Per Peterson: key requirements for new reactor designs, what about thorium?

NewImage

More from the reddit.com Science AMA Series with members of the UC Berkeley Department of Nuclear Engineering. Aside from “Can I still eat fish from the contaminated Pacific” the dominant questions seemed to be variations of “is thorium the answer?”. I was surprised – I thought obsessing on the LFTR concept was a niche nerdy issue. The slant no doubt illustrates that the reddit crowd doesn’t represent an Anglo cross section.

 I guess I’ll start things off. What type of reactors should we be building? I know a big deal a few years ago was made about liquid flouride thorium reactors. Is that the way of the future, or are there superior alternatives?

Prof. Per Peterson replies (emphasis mine):

I do not think that we have the basis to determine or select the best coolant or fuel type to use in future reactors. But there are some attributes which we do need to make sure are used in future reactors.

The first is to use passive safety systems, which do not require electrical power or external cooling sources to function to remove decay heat after reactors shut down, as is the case with the AP-1000 and ESBWR designs, and with all of the light water reactor SMRs now being developed in the U.S.

The benefits of passive safety go well beyond the significant reduction in the number of systems and components needed in reactors and the reduced maintenance requirements. Passive safety systems also greatly simplify the physical protection of reactors, because passive equipment does not require routine inspections the way pumps and motors do, and thus can be placed in locations that are difficult to gain access to rapidly.

The second is to further increase the use of modular fabrication and construction methods in nuclear plants, in particular to use steel-plate/concrete composite construction methods that are quite similar to those developed for modern ship construction. The AP-1000 is the most advanced design in the use of this type of modularization, and the ability to use computer aided manufacturing in the fabrication of these modules makes the manufacturing infrastructure much more flexible. In the longer term, one should be able to design a new reactor building, transfer the design to a module factory over the internet, and have the modules show up at a construction site, so the buildings are, in essence, 3-D printed.

The final attribute that will be important for new reactors will be to make them smaller, and to develop a regulatory framework and business models that work for multi-module power plants. While there will likely always be a market for large reactors, creating an ecosystem that includes customers for smaller reactors (inland locations served only by rail, installations needing reliable power even if fuel supplies are interrupted, mature electricity markets that need to add new capacity in small increments).

On thorium, a question:

Hello! What do you think is the most important advantage that thorium has over uranium as a “fuel?”

Prof. Per Peterson’s reply

The thorium fuel cycle has clearly attractive features, if it can be developed successfully. I think that most of the skepticism about thorium emerges from questions about the path to develop the necessary reactor and fuel cycle technology, versus open fuel cycles (uranium from seawater) and closed, fast-spectrum uranium cycles.

The most attractive element of the thorium fuel cycle is the ability to operate sustainably using thermal-spectrum neutrons. This allows the design of reactor core structures that use high-temperature ceramic materials like graphite, which have substantial thermal inertia and cannot melt. Because these ceramic materials also provide significant moderation, it is difficult to use them in fast-spectrum reactors and thus the most plausible fast-spectrum reactor designs need to use metallic structural materials in their cores.

So thorium reactors are compatible with higher intrinsic safety (cores which do not suffer structural damage even if greatly overheated) and that can deliver heat at higher temperature, which enables more efficient and flexible power conversion.

Molten fluoride salts are compatible with these high-temperature structural materials, and given their very high boiling temperatures make excellent, low pressure heat transfer fluids. In the near term, the largest benefits in using fluoride salts come from the low pressure and high temperature heat they can produce. This can be achieved with solid fuel, which is simpler to work with and to obtain regulatory approvals.

But molten salt technologies also have significant challenges. One of the most important is managing the much larger amounts of tritium that these reactors produce, compared to light water cooled reactors (the quantities are closer to what heavy-water reactors, such as the CANDU, produce, but methods to control and recovery of tritium are much different for molten salts than for heavy water, and key elements remain to be demonstrated).

To repeat a critical point “…largest benefits in using fluoride salts come from the low pressure and high temperature heat they can produce. This can be achieved with solid fuel…”. This summarizes why Prof. Peterson’s lab is focused upon developing the PB-AHTR design, which will also prove out many materials and technologies required subsequently to implement the more challenging Liquid Fuel molten salt reactor concept (such as LFTR).