INTRODUCTION

In December 1972, the U.S. Atomic Energy Commission published a set of projections of the expected growth of the U.S. nuclear power industry through the year 2000. These forecast a growth of installed nuclear electrical generating capacity from about 130,000 megawatts in 1980 to more than 1,200,000 megawatts in the year 2000. The corresponding number of nuclear power reactors is about 150 in 1980 and somewhere around 1,000 by the end of the century. It is beyond the scope of our study to evaluate these estimates in relation to possible future large scale development of non-nuclear sources of primary energy for use in electric power generation. But the AEC forecasts, should they be approximately realized, have implications for the future risks of nuclear theft. Thus we will want to determine the overall rates of production and flow of nuclear weapon materials associated with these projections.

It is also important in a long-term context to raise several basic questions. Will the inherent risks of nuclear theft depend significantly on the relative numbers of different types of nuclear reactors that might be built to provide the forecast amounts of nuclear power, assuming that the fuel cycles will have the same basic characteristics as the three we discussed in some detail in the preceding chapter? Are there opportunities for modifying these fuel cycles in ways that would be acceptable or advantageous from an economic and technological standpoint, and that would also reduce the risks of nuclear theft? And finally, are there likely to be new technological opportunities for producing nuclear power in ways that would reduce the risks of nuclear theft still further and yet be economically attractive? Our discussion of these possibilities is rather technical, but we hope the general reader as well as the expert will get a glimpse of future opportunities which, though speculative, we believe are worth considering seriously.