Nuclear v. renewables: accounting for intermittency
In what follows the term "baseload" is the minimum steady demand for (electrical) power over a period of hours or days. In terms of management of the Grid it's the power one needs to be generating from the sources that take longest to fire up and shut down. In Ye Olde Days this used to be coal; now it's coal (and biomass at Drax) and nuclear.
One then needs to have sources which can be turned up and down more quickly to make up the difference between the baseload supply and the actual demand at any hour, minute or second. In the UK these tend to be oil (I think there's still some about), gas, hydro and pumped storage. The latter two can be turned on and off very quickly, e.g. to match the sudden surge when millions of people put the kettle on after a popular TV programme!
Wind, solar, wave and tide don't fit at all well into this set-up: they are available when conditions are right for them rather than when they're needed. This has two consequences:
- to ensure availability one tends to have to have backup generation capacity from sources such as coal and gas to fill in when the wind's not blowing, sun not shining etc. Simplistically one needs exactly as much backup fossil-fuelled generating capacity as "renewable". Day-to-day one needs to call on backup capacity less if one has a range of different sources e.g. wind spread over a continent-sized area, tidal generators at different points around the British Isles etc, as the probability of all sources slumping together is less, but since there is still a certain probability that they will all slump together sometimes, having less than the total capacity of backup generation available would seem to be courting eventual disaster.
- however diverse the mix of intermittent sources they are always a more-or-less poor match for demand, so some other sources (typically fossil fuels) are required to fill in for them over time. (I'd be interested to know the actual figures.)
Of course nuclear, as a base-load source, also requires fill-in sources to match actual hour-by-hour, minute-by-minute and second-by-second demand. And this brings me (rather circuitously :-)) to my point: critics of nuclear / proponents of wind, solar etc criticise the cost of nuclear power stations but blithely compare the net cost of their preferred technologies without accounting for the costs either of the capacity needed to back them up, or of their intermittency. If we're serious about sustainable carbon-free energy then we need to account for the costs of matching supply to demand so we don't have to use any GHG-generating fuels. There are various ways of achieving this: storage schemes (pumped, compressed air, fleets of electric vehicles whose batteries can be used to put energy back into the grid when needed etc), smart-metering and demand management. But just as the costs of decommissioning nuclear power stations are built into the price of electricity they produce, so the cost of matching supply to demand should be built into the price of electricity from intermittent sources.
One way to do that might be through free market economics. At present when there's a surplus of energy, say because the sun is shining and the wind blowing, "renewable" generators actually get paid to not generate electricity! But if the owners of all generating plant were allowed to charge by the second for the energy they produce, what I think would happen is that when the sun is not shining and the wind not blowing then baseload generators (principally nuclear but also geothermal, incineration and river flow) would be able to name their price. When the sun is shining and the wind blowing of course the PVs and wind gennies have power to sell, but the baseload generators still need to get rid of the energy they are generating and would (and could afford to, with the profits from selling their energy in times of scarcity) engage in a price-cutting war to sell theirs, undercutting the intermittents and depriving them of income. To survive, the intermittents would have to invest in storage and demand management etc. so that they could compete with baseload generators, either by having stored-up power to sell at times they're not generating much directly, or by reducing demand at times they're not generating so much power. It would be interesting to see how competitive the intermittent generators, encumbered by these extra costs, would be with nuclear.
Of course we would like to generate all our power without producing GHGs and if we weighted the penalty for GHG-producing technologies appropriately it would become economic for baseload operators to also invest in storage, demand management etc. — though I think they would have to invest far less to achieve 100% supply-demand matching than the intermittents would.
What would get particularly interesting is if, instead of the baseloaders and the intermittents each having their own set of supply-demand matching resources, these resources were owned and operated by third-party operators who simply bought energy from the generators and sold it to consumers, taking a cut of the price to make a living. Then any promising new (not fully despatchable) generation technology could enter the market without having to invest in its own storage/demand-management technology and would be rewarded by not only how cheaply they could produce energy but on how well their supply matched demands. And it could spur the development of storage and demand-management technologies as profitable activities in their own rights.