Why EROEI Matters
Or why National Grid ESO’s obsession with intermittent renewables is leading us over the energy cliff
When I started writing about energy policy a few months ago, I naively thought that the Energy Return on Energy Invested (EROEI) would be a widely understood and discussed topic amongst energy policy wonks. Something happened last week that ended my innocence and now I feel compelled to explain why EROEI matters and demonstrate why our current energy policy is leading us over the energy cliff.
End of EROEI Innocence
Last week I wrote about the 2023 Future Energy Scenarios Report (FES) published by National Grid ESO. After I published the article, I participated in the “Energy System” explanatory webinar and one of the breakout sessions after the main webinar. In the webinar, I asked a question about EROEI, but time ran out before an answer could be given. So, I asked the question again in the breakout session. I was amazed to discover that the “Senior Electricity Analyst” facilitating the breakout session had not even heard of the concept of EROEI. He also explained that Levelised Cost of Energy (LCOE) was the only metric they used to help decide the generation mix. I did send him the seminal paper on EROEI by Weissbach in the vain hope that I might penetrate the citadel with some fundamental physics.
What is EROEI?
Starting with the basics, what is EROEI? In simple terms, EROEI is the ratio of useful energy out over the energy invested to get it. The energy in includes the energy required to extract fuels like natural gas from the ground. It also includes the energy required to extract, refine and process the minerals required to make things like solar panels and wind turbines. It also includes the embedded energy in necessary infrastructure such as the factories required to make solar panels and the pipework to transport natural gas. In his work, Weissbach produced a chart of the EROEI of many different electricity generating technologies (see Figure 1).
The blue bars represent the raw EROEI figures for each technology. The yellow bars recognise that some technologies are intermittent and their output needs to be “buffered” with some sort of storage that itself includes embedded energy. The resulting “buffered” EROEI is a real-world estimate of the performance of intermittent technologies. It can be seen that once intermittent renewables have been adjusted for intermittency that their EROEI falls well below conventional technologies. The only exception being hydro-power, which is very efficient, but its use is limited by geography. It should be noted that the buffering technology used by Weissbach in his analysis is pumped storage. He notes that other storage techniques like hydrogen electrolysis are much less economic than pumped storage.
Why is EROEI Important?
The reason EROEI is important is that if we spend too much time and energy gathering energy then we have much less time and energy available for other activities and society starts to degrade. This is best illustrated by the “Net Energy Cliff” described by Euan Mearns (see Figure 2).
Net Energy is the surplus energy available to society after deducting the energy used in energy gathering activities. If we have EROEI = 1, then net energy is zero. We are using as much energy to gather energy as the useful energy produced. If EROEI is less than 1, then we have an energy sink. It is possible to plot net energy as a percentage of total energy against EROEI. The blue area on the chart shows how the energy left for society varies with EROEI. As can be seen, when EROEI starts to get into single figures the energy available for society starts to plummet rapidly until it reaches zero when EROEI reaches 1.
Over time, society moved up the EROEI scale as we made the transition from wood to coal, then coal to oil and gas and then on to nuclear resulting in lots of surplus energy. This meant we could move from a largely agrarian society to one where we could invest in infrastructure such as better buildings, roads, bridges, railways, ships, aeroplanes and even space craft. We could invest time and effort into scientific research that produced new inventions that used more of that energy such as steam engines, internal combustion engines, electric motors, computers, rockets and so on. Surplus energy also allows more time and effort to be devoted to education and high culture such as art, theatre and music.
Unfortunately, lots of surplus energy also frees up time for luxury beliefs to emerge, even destructive belief systems that ignore (dare I say, deny) the fundamental physics that allows those beliefs to exist in the first place. We now have a situation where low-EROEI renewables are being placed at the forefront of energy policy regardless of their impact on the energy system and the wider economy and society.
Electricity System EROEI
To illustrate the slippery slope we are sliding down I thought it would be useful to plot how the EROEI of the UK electricity generation system has varied over time and where the GB generation mix planned for 2050 in the FES report puts us on the energy cliff.
