Slashdot Mirror
Small, Modular Nuclear Reactors — the Future of Energy?
cylonlover writes "This year is a historic one for nuclear power, with the first reactors winning U.S. government approval for construction since 1978. Some have seen the green lighting of two Westinghouse AP1000 reactors to be built in Georgia as the start of a revival of nuclear power in the West, but this may be a false dawn because of the problems besetting conventional reactors. It may be that when a new boom in nuclear power comes, it won't be led by giant gigawatt installations, but by batteries of small modular reactors (SMRs) with very different principles from those of previous generations. However, while it's a technology of great diversity and potential, many obstacles stand in its path. This article takes an in-depth look at the many forms of SMRs, their advantages, and the challenges they must overcome."
Distributed power is how our grid should be set up. Also, being self-contained, these would allow us to put them closer to the actual users and cut transmission losses and costs. Why the hell aren't we doing it yet?
-SaNo
No sure if anyone has, but India is aggressively pursuing it.
From the summary: "...It may be that when a new boom in nuclear power comes..."
...that's unfortunate phrasing.
The best demonstration of Liquid Fuel Thorium Reactor (LFTR) was by ORNL in the 60's. They had a prototype molten salt reactor using U-233. This is the fissile component of the Th-232/U-233 fuel cycle. The breeding of TH-232 into U-233 was simply omitted as unneeded complication for this prototype. This was intended to prove / debug the molten salt reactor, it was very successful in key ways.
India has been working on solid fuel thorium reactors, this is an attempt to re-use our experience with U-235 reactors technologies. It is doubtful that this would ever be competitive with a clean LFTR design.
In the US, the regulatory hurdles for LFTR are very high, unless you bypass them by selling your design to the military, which has the option to bypass these regs. This is why Flibe Energy is planning to sell their LFTR to the military first. It is a lot easier to change the regulatory environment if there is clearly functional and safe product being used by the military.
Th-232/U-233 was investigated as a nuclear fuel back in the 60s because there was widespread fear that uranium would prove to be scarce and prohibitively costly. That didn't turn out to be the case -- uranium is cheap (relative to the costs of the plant) and abundant. Light-water reactors fueled by low-enriched uranium oxide fuel pellets are well understood by utilities and regulators. Utilities are notoriously conservative and risk averse, so "amazing new technology" makes them nervous.
Molten salt reactors are essentially unproven in large scale testing. Yes, I am very much aware of the Air Force's experiments at Oak Ridge. But the fact remains that nobody has built a large one, and nobody has run one for long periods of time. On-line reprocessing is a clever idea that has never been demonstrated in a reactor. And U-233 most certainly can be used to make a weapon... so the proliferation resistance argument is a bit overblown.
Solid thorium oxide fuels were used at the Ft. Saint Vrain reactor in Colorado in a gas-cooled reactor. That's another promising technology that isn't going anywhere because the U.S. Nuclear Regulatory Commission has basically said "we know how to license and regulate LWRs. We don't have the manpower or resources necessary to do the same for a host of advanced concepts." And the utilities have basically said "we know how to run LWRs with better than 90% capacity factors. We're skeptical that you can do the same with an advanced non-LWR."
So yeah, we're gonna build Voglte-3 and 4 (AP1000 PWRs). We're gonna build Summer-3 and 4 (AP1000s again). Beyond that, the financing is the bottleneck. Until the economy picks back up, no utility is going to try to finance the overnight cost of a large-scale reactor. The SMRs that will be licensed in the next decade are all small PWRs: NuScale, mPower, Westinghouse SMR. GE isn't pushing PRISM (a sodium cooled fast reactor) in the US. Hyperion Power Generation is a joke with no realistic licensing strategy. The Traveling Wave Reactor is a pipe dream due to fuel cladding limits. It'll be advanced LWRs for the next two to three decades.
Something which doesn't often get discussed, but which I learned about a couple years ago - a number of knowledgeable people have said that what really killed nuclear power in the United States was the Shoreham Power Plant.
This was a nuclear power plant built in Long Island, New York, for about $6 Bn. The plant passed certifications and inspections and was all ready to go into commercial operation. However, because of politics, the plant was never able to get the go ahead from the State of New York to operate. The governor, Mario Kuomo, basically vetoed an *already built* power plant.
As long as the laws are such that investors can't get reasonable assurance *before* they spend all the money to build the plant, that they will be definitely allowed to operate as long as the plant meets relevant technical standards, the *politics* of the situation make the plants not viable.
Without such political uncertainty, nuclear plants are, generally, good investements, economically. A nuclear plant (depending on how much power it produces), should produce more than enough power to pay for itself in the course of 60 years, if it's allowed to operate.