As the capital cost of wind and solar photovoltaics (PV) falls, some say it makes sense to oversize their installed capacity, to ensure that a greater proportion of energy demand can be met during lower wind or PV-availability periods. Even if it means that, at other times, there would be too much power output and a need to dump, or curtail, it. As I noted previously, the UK Energy Research Centre (UKERC) made that point years ago — curtailment of excess output is not necessarily economically irrational. “Some level of curtailment may be both economically rational and sensible from a system operation perspective — so, in isolation, a degree of curtailment is not necessarily an indicator of the unsuitability of any particular form of variable renewable generation,” the centre said.
A recent article on The Conversation website, also developed at energypost, claims that this could be the case with renewable capacity doubled — presumably over what’s needed to mostly meet average annual demand — and significant curtailment of output accepted. It says “overcoming the natural variability of solar and wind can be accomplished at costs below current grid costs (so-called ‘grid parity’) by overbuilding solar and wind resources and adopting a grid operating strategy of allowing about 20% to 40% curtailment of excess energy generation”.
However, curtailment is still a waste of potentially valuable green power, even if it is only valuable if it can be used at some other time. Hence the attraction of storage, and, in particular, “Power to Gas” (P2G) conversion of surpluses to hydrogen, which can be stored if necessary over long periods and later be converted back to power to meet demand peaks and/or long lulls in wind and solar output.
The Conversation article, which is based on a US solar PV study, argues that curtailment will be cheaper than storage, which would obviously be true if demand could still be met, but it is less clear if it can’t, e.g. at peak times or during long wind and solar lulls. It also mentions demand side management (DSM) in passing i.e. delaying peaks. That may also be a cheaper option than storage/P2G if any residual demand at that point can still be met. But if not, then the value of power from backup generators fed from stores, or from grid imports from areas where there is excess green power at that point, would be high.
The optimal economic mix of curtailment, storage, imports and DSM will vary by location and time of day but, although storage/P2G is currently relatively costly, it looks like that will change, so curtailment may become decreasingly attractive. However, compensating for that, as PV and wind get increasingly cheap, the cost of providing overcapacity will fall.
Even so, no amount of overcapacity, curtailment, or DSM will help when there is no wind or sun. What’s more, when there is some input, as capacity/load factors improve (63% is claimed for GE’s new offshore wind turbine) there will be fewer times of low generation, so less need for overcapacity — and compensatory curtailment.
Nevertheless, in some high renewables scenarios, there would be a large wind and PV capacity and some of the output would be used to make hydrogen for later use for balancing, and for vehicles and heating, this share being temporarily reduced when demand for power was high. So there would be high capacity, but no need for curtailment.
However, there are issues around the economics of just using expensive electrolysers part time. It’s a little complicated since, firstly, electrolysers run more efficiently part-loaded and secondly, the best time to run them is when there is cheap surplus power. However, some reckon we should have full-time dedicated hydrogen production from electrolysers, using continuous renewable power, not just surpluses. For example, Navigant says “once all the demand for direct electricity is satisfied with renewable energy, you can build additional wind turbines and solar panels specifically dedicated to producing green hydrogen”.
A possible problem with this idea is that the power input from the extra renewables, just like the directly used renewable power, would be variable so it wouldn’t support continuous hydrogen production. We are back where we started…Though it may be that some surplus power could be stored short-term in batteries to be used later, when there was a lull in power availability, so that hydrogen could be produced on a continuous basis.
That may be too complex and, in any case, it may not be necessary. There are more efficient electrolysis technologies emerging that may change the whole thing, some with new feedstock inputs, others with new synfuel outputs, potentially opening up new markets and new balancing options. There certainly should be plenty of demand for hydrogen, however or whenever it is generated. For example, given the need for extra heat in winter, if too much hydrogen is made in the summer it could be fed to inter-seasonal cavern stores. There would also be more continuous demand for hydrogen, for example for vehicles and to replace the use of fossil hydrocarbons in industrial processes such as paint manufacture. With a lot of cheap low-carbon power potentially available, all sorts of new batch production chemical process options may also be developed. So there may be new possible uses for hydrogen, short- and long-term, continuously produced and/or batch produced.
Is any of this realistic? The UK Committee on Climate Change (CCC) is not too optimistic about either P2G approach — using surpluses or going for continuous supply. “While there is some opportunity to utilise some ‘surplus’ electricity (e.g. from renewables generating at times of low demand) for hydrogen production, our modelling shows that the quantity is likely to be small in comparison to the potential scale of hydrogen demand,” the committee says. “Producing hydrogen in bulk from electrolysis would be much more expensive and would entail extremely challenging build rates for zero-carbon electricity generation capacity.”
That seems very conservative. But then the CCC only proposes a relatively limited amount of new wind and PV capacity, and most of its hydrogen (225 TWh) is derived from steam reformation of fossil gas, coupled with carbon capture and storage (CCS) to make it low carbon. “If all hydrogen in our scenarios were produced via electrolysis this would increase electricity generation by over 305 TWh,” it says.
Well yes, that’s roughly what The Conservation article implies but, as argued above, rather than leading to massive power surpluses at times, renewables could be used, via P2G conversion, for grid-balancing and a range of other valuable new end-uses. The CCC does see 70 TWh of its fossil gas-derived hydrogen being used for transport, mostly shipping having been converted to ammonia, and 120 TWh for industry, but only has 2 TWh assigned to peak power support.
With much more being available via P2G from the much larger renewable capacity proposed in The Conversation article, much more could be used for balancing, without the need for the use of fossil gas, or any need for CCS. Assuming, of course, that P2G/electrolysis does get cheaper. If that does happen, there will be lots of interesting trade-offs and options to consider as we seek to find an optimal mix for balancing and new end-uses.
The bottom line? Overcapacity may be sensible, especially if the resultant surpluses are converted to hydrogen to help with balancing and/or for heating, vehicle and industrial applications. If the cost of P2G electrolysis is low, overcapacity and P2G combined would limit the need to have fossil backup plants to deal with renewable variations, and also reduce the economic problem faced by renewables — that they cannot always generate power at peak demand times, when prices are high, and may generate at times when demand and prices are low. Instead, with P2G, they may be able to expand even more, while providing their own balancing and finding a range of new markets, so helping to cut emissions even more. Sounds too good to be true? In my next post I look at some of the problems with P2G and hydrogen gas distribution.