Engineering Net Zero Part 9: Strategies for Clean Energy

David Simmonds builds on his ENZ series, visualising the UK power system, and explores the unintended consequences of current strategies for 2050

THE National Energy System Operator, NESO, last year presented two challenging Clean Power options for 2030,¹ and three Future Energy Scenarios for 2050,² based upon renewables and widespread electrification of industry, heat, and transport. Two questions arose: can we afford them and, are they leading to hidden costs and unintended consequences, such as oversizing of our power grid, retaining a reliance on imports, and prolonging high electricity prices? In my view, the answers are no and yes, respectively.

I discussed many of these challenges in my series Engineering Net Zero³ which concluded with Part 8, the cover feature of last April’s TCE. Since then, doubts on achieving our targets for 2030 and 2050 have grown, prompting me to author three new online features, Parts 9A, 9B, and 9C, which, in some detail, visualise and quantify the challenges facing our future system, concluding with an alternative which addressed these challenges. This, Part 9, summarises that analysis.

Visualising the future power system

Today’s power system relies upon “dispatchable power”, which can be turned on and off to match demand. From the 1950s we added expensive nuclear power which operates as a baseload supply, while recently we started to introduce renewables, solar and wind, which, although cheaper, are intermittent and not dispatchable. Both nuclear and renewables have replaced fossil fuel generation, and NESO rightly plans to pursue renewables, electric vehicles, and heat pumps to meet our net zero goals but can they reduce costs? Individually they can, but the variation of demand across the day, the intermittency of renewables, and the growing seasonal demand from heat pumps currently met by our gas system, means we require system balancing to avoid power cuts. In Part 9A I presented a hierarchy of system balancing technologies, visualising them in a “Power System Onion” (see Figure 1). This displays measures to manage periods of oversupply (when the wind really blows) in green, with dispatchable capacity measures to offset shortages in red. The hierarchy works out from the centre, starting with short-duration/low-capacity technologies such as demand side response (DSR) measures which call on consumers to directly tailor their own demand. The next layers include domestic and industrial-scale batteries balancing demand across a few days. Hydro follows, typically working on a week to ten-day cycle, but, at best, these short-duration technologies will contribute only 1–5 TWh storage capacity. 

The outer layers of the onion are longer-duration measures essential for balancing seasonal demand. By 2050 with widespread electrification, power demand will have doubled from that today, to +/-600 TWh, and, by then, the Royal Society and others estimate we will require some 80–100 TWh energy storage to manage seasonal variation, way in excess of our battery/hydro potential. I term this the 15–20% demand challenge, 15% in an average year, 20% in a harsh year.

For both 2030 and 2050, NESO utilises exports and imports to balance ten of the last 15%, leaving the final 5% to fossil gas generation in 2030, and hence their 95% clean power target. To balance the final 5% in 2050, NESO looks to the “hydrogen cycle”, using surplus power to produce green hydrogen from electrolysis, which is stored and used to generate power in times of shortage. In a harsh year we would need more hydrogen, and recognising the 20% challenge, Imperial College in its Hydrogen Futures report⁴ notes that “If you are going for an 80% decarbonisation target, you probably don’t need a lot of hydrogen. But once you go for 100%, you do”. How true, so while a lot of focus is being placed on short-duration measures, long-duration balancing is crucial, and, as we will see, is already increasing the cost of our electricity.

Quantifying the 15–20% challenge

In Part 9B I looked to see how we could better quantify the measures needed to create the “skin” of the onion. Both demand and renewables supply have a sawtooth nature, such as shown in Figure 2, adding to the difficulty in understanding system balancing.

I recalled my experience planning the Dutch gas grid, supported by the massive Groningen field, in the 1990s, where gas demand had a similar sawtooth profile. We made sense of that data by creating a load duration curve (LDC) as shown in Figure 3. Daily demand is stacked in the LDC by magnitude, with day 1 being the highest demand day of the year and day 365 the lowest, producing a smooth, but sharply sloping curve. We generated this data through stochastic modelling of demand using historically trended weather and probabilistic equipment reliability data. Monte Carlo modelling allowed development of a high load or design case to size capacity measures, such as underground gas storage (UGS) facilities and other external measures. The shape of the LDC demonstrates the seasonal demand for gas, a direct equivalent to our 15–20% challenge.

Recognising that within our future power system both supply and demand are weather dependent, a simple load duration curve does not apply. In Part 9B I went on to propose a stochastic load duration difference curve (LDDC) capturing the difference between daily demand (net of in-day measures) and non-dispatchable or must-take supplies (typically renewables and nuclear) for each and every day, stacked by magnitude. This results in a sideways S-shaped curve as shown in Figure 4. Days when demand is higher than supply, requiring additional dispatchable power, appear on the left, while days of surplus non-dispatchable power, requiring oversupply measures, are registered on the right.

Like the LDC, the LDDC can be used to scope measures needed to balance the system. Short-duration measures should be able to balance the system much of the year, typically days 90–270 on the stacked LDDC. The 15–20% challenge is represented by the outer edges of the LDDC, with capacity measures on the left (typically days 1–90) and oversupply measures on the right (days 270–365). The area under the curve on the capacity measures side is the annual balancing load or required stored energy, which by 2050 will be that forecasted 80–100 TWh. The chart also shows the trend of variable pricing, increasing when capacity measures are in operation and dropping significantly at times of oversupply. Further, the LDDC can present design cases for cold winters and windy summers.

