Monday, August 1, 2016

Nuclear Power Revisited

We face some major energy policy choices, and a need to move away from fossil fuels, with nuclear energy and renewables often being presented as solutions, but also as polar opposites.
Assuming energy demand can be managed appropriately, it seems likely that renewables can supply all human energy needs. So why is there still support for nuclear power?

The simple answer is that it is well established with powerful supporters, and it does deliver energy, about 11% of global electricity, with relatively low carbon emissions and relatively reliably. These ‘relative’ qualifications are however important in making comparisons with renewables. The load factors for UK nuclear plants have been rather poor in recent year, around 60%, due in part to unplanned shut downs, so that wind turbines, with load factors of 30-40%, depending on location, do not look look too bad by comparison.  Higher nuclear load factors are now more common in the UK and elsewhere, and new plants may be able to get to 90%, but wind technology is also improving, with offshore load factors being higher. 

Nuclear will no doubt always win on this measure, but the comparison turns in favour of wind when we look at the embedded energy and the Energy Return on Energy Invested ratio. It is around 15:1 for nuclear plants and up to 80:1 for wind turbines on good sites: www.routledge.com/books/details/9781849710732/. The figures for wind and other renewables are likely to improve, while those for nuclear are likely to get worse as reserves of high grade uranium ore deplete. More energy is then needed to mine and process the ore to make reactor fuel. Renewables like wind and solar do not need any fuel. Given that most of the energy used for the nuclear fuel production still comes from fossil sources (e.g. diesel for strip mining bulldozers and trucks in remote areas), emissions will rise as ore quality drops. At present carbon emissions for the nuclear cycle are said to be similar to those for some renewables (though views differ), but, unless nuclear and/or renewable energy can be used for fuel processing, they will rise to be much more, with, in any case, ever diminishing energy returns on energy investment: www.wiseinternational.org/nuclear-energy/studies-reports

The economics of nuclear suffer not just from the need to supply fuel but also due to the need to deal with wastes, as well as the cost of maintaining safety and security, problems which are either absent or relatively small for most renewables. At present, in the UK, on shore wind projects are going ahead with strike prices much lower than proposed new nuclear projects. Some PV solar projects are also cheaper and both are likely to be substantially cheaper in the years ahead as they move down their learning curves. By contrast nuclear costs appear to be continually rising, with in effect negative learning curves, at least in the US and France:  www.sciencedirect.com/science/article/pii/S0301421513011440 
That may not always be the case everywhere, as was claimed recently: www.sciencedirect.com/science/article/pii/S0301421516300106  
Though that study has been challenged as unrepresentative and flawed: www.sciencedirect.com/science/article/pii/S0301421516301690 www.sciencedirect.com/science/article/pii/S0301421516301549

It may be that new nuclear technology will emerge that improves on this situation, but that is far from certain. Some want to look again at fast breeder reactors and the use of molten fluoride/thorium systems, but there are many unknowns. Certainly early breeder and experimental thorium reactors proved to have problems: http://thebulletin.org/thorium-wonder-fuel-wasnt7156 . New technology may limit some of the problems, but the economics are unclear, and, even if prototypes prove to be viable, commercial-scale projects are likely to be decades away. Some also look to the idea of developing new types of small reactors. In the past attempts have been made to improve the economics of nuclear plants by going for larger scale units, with little success. There would seem to be no reason why going for smaller scale units would be any more successful: http://ieer.org/wp/wp-content/uploads/2013/08/SmallModularReactors.RevisedSept2013.pdf www.sciencedirect.com/science/article/pii/S0301421513011440

Fusion remains the ultimate dream. If it can be successfully developed, it would avoid the fuel resource limits of fission: without breeders, there is not enough uranium to run fission plants for more than a few decades, depending on the number of plants in operation. However, despite very large scale funding, there is some way to go to viable fusion technology, and, although break-through are always possible, at present it seems that, even if all goes well with experimental tests, workable commercial-scale fusion reactors will not be available until the second half of this century.

Moreover, even if it can be developed successfully, fusion may have limits. Quite apart from the unknown cost, there would be safety and security issues. Plasma based fusion plants would contain radioactive tritium gas at very high temperatures, and there is a risk of leaks or even explosive loss of containment.  They may not generate very long lived radioactive wastes, but activated components would have to be stripped out regularly and stored somewhere. Fusion reactors are likely to need lithium, to make tritium, and reserves of that are finite, and are already in demand for the batteries of electric vehicles. Looking far ahead, it may be possible to find fuel for fusion plants on the asteroids or elsewhere in the solar system. That is fortunate since, if we are to engage in extensive interplanetary travel, we may need fusion energy for propulsion. For the nearer future however, on earth, fusion seems almost totally irrelevant. Unlike renewables, which are available now, for the foreseeable future it can make no contribution to dealing with the urgent problem of climate change. 

Why not both?

While that may be true, couldn’t nuclear fission support renewables, at least in the interim while fusion and/or more renewables are developed longer term? This seem unlikely on any significant scale for a range of practical reasons. Nuclear plants are usually run 24/7 to recoup their high capital costs, although their output can be varied to some extent, so in theory they might be able to balance the variable output from renewables. In practice however this would be difficult for the regular short variations associated with renewables: there are operational and safety constraints limiting nuclear plants to relative slow, infrequent ramping up and down. Some of them can follow the slow daily cycles in demand, but they could not balance the rapid and frequent, minute by minute, variations in wind and solar availability.  Basically, they are inflexible. So rather than complementing renewables, they can play no real role in flexible energy system that will be needed. 


Maintaining costly nuclear plants to provide back up when wind and solar were low for long periods would clearly not make sense since there are other much cheaper options for that occasional standby role. Moreover, if there was a large nuclear element on the grid, then at times of low demand (at night in summer), its input to the grid would conflict with any renewable energy input that was available. One or other, or both, would have to give way, dumping power wastefully. It is conceivable that storage and exports could compensate for this and that smaller more variable nuclear plants may emerge, possible feeding waste heat to nearby users, although there would be safety and security issues with locating mini-nuclear plants in or near cites, as has been proposed. For the present, the current generation of large inflexible, usually remotely sited, nuclear plants and widely distributed variable renewables do not fit well together on the same grid. The two options are not compatible at large scale.