Tuesday, April 26, 2011

Will Fukushima increase interest in small modular reactors?

The combined effects of the earthquake and tsunami illustrate the risks of putting all your eggs in one basket

small reactorsThe architecture of energy distribution in major industrial nations like Japan, the U.S., France, the U.K. , and Germany is based on very large power plants, often several at a single site, which are tied into huge transmission and distribution (T&D) grids.

The vulnerability of this model was tragically illustrated March 11, 2011, as six nuclear reactors representing about 10% of the nuclear generated electricity in Japan were permanently taken out of service by a single natural disaster.

In addition to the loss of electricity, radiation hazards from some of the crippled reactors have created significant safety issues of workers at the others.

In engineering parlance, the decision to build all six reactors at Fukushima at a site on the earthquake prone coastline created a single point of failure. TEPCO's decision not to take historical records of tsunami events into account in setting the height of the protective sea wall sealed the fate of the reactors.

When the earthquake struck, the reactors shut down as planned and emergency generators switched on as expected in such an event. However, the 15 meter high tsunami breached the 5 meter sea wall sweeping away the fuel tanks for the diesel generators and the electrical switch gear needed to deliver their power to reactor cooling system pumps.

Are resilient networks of SMRs an answer?

What if industrialized countries began to mitigate the risk of similar future events by building resilient networks of small modular reactors? What if instead of building a single 1,000 MW nuclear reactor, a utility laid plans to build six-to-eight small modular reactors (SMRs) in locations near major demand centers for electricity?

A resilient grid of SMRs built in a distributed network would be much less susceptible to damage from natural disasters or man-made disruptions. If one SMR goes out of service, it doesn't create a regional blackout for everyone else in the utility's service area.

Another issue that comes to mind is whether continued reliance on traditional light water reactor (LWR) designs is the only feasible path forward for SMRs? A lot of emphasis, perhaps too much, has been made on the production of hydrogen when fuel assemblies with their zirconium cladding are uncovered from cooling water. At Fukushima three of the six reactors suffered significant damage to their secondary containment structures from hydrogen explosions.

Design legacy of the Integral Fast Reactor lives on

ANLWestWould reactors built with different fuels, and metal cooling systems, offer advantages to utilities thinking about reliability and safety when considering an SMR?

I've been exchanging emails with Irfan Ali, CEO of Advanced Reactor Concepts (ARC), a Reston, VA, firm that is developing a 100 MW SMR. It is based on the design concepts of the Integral Fast Reactor which was demonstrated at the Argonne West site of the Idaho National Laboratory.

In a white paper released in April 2011 to address the issues surrounding Fukushima, Ali says the time is right for an objective assessment of alternative energy distribution architectures. The Fukushima reactors were 40 years old and built to standards that would not be accepted in today’s regulatory environment. The reactors provided a significant portion of the electricity used in Japan, 6 GWe of the 45 GWe that comes from nuclear reactors in that nation when all of its plants are online.

What about emergency shutdown?

The U.S. Nuclear Regulatory Commission is still grappling with the challenge of how to conduct a safety review of a SMR using sodium cooling systems and uranium alloy fuel. Preparing for that review, it is fair to say the NRC will find some of its wisdom about LWR pumps and cooling system is not directly relevant to some aspects of the new designs. There will be a steep learning curve for the agency,

Citing the design of the sodium cooled ARC-100 design, ARC’s Ali discusses the advantages of natural circulation pathways that carry decay heat away from the fuel rods. There are no pumps hence the lack of a need for electricity to run emergency cooling systems. Cooling loops are backed up by air circulation outside the containment structure.

Also, he explains the fuel is a uranium metal alloy rather than uranium oxide as used in LWR designs. The steel cladding doesn't present a risk of hydrogen production.

Perhaps the most novel element of the reactor is what happens if heat transfer from the reactor to the turbine is stopped by accident or other interruption. In effect, the reactor sits on its hands and does nothing.

In technical terms, the passive feedback mechanism which shuts the reactor down is based on the physics of the design. Rising coolant temperatures cause structural elements to thermally expand which allows more neutrons to leak out of the core rather than be absorbed by the fuel creating fission. As a result, high heat causes the neutron chain reaction to shut down.

A key issue at Fukushima is the widespread release of radioactive from the turbine buildings and spent fuel pools. In a sodium cooled reactor using uranium alloy metal fuel, the iodine is chemically bound inside the reactor limiting the potential for radioactive releases.

Increasing interest in mitigating risk with SMRs

The ARC-100 reactor design concepts contain intriguing safety measures which might benefit highly industrialized countries seeking a more resilient power grid. Similar benefits might come from other SMR designs including those that use conventional LWR designs. It depends in part on the pace of advancement in fuel cladding materials science.

The key idea is to find ways to avoid future consequences of having too much electrical generation capacity invested in a single site. This is especially important in areas where there is a potential for earthquakes, tsunami, and other natural disasters or man-made disruption. SMRs buried underground add the natural containment of that design paradigm to their protective envelope.

