The long and winding road to energy net-zero 2050

For a power plant due to be mothballed amid a long-term plan to abandon electricity generation from coal, West Burton A weathered last winter surprisingly well. And following a letter from secretary of state for business, energy and industrial strategy Kwasi Kwarteng to its owner EDF Energy telling them it would need to be on standby for a while longer, it is probably going to have another good winter.

Amid generation shortages in the winter of 2021, the electricity system operator (ESO) arm of National Grid was forced to call West Burton A back into action to the point that for a while it would match wind generation: 5.5 per cent of the nation’s electricity mix. Though it was the lack of wind that made the headlines, according to the spring 2022 Energy Trends report published by Kwarteng’s department BEIS, the revived dash for gas during late 2021 also had its roots in the problems facing the ageing nuclear fleet as well as decisions by fossil-fuel extractors themselves.

Seasonal demand was increasing just at the point when nuclear operators performed maintenance on their fleet, which cut output by almost 10 per cent to levels last seen in 1982. It was a similar story for North Sea oil and gas, who cut production by close to 20 per cent during the summer of 2021 while global gas prices also continued their climb, hitting 300p at the end of last year. Electricity prices followed almost in lockstep, pushing close to £300/MWh from a low of less than £50/MWh only the year before when, amid Covid lockdowns, gas and oil prices had slumped. Day-ahead contracts for both gas and electricity hit new highs at the start of July on the threat of strikes at Norwegian oilfield operator Equinor.

The dramatic rises seem counter-intuitive at a time when renewables deliver a substantial and increasing chunk of power and at successively lower prices. According to the BEIS Energy Trends reports, renewables’ share of electricity increased in the last quarter of 2021 to almost 43 per cent compared to 2020 and increased again to 45.5 per cent in the first quarter of this year.

The wind farm projects expected to come onstream by 2025 have agreed to deliver power at just under £40/MWh and will, if they are in the contracts-for-difference (CfD) scheme that sets a minimum price for their electricity, wind up paying back the money from higher prices to the publicly owned Low Carbon Contract Company normally used to offset payments for the following quarter. The first quarter of 2022 saw the scheme return money to electricity suppliers for the first time.

The prospect of starting off their CfD arrangement by paying out has seen a number of generators delay their use of the 15-year contracts on their latest farms in order to benefit from the full market price of energy in the short term: they can hold off potentially for as long as three years. According to management consultancy Cornwall Insight, 90 per cent of the hourly periods used to calculate electricity prices from the start of 2022 up to mid-May were above the strike price. On average the wholesale price was more than double the agreed CfD price.

Though its structure makes it vulnerable to larger swings, the UK’s market structure is not all that different from that in many other developed nations in that a ‘merit-order stack’ determines which energy source is prioritised at different network loads. Operators will naturally favour cheap supplies over more expensive options up to the point where they run out, forcing them to use the next more expensive available option until they have met their needs.

Studies such as one by Michael Grubb, professor of energy and climate at UCL, and colleague Paul Drummond published in 2018, have shown that it is a combination of this effect and the way in which prices are set nationally that means it is the marginal cost of that final, last-resort fuel that controls the price seen by consumers, not protection by a price cap.

In March, Cian McLeavey-Reville, senior director of market development at National Grid ESO, explained in a seminar organised by the company on the possible restructuring of the way electricity capacity is managed in the UK that a variety of perverse incentives and shortcomings have become embedded in today’s energy market. “We don’t believe that the current market was designed for net-zero and left unchanged will result in excessive and unnecessary costs for consumers,” he claimed.

If we discount the pricing problems faced by the consumer, the target set by Prime Minister Boris Johnson of a fully decarbonised electricity grid by 2035 – ahead of a net-zero energy sector in 2050 – is not unrealistic from the perspective of building out more low-carbon capacity. Fintan Slye, National Grid ESO executive director, had claimed a few months before the announcement of the target: “We’re confident that by 2025 we will have periods of 100 per cent zero-carbon electricity, with no fossil fuels used to generate power in Great Britain.”

One reason for this optimism is electricity’s role in the current energy mix. At the end of the last decade, gas and oil for heating and transport were responsible for more than three times as much energy as electricity, for a total of almost 1400TWh. That imbalance makes delivering a zero-carbon mix by 2050 more of a stretch compared to net-zero electricity by 2035. However, few expect to have to do it by boosting electricity generation threefold in as many decades.

Scenarios considered by National Grid ESO in a series of annual studies assume efficiency improvements will bring the total required by 2050 down significantly. In the most optimistic forecast, tagged Consumer Transformation in the 2021 edition of its Future Energy Scenarios reports, total energy delivered in 2050 would be 600TWh, almost all of it electrical: double 2020’s electricity usage. A less optimistic but net-zero System Transformation option calls for 800TWh of annual capacity, which equates to an average power delivery of around 90GW. The biggest cuts to energy use would come from substantially electrified transport with industrial and residential energy use falling by around 25 per cent.

