Emissions, Nuclear, Reactors

Study Questions Value of Uranium Recycling

Issue 4 and Volume 108.

By John C. Zink, Ph.D., P.E.,
Contributing Editor

At the dawn of the nuclear age, Enrico Fermi insisted on developing a breeder reactor because he was afraid there was not enough uranium in the world. During the nuclear plant ordering boom 30 years ago, uranium prices soared as companies paid about any price to lock up uranium supplies for their planned reactors. Now, a new study finds that uranium is so plentiful and so cheap that recycling spent fuel will not be economical in the near future.

In December, the Project on Managing the Atom at Harvard’s Kennedy School of Government issued the final report from its four-year study on the economics of nuclear fuel reprocessing. The 127-page report, “The Economics of Reprocessing vs. Direct Disposal of Spent Nuclear Fuel,” was written by Matthew Bunn, Steve Fetter, John P. Holdren and Bob van der Zwaan, and is available online at http://bcsia.ksg.harvard.edu, under “Publications.”

The authors recognize that economics may not be the only factor controlling the ultimate decision as to whether or not to reprocess uranium, but they maintain that economics is important in the emerging competitive environment of electricity generation. Furthermore, “if reprocessing is being done to achieve objectives other than economic ones, it is worthwhile to know how much one is paying to achieve those other objectives.”

The authors first try to establish a baseline of expected costs for the various steps of the fuel cycle, from uranium purchase through waste disposal, including such things as reprocessing costs and fabrication costs with and without recycle. They recognize the difficulty in determining accurate data for some steps, such as reprocessing and waste disposal under the two scenarios. As a result of these uncertainties, the authors concentrate on a breakeven-price sensitivity analysis. For example, assuming a fixed uranium price of $50/kgU, which is slightly higher than its recent market price, and with all other parameters frozen at the study’s assumed base prices, the authors conclude that the cost of reprocessing would have to drop by more than half to make recycle economical. That is, their base case cost for reprocessing, which is $1,000/kgHM (kg of heavy metal content), would have to drop to $420/kgHM. Similarly, changing one variable at a time, waste disposal costs would have to increase from $200/kgHM to $630/kgHM, or enrichment would have to go up from $100/SWU (Separative Work Unit) to $1,200/SWU.

The bottom line is the cost of nuclear-produced electricity. The authors attempt to partially address this issue with a sensitivity study that holds all parameters constant except the reprocessing cost. In this case they find that, at a uranium price of $40/kgU and a reprocessing cost of $1,000/kgHM, nuclear generation cost would increase by 1.3 mills/kWh. If the reprocessing cost jumps by 50%, the reprocessing penalty would increase by nearly 100%. While 1.3 mills/kWh does not seem like a large number, the authors point out that the entire back-end of the fuel cycle under the current once-through regime costs only about 1.5 mills/kWh. Thus, recycling would increase back-end costs significantly. The authors calculate that a rise of even 1.3 mills/kWh equates to a fuel cost increase of about $1 million a year for a typical nuclear plant.

Using the price of uranium as a dependent variable, the authors determine what the uranium market price would have to be for recycle to break even with the once-through fuel cycle under certain scenarios. For example, at a reprocessing price of $1,000/kgHM, using the base case assumptions for all other parameters, they calculate the breakeven price of uranium to be about $220/kgU. This is about six times the current market price.

Perhaps the most important assessment of the breakeven analysis, which lends some perspective to the authors’ conclusions, is their observation that, “the parameters that have the largest impact on the outcome are the reprocessing price, the difference in price between disposal of spent fuel and high-level wastes from reprocessing, and the price of MOX fuel fabrication.” It so happens, of course, that these are the very parameters for which the values are most uncertain.

In order to address the apparent shortcoming that the shakiest numbers are controlling the results, the authors examine both their assumptions and the industry cost structures for each step. This is not easy, as actual costs are generally held to be proprietary. Gathering as many data as they can from public sources regarding costs at existing facilities around the world, they conclude that, although there are major uncertainties in their chosen values, their assumptions are conservative (i.e., tend to favor recycle). Furthermore, they argue that it is not likely there will soon be major technological changes that will have substantial impact on the cost of reprocessing, disposal or MOX (mixed-oxide fuel) fabrication. Therefore, the authors argue, their conclusion that recycling nuclear fuel is not economical is a robust conclusion and should hold for the next several decades.