I’ve been following lots of analysis around the recently-closed auction to procure guarantees of electricity capacity for next winter here in the UK (winter being when demand peaks). Much of what I’ve read has been critical, not least because much of what I read comes from those – like myself – who are strong supporters of a lower-carbon future, with a system increasingly making use of innovations such as storage and demand response.
I’ll review these criticisms below. However it’s always useful to consider issues from other angles too, not least the political viewpoint (this being a policy mechanism). That’s led me to lay out my thoughts on what the capacity market is – and, importantly, what it isn’t. I’m sure there will be other views, including experiences from capacity markets elsewhere, so I’d certainly welcome any discussion or criticisms!
The Capacity Market (CM) auction is designed to ensure that enough electricity capacity is always available, even in case of unexpected power station outages or peaks in demand. Or, as the newspapers like to breathlessly phrase it, to ensure “that the lights don’t go out”. The backdrop is that the UK has pledged to shut of coal capacity by 2025 and also faces an ageing (i.e. due to close) set of nuclear plants.
Roughly speaking, the CM operates like this:
National Grid say how much capacity they need to guarantee and over which time period. Through an auction system, sources of supply (i.e. generators, interconnection, storage) bid the capacity they will guarantee and the price they want to receive for doing so. Demand-side response can bid too (displacing generating capacity by guaranteeing to reduce the required demand). Stacking up bids from low price to high until the required demand is met, the auction produces a clearing price.
A (provisional) summary of the most recent results can be found here. The summary is:
- 54.4 GW of capacity is guaranteed for the winter of 2017/2018
- Price: £6.95 kW / year; so a total cost to the National Grid – and ultimately customers – of £338 million (i.e. 6.75 £/kW x 54.4 GW x 1,000,000 kW/GW)
- Just over 40% of the contracts go to combined-cycle gas, the biggest source of capacity, with other conventional and existing sources (coal, nuclear) also doing well out of it.
- Demand-side response (reduction) managed only 209 MW (<0.5%)
- Interconnection provides 2.4 GW (4.3%)
- Storage provides 2.7 GW (5%); but most of that is existing pumped hydro
Here’s the list from the summary I linked earlier (CMU = Capacity Market Unit):
I’d sum up these up as follows:
- The price was too low to allow new capacity (or demand reduction) solutions to compete with existing power plants.
- The price is so low that it suggests no great effort was required to persuade those existing power plants to guarantee their availability. They would have been there anyway, so this is just giving them money for nothing.
- The majority of those existing power plants are fossil-fuelled. The process could have been designed to favour low-carbon solutions.
I wouldn’t argue with any of these points. In future, I’d be surprised if the CM rules aren’t evolved to address them. However to better argue the case for that, it’s important to review where we are just now – in a world which is in transition – and to recognise that policymakers face other issues too.
Capacity or Flexibility?
The UK, like many countries, has been building out substantial generating capacity of both solar PV (>11 GW) and wind (>14 GW) As a proponent of renewable power, I’m delighted by this and supportive of more, particularly since their costs continue to fall. However neither can, unaided, guarantee to be available at times when UK demand is at its highest.
On a winter’s evening in the UK (when peak demand occurs) there is no sun. There may also be very little wind. As an example: on January 22nd 2017 from 5-7pm, wind only generated between 500 and 550 MW; from an installed capacity of well over 14,000 MW.
So if we want to keep building more of this clean and increasingly cheap (but inflexible) capacity, we either need to be able to store and time-shift large amounts of it, reduce demand when there’s no wind, import power from overseas, or have enough other generating capacity in the system which we can guarantee to be available when we need it. Hydro would be nice for the latter, otherwise we need sources of capacity that use fuels, either carbon-emitting or nuclear. (I’ve reviewed the fundamentals of this balancing challenge elsewhere)
In practice, we’ll be well advised to consider all of these. That means building a system with capacity that is both flexible and available. To do so will require innovation and political support (guided by scientific and economic evidence), but also money (investment).
The solutions mentioned above do, to varying extents, all appear in the capacity market results I showed earlier. They might not be in the relative proportions you or I might prefer (especially if you are selling one of them).
The mix of solutions does differ for secured capacity further into the future. Here are the results for an auction that took place at the end of last year, covering the winter of 2020/2021:
It was a more expensive result, clearing at a price of £22.5 kW/yr (meaning a total cost approaching £1.2 billion). Gas is still the main guarantor of capacity, but storage is up at 6% from 5% (with 500 MW from battery) and DSR is at 2.7% (rather than <0.5%). Interconnection is essentially unchanged. Notable also is that coal/biomass is down at less than 12% (from over 19% in the auction for next winter), in line with the UK’s policy of phasing out coal. Progress?
But why have policymakers not stepped in to favour and drive faster growth of “future, clean” things like storage and demand response in the capacity market, over “old, dirty” things like coal plants or “old, inflexible” things like nuclear?
The answer lies in what the capacity market is (rather than what we might prefer it to be).
The CM as Insurance
The UK government describe the most recent capacity auction as “providing guaranteed electricity capacity at a low cost to bill payers.”