Because of its heavier reliance on the use of hydrogen, National Grid’s System Transformation scenario is not far from the near-700TWh picture presented by the UK government in its 2020 Energy White Paper. In the scenarios modelled by BEIS, around 9 per cent of the energy required would continue to come from methane coupled with carbon-capture and storage, pushing green hydrogen to less than 5 per cent. Nuclear would be responsible for 17 per cent and potentially higher than 20 per cent under alternatives. The projected share for nuclear has now risen to 25 per cent of the total with a peak capacity of 24GW announced in the British Energy Security Strategy launched in April.

According to the June 2022 BEIS Energy Trends report, renewables capacity increased by 1.8GW in 2021 to surpass 25GW, with another 1.5GW added in the first quarter of this year alone. Most of the new generation came in the form of offshore wind. And the government’s focus in renewables is now very much on offshore wind, setting a target in the spring of 50GW by 2030, four times the level at the end of 2020. Onshore wind and solar do not have targets. And the former is increasingly becoming mired in planning disputes, though public opinion overall is broadly in favour of the technology and it is the cheaper option. On a cost-per-kWh basis, according to estimates from IRENA (the International Renewable Energy Agency), it can be half that of offshore turbines, which fell to around 9 US cents per kWh by 2019 based on projects around Denmark, which started building these farms in 1991.

Were wind the sole contributor to power delivery, the 2050 target would call for a nameplate power capacity of around 200GW and perhaps a little higher. This figure reflects the lower load factor that renewables have compared to nuclear or gas: a factor of their variability. That would reflect a build-out from 2030 of some 7GW per year, though that is only double the increase seen in 2018.

Trading as much as a third of that for nuclear and gas would bring the increase needed in terms of nameplate capacity down to the same pace as that envisaged for the government’s current targets. However, it is not likely to be the lowest-cost option. The strike prices for nuclear are set today at higher levels – now more than £100/MWh – than those projected for renewables. Construction costs are also likely to be high for nuclear though, once load factor is taken into account, some offshore wind works out more expensive than the £23bn currently estimated for the 3.2GW Hinkley Point C reactor. Novel floating farms such as Hywind Scotland off the coast of Peterhead came with a construction price tag that worked out at almost £9m/MW of nameplate capacity. However, it also recorded an average capacity of almost 60 per cent in 2020 compared to the mid-30s average seen for older offshore wind turbines. Many other fixed-turbine offshore projects that can stand in shallower North Sea waters have fallen to below nuclear, even taking into account the lower load factor. In the meantime, Equinor expects the Tampen floating-wind project in Norway to cut costs by 40 per cent compared to its Hywind Scotland.

The market is also increasingly in favour of borrowing cash to fund wind farms, with a large chunk now coming from pension funds, though this may change if the cost of capital increases in the wake of quantitive tightening by central banks. Though the Scottish government aimed for 11GW of offshore by 2030, up from 1GW in 2020, the ScotWind auction in January 2022 wound up with 17 fixed- and floating-turbine projects for almost 25GW across 7,000 square kilometres of sea area.

This is where the details start getting in the way. A continuing problem is moving wind-generated energy south of the border thanks to congestion on the transmission network that already affects Scotland’s onshore farms. A particular problem lies in the B6 boundary that divides the England and Scotland grid sectors. A report by Aurora Energy Research published in February pointed to National Grid’s plans to add 7GW in transmission capacity across the B6 boundary by 2035 but added “this will not keep pace with the expected build-out of Scottish wind capacity in a net-zero world”. The result is that wind farms have had to suffer additional curtailment – and be forced to dump power – even when that electricity is needed elsewhere in the country. In compiling a report for generation operator Drax, consultants Lane Clark & Peacock found 80 per cent of the energy forced to be dumped through curtailment of renewables was generated in Scotland.

At the March National Grid ESO event, McLeavey-Reville showed a graph that demonstrated how the costs caused by dealing with transmission constraints had risen eight-fold since 2010. The company’s current modelling points to those costs alone possibly hitting £2.5bn by the end of the decade. The ESO has advised Ofgem to change the way electricity is sold so that it reflects where the usage takes place to try to better control regional demand and prevent perverse incentives in the current structure that can lead to power being imported through international interconnectors even when there is cheap power available domestically. However, at the event, some such as Cathy McClay, trading and optimisation director at Sembcorp Energy UK, pointed out that the boundary problem may simply need a lot more high-voltage cables to move energy where it is wanted across Great Britain.

The answer recommended by Aurora and a growing community around renewables is to invest far more heavily in long-duration storage to be able to support many hours of delivery. Right now, according to National Grid ESO, storage accounts for just 1 per cent of the electricity delivered to users over its grid; much of it still comes from the pumped-hydro plants in Scotland and Wales originally built in the 1970s to support an earlier planned transition from fossil fuels to nuclear. The new storage capacity, which represents about a quarter of the current total, is mostly based on a growing collection of lithium-ion battery installations and used by the ESO to prevent short-term stability problems in the grid. Drax, however, wants to expand its pumped-hydro installation at Cruachan in Scotland with 600MW of storage.

