Labour’s announcement on Sunday of its determination to entirely decarbonise our electricity supply is something, if I were still a civil servant, I would describe it as ‘bold’. Very bold, in fact.
But something being bold doesn’t mean it can’t be done, and in considering Labour’s proposal we should look at two things: firstly, is it physically possible – as in, can you design a power system that delivers the outcome without invoking the technological equivalent of magic. Secondly, is there a way of doing it cheaply?
The physical challenge
On the first question, it’s important to recognise that even existing plans would see us quite a long way down the road to decarbonising power by 2030 anyway – the Government already has a 2035 target for this. So Labour’s policy is about accelerating existing plans, rather than coming up with a totally new one.
To get a sense of the challenge, let’s turn to the following chart from the Climate Change Committee’s advice on Carbon Budget 6:
To get to zero-carbon power we need to strip out all of that purple fossil fuel bar by 2030, which represents electricity generation totalling about 32TWh – or the demand of eight million homes. However, you can’t just do a one-for-one replacement with renewables, you have to replace it with generation that has the same characteristics to ensure system security.
For reference, 32TWh is about the annual output of 4GW of combined cycle gas turbine power station, or one-and-a-half West Burton B plants. So you might think that you could build perhaps two more gas power plants, capture their carbon and put it into a store, and you’d be done.
But, again, it’s not quite that simple. That 32TWh conceals a very large number of plants that aren’t being run flat-out, as in our 4GW hypothetical. As the below chart from National Grid’s Future Energy Scenarios shows, actual unabated fossil capacity in 2030 is in the region of 20-40GW. All of these plants are being run at very low load factors, even compared to the present day. Realistically, you need to replace them like for like.
So what are your options?
Labour’s announcement makes reference to nuclear, but it’s difficult to see any additional very large-scale plant coming on stream by 2030 beyond Hinkley C. You might be able to get 500-1000MW of small modular reactors, which in theory take less time to develop, although most commentators assume they won’t come online till after 2030.
But – if you really went for it – there’s no reason you couldn’t build at least one before 2030. So let’s assume you can have one 500MW SMR running 90% of the time. This is about 4TWh. You’ll also want to deploy carbon capture and storage on as many gas assets as you can. Let’s assume 5GW of that 40GW can run flexibly and the economics of CCS stack up for it.
At the 40% load factor for natural gas CCS National Grid assumes, that gives you 17.5TWh. Only 10.5TWh to go!
This last bit is the hardest, and will need a mix of solutions, as it’s the assets that will deliver system security. National Grid already assumes a nice big chunk of storage, primarily batteries, which provide rapid response. But we need other assets capable of maintaining output during a significant dip in wind. I’m going to assume 5GW of this is a long-duration storage solution like liquid air, and 30GW is hydrogen turbines. Let’s assume a round-trip efficiency (RTE) of 50% for both.
These assets would be running at load factors of about 4%. You’d need storage for hydrogen and a transport network too. But not that much storage – assuming this underground facility can be pressurised up to about 40-50 bar, it would be enough. You also need the power to make that hydrogen. Assuming 50% RTE and 50% load factor for offshore wind, this is about 2GW of dedicated offshore wind.
This very quick mental model of the power system indicates that Labour’s plan is physically doable, even if there are some supply chain constraints to solve – for example, hydrogen turbines don’t yet exist in the UK at a meaningful scale, and we’d need to build a lot in the six years from a 2024 general election.
The cost of decarbonisation
This brings us neatly to our next question: how on earth could we deliver all this cheaply?
To answer this, I want you to throw all the technologies I reference above out of your head. There’s a terrible temptation to treat the energy system like some kind of enormous train set, where you can deploy your preferred TT gauge assets wherever you like – and that because something is physically possible must mean it’s the right solution.
But, of course, the real world is much more unknowable than this. What you want to do is now that you’ve checked something is possible in theory is to lay the conditions for companies to deliver it in practice. There are two key components to think about here.
The first is defining the outcomes you want to deliver rather than the technologies you prefer. We want a stable, secure and cost-effective zero-carbon power system by 2030. So we should structure our policy to deliver that, rather than a certain number of, say, hydrogen assets.
This means markets. Many of the right kind of markets are already in place – the Capacity Market is responsible for delivering security of supply, we have routes to market for various low-carbon technologies and we’re evolving the digital side of the system to enable more consumers to make money out of their new electric cars and heat pumps. These markets are somewhat fragmented – bitter experience has taught policymakers that trying to do everything via a single wholesale price can lead to perverse outcomes – but they are preferable to Whitehall deciding the precise model of wind turbine to build on a village green.
Long-term contracts dominate both energy markets and energy policy, and for good reason: large capital-intensive assets can only be cheaply financed with a degree of revenue certainty. Getting the design of the markets that award these right is critical, and ideally should be something handled by a non-government entity. A good option may be to split the wholesale market into a short-run market for plant you can switch on and off at will, and a long-term market for assets like wind turbines and nuclear.
The alternative to markets for replacing our 32TWh of generation is likely to be government contracting for some kind of strategic reserve; given the short running period of our hypothetical hydrogen turbines, this might be the quickest way of getting them built. But speed does not translate well to low cost, and we should be careful not to over-privilege certainty of delivery over price.
The second component is that actually building things in the UK is hard. There’s a good reason that the Government’s proposed relaxation of planning constraints, both within investment zones and for onshore wind, was met with significantly wider acclaim than its tax cuts.
Put simply, there is no way of decarbonising our power system by 2030 – which will require new pylons as well as generation – without making it considerably faster and cheaper to get permission to build. At the moment delivering a new onshore or offshore wind farm can take five to seven years, which would eliminate any possibility of achieving Labour’s target.
Likewise, getting the vast amount of wind out of Scotland to England, where the bulk of our island’s demand is, requires reinforcing connections over the border, and potentially upsetting people who’ll be able to see them. It took nearly a decade for the Beauly-Denny transmission line to be built in the face of ferocious local objections. New infrastructure just can’t take that long again. While this may run counter to the instincts of some in the Labour Party, it’s a truth they should embrace.
Click here to subscribe to our daily briefing – the best pieces from CapX and across the web.
CapX depends on the generosity of its readers. If you value what we do, please consider making a donation.