Atomic Show #278 – Micro-Modular Reactor (MMR) project partners USNC, GFP and OPG
Global First Power (GFP), Ultra Safe Nuclear Corporation (USNC) and Ontario Power Generation (OPG) recently announced that they had formed a joint venture called Global First Power Limited Partnership. That venture will build, own and operate an installation called the Micro Modular Reactor (MMR™) at the Chalk River Laboratories site.
Mark Mitchell and Eric MGoey joined as guests on Atomic Show #278 to provide depth and background information about the technology and the project that was not included in the press release.
Mark is USNC’s director for the MMR project. Eric wears two hats, one at GFP and one at OPG. For GFP, he is the director of outreach and communications. For OPG, he is the director of remote power generation.
We talked about the project’s genesis and the joint venture’s mission of proving through doing that the system design can be licensed, manufactured, assembled and operated in a cost-competitive way.
Eric provided a brief overview about OPG. He explained that it is committed to providing clean, reliable power both to grid-connected customers and to customers in areas that are not connected to the grid. He described how OPG has a current charter to serve markets throughout Canada and into the United States, and how it hopes that the MMR project will open new markets to the company.
For this first of a kind project, the MMR is a 15 MWth, 5 MWe power system with essentially two main plants. The nuclear plant is a helium-cooled, fission reactor-heated system that circulates helium through a heat exchanger. The adjacent plant is a conventional steam plant that circulates water through a heat exchanger/boiler and a steam turbine/condenser.
Between the two plants is a molten salt heat storage system that acts to buffer heat supply and steam demand. It gets heated by helium that has passed through the reactor. Hot molten salt transfers heat to boil water, creating high pressure steam to turn the turbine.
This arrangement allows the supplied grid to rapidly respond to load changes while enabling operators and control systems to vary reactor power output in a more gradual and efficient manner.
The reactor heat source differs from other high temperature gas reactors. It uses the same Triso coated particle fuel often chosen for gas cooled reactors and some molten salt cooled systems. Instead of using a random graphite matrix material to produce fuel elements from Triso particles the MMR uses USNC’s patented Fully Ceramic Microencapsulated (FCM) fuel.
That innovation replaces random graphite with densely packed silicon carbide (SiC) as the matrix used to produce fuel elements. According to corporate literature on this feature, FCM fuels can retain fission products without failures at temperatures approaching 2000 C.
MMRs are designed to operate for 20 years between fuel system replacements.
While we talked a bit about the technological specifics, most of my conversation with Mark and Eric revolved around business considerations, the importance of developing manufacturing competence, the importance of effective cost controls and the importance of transparent engagement with regulators and potential customers.
Your participation in the comment thread is always welcome. If questions arise that need more details, I will seek assistance from the show guests.
I hope you enjoy listening.
Podcast: Play in new window | Download (Duration: 40:09 — 46.1MB)
Subscribe: Google Podcasts | RSS
Thanks again Mr. Adams:
It doesn’t take much of a memory to remember times when diesel fuel was not considered as inexpensive. Despite recent changes in the oil industry, I don’t think the idea of “peak oil” should be entirely discounted. As the use of oil in China, India and the continent of Africa are expected to increase, it seems logical that there would be a similar increase in the price of diesel fuel over the next twenty years.
Over such a time span, how do you put 20 years worth of diesel fuel into today’s dollars? It seems like the fixed cost of small reactor producing power for a mining community over that twenty year period would be a large decrease in economic risk. Did I understand this wrong? Any response is appreciated.
If the initial load of fuel lasts for twenty years. would that fuel last somewhat longer if the load to the reactor is less over those twenty years? Are there other components that will necessitate the shutdown of the reactor after that 20 year time period?
During the 20 years of operation the nonnuclear components will be inspected on a regular inspection frequency. I suppose that the nuclear components will also designed in an inspectionfriendly way.
Wonderful episode as always!
I have a question about reactivity and the fuel loading. For this reactor to have a twenty year lifetime it must start with a huge amount of excess reactivity correct? Or is it operating as a breeder/burner? Is there some other fancy new trick that I am missing?
Lasercowboy – you’ve identified one of the things that make solid fuelled reactors complicated to design, as I understand reactor engineering. (I’m not a nuclear engineer, but have read a textbook and am watching the MIT Introduction to Nuclear Engineering course lectures, free on the internet. ocw.mit.edu/courses/nuclear-engineering/22-01-introduction-to-nuclear-engineering-and-ionizing-radiation-fall-2016/lecture-videos/ )
Yes, you need to have enough reactivity in the uranium load to still be useful at the end, and the reactivity of the uranium does drop as the fissionable atoms get used up. The solution is to add isotopes of other elements that absorb excess neutrons when the fuel is fresh. Once the “burnable poison” atom absorbs a neutron it becomes another isotope that no longer absorbs a neutron. That atom has been “burned up”, and the reactivity of the fuel has fallen as well.
That process balances the total reactivity as the fissionable atoms get used up. This process has to be calculated for all the isotopes in the reactor, including all the fission products. This is part of what makes designing solid fuels difficult, and I don’t think any reactor relies on just the fuel design to keep reactivity under control. I expect that’s a good part of what all the stuff in the reactor control rooms is all about.
“Yes, you need to have enough reactivity in the uranium load to still be useful at the end”
The other solution is to design the reactor so it can be refueled while running like the CANDU or the liquid fueled reactors
Online refueling is one of features usually included in pebble bed HTRs.
