All entries for Friday 14 October 2005
October 14, 2005
Chris recently raised a debate on energy sources and the future of gas supplies. One of the main points that seem to have spun off from this is rather predictably the nuclear debate. In this post in response to some comments, I will attempt to outline why I feel that nuclear fission is the most viable stop–gap until fusion power can be fully developed to commercial levels hopefully sometime in the middle of this century.
The first point I wish to make is to alleviate a public misconception about nuclear fission – the association of nuclear power generation with nuclear weapons and explosions. Now I know that many people who will be reading this will already be well informed on this so if you don't want a basic talk through the differences then skip the next couple of paragraphs. Now, in order to understand why nuclear reactors cannot explode it is necessary to understand some basic things about nuclear reactions. Fission (the splitting of atoms to create smaller atoms and release energy) releases energy because unstable nuclei (uranium 235 in power generation) split to form two stable nuclei, releasing energy because of the difference in something called the binding energy per nucleon between the unstable and stable nuclei. The reaction is initiated by the absorption by the uranium nucleus of a neutron. The fission process for uranium 235 releases on average 2.4 neutrons per fission reaction (called the reproduction rate), so you can see that in theory the reaction rate would increase with each fission reaction starting another 2.4 fission reactions. This is what happens in a nuclear explosion – a neutron source is put into contact with a mass of fuel greater than the critical mass (minimum needed to sustain a reaction) and the reproduction rate explodes at very high rates. The usual fuel for this is Plutonium 239.
However, naturally occurring uranium is only 0.7% Uranium 235, with the rest being a stable isotope Uranium 238. Thus, most of the neutrons are absorbed by this and do not initiate further reactions, therefore the reproduction rate is much less than one and a chain reaction does not occur. For power generation the target reproduction rate is obviously exactly 1, because then the reaction is happening at a constant rate. In order for this to happen, we first enrich natural uranium to higher levels of Uranium 235 so that a chain reaction can occur (but nothing like the enrichment levels in nuclear weapons). Furthermore, we utilise control rods (usually graphite) to absorb excess neutrons as well, which we can move in and out of the nuclear fuel to control the reproduction ratio to exactly 1. The reaction is very stable and simply controlled, and even if the control rods are completely removed the reaction will not explode out of control because the fuel is not rich enough. Instead, what happens is something like happened at Chernobyl, where the control rods were almost completely removed in an experiment where the safety features were disabled and the reaction rate increased dramatically, causing the temperature to rise and the reactor to fail. This incident was a meltdown, not a nuclear explosion. The whole incident was a catalogue of errors which went right back to the design of the reactor itself, and would never happen in a more modern reactor. For a more in-depth look, check out this site. Essentially, it was a catastrophic operator error incident that actually caused the reactor to fail, something I would not unfairly liken to trying to get a child to loop the loop in a 747 jet (in case you're wondering, such a thing is technically impossible – either the wings would fall off or the plane would stall).
There are three other concerns raised by most members of the public – costs, long term waste disposal and leak risks from accidents or terrorist attacks. Long term waste disposal is perhaps the single largest of these. High level waste is actually produced in very small quantities, and is difficult to deal with. The best idea I have seen to date is a system involving a facility called a geological repository. The waste is safely buried in a stable rock formation deep underground in a deserted area, a long way from any water tables or risks of geological movement. This is a very high initial capital cost solution in one respect, but overall I believe the economies of scale of having a one–off disposal cost at a site that could hold many thousands of tonnes of nuclear waste would pay off very well. The threat of radiation is solved by the natural barrier of hundreds of metres of rock, and let us not forget that the earth itself is highly radioactive underground. The best place to put the stuff is back where it came from, where it can't hurt anybody and can't go anywhere. If anyone has reasons as to why this would be unsafe, please let me know. To give you an idea of quantities, low level waste accounts for 90% of the volume of radioactive material, intermediate level waste a further 7% and high level waste 3%. This equates to about 3 cubic metres a year for a typical large nuclear reactor. Compare this to the hundreds of thousands of tonnes of CO2 a comparative fossil fuel station emits, and we start to see an argument stacking up in favour of nuclear power. High level waste is solidified by mixing with glass–forming compounds, which are then poured into stainless steel containers and welded shut. After a period of approximately 40 years (during which these "vitrified" wastes are held in dry concrete casks or underwater in specialised ponds), the material has dropped to 0.1% of its initial activity. It is then suitable for storage deep underground. In the US, a facility in Nevada for long term storage has been costed at about 0.1 cents per kWh, which is comparatively low to the consumer price, which is in the order of 100 times this, thus the long term storage represents about 1% of cost.
Only about 3% of spent fuel from a reactor is actually waste. In Europe (primarily at Sellafied in the UK and a similar site in France), most fuel is reprocessed with the 3% waste removed, and the remaining 97% returned to rods to be used again as fuel. In this way, we both extend our fuel supplies dramatically and at the same time cut down on the amount of high level waste that requires burial. For more information on wastes check out this site (my specific data on wastes has largely been sourced from this article).
