Nuclear power plants aren't just for base load

It is more than 2 weeks since my last post which is very poor - I do apologise. I plead pressure of work! Not only have I started a fascinating but horribly time consuming paid research project but I am also preparing for my first radio broadcast - I and my friend Roger Frost are putting together a fortnightly science program on Cambridge 105 community radio station. Fingers crossed you will be able to catch our debut on Sunday at 1.30pm. Wish me luck!

Back to this blog. I did promise to comment some more on the Adam Smith Research Trust report 'Renewable Energy Vision or Mirage'. The report looks at whether or not renewable energy sources are a good way to address our energy security issues - with cheap fossil fuels becoming more and more scarce. The overall conclusion is that the only renewable technology with significant capacity is wind, but this is too expensive as well as being too unreliable to be useful, at least without an energy storage capacity which we don't have and won't have any time soon. Instead, the report advocates heavy investment in nuclear power. You may recall from my last post that the 2050 pathways model predicts that a mainly nuclear strategy would be slightly cheaper than a mainly renewables strategy, purely on cost grounds. Neither that nor the Adam Smith report discuss environmental costs, which are an entirely different issue.

The most interesting thing in the report for me, though, is the fact that nuclear power stations can react to changes in power demand, just as fossil fuel power stations can, at least as long as the changes are fairly gentle. Previously I was under the impression that nuclear power can only run at a steady rate because changes in load damaged the fuel rods. This would mean they are only good for base load, not load following demand. It turns out that this is not exactly true any more, because a great deal of effort has gone into designing power stations that are more adaptable.

In summer time, average UK electricity demand varies between about 20 GW at six in the morning to about 40 GW at noon. In the winter the load is higher and the peak time is early evening rather than noon but the range is still about 20 GW between the highest and lowest daily demand. So can nuclear power stations ramp up and down between 50% and full capacity on a daily basis? France gets about 75% of its  energy from nuclear power and that wouldn't be possible if they weren't capable of some load following. It turns out that French power stations can and do load follow,  typically down to 50% load or even less for some time of the day, taking 30 minutes to an hour to ramp up or down (less than 1.5% of capacity per minute). This is achieved by some very clever design involving patterns of control rod movements to minimise temperature gradients in the fuel pellets. Without care you can get up to 450°C difference between the centre and edge of a 9mm pellet. As you can imagine, this leads to different thermal expansion between the middle and the edge and the pellets can crack -  which is bad news because you get leaks of corrosive gases into the fuel casing which damages the rods. So keeping the temperatures nice and even is very important. Also, there are some subtle but significant effects from changes in load due to the delayed response of some of the fission decay products. For example after a decrease in reactor load, the concentration of Xe¹³⁵ eventually peaks 7-8 hours later. Xe¹³⁵ absorbs neutrons strongly so it tends to slow the fission reaction in the core and adjustments to the control rods are needed just to keep going, never mind if there is a need to ramp the power up again. So the schedulers try to avoid power cycles with 7-8 hours between drop and peak [1].

In any case, these technical difficulties can be avoided or overcome and nuclear power plants can load follow to a considerable degree. The main reason we don't do this in the UK is that it makes the power stations more expensive.  For gas and coal power stations, a lot of the cost is in the fuel and if they aren't running they aren't consuming fuel so there is money saved. However, for nuclear power the fuel accounts for at most 16% of the cost of the power station and they cost money whether they are generating or not. Starting from a typical load factor of 85%,  if this was reduced to 50% the cost of nuclear power would go up by more than a third.

If we do choose a mainly nuclear power strategy for our electricity supply we have to accept it won't be cheap.


[1] Technical and Economic Aspects of Load Following with Nuclear Power Plants (NEA, 2011)