The legendary clock of the Bulletin of Atomic Scientists, marking the time left for an estimated nuclear catastrophe, has sensitised generations of neutron scientists and others to the possibilities of nuclear accidents. Current public interest in nuclear power, encouraged and promoted by political and economic interests, has led to a revival of the debate. Does the solution to global warming and power shortages lie in a vast expansion of nuclear power plants connected to nation-wide electrical grids?

Most scientists have now identified the major risks for survival – climate change, depletion of resources, and a shortage of energy. Some of them profess an article of faith that such problems can be solved by more effort and technology. It is indeed likely, to some extent, that the large scale expansion of nuclear power will help to mitigate the consequences of these phenomena. However, this technology is fraught with such uncertainties, and unpredictable, potentially catastrophic consequences, that a choice is very difficult.

The three classical risks associated with the nuclear option are:  the link between the nuclear fuel cycle and the proliferation of nuclear weapons, the possibility of nuclear accidents and the problems of the disposal of the nuclear waste. Recent political developments have shown that proliferation is a genuine danger. Israel, North Korea and Pakistan have been able to acquire these weapons. It is difficult to count on the sagacity or commonsense of political leaders. Dr Strangelove is an imaginary character, but Dr Kissinger is not.
The current arms race in North-East Asia and in the Middle-East, mostly based on massive sales of non-nuclear weaponry, tempts a weak power to use nuclear weapons when hemmed in by threats of destructive bombing, a favourite option of the sole superpower.

The extremely low probability of nuclear accidents in industrial power plants, ensured by extreme attention to detail, is however, offset by their potentially catastrophic consequences, and it is not clear that it will be possible to install an international regime strict enough to obviate all risk of such incidents.

 Finally, we do not know enough of the risks in storing and disposing of a nuclear waste product with an activity  increasing with age and peaking around an estimated thousands or tens of thousands years in the future. If all these factors, and the expense, are taken into account, it is indeed a very risky technology.

These risks and uncertainties have led to considerable hesitation among industrial investors. They are asking for heavy loan guarantees, regulatory insurance and faster licensing procedures. These complex plants take a very long time to build. It is twenty-five years since a completed power plant has been delivered in the US and the infrastructure is no longer there or is obsolete. Either it has to be imported or rebuilt, which may have limitations.
The special feature of this technology arises from the mix of three factors – economic, technical and political, to a greater extent than in other applications.

A great expansion of the fuel cycle is essential to fill the energy requirements, and that makes nuclear weapons so much more available. The number of reprocessing plants required is difficult to estimate. The problem of dealing with the spent fuel has its own economic, technical and political facets.The notion of depositing it in deep mines has been tossed around since 1957, and it has not been satisfactorily implemented. Our knowledge of the geological migration of these disturbed zones in the future is incomplete.

The production of 1kg of nuclear fuel suitably enriched requires 11.8 kg of natural uranium and 10.8 kg of depleted uranium, containing mostly U-238, are left behind. In the US, about 0.5 million metric tonnes are stored. It has three uses – to produce plutonium in reactors, as dense armour penetrating shells when alloyed, and in Boeings as ballast. The US used 320 tonnes of this toxic material in the Gulf War.

Finally the risk of nuclear accidents has new features. The probability can perhaps be deemed fairly low, but the very high consequences may make them unacceptable. There are three factors which determine the scenario. a) design defects may lead to failure, b) maintenance of ageing parts and c) the unpredictability of human behaviour under stress.

In the wake of a poorly-designed experiment in the reactor at Pripyat, Ukraine, on April 27 1986, the chemical explosion blew off the covers, releasing more than 50 tonnes of radioactive elements to Sweden, 800 miles away from the plant.

The issue of nuclear weapons is the strongest argument against nuclear power. Non-proliferation treaties notwithstanding, the real risk lies in the conflict between the haves and the have-nots. Nuclear weapon states want to maintain these toys while denying the legitimate claims of other regimes to access fuel cycle technologies for civilian use.

They have the obligation to achieve nuclear disarmament under treaty, and if they do so seriously, the international climate could change in favour of an effective and joint effort to tackle global warming and similar problems, otherwise, not. The biggest impact on carbon emissions is likely to be secured by simply increasing the efficiency of energy utilisation in houses, commercial establishments and in industries. The production of electricity, energy costs for transportation and heating are clear targets.