343

concludes*/:

edures, and practices

-

and above all

-

revent nuclear accidents as serious as Three Mile Island, amental changes will be necessary in the organization, in the attitudes he Nuclear Regulatory Commission and, to the extent that the itutions we investigated are typical, of the nuclear stry.

SSMENT OF SIGNIFICANT EVENTS

The accident at Three Mile Island (TMI) occurred as a
a series of human, institutional, and mechanical failures.
Equipment failures initiated the events of March 28 and
ed to the failure of operating personnel (operators, en-
and supervisors) to recognize the actual conditions of the
heir training was deficient and left them unprepared for
s that took place. (See finding F.) These operating
made some improper decisions, took some improper actions,
d to take some correct actions, causing what should have
nor incident to develop into the TMI-2 accident.

lemental Views" from Commissioners are available and will be in the permanent edition of the Commission's report.

27

stuck-open PORV did not state that unless the PORV block closed, a LOCA would exist. Prior to TMI, the NRC had pai insufficient attention to LOCAS of this size and the proba their occurrence in licensing reviews. Instead, the NRC 1 most of its attention on large-break LOCAS.

-- a maj

4. The high pressure injection system (HPI) safety system came on automatically. However, the oper conditioned to maintain the specified water level in the and were concerned that the plant was "going solid," that with water. Therefore, they cut back HPI from 1,000 gallo per minute to less than 100 gallons per minute. For exter periods on March 28, HPI was either not operating or opera insufficient rate. This led to much of the core being und extended periods on March 28 and resulted in severe damage core. If the HPI had not been throttled, core damage woul been prevented in spite of a stuck-open PORV.

5. TMI management and engineering personnel also ha difficulty in analyzing events. Even after supervisory pe took charge, significant delays occurred before core dama fully recognized, and stable cooling of the core was achi

6.

Some of the key TMI-2 operating and emergency p in use on March 28 were inadequate, including the procedur LOCA and for pressurizer operation. Deficiencies in thes procedures could cause operator confusion or incorrect act

*/

For a definition of loss-of-coolant accident and oth technical terms used in the Commission's report, see the the back of this volume.

ing on rising pressurizer level. The Davis-Besse operating at only 9 percent power and the PORV block i approximately 20 minutes after the PORV stuck open. is investigated by both B&W and the NRC, but no ling attention to the correct operator actions was lities prior to the TMI accident. A B&W engineer had ternal B&W memorandum written more than a year before that if the Davis-Besse event had occurred in a ig at full power, "it is quite possible, perhaps core uncovery and possible fuel damage would have

An NRC official in January 1978 pointed out the erroneous operator action in a TMI-type incident. notify utilities prior to the accident.

A Tennesse Valley Authority (TVA) engineer analyzed rising pressurizer level and falling pressure more >re the accident. His analysis was provided to B&W, visory Committee on Reactor Safeguards. Again no s given to utilities prior to the accident.

ontrol room was not adequately designed with the 1 accident in mind. (See also finding G.8.e.) For

Burns and Roe, the TMI-2 architect-engineer, had cally evaluated control room design in the context of ent to see how well it would serve in emergency

The information was presented in a manner which could

rs:

(i)
Over 100 alarms went off in the early
stages of the accident with no way of suppressing
the unimportant ones and identifying the important
ones. The danger of having too many alarms was
recognized by Burns and Roe during the design
stage, but the problem was never resolved.

C. After an April 1978 incident, a TMI-2 cont operator complained to his superiors about problems with room. No corrective action was taken by the utility.

9. In addition to the normal instrumentation prese control room at the time of the accident, TMI-2 was equip special data recorder that B&W had temporarily installed plant start-up and never removed. This data recorder, ca reactimeter, preserved a large amount of information usef post-accident analysis. This type of data recorder was n as standard equipment by the NRC.

10. Those managing the accident were unprepared for significant amount of hydrogen generated during the accid Indeed, during the TMI-2 licensing process which concentr large-break LOCAs, the utility represented and the NRC ag in the event of a large-break LOCA, the hydrogen concentr containment would not be significant for a period of week first 10 hours of the TMI accident (a small-break LOCA), hydrogen was produced in the core by a reaction between s the zirconium cladding and then released to containment t burn or an explosion that caused pressure to increase by per square inch in the containment building. Thus, TMI i a situation where NRC emphasis on large breaks did not co effects observed in a smaller accident.

11. Iodine filters in the auxiliary and fuel handli buildings did not perform as designed because the charcoa capacity was apparently partially expended due to imprope before the accident. Required testing of filter effectiv the fuel handling building had been waived by the NRC. no testing requirements to verify auxiliary building filt effectiveness.

T

12. The nature and extent of damage to the core is to be known with assurance until the core materials are r and carefully examined. However:

a. We estimate that there were failures in th around 90 percent of the fuel rods. The interaction of t cladding with water generated somewhere between 1,000 and pounds of hydrogen gas and converted 44 to 63 percent of zirconium to relatively weak zirconium oxide. As a resul oxidation and embrittlement of the fuel rod cladding, sev