According to the Lane Clark & Peacock (LCP) report, 600MW of power would cover almost half of the curtailment situations the team identified in 2020 and 2021. In principle, a greatly increased storage pool could overcome the need to use natural gas at all.

Several groups have calculated the mixture of renewables most likely to be able to support net-zero when coupled with storage. Stanford University’s 100 Per Cent project, headed by Professor Mark Jacobson, estimated in 2017 that a 60/40 split between wind and solar would likely work best. Bruno Cárdenas and colleagues at the University of Nottingham found a mix of 85 per cent wind and 15 per cent solar would deliver the lowest-cost option: it is the one that minimises the amount of storage capacity required.

However, among policymakers, large-scale storage is a relative latecomer. The 2020 Energy White Paper made almost no mention of energy storage. A year later, BEIS and National Grid ESO began to look more closely at a much expanded role for storage to deal with the shortfalls that will inevitably occur when renewables cannot meet peak demand as well as to maximise the amount of energy recovered from renewables. Following a call for evidence last autumn, BEIS expects to update its policy on long-duration later this year.

With the installation of plants like Drax’s expansion at Cruachan, wind operators would be able to send their peak power, in principle, to large pumped-hydro stations in Scotland optimised for long-duration storage. Even then, there remains the issue of later shipping that energy to England and Wales and what pricing signals would allow for it to happen at the right times in a grid that is largely built around managing very short-term storage.

If the recommendations made by National Grid ESO to move to location-based pricing go through, it is possible that this would shift the investment decisions made by both generators and major users. Industrial and data-centre users could decide to set up shop close to Scotland’s wind farms while the renewables projects themselves decide to risk greater NIMBYism in the south in order to avoid curtailment if and when they get their installations approved. Another consequence, however, might be the more widespread deployment of floating wind platforms around the coast. A 2017 UCL study found this would reduce storage levels by taking advantage of differences in wind around the island of Great Britain and potentially could lead to generation more often being close to large users.

Storage operators have a lot of choices. They can turn to pumped-hydro, compressed or liquified air as well as batteries, though lithium-ion tends to become very expensive for durations beyond four hours according to analysis performed by engineering group Jacobs. Pumped-hydro is not only the most mature storage technology, it also works out the cheapest on a per-kWh basis in most studies. The bad news is that pumped hydro has the highest capital cost. Not only that, LCP on behalf of Drax argue that pricing distortion in the current market makes these large plants even more difficult to finance because of problems such as paying twice for connections to the grid on the basis they both absorb and deliver power.

Compressed air has a far shorter track record and costs more to operate, at least at the moment, but it has the distinct advantage of being easier to site, not just the transmission grid but the distribution network. That could help ease problems with network constraints and deliver savings in terms of the grid infrastructure they need around them. Consumers’ own batteries – both fixed and sitting inside electric vehicles – would in a grid managed around storage provide a large distributed reservoir even if it only acts ‘behind the meter’ to support their own management of pricing peaks.

From pumped hydro to batteries, electrical storage has an underlying shortcoming in a renewables-heavy grid: fixed capacity. Large installations distributed around the country could possibly deliver energy on bad days for a few days, and proposals such as the experimental sand battery operating in Finland can last for weeks. But to support big inter-seasonal changes, such as those seen between 2020 and 2021, the UK system will need something else to make up the difference. This is likely to be some form of gas, which under existing government plans is a mixture of methane and hydrogen. The question is where does that gas come from? It might not need to be imported at all but it could continue to play a major role in determining how much energy costs to consumers; decisions taken in the near future may ripple out across the decades.

One answer is to convert wind and solar that cannot be used immediately or stored as potential energy into hydrogen. The round-trip efficiency of hydrogen electrolysis and burning is around 30 per cent compared to 80 per cent for pumped-hydro and as much as 90 per cent for batteries. As hydrogen will be an important feedstock for synthetic fuels and chemicals as well as a methane substitute, it may prove to be commercially viable in a renewables-dominated grid and provide a way to profit from over-generation from renewables.

“If hydrogen starts being diverted as a fuel either for transport or heating rather than as a pure form of storage, then the demand for it will be greater and we will need a larger capacity to store the gas,” says Cárdenas. “So in a way, our numbers are conservative in the sense that what we’re saying we will need is never going to be too much.”

Competition will come from methane produced from waste using anaerobic digestion that is, under current approaches to carbon accounting, a net-zero fuel. By 2050, capacity for biogas could deliver the quantity of methane anticipated by the government for fuel. However, it may make more sense to divert it into chemical production that needs the carbon, with hydrogen taking its place in electricity generation.

What is less easy to predict is how changes in pricing structure will deal with the distortions we are experiencing today in the electricity market and what unexpected distortions the location-based pricing proposed by National Grid ESO will introduce. How that pricing policy is set may prove to be more important than the sticker price for a hundred gigawatts or so of reliable, clean generating capacity. The devil is in the detail. And if there is one thing privatised energy markets have, it’s a lot of detail.