Three Mile Island

From Resources -- Environment & Policy by John Fernie & Alan S Pitkethly, Harper & Row, London, 1985.

Three Mile Island

The gravity of the nuclear industry's problems stems from the accident at Three Mile Island (TMI), Pennsylvania, on 28 Mar 1979. Already by 1979, sluggish electricity demand and high capital costs had caused utilities to revise their nuclear programs. It has been shown in Fig 10.1 that the number of new orders began to dwindle by the mid 1970s and previous obligations were reneged. TMI accelerated this process. It showed that after nearly four decades of experience, the nuclear industry had still not perfected an adequate reactor design. A Group 1 accident -- a complete meltdown with failure of the backup safety systems -- was calculated a near impossibility: 1 chance in 200 million reactor years. TMI came within 1 hour of this unlikely event. The chaotic state of affairs in the wake of the accident did not inspire confidence in the operator of the plant or the regulatory agency (NRC). Conflicting reports on the extent of the accident, media over-reaction and a lack of co-ordination concerning evacuation plans left the public in a state of confusion [44]. This crisis in institutional confidence has baulked the operator -- Metropolitan Edison -- from re-starting Unit 1 on TMI, which was down for repairs when difficulties began with Unit 2.

It was not until mid-1982 that the full extent of the damage to Unit 2 was ascertained, when a specially designed camera was used to investigate the core of the reactor. Around 90% of 37,000 fuel rods were damaged when temperatures had reached 5,000 degrees F in the core [45]. This news meant that the cleanup will take 5 years, longer than expected, and will cost $1 billion. General Public Utilities Corporation (GPU), the parent company operating the plant, sued the reactor suppliers and builders Babcock & Wilcox (B&W) for this amount, plus billions in economic losses because of the cost of alternative supplies of power. But this dispute was eventually settled out of court to avoid any further embarrassment to the US nuclear industry.

The history of events leading up to the accident is revealing. Toledo Edison's Davis-Besse station had encountered similar problems 18 months earlier and investigators for NRC and B&W had reported the seriousness of the incidents to their superiors, but no action was taken [46]. In short, it was proposed that new operating guidelines should be sent to utilities with B&W reactors instructing them not to shut off the Emergency Core Cooling System (ECCS) unless the pressuriser level was increasing and the primary pressure was 1600 psi and rising.

When the same fault manifested itself at TMI 2, not only did it take the operators several hours to realize what had happened but their difficulties were compounded by the shut off during routine maintenance a few days earlier of valves carrying water from the emergency feed-water system to the steam generators. What started as a minor fault -- clogging of the main feed-water pipes -- almost ended in a meltdown. The closure of these valves meant that the emergency feed-water system could not cool the steam generators which began to boil dry. This in turn led to an increase in coolant pressure and temperature and the reactor was scrammed. However, the operators were unwilling to activate the ECCS fearing that an over cooling accident might occur!

Slowly the operators began to realise what had happened but much of the damage had already been done. The zirconium allow cladding around the fuel rods reacted with the steam releasing hydrogen and gaseous fission products. NRC experts disagreed about the chance of an explosion so evacuation plans were put into force. Eventually the hydrogen was bled out of the coolant and burnt off. Disaster was averted and radiation emitted from the crippled reactor was within the prescribed limits.

This is the context within which GPU had initiated their civil action. However, the Kemeny Commission was critical not only of the suppliers and operators of the reactor but also of the NRC for its inability to resolve small safety problems before they snowballed into major ones. To redeem itself, NRC ordered a moratorium on licensing until August 1980 while it undertook safety reviews and revised standards. All operating plants were studied according to performance (a report card) on matters such as equipment, operating procedures, radiation and environmental protection, emergency planning and fire/security protection in order to focus inspections on plants with below average performance.

Since the accident all operators have taken extra training and federal licensing has been toughened. NRC has required all plants to install safety vents to release hydrogen accumulations and to put in instruments to measure water levels in reactors to prevent further TMI 2 type accidents. These are only the most urgent of over 300 actions deemed necessary by the agency in response to TMI 2 [47]. The industry has also responded to the accident by forming the Institute for Nuclear Power Operations (INPO) to upgrade training of operators; and plant operators must belong to INPO to qualify for insurance to help pay for substitute power if a reactor is put out of action for a long period. Furthermore, the Electric Power Research Institute has created a Nuclear Safety Analysis Center to examine power plant incidents to review design changes to minimise safety problems [48].

Despite these improvements, TMI 2 is fresh in the memory of public and industry alike. High capital costs and sluggish electricity demand was undermining the need for new nuclear plant; TMI 2 has eroded financial and public confidence in the industry. No new order will be forthcoming until this confidence can be restored.


[44]
For a detailed account of the history of TMI 2 see B Keisling, Three Mile Island: Turning Point, Veritas Books, 1980
[45]
Arizona Republic, 22 July 1982.
[46]
See M Gray and I Rosen; The Warning, W M Norton, 1982m for another review of TMI 2, especially the NRC responses to the incidents before the accident and the accident itself.
[47]
Wall Street Journal, 26 Feb 1982.
[48]
American Petroleum Institute, Two Energy Futures: a national choice for the 80s, API 1981, p 97

Kym Horsell /
Kym@KymHorsell.COM

ADVISORY: Email to these sites is filtered. Unsolicited email may be automajically re-directed to the relevant postmaster.