In Part I we made the tentative suggestion that the cost of accidents was low, indeed negligible, when compared with the value of the electricity generated. This introductory conclusion was based on a scratch calculation drawing on upper-case assumptions, multiplying the (ill) chance of a disaster of 1 in 100,000 years of a reactor's operation by damage costing €1,000 billion. The cost of such an accident amounts to €1 per MWh generated. Setting aside the back of an envelope, which lends itself to quick, simple calculations, we shall now look at the matter in greater detail, from both a theoretical and an empirical standpoint.

Calculating the cost of major accidents

Is the cost per MWh of an accident negligible, at less than 10 euro cents? Or is it just visible, at about €1, or rather a significant fraction of the several tens of euros that a MWh costs to generate? How does our scratch calculation, in Part I, compare with existing detailed assessments and how were the latter obtained?

Risk is the combination of a random event and a consequence. To calculate the risk of an accident, the probability of its occurrence is multiplied by the damage it causes. Much like many other forms of disaster, a major nuclear accident is characterized by an infinitesimal probability and huge damage. A frequently used short-cut likens the former to zero, the latter to infinity. As we all know, multiplying zero by infinity results in an indeterminate quantity. The short-cut is easy but idiotic.