I believe that quote neatly captures the purpose of the capacity market, at least from a policy perspective: it’s an insurance policy, nothing more. One designed to cover the next few years (thus far, up to 2020/21)
It doesn’t exist to drive the adoption of particular technologies and it doesn’t exist to increase system flexibility or decrease carbon emissions, however much we might like it to be so. It exists purely to provide an assurance that “the lights won’t go out” – and to do that with as little immediate impact on consumers’ bills as possible.
Which brings us back to the three criticisms I highlighted earlier: the view that the capacity market is handing money to fossil fuel plants that would probably be available anyhow, squeezing out new and low-carbon alternatives.
If we think of the capacity market as an insurance policy, then the word “probably” in that last sentence is key. My house probably won’t burn down and I probably won’t crash my car. But I still pay for insurance for both because, were either instance to happen, the outcome is far too dire for me to face.
In an electricity system – and for politicians – the risk of power cuts is up there with the risk of your house burning down. Even without power cuts, unfortunate circumstances combined with tight supply can produce very expensive spikes in price as the system operator scrambles to balance supply and demand at short notice. These spikes eventually arrive on customers’ (voters) bills.
Avoiding these outcomes is important for anyone, not least those of us who favour the transition to a low-carbon energy system. Although the trigger points for “system stress” events recently have tended to be thermal plant failures, low wind availability has been a contributing factor too (in failing to provide alternative capacity at the right time). So we need to be sure that other capacity (or demand reduction) is guaranteed to be available.
That’s certainly true now, making talk of blackouts in the UK fanciful.
However the capacity market isn’t about the situation now, its singular aim is to be absolutely sure that the same remains true in the future.
So, would the fossil plants that we currently call on be available in future, even if there were no CM payments?
In the near future, no doubt the vast majority would (as the low clearing price for 2017/2018 suggests). What about in the winter of 2020/2021? Probably? All of them? Which ones might not? Only they really know for sure. The real policy question is whether we are prepared to take a risk or whether we choose to insure against it (through capacity payments).
Let’s say we accept paying to insure against any capacity risk. What if we now skew the market, for example by excluding some of the “least desirable” existing capacity from the CM mechanism or by carving out a portion of the market just for newer solutions such as storage or DSR? Surely solution providers would rush to fill any such market carve-out, providing a big boost to the growth of those technologies? I’ve no doubt that’s true and it would help build a more flexible and future-proof system more quickly.
At this moment though, they are more expensive ways of achieving capacity guarantees – otherwise they’d already be bidding for and winning more share in auctions based on price. So by guaranteeing to include them, we would make our “insurance” more expensive.
Perhaps more storage and DSR will offset the cost of their capacity payments by providing other savings within the system? Almost certainly, as various reports such as this recent one have described. However “quantifying these benefits is very complex“, so how to price them into our insurance policy? Will these benefits accrue at the same time as their costs (the capacity payments), or will policymakers have to explain to consumers how and why there’ll be “jam tomorrow”.
Unfortunately, the political context is that rising electricity prices are once again under the spotlight, as are the costs of subsidising the shift to a low-carbon system.
I’m also dubious about how true it is to say that capacity payments can simply be pocketed in full as extra profit by existing generators. In competitive electricity markets, is this really possible? Or will the extra money power plants get for providing guaranteed capacity be offset by lower prices when bidding into competitive markets for energy supply; bringing profits back into balance? In other words, simply a shift of business model from variable to fixed revenue sources. In the absence of a parallel universe in which they didn’t receive capacity payments, evaluating this is difficult!
Similarly, it’s important to note that having carbon-emitting power plants guaranteeing capacity doesn’t necessarily mean a rise in carbon emissions. Being available to guarantee capacity doesn’t guarantee anything about these plants being called on to generate energy. In fact, guaranteeing capacity for the most challenging times of demand will help us to keep increasing the proportion of solar and wind in the mix. We’ll generate more energy from those and need to call on fossil plants less. That will drive emissions down. That’s what’s been happening, particularly with respect to coal.
I’d love to see faster growth of solutions such as storage and DSR, providing vital balance to the continued growth of variable renewables. In the long-run, I want to arrive at a smart, low-carbon system that is fully dependable because it is flexible; through a mix which relies on these solutions much more than it does on fuelled power plants.
However that is going to be a lengthy transition. Along the way, guaranteeing supply at any point in time remains a vital system requirement. It may require “imperfect” solutions, such as paying an “insurance” to ensure conventional plants remain available (even though being used less) while we build out better alternatives.
It’s easy to criticise the CM because it doesn’t explicitly favour the flexible or low-carbon solutions of the future, but I think that misses the reason for its existence: peak capacity insurance at the lowest cost. Those are key and hence indirect enablers for the electricity transition to continue apace without losing the support of those footing the bill.
So it’s important to recognise a difficult balancing act for policymakers: between speeding the adoption of solutions which are “the future” (long-term gain) while reducing the impact of new, initially more costly technologies on consumer bills (short-term pain).
No doubt we’ll see the capacity market process evolve going forwards. At the moment we have a capacity market which guarantees that peak demand will be met, but in which increased flexibility is an occasional by-product rather than a core goal.
Let’s hope that, in future, we create a market which delivers least-cost flexibility, combined with the same continued guarantees that peak demand will be serviced. The more we can quantify and demonstrate the cost-effectiveness of new flexibility solutions to policymakers and consumers alike, the quicker we’ll get there.