The Fukushima accident raises again the issue of the social and economic viability of nuclear technology. To re-evaluate this viability, we analyse the past process of internalisation of external costs of nuclear energy, which present the specificities to be chanted by accidents and has had a constant effect of complexification. This process has provoked a de-organisation of the classical learning process reflected in constant cost increases and the change of social preferences, to end up by the lack of competitiveness before climate policies. Independent institutions of safety regulation have become essential elements of the social embeddedness of nuclear technology at the expense of technology stability and standardization, condition of its competitiveness. In this perspective, the paper argues that the new sequence of social costs' internalization opened by Fukushima will have limited effects on costs, because of anterior steps of safety improvements. Nuclear technology complexification reaches its asymptote : it is being to overcome the challenge of "learning by major accidents". On the other hand nuclear institutions must be re-designed in such a way that it could guarantee maximum safety records and minimum residual risks by going to the other root of the safety issue, the degree of independence and capabilities of the safety authorities in every country, what cannot be decreed. It is nevertheless at this price that could be preserved the global public good of the social acceptance of nuclear technology by limiting drastically chance of new accidents.
展开▼