This note argues that ‘net zero’ energy is likely to be most cheaply achieved by a huge expansion of renewables combined with hydrogen as a storage medium.

In particular, I look at the first stage of this strategy: the building of sufficient renewables capacity to provide all UK electricity, rather than all energy. I use data from the month of September 2019, showing that a 6.2 times expansion of wind energy supply would have created a sufficient electricity to at least cover current needs about 62% of the time. At times of surplus, up to 30 gigawatts of electricity is assumed to be converted to hydrogen. This hydrogen is then used to make electricity in the 38% of half hour periods when renewables supply is insufficient, through either combustion in a hydrogen CCGT units or the use of fuel cells. The supply from a 6.2 times multiple of current wind energy would have covered total electricity demand in each half hour of the month. No other capacity would be required, either from fossil fuel or, indeed, other renewables.

I use projected 2025 costs to assess the financial implications of this. The recent offshore wind auction produced prices as low as £39.50 (in 2012 money) per MWh. After applying CPI inflation to this number, the price would be about £51 in 2025. I then use estimated costs to calculate the price for converting surplus electricity to hydrogen and then back to electricity. Using these estimates, I suggest that the cost of fully renewable electricity system is only slightly more than today’s electricity supply pattern, updated to 2025 prices.

Finally, I postulate that this calculation is too pessimistic and that cost trends in renewables will make massive overbuilding of renewables cheaper than any alternative by 2025. Specifically, the expansion of renewable electricity as a source for replacements for liquid fuels will aid the economics of the proposed approach.

I believe the analysis contained in this article is the first attempt to estimate the financial implications of the strategy of moving to full reliance on renewables in the UK. It uses many uncertain estimates, strong assumptions and incomplete logic but I believe helps us begin to look at the impact of a radically different strategy for decarbonisation.

Introduction

Complete decarbonisation of the energy system is a fiendishly difficult challenge. I believe the only way of achieving it is through a huge expansion of renewables. The intermittent large surpluses of electricity will be converted to hydrogen via water electrolysis.

The hydrogen can then be used to generate electricity when renewables are not providing enough as well as providing fuel for home heating, energy for industrial processes such as steel-making and a core ingredient for the manufacture of synthetic fuels that will replace fossil sources.

UK commentators are sceptical about this path. They tend to prefer a mixture of a much smaller amount of renewables, combined with gas power stations plus CCS. The problems with the conventional approach are three-fold: first, it does not fully decarbonise the electricity system because of the loss of methane and of CO2 to the atmosphere. 10% of emissions will probably never be captured. More methane escapes during the gas production process than previously estimated. Second, we cannot be sure that CCS will work, either technically or financially. It certainly hasn’t on the first power stations on which it has been tried. Third, the strategy is a poor route to full decarbonisation of the wider energy system because it doesn’t link electricity outputs to gas and liquid fuel networks.

These problems mean that we need to consider alternatives. This article tries to start the process of such consideration. It doesn’t present a definitive answer but does suggest a method for assessing whether ‘massive overbuilding’ of renewables might work. I think it is the only way of dealing with the intermittency of wind and solar and, second, the need to continue to have substantial stored sources of non-electric energy.

The approach

I assess the possible costs of substantial expansion of renewables by contrasting two potential routes forward: the government’s route and a plan which sees enough renewables installed to cover all needs for electricity in September 2019.

Chart 1

The bizarre nature of real-time electricity reporting in the UK requires an investigator to make choices. Only large wind farms are connected to the main high voltage transmission network (‘the National Grid’). Other wind farms, and solar parks, do not have their output recorded immediately in a public database.

My work uses the public data provided by the Balancing Market Reporting Service (BMRS). I used the figures for 1-30 September this year. In the analysis that follows I just use extrapolations of electricity supply based on the data provided by BMRS about grid-connected wind.

My analysis refers to a potential 2025 situation and it assumes that demand remains constant between September 2019. This is unrealistic because the requirements from electric cars are likely to produce an increase in usage, although the growth in EVs is against a wider UK background of falling electricity demand as energy efficiency improves and de-industrialisation continues. Any lack of realism of the central assumption that demand will not change does not adversely affect the conclusions.

Chart 2

The following slide shows how grid-connected wind varied across each half-hour period in September 2019 and compares this figure with the total recorded demand for electricity.

Chart 3

September 2019 was a reasonably typical month in which about 20% of electricity demand was met by grid-connected wind. (But also noting that wind and solar that are not grid connected reduce reported levels of electricity use). The percentage varied from about 47% down to around 2%.

Many outline plans for the UK envisage an expansion of wind supply, particularly offshore, so that it covers a much larger fraction of monthly demand. Chart 4 shows the impact of doubling the amount of grid-connected wind. The amount of new wind power is restricted so that output will rarely exceed the total demand for electricity. Having double the amount of wind would produce an average supply of 40% of overall need, and a maximum of 94%.

The assumption of the analysts, such as the Committee on Climate Change, is that the remainder of energy demand will be provided by gas-fired power stations that collect and store the CO2 from the flue gas. (However I believe that nowhere in the world does a gas-fired power station collect and store CO2 currently).