It's worth mentioning as an aside how long known uranium ore supplies would last. It is estimated currently in the 100's of years (although I couldn't find specific data); and that is without fuel reprocessing and the use of breeder reactors, which would increase reserves by a factor of 60–70. In short, fuel supply is not a problem for the foreseeable future of nuclear power.
Cost issues are also another commonly raised concern, with many opponents to nuclear power arguing that decommissioning is not taken into account and is hugely expensive. I already gave some ballpark figures for long term storage costs for nuclear wastes (although this is the storage cost itself; the solidification process and 40 year storage period is not costed into this). Exact costs are hard to find and calculate, as of course it is a long term decommissioning process, and varies depending on reactor type. Furthermore, there are multiple options to take in the decommissioning process, and the whole task itself is still one which we are gaining in experience all the time, making long term reductions in costs even harder to calculate. Because of this complexity, facts are often misreported and twisted by opponents of nuclear power. At present, nuclear generators in the US are collecting between 0.1 and 0.2 cents per kWh to cover decommissioning costs, again not a huge sum. This site explains decommissioning including costing in much more depth that I will go into here, but suffice to say that if you want to know more about decommissioning then it's a pretty good introduction. Another cost analysis is available from the same website here if you wish to compare generating costs with rivals.
So what about safety? Well again I refer you to my favourite website for more information, but I'll summarise here. All modern nuclear power stations must have containment systems built around the reactor (Chernobyl did not have this). Current standards state that reactors must be built with a one in 10,000 year core damage frequency, however modern designs exceed this by a factor of between 10 and 100. Designs in the next decade are likely to see this increase by a factor of 10 again. In 12,000 years of combined reactor operation, there have been two major accidents; one at 3 mile island in 1979 where the leak was contained by the built in containment around the reactor, and the Chernobyl incident discussed previously. Other factors need to be considered against other industries too. For example, every year around 1,000 people die mining coal. Long term damage to the environment of CO2 and the health implications that this will cause to mankind – increased environmental disasters killing many people etc etc. Quoting directly from the above site, optimum safety is achieved by:
- high quality design and construction
- equipment that prevents operational disturbances developing into problems
- redundant and diverse systems to detect problems, control damage to the fuel and prevent significant radioactive releases
- provision to confine the effects of serious fuel damage to the plant itself
The safety systems include a series of physical barriers between the radioactive reactor core and the environment, the provision of multiple safety systems, each with backup and designed to accommodate human error. Safety systems account for about one quarter of the capital cost of such reactors."
Other safety features of nuclear power stations include things like automatic shutdown in periods of seismic activity. Post the WTC attacks, various studies have concluded that nuclear power installations are amongst the safest civil buildings and most resistant to terrorist attack. Even when models were generated of the largest civil aviation aircraft impacting at high speed while fully fuelled, there was no perceived risk of a breach to either the reactor fuel or storage for wastes, or transport casks.
It's also worth putting the severity of nuclear incidents into perspective. In the case of the Chernobyl disaster (by far the most catastrophic on record), the incident killed 31 people, 28 of which died within weeks of radiation exposure. It also caused radiation sickness in between 200–300 emergency workers. About 130,000 people received doses above internationally accepted limits. About 800 cases of thyroid cancer in children have been linked to the incident, of which most have been cured, around 10 have been fatal. No detected increase in other cancers has yet shown itself, although there is an expected rise. The WHO is still closely monitoring the after–effects to this day. We can conclude that while a terrible and tragic incident (and even more importantly, avoidable), it's hardly a doomsday end of the world type scenario as is often predicted by those who wish to over–dramatise the effect of nuclear catastrophe to strengthen their anti–nuclear case.
Perhaps the most pertinent question of all though remains – do we need it? Well, the UK has a base demand of approximately 47GW of electricity. This has to be generated somehow. By the government's own targets, even by 2050 so–called renewable energy sources, which are comparatively new and untested on the scales necessary to satisfy such base loads, potentially expensive and unreliable and also require backup generation, will only satisfy 60% of UK demand, and compared to many other parts of the world this is a pretty high target, although not the highest. In the meantime, where are we going to get our energy? Renewables cannot satisfy the interim demand, and in my judgement cannot meet long term demand and expansion either, certainly not without large areas put down to expensive wind farms or damming off rivers etc. Biomass schemes are new, not accelerating in growth fast enough and in any case are unlikely to be able to sate our generating demands (although I believe the future of transport lies in bio–fuels, I may make a post on this in the near future). Our options remain to be nuclear or further depletion of fossil fuels, along with associated CO2 emissions. These are the only choices for the short to medium term, regardless of what alternatives you might have faith in. I have put the case for nuclear here. What's your choice?