Examining Figs 10.6 and 10.7, it is apparent that the most frequent abnor­mal events are likely to have zero health consequences to the public. However, these events can be very expensive to the operating organization, because they often result in power decreases or reactor shutdown. (Most of the electricity production cost arises from debt retirement of the plant capital cost; operating costs are a relatively small part of the total.) Plant management has a very strong motivation toward reduction of these minor malfunctions. Obviously, this reality exerts a strong positive effect on overall plant safety motivation. The mid-frequency range of malfunctions also exerts a strong positive influence on plant safety — there is a possibility that such accidents might result in some damage to the plant, or at least may lead to extended outage time for inspection, repairs or plant modifications. This is the frequency range in which probabilistic safety analysis is most effective, as we will see in Section 10.4.

The ‘disaster range’ of accident events shown in Fig. 10.6 is the range that concerns safety regulatory agencies the most; it is the range in which the reactor accident might lead to human fatality. In common practice, this is the frequency range (~10-4 per year and lower) over which the special safety systems provide the primary defence against health consequences. In US practice, this range is identified as the ‘severe accident’ range, which includes at least some degree of reactor core disruption.

In the early history of the nuclear industry, this disaster range received an undue amount of attention on the part of designers and regulators. The reason was that accident models were very simple and extremely conserva­tive, so that the predicted consequences were correspondingly large. During the past several years, more exact and realistic predictions of consequences have been made, and so the predicted consequences have become much smaller. Nevertheless, regulatory agencies have tended to continue applica­tion of the very rigid and conservative acceptance criteria that were devel­oped during the time when analysis models were crude and overly conservative. The nuclear reactor safety field is now in transition toward more realistic modeling. It is expected that these plants will eventually be proven to be considerably safer in terms of human health than they were originally thought to be.

In the end, then, the question of whether or not nuclear energy will be installed on the very large scale needed to replace today’s energy supply from oil, natural gas and coal will probably be determined by a balance of fear — between fear of the technology and the fear of falling short of the high level of supply required to maintain the people of the world in good health and spirit.