Chart 4

In the rest of this article, I will compare the first two scenarios (staying at today’s level of wind energy or doubling it) with a more radical approach that multiplies the amount of wnd electricity by 6.2 times. This would take grid-connected wind up to over 100 gigawatts from about 18 gigawatts today.

Why have I chosen a 6.2 times multiple? This is how much the UK would require to meet all its electricity demand over the course of September 2019. I have used the assumption that the conversion of electricity to hydrogen will be approximately 80% efficient in 2025 and, second, that converting ot back to power - through turbines or fuel cells – will deliver about 60% of the energy value of hydrogen. Both these numbers are slightly above today’s figures but technical progress is very likely to take efficiency to higher levels over the next few years.

Chart 5

To cover September’s demand with grid-connected wind in 2025 will require enough turbines to provide about 124% of demand. The excess is required because of the efficiency losses turning power into hydrogen and back again. (The overall loss is 52% of the power used, meaning a round-trip efficiency of 48%).

The electricity system will operate with a simple decision rule. If demand is less than supply, surplus electricity will be converted to hydrogen via water electrolysis. In the opposite situation, stored hydrogen will be used to generate electricity.

Chart 6

The system is assumed to have 30 gigawatts of electrolyser available. This means that enough electrolyser capacity is available to use surplus power at almost all times. Only about 5% of the surplus wind electricity is not used for electrolysis.

Chart 7 shows how much electrolysis capacity would be used over the course of the month.

Chart 7

The overall pattern of supply is laid out in Chart 8. Overall demand for the month is about 19.8 TWh with approximately 16.1 TWh met directly from wind. The remainder is provided by electricity generated from stored hydrogen that was created by electrolysis earlier in the month.

Chart 8

How much hydrogen storage capacity would this month’s pattern of demand and supply required. The first thing to note from Slide 9 (11) is that if the UK had started with no hydrogen in storage on 1st September it would have been unable to meet the needs for the gas from about the 18th to the 27th. At the bottom of this period, the UK would have been short about 1,000 gigawatt hours, or one terawatt. This is an illustration of the necessity to have storage at the beginning of the month that is sufficient to cover periods of low wind power production

Chart 9.

The results

Slide 10 (12) gives some of the key figures used for the financial assessment. The most important are probably the costs of wind energy and those of CCS and gas power production.

Chart 10

The latest offshore wind auctions (September 2019) produced a low price of £39.65 per megawatt hour for a project on Dogger Bank that is scheduled for completion in 2023/24. This price was offered in 2012 price and since there has been inflation since then the actual price paid will be higher. By 2025, 2% yearly inflation will take this number to just over £51 per megawatt hour and I have used £51 in my assessment of the underlying cost of wind power by 2025.

The price assumed for new CCGT power stations with full scale carbon capture and storage is £89 per megawatt hour. This number is taken from the Net Zero report of the Committee on Climate Change of May 2019. The figure there of £79 appears to be in 2019 real numbers, and I have inflated this figure by 2% a year (the target for CPI inflation) and rounded the result to £89.

Electrolyser costs in 2025 are estimate at around £500 per kilowatt and the running cost £10 per kilowatt per year. An 8% cost of capital is used.

I have then calculated the full cost of all electricity delivered in the month of September 2019 using the figure of £51 for wind and £89 for gas with CCS. I do this calculation for three scenarios: a mixture of 20% wind and gas with CCS, a doubling of wind and gas with CCS and lastly, a 6.2 times multiple of wind with surpluses held as hydrogen. (I stress that this is a hypothetical exercise because the UK is very unlikely to be just wind and gas powered in 2025).

The costs of a renewables plus hydrogen route are given below

Chart 11

The spreadsheet analysis shows that the 2 times wind route is likely to be the cheapest option in 2025, if the CCC is right about the price for gas power with CCS. But this second scenario is only about 3% cheaper than my proposal of wind plus hydrogen to cover all electricity needs.

Chart 12

Even a small reduction in the cost of wind to £48 would mean that full decarbonisation using wind (or other renewables) and hydrogen would be cheaper than any other route.

Chart 13

In addition, it is unclear whether the CCC has included any estimate in its gas costs for the impact of the uncaptured CO2 at the power plant or the methane escaping from production wells and pipelines.

Moreover, in one sense my proposal is unduly conservative. The ideal route for the UK and other economies to follow would be to use hydrogen not just for power generation but also for heating buildings and for creating synthetic fuels that substitute for fossil oils and gases. If this direction was taken, the UK could run its electrolysers at much higher rates of capacity utilisation, bringing down the hydrogen costs per kilowatt hour.

Can the North Sea provide more than 100 gigawatts of turbine capacity within British waters? Yes, almost certainly. Shell has estimated that 900 gigawatts across all national zones is possible and the UK has a large share of shallow water sites, such as Dogger Bank. The economics of onshore would be even better if the government were to encourage development on western coasts. Similarly, larger scale development of solar, which is now cheaper than offshore, would similarly help.

The economics of using what I call ‘massive overbuilding’ clearly needs more work. However it does seem a highly competitive route to full decarbonisation without any of the problems caused by the need for carbon capture and offering a low cost route to synthetic fuel manufacture.