Consequently, he was not aware that the control
. room personnel had been attempting to draw a bubble in the pressurizer since about 11:00 a.m.
The investigators stated that Chwastyk's actions following the p essure event did not actually represent a chenge in cooling strategy.
On a related matter, the investigators stated their belief that the so-called "think tank" in the Unit 2 control room did not function in the manner described by Gary Miller.
In their opinion, a large number
-of control room personnel assembled as an analytical body eniy twice on March 28:
at 8:00 a.m., and again around 11:00 a.m. when the decision was made to pressurize.
They believe that throughout the first day, the think tank actually functioned as small groups of two or three people c'onferring with Gary Miller about specific problems.
They concluded that the general atmosphere in the control root prevented calm, detached discession of the situation.
-