Changes will also be needed in the way rate structures are set for resilient networks of SMRs compared to rates for single large plants. Perhaps the lure of lower costs for T&D architectures will be an incentive for utilities to speed up their assessments of SMRs in the wake of Fukushima?

Prior coverage on this blog

# # #


Alan said...

You said the 6 reactors represented 14% of nuclear generated electricity. My calculations based both on capacity and generation get closer to 9 and 8%. Not sure what the reason for the difference is.

Indeed, it would appear to an outside observer that TEPCO did not take into account historical tsunami events. I'm glad that other commentators are recognizing this apparent issue. TEPCO, among other, is still backing the position that this was an event that can't be planed for, or is otherwise extraordinary and thus was reasonably not prepared for. I've been hearing that politicians in Japan have been bomblasting this stance, including the prime minister. Granted, I think there is a greater social issue at play, whereas the design values for external events have to be set at something, and whatever it is, there remains a frequency of events surpassing it. We do, in fact, need to talk about how many meltdowns and what consequences of them are acceptable and how we assure that the safety standards have assured that, yes with PSA, but trust is more important. All risk assessments are useless without trust.

I'm less familiar with the ACR-100 design than other SMR designs, but the issues can be hairy for all of them. Grouping multiple reactors at a single site could be problematic, which is a major detail of SMR deployment scenarios. The inherent safety made possible by SMR designs should be a major part of the discussion as well. It's important to look to those and ask what it buys us compared to typical LWR accident scenarios.

Anonymous said...

This is posted anonymously, but I'll identify myself: E. Michael Blake.

The reference to what the Fukushima reactors (I guess the six at Daiichi are meant) could generate has perhaps been confused by the fact that the World Nuclear Association has already removed Daiichi-1 through -4 from its commercial operation list. The data in the parentheses after the 11th paragraph appear to refer to Daiichi-5 and -6 and Daini-1 through -4. The six Daiichi reactors had a total capacity of about 4.5 GWe, and the four Daini reactors had a total of about 4.3 GWe. Daiichi-5 & -6 plus Daini total roughly 6 GWe.

djysrv said...

I revised the number to be approximately 10% of nuclear generated electricity (4.6 Gwe)

robert_steinhaus said...

Small is good because the US has lost the industrial infrastructure to build reactors much larger than naval reactor size (heavy forged steel reactor pressure vessels). If a new LWR reactor is larger than around 400 MWe, and current commercial designs are up around 1600 MWe, then the US requires help from a foreign manufacturing partner [1] to build current commercial nuclear reactor designs. The impact of regulatory "improvements" introduced "just to be safe" by NRC has priced up conventional nuclear to the point the wealthiest nation on the planet can't afford to build it BIG anymore.
[1] - http://bit.ly/dyEaMG
("Regulatory ratcheting" has increased the cost of a nuclear power plant by a factor of 4-5x over and above inflation)

In a non-USNRC nuclear universe, where all that occurs is not paced by and limited to what NRC has a chance in its schedule to review, there is still advantage in BUILDING BIG. Molten Salt Reactors and Supercritical CO2 Brayton turbine-generators could scale BIG (~10 GWe) very economically for underground installations in major cities in the developing world. Current NRC regulatory obstacles have reduced America's hope of nuclear innovation to "small is better than none".

Anonymous said...

The time for SMRs is clearly now. With the price of new large LWRS almost incalculable (since raw materials costs are so variable, loan guarantee programs don’t always work to the borrower’s advantage, etc), smaller reactors with associated smaller costs provide a less risky path to new nuclear. The issue of not being able to certify these designs (like the ARC reactor) in this country really rankles me. The NRC may not have the technical expertise or bandwidth to proceed with design certification at this time but that expertise does exist within our national lab system (Idaho, Oak Ridge, etc). The ARC reactor is based on the EBR-II design which ran for 30 years at Argonne National Lab. Can we not take advantage of that expertise to help the NRC build a regulatory framework by which to evaluate these advanced design, non-LWR SMRs? It doesn’t make sense for us not to help this American-designed, built and tested technology get off the ground and to the market-place to compete with SMRs from Korea, Russia, etc. Having a design certification from the country of origin (US NRC) would be invaluable in marketing these SMRs both here and abroad. In the wake of Fukushima, we have an opportunity to jumpstart a new American industry – let’s not waste it.

Gunnar Littmarck said...

SVbr-100 is one of my favorites, with supercritical CO2 and closed Brayton cycle, it can be SVbr-150. Reactors should probably not be larger than they can be built in a factory, if we are creating a global welfare by 2050 .... requires 4-6 times as much global electricity production. My most favorite is David LeBlanc's LFTR, but no one seems to show up capital to build it yet. (where is Bill Gates?)

//Gunnar Littmarck

Joel Riddle said...

Wednesday April 27's rash of tornados in the TVA service region might be a further (possibly equally strong) example of how a distributed grid of smaller generating and transmission assets could provide more robustness than the currently entrenched generation and transmission paradigm.