Dawson replies: Physics Today contacted Ernest Moniz, cochair of the MIT study, for a response.
Moniz comments: Both the Norbeck and Clark-Phelps letters contain some valid observations, but they fail to note additional key considerations.
Edwin Norbeck’s assertion that the nuclear waste management problem “should be regarded as insignificant” underestimates both the technical and political challenges if considerable growth in nuclear deployment is to be realized. The MIT report states that “geological disposal is technically feasible but execution is yet to be demonstrated or certain.” The issue of execution should not be minimized. Second, he emphasizes the economic value of actinides and fission products in the spent fuel but does not acknowledge either economic or proliferation concerns. Several countries today separate plutonium and uranium from irradiated fuel for recycling. However, this is not economically competitive, and an accumulation today of about 200 tonnes of separated plutonium is a clear proliferation risk (the International Atomic Energy Agency defines “significant quantity” as 8 kg). Further partitioning of commercial-reactor spent fuel to extract specific fission products is not currently performed, has been explored for some elements, and faces major economic hurdles.
Robert Clark-Phelps correctly asserts the importance of comparative studies of alternative greenhouse-gas reduction strategies. However, all carbon-free technology pathways are likely to be needed in a robust response to the daunting greenhouse-gas challenge, and comparative studies should be grounded in objective, in-depth multidisciplinary analyses of each pathway. Nuclear power, the starting point for the MIT group, is in many ways the most contentious, but all others face significant challenges at the terawatt scale. Wind, for example, has a substantial tax credit, and its deployment should continue to show strong growth, but many issues, such as intermittency, long-distance transmission, energy storage, and public acceptance, must be resolved to enable truly large-scale deployment.