[Update 20 July 2023:] Updated historic mix to use data from DUKES instead of Energy Trends Table 5.3 because DUKES looks at electricity generated not fuel consumed to make it more consistent with FES approach. This does not change overall conclusion but makes the change since 1998 more pronounced. [/Update]
For the historic generation mix I used Table 5.6 Digest of UK Energy Statistics (DUKES) report published by DESNZ. For the FES generation mix I used their Leading the Way scenario which removes all unabated hydrocarbons from the generation mix by 2035, in line with Government policy. I then created a weighted average EROEI for each mix using the buffered values in Figure 1 above. Weissbach did not cover BECCS in his report, so I used the Royal Society of Chemistry estimate for BECCS using wood chips from Louisiana at 0.7, a net energy sink. For Gas with CCUS I scaled down the EROEI by 20% to account for the reduction in CCGT turbine efficiency with an allowance for pumping and storing the CO2. For waste, I estimated an EROEI of 3.5 in line with biomass and for marine and “Other” I used an EROEI of 10. Waste, Marine and Other represent a tiny part of overall generation and so the weighted average is not sensitive to the value chosen. The results are shown in Figure 3.
As can be seen, EROEI has fallen from 40.8 in 1998 to 33.6 in 2008. This reflects much less nuclear power in the mix as plants were phased out and replaced by gas. The EROEI fell again to 25.4 in 2021, this reflects a sizeable increase in wind and biomass at the expense of coal, partly offset by a slightly larger share of nuclear even though overall nuclear output fell alongside overall output. The share of gas in the mix also fell slightly.
The weighted average EROEI of 10.7 for the Leading the Way scenario in 2050 is teetering on the brink of the energy cliff. This reflects the 86% of the generation mix coming from wind, biomass, solar and BECCS all with buffered EROEI of less than 4, with energy sink BECCS at 0.7. The overall weighted average is bolstered by the 9.3% of generation coming from nuclear power with an EROEI of 75. National Grid ESO describes the planned 16GW of reliable nuclear baseload generation as “challenging” to deal with when it is windy on low demand days. Their main solution for buffering intermittent renewables is hydrogen which was described by Weissbach as much less economical than his preferred solution, so it is entirely possible that the buffered EROEI I have assumed here is too high. Moreover, this analysis does not consider the embedded energy needed to build the extra grid connections and transmission capacity that will be required to bring power ashore from remote Scottish offshore windfarms and send it to where it is needed in England. Nor have I considered all of the extra effort to design and implement the Locational Marginal Pricing (LMP) scheme they seemingly favour to make this system work.
I also added my own generation mix suggestion for 2050 into Figure 3. It comes out with a weighted average EROEI of over 70. The average brought down by the flexible generation required because nuclear power is not yet particularly good at load following. I assumed this flexible generation would include the same amount of hydro, waste and other renewables as the FES Leading the Way scenario with the balance, representing almost all of the additional requirement, coming from Gas with CCUS. Ironically, nuclear power enables the luxury of Carbon Capture and Storage. This is a way to achieve a decarbonised grid with the high EROEI required to enable society to flourish.
I find it very worrying that the “experts” don’t know what they don’t know about EROEI and its impact on society. National Grid ESO are recommending and planning for an energy scarce system with inherently low EROEI that is teetering on the edge of the energy cliff. This risks destruction of society as we know it. It is even more troubling that the overall system EROEI is bailed out by the relatively small share allocated to nuclear power that they seem to view as an inconvenience in their overall plan. At best this is an example of institutionalised unconscious incompetence. If a low EROEI system is a conscious choice, then it is bordering on criminal negligence.
If there are any policymakers reading this, it is clear that a significant intervention needs to be made at National Grid ESO before they lead us over the energy cliff to economic and social disaster. In fact, I would encourage readers to contact their local MP, send them this article and ask how they are going to influence Government to ensure that physics triumphs over ignorant dogma.
If you have enjoyed this article, please share it with your family, friends and colleagues and sign up to receive more content like this.