Nuclear plants are designed to run flat out, in part to recoup their large construction costs. Their output can be varied a bit, but this entails thermal stresses and potential safety issues with the build up of active xenon gas that is released when fission reactions are reduced. It needs time to decay. That limits how often and how quickly the plant can be ramped down and then back up so as to match changes in energy demand (“load following”) and the varying output of renewables. So basically nuclear plants are inflexible. Do they have any role for balancing variable renewables? 

It is sometimes claimed that some reactors can do better than others. For example, EDF says its reactors can vary outputs to match renewables by 80% twice a day. That’s fine for dealing with the daily cycles in energy demand, but what about cycling them more often?

A new study by Craig Morris says that, although some individual French and German nuclear plants can (and do) ramp up and down rapidly, they can’t do this often and he notes that, in such cases, most of the rest of the nuclear fleet continues to run inflexibly, so as to avoid economic losses. If more nuclear plants had to operate flexibly, the nuclear fleet economics would suffer, posing a serious problem for France if it wants to retain a 50% nuclear element while expanding renewables. So Morris concludes that, while ramping is possible, so far, as a whole, “the French and German reactor fleets, held to be the most flexible world wide, do not seem to have ever ramped by more than a third in a day, which is less than gas and coal”, and are unlikely to be able to do this more.  In which case, their continued use, as renewables expand, will impose increasing costs on the energy system, with green power curtailment losses mounting more than they would do if nuclear was not on the grid and more effective balancing measure were favoured.

In exploring this issue Morris argues it’s important “to make a distinction between ancillary services (to support grid frequency) and proper load-following”. The former are limited to a small percentage (generally 5% or less in the literature) of power output adjustment; such changes are indeed frequent in the German and French reactor fleets. Load-following is potentially much larger, “so the question is what the maximum upward and downward ramp could be – and how often it could occur both per day and over a reactor’s service life”. And his simple message is that they can’t all do it much or often without major problems.

Basically, Morris says nuclear just gets in the way: a mix of nuclear, wind and solar will be the most expensive option, unless future nuclear reactors can ramp like current open-cycle gas turbines. Which is unlikely. Certainly for the large reactor designs currently being built or proposed.

The report doesn’t look at small modular reactors (SMRs). Some may be more flexible. For example, NuScale’s mini PWR system is planned to have six 50 MW modules, each of which can be ramped up and down (or not) separately, so spreading the strain and cost. Since it uses well established PWR technology, that is the most developed SMR option so far. Most of the other SMR options are further off and, if they eventually reach commercial deployment, they are all likely to have to be in or near cities, so they can supply heat to them to offset their cost. Some may be able to vary their output if run in CHP mode – by changing the ratio of heat to power output. The molten salt reactor design is claimed to be able to do this. But, if they ever become a reality, would we really want to operate these small near-urban nuclear plants in this mode?

Meanwhile, given the load following issue and existing reactor types, it is hard to make sense of the current grid balancing situation in the UK. None of the UK nuclear plants load follow, but they have been included in the capacity market, with extra contracts for being available to cover when there is a supply shortfall, e.g. when renewables are low and/or demand high. The 24 GW of gas capacity that has been contracted in the latest T4, four years ahead, capacity market auction round can do that, it’s very flexible. So, of course, can the smaller amount of storage capacity that has won contracts, at least for a short period. Even the few coal plants still left can do that. But the near 8 GW of nuclear? It’s meant to be run as fixed base-load. Is its output actually kept a bit low just in case its full output is ever needed? Unlikely.

There will be a bit of flexibility, as Craig Morris suggests, for frequency balancing, as with all large power plants on the grid; see my last post. Is it that we are paying nuclear for? At £8.40/kWh pa, under the overall T4 contract.  A nice little earner. Certainly, in terms of  power, with around 42 GW of flexible capacity in all contracted under T4 (leaving out nuclear) there should always be something available to meet power shortfalls, even with the current 35 GW or so of renewables, which might sometimes not be delivering much. And with all the fossil plants and pumped hydro/battery storage also being available to back up nuclear should it go off line, as has happened quite regularly.

Renewables will continue to expand of course – by 2035 there might be 45 GW. But just in case you thought that balancing some of that with nuclear might be possible in future, the Hinkley nuclear EPR plant is not scheduled to load-follow. And it seems unlikely if any of the other proposed new large nuclear plants (Wylfa, Oldbury, Moorside, Sizewell, Bradwell) would do – it would undermine their already precarious economics. Though as now, they may be added to the capacity market, to be there for background support, if that makes any sense. A more cynical view is that, as now, this inclusion is just a way to provide nuclear with an extra subsidy, which, like the rest of the contracted capacity, is paid for by a surcharge on consumers’ bills.

Some say that the whole capacity market is a bit of a con. Here’s an overall Capacity Market policy critique. Certainly we do need balancing capacity, although storage and demand side management, and maybe interconnector imports, ought to be preferred and expanded. Since they are flexible, gas plants can also be useful and relatively cheap to run, although it does still seem odd to subsidize them for this purpose – but that’s partly since renewables are getting so cheap there might otherwise not be enough of them left to provide balancing. Longer term, gas plants could also use biogas and syngas, instead of fossil gas, so there is still a case for them for balancing. However, subsidizing nuclear even more via the Capacity Market payment seems really odd.

Looking more broadly, beyond balancing, it is also sometimes claimed that we will need nuclear in the UK and globally, since renewables can’t be scaled up fast enough to deal with climate change, whereas nuclear allegedly can. That has taken a bit of knock from a paper by Amory Lovins et al., which challenges the way historical data has been used and points to the rapid current growth of renewables.

That’s also the case in Africa, where, as I report in my next post, new approaches are being adopted to accelerate renewables deployment, leaving nuclear, at best, on the margins.