On Feb 9, 2015, I had the opportunity to visit the faculty and students at the University of California Berkeley. Prof. Per Peterson invited me out to give a colloquium talk and to see some of the interesting work that his colleagues and students were doing in advanced nuclear technology.
One of the primary research projects underway in Berkeley’s nuclear engineering department is the Pebble Bed Fluoride Salt Cooled High Temperature Reactor (PB-FHR) (10.6 MB PDF).
This innovative reactor design combines the high integrity solid fuel form based on TRISO coated fuel particles that has been developed for high temperature gas reactors with the useful heat transfer capabilities of a fluoride based molten salt coolant. The TRISO particles are compacted into 3 cm diameter spheres and the molten salt remains free of fission products.
The pebble diameter is one half of the 6 cm diameter that used in traditional gas cooled reactor designs like the Chinese HTR-10 or the now mothballed South African PBMR project. The combination of smaller pebbles and molten salt heat transfer enables the PB-FHR to achieve a far greater power density than is possible in a gas cooled pebble bed reactor.
According to Prof. Peterson, one of the characteristics that he likes about the high power density design is the fact that each individual fuel element will reach its full depletion in about a year and a half of core residence time, which speeds up the process of fuel qualification compared to fuels designs that take 3-5 years to achieve full burn-up. As is the case for several other pebble bed designs, the PB-FHR will have the capability for continuous refueling with a circulating bed.
Another intriguing aspect of Berkeley’s PB-FHR design concept is the proposed power conversion system. It will be an open Brayton combined cycle gas turbine that has the capability of co-firing with natural gas in a reheater stage. If operating as a baseloaded plant on just nuclear heat, each module will produce 100 MWe; with maximum gas reheating, the peak power output per module would be 242 MWe. The overall cycle thermal efficiency is nearly 45% in the full refiring mode.
The current concept for deployment of the PB-FHR includes a site layout plan with space for 12 individual units, which would end up being a large generating station with 1200 MWe baseload nuclear generating capability and a peak capacity of 2900 MWe with 1700 MWe being produced by burning natural gas.
During the day before the 4:00 pm colloquium, I had the opportunity to visit with the students who were involved with various aspects of the design.
Aside: Unfortunately, I did not take notes in sufficient detail to provide the names of the students that presented the material, so if any of the presenters read this and want to get credit for what they told me, feel free to get in touch. End Aside.
The first student presenter described the gas turbine power conversion system. The Mk-1 conceptual design uses a modified GE 7FB gas turbine. That machine has a compressor that produces an 18:1 compression ratio.
The primary modification is piping that replaces the existing combustors. The piping extracts compressed gas and routes it through a pair of tall coiled tube air heaters that have molten salt flowing through the pipes. That heated air is then injected into the high pressure expansion stage of the turbine.
A second major modification is the insertion of low pressure extraction and injection nozzles and a combustor for gas co-firing. This modification requires the insertion of some extra length between the first two stages and the rest of the power turbine.
After being heated in the annular pebble bed core, the molten salt enters the coiled tube air heaters at a temperature of approximately 700 C and leaves at approximately 600 C. The compressed air side of the heater enters with a temperature of approximately 420 C and leaves with a temperature of 670 C. When gas co-firing is used, the gas coming from the first stages is heated back up to 670 C and expanded through the remaining low pressure expansion stage.
Because the turbine will operate at substantially lower temperatures than the fossil fuel only version of the 7FB, it might be economical to also modify the first stages to remove air cooling and modify blades to use lower cost, lower temperature materials.
Aside: Another possible economy measure would be to locate existing commercial turbines that require less modifications to take full advantage of the differences in cycle temperatures available when using nuclear heat. End Aside.
The next student presenter described part of what he and his team have done to model the circulating pebble bed. They discovered a supplier of tiny polypropylene spheres that serve as reasonable analogs for the graphite coated pebbles. They’ve created several mock-ups that allow their spheres to be circulated through shapes that represent the shape of the core structures. They use different colored spheres to allow visual detection of the ways that the spheres flow and mix depending on where they enter the annulus.
In the mock-up, the plastic spheres are surrounded by air and flow down; in an salt cooled core with graphite pebbles, the pebbles will float and flow upward. The sections used in the mock-up have more surfaces than a complete core, so additional corrections have to be made for the friction caused by the pebbles rubbing on those surfaces.
Since the colored balls provide good visualization but are not easy to track with automated sensors, one student devised a way to tag the plastic balls with tungsten wire so that x-ray equipment could be used to trace the balls as they flowed through various mock-ups. Finding a suitable suppliers for those tagged balls was apparently a valuable engineering learning experience.
The final lab I visited brought back memories of the integral thermal flow testing loops I’ve seen at B&W mPower and at NuScale. As is the case for those systems, the decay heat removal concept for the PB-FHR is natural circulation using convection heat flow and differential densities to drive the fluid around a heat loop.
Though the light water cooled systems at B&W and NuScale require detailed scaling, high quality engineering, and a substantial heat input in order to provide a good simulation of conditions that would be experienced in a real system, the designers can at least use the identical fluid expected in the real plant.
That is not achievable for a university that is striving to produce a useful model of a molten salt system that would normally operate at temperatures of 600-700 C. (Nearly 1300 F) The creative solution here is the use of a heat transfer chemical that has similar fluid characteristics to hot salt at a much more modest temperature.
Nicolas Zweibaum, the design coordinator and current program manager/operations coordinator of the UCB Compact Integral Effects Test (CIET) facility, gave me an excellent overview of the equipment installation. It uses detailed scaling and a simulant fluid known as Dowtherm to provide an experimental model useful for validating thermodynamic flow models and to test theories of behavior under natural circulation conditions. This facility is in the early stages of testing and validation. While I was there, researchers were running an experiment and collecting data for analysis.
I later heard that the researchers that day learned a valuable lesson about data recording and hard disk reliability.
Throughout my visit, I enjoyed the opportunity to talk with both students and professors who were excited about their professions and their visions for ways to improve technology. During my talk at the colloquium, I attempted to share some lessons learned about the importance of understanding the cost implications of design choices and the difficulties one can experience during attempts to execute what seem like great ideas on paper.
The questions indicated that my talk had the desired effect of stimulating thought, not discouraging innovation.
If you are intrigued by the PB-FHR concept and would like to learn more, please download a copy of the Technical Description of the Mark-1 PB-FHR Power Plant report. It is an encouraging piece of work for a small team of graduate students and their faculty advisors.
The day at UCB was capped with a stimulating dinner conversation with Prof. Peterson and two of the colloquium attendees, both of whom are working on the ThorCon team. Before I share anything about that conversation, I need to obtain some permissions. (I think other media sources would call that a “teaser.”)
Update: The lecture was streamed live on UStream. Here is the archived version.
Confession: I erred by not planning a talk that would fit inside an academic period. Since the talk started at 4:00 pm, I sort of expected that people could stay for a while if things got interesting. Neither of the institutes of higher education that I attended scheduled any classes in the evening. We weren’t lazy, early morning classes were a lot more commonly filled up than at most universities.
Lesson learned. I’ll keep it shorter for my next college or university invited talk. (Any takers? Please use the contact link at the bottom of the page to schedule a talk.) End Update.