The nuclear option: AP1000 or bust?
I’m going to relax my prohibition on discussion of nuclear energy, and offer a couple of points to start things off. I expect everything to go in circles as usual, but there’s plenty of room on the comments page, after all.
First, I have a piece coming out soon in the National Interest, arguing that 2011 marked a watershed in the development of energy sources to replace fossil fuels, with nuclear power finally ceasing to be a relevant option, and solar PV finally becoming a serious contender. I’ll add a link when it appears.
I’m sure not everyone will agree, so I’ll also offer a subsidiary claim which, if accepted, would at least simplify the debate. The claim is that the only nuclear technology with a serious chance of substantially reducing CO2 emissions before 2030 is the Westinghouse AP-1000.
To clarify my reasoning, I’ll begin with a definition. I’m going to define “substantial reduction” to require at least 100GW of operational installed power. If that exclusively displaced coal, it would save around 600-700 million tonnes of CO2 a year, a bit more than Australia’s current emissions, or about 2 per cent of the total. That seems like vey modest goal.
Next, I assert that achieving that outcome will require at least 20GW of capacity to be operational or very close by 2020. Otherwise there simply won’t be any chance to achieve the economies of scale needed to make nuclear economically viable, and the operational experience required to convince decision-makers to invest the necessary capital.
That in turn means that the decision to build the power stations has to be taken very soon. Even in China, where construction schedules are much faster than elsewhere (with implications for safety standards that have been seen in other sectors), a power plant takes a minimum of five years from the first concrete pour, and planning and site preparation takes two or three.
My final point is that any option that can’t make a real difference before 2030 should not play a major in current policy discussions. Obviously we need substantial reductions in emissions well before then, and that should be the primary focus.
The case in favor of the AP-1000 as the only feasible nuclear option is based on two simple facts.
(i) The US NRC has just approved it for the only four plants currently under construction in the US
(ii) China has adopted it as the main design for its program.
The argument that there are no other serious contenders remaining can be approached in two ways. First, by considering the options. These can be divided into three groups
(i) Currently existing competitors, such as CANDU, Areva and so on. None of these have any real chance of achieving the necessary scale, either now or in the future. The only possible competitor is the CPR-1000, a Chinese adaptation of a French design, which is at least a generation behind the AP-1000
(ii) Small modular reactors, still being touted despite having gone nowhere. The most promising version, the ‘pebble-bed’ design being developed in South Africa, was abandoned a couple of years ago. The only contender that has any serious chance of being operational by 2020 is a cut down version of the AP1000. Apart from that, there are a bunch of designs with no prospect of being built, even as demonstration projects, much before 2030.
(iii) Breeder reactors, like small modular reactors, only with a much longer track record of consistent failure. The only actually existing technology, sodium-cooled fast reactors, has been tried in quite a few countries, but has invariably failed. There are a handful of efforts still continuing, but none on a scale large enough to make a difference by 2030. Then there are a bunch of conceptual designs, with no firm plans for construction, even of prototypes.
So, if the AP-1000 is the only real contender, how good are its prospects? Not very good, in my view. The claim that modern designs will produce big cost reductions is so far, just a claim. Even a substantial deployment in China is unlikely to allay safety concerns, given China’s poor track record in this respect. So, a lot depends on the US projects being completed on time and on (or under) budget. Maybe this will happen and maybe not. Even if the AP1000 is as successful as possible, nuclear is still a backstop option at this point. Our primary hopes have to be placed on some combination of energy efficiency, renewables and lower energy consumption.
A final point about the Australian debate. Given the most optimistic possible projections, the AP1000 design might have achieved a couple of hundred reactor years of successful operation by the late 2020s. That would be a good time for Australia to start looking at the technology. The idea that nuclear power is a relevant option for Australia any time soon is just silly.