@James R. Baerg
“The other solution is to design the reactor so that it can be refueled while running like the CANDU…”
Records for continuous operation:
1. 940 days UK’s Heysham II, ending September 2016 (AGR).
2. 894 days Canada’s Pickering 7, ending October 1994 (CANDU)
3. 739 days USA Lasalle, ending February 2007 (BWR)
So, the online refueling capability of the CANDU with minimal excess reactivity does not necessarily translate into higher availability relative to LWR with significant excess reactivity. Some training material I have shows the CANDU having something like 10 days of excess reactivity in a Liquid Zone Control system (tanks of light water ‘poison’ in the calandria), so if online refueling machine is broke down they can maintain full power for about a week by replacing the LZC H2O with D2O (or void?).
TLDR: online refueling does not translate into uninterrupted operations.
The COG says (http://www.candu.org/Pages/RESOURCES.aspx) that CANDU “uses about 15% less uranium than a pressurized water reactor for each MW of electricity produced.” – I argue that is splitting hairs: maybe true some days and not others. If a CANDU discharges 1 ton of natural fuel after 5GWD/MTU and the LWR discharges 1 ton of 5% enriched fuel after 50GWD/MTU while the enrichment plant discharges 9 tons of depleted tails, then CANDU and LWR fuel economy is essentially the same.
Jaremko made a decent point about excess reactivity in burners vs. breeders… No doubt that fuel resources can be stretched by under-moderating LWR and HWR, at the cost of needing to reduce the heat rate. If you read the PRISM materials of a certain vintage they discuss the need to periodically offload fuel as excess reactivity rises with core burnup. Even if ‘isobreeding’ (infinite doubling time) is achieved, the solid fuel will have to be discharged before mechanical end of life. For steel cladding in sodium that end of life could be double or triple the duty we see LWR zirc clad. Still, got to throw the fuel out at some point.
So this micro-reactor will perform online refueling? That seems rather difficult to do at a small scale and the sort of thing that would require extensive resources to be brought to the far north/remote mine/wherever.
I was under the impression from the podcast that the core was loaded up once at the factory and then left to operate for 20 years.
I hope all understand that virtually any solid fueled reactor can be operated for 20 years without refueling. The naval reactors are a good example of this; their reactors reach end of life after ~4 Effective Full Power Years. Those 4 EFPY are not spent in 30 years. The sub is rarely underway at full steam and is often in port. Similarly, a commercial PWR with an 1.5 year cycle could be operated for 3 years at half power and 6 years at 25% power, but we don’t do this because it is sub-optimal in accordance with the time value of money. So any of these concepts where small burner reactors leave the factory with a 20 year gas tank have a lot of money tied up in there cores and take a long time to realize ROI.
Oklo has a method of controlling reactivity of a fast reactor with a 20 year life that I had to read multiple time to visualize (read Oklo NRC submission to see what I read).
Oklo surrounds their solid core with specially made rods that rotate through 180 degrees during its 20 year life (that’s really slow). My own picture of the rod is a half-black side and a half-silver side. The black side absorbs neutrons while the silver side reflects neutrons. At startup the black absorber side points toward the core to absorb the excess neutrons. After 20 years, the silver side points toward the core reflecting neutrons back into the core. In between, the core sees part black and part silver.
Quick shutdown is handled by conventional rods not the black and silver rods.
Automatic load following is not done with the black and silver rods (just long duration reactivity control).
Excellent; you now kind of understand leakage from the radial boundary and how it is a part of the neutron balance. These radial ‘control drums’ are a concept found on other [space reactor] concepts including NTP types.
Note that nothing actually ‘reflects’ neutrons. It is an unfortunate description of core boundary condition that we are stuck with. Placing an absorber on the radial boundary condition of a fast reactor doesn’t seem more worthy than placing a ‘D’ shaped void along the length of a steel ‘reflector’ drum – IOW use a half-hollow drum like the other concepts – not my idea.
Re: Note that nothing actually ‘reflects’ neutrons.
The phenomenon seems to be more like “refraction,” IMO.
Keep in mind that the longer core lifetimes of many of the advanced reactor design will rely upon, in part, on high-assay low-enriched uranium, HALEU, fuel with enrichments close to 20%.
Read that DOE achieved a record burnup rate of 25% on 20% HALEU over six months with a uranium oxycarbide TRISO pebble system. Now present article sez with only 12% enrichment, the USNC system will operate 20 years, achieve a temp of 2000 degrees, and, although I see no figures on maximum burnout, the 20 years should either be sufficient for significant breeding, or long residency to convert a large quantity of thorium to U-233. What is the DOE and/or me missing about the fuel economics or overall desirability of on system versus the other? Note: any chance of promoting helium as a reliable and cost cost effective coolant seems to rely on completely sealed primary coolant pipes in a small modular scheme, as opposed to the MIT pebble bed design being tested in China.
Recent Comments from our Readers
@Cyril R What was Tesla’s learning rate starting at the first Roadster? How much do you think that first unit…
A new engine or turbine product line doesn’t just cost triple a unit. That’d make it pointless. Yet this is…
Cyril First of a Kind (FOAK) applies to products whose parts and method of assembly are new, not just products…
The problem with the FOAK argument is that FOAK LWRs were built half a century ago for under $300/kWe. And…
I kind of wonder if there aren’t some smart Canadians looking across the border and rubbing their hands with glee.…