In this example I have applied NESO’s balancing measures for 2030, which, for capacity, first consider imports followed by fossil gas generation. For oversupply, they consider the interconnectors for export (typically 15%), followed by, where unavoidable, costly curtailments (the compensation paid to renewable operators to turn off their windfarms when the power they are generating isn’t needed). None of these measures are cyclical or offer energy storage and hence the discontinuity X in the Power System Onion. Given that 2030 is imminent, we may have to accept this expensive soft fix, exporting cheap surplus power and importing expensive back-up power.

As imported power is marginal to European demand, and with some interconnected EU countries still burning coal in 2030, imports will likely come with a high embedded carbon footprint. Consequently, I recommend NESO reassesses both the cost and environmental impact of imports before turning off our fossil-fuelled plants. In their options report, NESO indicated that Hi-Dispatch, with carbon capture on fossil gas, was cheaper than its Hi-Flex base case (with more renewables). So, wider application of carbon capture, which can also offer synergies with industrial clusters, may provide a cleaner way to achieve the government’s target.

Certainly more effective solutions are required by 2050, and as we have seen, NESO applies the “hydrogen cycle” for the final 5%, but they still rely on imports and exports for most of the tranche, as shown here. In Part 9C, I questioned the continued use of interconnectors for balancing, particularly as the export/import strategy exacerbates today’s UK/Europe electricity price differential. Further, even in 2050, imports will still incur an embedded carbon footprint as Europe will retain some dependency on fossil-gas generation. Renewables are meant to reduce our reliance on gas imports, but it would appear current scenarios for “Clean Power 2050” replace these with power imports, offshoring emissions, and increasing our exposure to harsh winters.

I also considered “grid efficiency”. First, regarding our high-voltage (HV) grid, it must be designed to transmit surplus peak power from Scottish windfarms across the country for export from southern England. This requires oversized cables and pylons, which operate at low utilisation. Secondly, the National Infrastructure Commission, NIC, rightly points to the need to upgrade our low voltage (LV) distribution grid to meet peak consumer demand from heat pumps over winter. This requires upgrading local grids and in-street cables for coincident peak power demand, again, a rarely needed design case. In Part 9C, I also exemplified the inefficiencies by considering a standalone 10 kW heat pump. This requires over 20 kW of installed generation capacity for it to operate throughout the year: 10 kW of renewable energy generation capacity for normal operation and 10 kW of standby capacity for when the wind does not blow.

The unintended consequences of Clean Power 2050 will add cost and incur public angst, so can we do better?

An alternative strategy for 2050

In Part 9C, I developed an alternative strategy, one which I believe can reduce costs across the board. This considers wider application of the “hydrogen cycle” for the complete 15–20% challenge, alongside deployment of hybrid technologies for heating and transport; here I call it “Clean Energy 2050”. Yes, green hydrogen is both expensive and inefficient, but within the cycle it is produced using cheap power which, otherwise, would be exported.

I previously provided the rationale for hybrid electric vehicles in Part 2 and heat pumps in Part 3, as they offer consumers choice and flexibility, and as shown in Part 8 they only use limited amounts of hydrogen. Recognising the importance NESO places on DSR measures for day-to-day balancing, I now realise that hybrids offer long-duration DSR. I am not the only one as, subsequent to authoring Part 9C, I attended IGEM’s 2025 Sir Denis Rooke lecture presented by Chris Stark CBE, head of Mission Control for Clean Power 2030. In his talk, Stark raised his concerns for long-duration storage and cited hybrid heat pumps as a possible solution to assist 2050 power demand. Others at the meeting proposed that heat pump incentives should be extended to hybrid solutions.

The Clean Energy 2050 scenario, represented in Figure 7, reduces both peak power demand and exports, and will decarbonise our total energy stsyem. In turn, HV and LV grid infrastructure does not have to be oversized, minimising pylons and local disruption, and cuts back on copper usage. Development of offshore power hubs, equipped with electrolysers, would allow green hydrogen to be piped directly to offshore storage, while power can be collected and routed offshore to English markets, minimising overland transmission. With hydrogen likely becoming an international commodity, larger offshore storage facilities would also allow the UK to benefit from timely imports to offset harsh winters. Clearly, the alternative relies on conversion of the gas grid to hydrogen service, and critics look to transmission inefficiencies and the safety of hydrogen. In terms of energy equivalents, hydrogen use in hybrid heating will be less than a quarter of that today for natural gas and so, despite its lower calorific value, the transmission inefficiencies can be managed. Meanwhile, the Health & Safety Executive⁵ and others are addressing safety concerns as part of their ongoing work assessing hydrogen’s role in heating. Finally, coming back to that heat pump example, a 10 kW hybrid heat pump will only require the first 10 kW of generation capacity, as the back-up is provided by the gas grid.

I am not in a position to model or cost Clean Energy 2050, but after decarbonising the power grid by 2030, ideally through Hi-Dispatch, it decarbonises our wider energy system by 2050. Can we afford it? I don’t know, but we certainly can’t afford a strategy which offers consumers limited choice, leads to inefficient power grids, still relies on power imports, and exposure to expensive and uncertain decommissioning of our gas grid. Policymakers face significant challenges, so our community must support Mission Control with its planning for beyond 2030, and not sleepwalk into those costly unintended consequences.

References

1. Clean Power 2030: https://www.neso.energy/document/346651/download
2. Future Energy Scenarios for 2050: https://www.neso.energy/publications/future-energy-scenarios-fes
3. Engineering Net Zero series: https://www.thechemicalengineer.com/tags/engineering-net-zero-series/
4. Hydrogen Futures report: https://www.imperial.ac.uk/Stories/hydrogen-futures/
5. HSE: Hydrogen Safety: https://solutions.hse.gov.uk/safe-net-zero/hydrogen-safety