Letters to Jesse 4: A Second Manhattan Project?

Dear Jesse, The conventional view is that it would take a long time to develop Generation IV nuclear technology. This is mistaken because the Indians expect to complete a commercial Generation IV Fast Breeder Prototype Reactor in 2011, and then begin to build standard production reactors immediately after. They currently expect to complete at least 4 commercial fast breeders by 2020, and more later.

The long gestation period view assumes that the development of Generation IV technology would be conducted with business as usual approaches. But if we think that the fate of human society would rest on the pace of a Generation IV development project, would a business as usual approach make sense? Alternatives would be a simi-Manhatten project model and a mini-Manhattan project approach. The difference would have to do with time scale, with the Simi-Manhattan project approach trying to bring in everything in a two to three year time range, while the mini approach might take 5 years. The mini approach might cost $20 billion, perhaps twice the cost of the business as usual approach, but at the end of the five years a saleable product, and a factory to build it would be ready.

Let me illustrate what I mean by the Manhattan Project approach. The Mahnattan Project was a massive research, development and production project conducted during World War II. The aim of the project was the development of deliverable nuclear weapons. That goal was meet. Rather than develop one single approach to the project, and perfect it, project scientists undertook to develop parallel approaches to project goals. Scientists identified two fissionable materials that could be used in Nuclear Weapons, U-235 and Pu-239. Rather than settle on one approach, they decided to develop two weapons, each using one of the fissionable matereials. The method of producing Pu-239 was deemed very dangerous, and the production facility was located in a desert in Washington State. Production at the Washington State site was to be accomplished through the use of 3 large, experimental reactors of a type never built before. Construction of the reactors began in August 1943, The first was finished in September 1944, and the final reactor was completed by February 1945. The entire 3 reactor project was completed in 18 months. Despite questions about the safety of their design, the Hanford reactors never had a serious accident. Their designer, Eugene Wigner was trained by a chemical engineer who had done notable chemistry research in the 1930's.

A further research reactor was build in Oak Ridge with an overlapping time schedule to the Hanford Reactors. The X-10 Graphite Reactor, was intended to produce plutonium for the research required to waponize it. The designer of the Graphite Reactor was a young scientist, who had recently acquired a PhD in biophysics from the University of Chicago. Despite the fact that the youthful Alvin Weinberg had more training in biology and mathamatics than in physics, and had no engineering training at all, he was able to design a reactor that was built in 10 months, and performed flawlessly, and proved a valuable research tool.

Thus from December 1942, when Enrico Fermi's Chicago pile went critical, and November 1944, the design of reactors leaped forward by would have required a business as usual approach a generation to accomplish. Further more, the designers of these reactors would have been viewed as completely unqualified to perform this task, because they lacked the proper educational background.

In addition to the development of reactors to facilitate the production of a plutonium based nuclear weapon, the project to develop a uranium based weapon had an equally remarkable history. Three separate uranium enrichment projects were developed in Oak Ridge. The Y-12 project developed and used electro-magnets in devices called calutrons to seperate the uranium isotopes. The calutrons required a huge amount of copper wire, and when copper was in short supply, the Manhattan project borrowed 14,700 tons of another electrical conductive metal, silver, from the United States Treasury to wire the magnets. A second uranium separation process was housed at K-25, which when finished was the largest building under one roof in the world.

The K-25 project would have cost $8 billion today. While it was being built, scientists and engineers did not know if they could make the gaseous diffusion method work. Again, a huge investment produced in months what a business as usual approach would have required years to accomplish.

I would argue that given the dual crises of CO2 emissions/Anthropogenic Global Warming and Peak Oil, and the potential for Generation IV nuclear technology, a rapid nuclear development program is demanded.

If a Manhattan project type endeavor were undertaken, regulation would be expidited but safety not compromised. The NRC would work alongside reactor researchers, establishing reasonable safety standards, and passing them on. During the development period the NRC should determin that reactor developments are meeting all NRC safety goals. The complete design should already have an NRC license, even before the prototype is built.

In the Simi-Manhattan project alternative design approaches would be researched in parallel, while in the mini approach they might be investigated sequentially. Both would involve spending at a robust level. There are shortcuts to development including licensing sucessful technology. This might include licensing Russian BN-600 technology, Indian Fast Breeder Prototype Reactor technology, in addition too drawing on American Experimental Breeder Reactor-II (EBR-II) technology and experience. I am not a big fan of the LMFBR type, but it is probably inevitable that we are going to build some, and if we do, we might as well develop and build them fast.

As it was being forced by the Ford administration to wind down LFTR/MSR research, Oak Ridge National Laboratory MSR project leaders prepared a detailed developmental program for LFTR technology that would lead to solving all known developmental problems that might impede the construction of LFTR prototype (ORNL-5018).

That document assumed a business as usual approach, and suggested development plans that would take a generation to realize. How much would it cost?

According to ORNL-4812, up to 1972 ORNL had spent $130 million dollars on MSR development. In 2009 terms this was less than than one billion dollars,

In 1980 the ORNL staff estimated that a commercial DMSR could be developed for $700 million (about 2.5 billion in 2009 dollars). Given another 2.5 billion for the development of the LFTR prototype we would have a total investment of between 5 and 6 Billion 2009 dollars investment. At that point there would be a product ready to go on the assembly line. Thus the total investment in the LFTR would be comparable to the Federal investment into the LWR. It would be one fourth the investment made so far in unsuccessful American LMFBR technology.

My analysis suggests that with factory production and by recycling coal fired power plants, modular LFTRs can come online for an investment as small as a dollar a watt. Let us assume that the actual cost is twice that. We still have a price for LFTRs that is lower than the 2009 price for windmills, even with a capacity factor no better than the windmills, the LFTR would be a far better buy because of its superior flexibility.

It would be nice to imagine a private enterprice investing in the LFTR. Is it possible? $5 billion would not be unreasonable for a private business invest in LFTR development. There are American businesses that are capable of writing a$5 billion check for LFTR development today. Consider the €11 billion plus that Airbus invested in the development of the A380 aircraft. At a cost of $327 million, the A380 would be, if anything, more expensive than the modular LFTR. In fact it is doubtful that Airbus will ever recover the Airbus 380 development cost, while the LFTR potentially could be quite profitable.

Compaired to the cost of renewables, the Manhattan project approach would be an incredible bargain. For example, the German newspaper Die Zeit recently reported that the costs of photovoltaic instalations built in Germany up to 2008

will amount to even more than 30 billion Euros.

And how much electricity will German consumers get for their investment? A recent estimate reported that in 2008. German PVs produced 4,300 GWh, about half the power output of one conventional nuclear reactor. 30 billion Euros would pay the development of both Sandia's "Right Size" Reactor, a small, factory built Fast Breeder Reactor, and the the Liquid Fluoride Thorium Reactor, a very safe, factory build reactor.

Eventually, the LFTR will prove to have significant advantages over the Fast Breeder Reactors. First the core of the LFTR is smaller, hense the structure ment to house the LFTR core will be smaller, and lower cost. Secondly the LFTR has safety abvantages over the fast reactor. Even if the fast reactor proves in practice to be as safe as the LFTR, that safety is not entirely inherent, and will come at a cost. Finally, fuel reprocessing with the fast reactor, will be far more expensive than with the LFTR.

Given the very great importance of a rapid, and massive world wide deployment of low cost nuclear technology capable of safely meeting human energy needs. a Manhattan Project type approach to facilitate the development of promising nuclear technology seems more than warranted. Indeed given the potentially disastrous consequences of failing to safely meet human energy needs, the rapid development of promising technology, is an imperative, not an option. - Charles