Assessment of the global transition to low-carbon energy from the lens of EROI
Can we utilize low-carbon energy sources to meet our standard of living while addressing the energy workload required for climate mitigation simultaneously?
Introduction
It has been well established that greenhouse gas emissions (GHG) are the highest driver of climate change. Emission intensity, which is measured as kg CO2e/kWh of electricity produced, is typically used to evaluate low-carbon footprint options in the global energy sector. However, focusing solely on emission intensity for evaluation leads to carbon tunnel vision, overlooking deeper issues with the transition.
To fully assess whether low-carbon energy sources (like renewables) can support both our standard of living and the energy demands of climate change mitigation, it is essential to consider another critical metric: Energy Return on Investment (EROI).
EROI is defined as the return generated on the energy invested to obtain an energy source. This metric was devised by Charles A.S. Hall in 1984 to compare the performance of various energy sources. Naturally, a higher EROI is desirable.
To begin simply, a positive EROI value means that fuel is a net energy source, and a negative value means it is an energy sink. An EROI of 1:1 means our energy returned is equal to energy invested, and thus, we do not have any spare energy left to put to work.
As EROI increases, we get surplus energy that can be put to further use. To better understand how an EROI above 1 benefit us, I am paraphrasing from Charles A.S Hall 2013 interview with Scientific American, where he explains the concept of EROI in the context of oil extraction.
If you have an EROI of 1.1, you can pump the oil out of the ground and look at it. If you have 1.2, you can refine the oil and look at it. At 1.3, you can move the oil to where you want it and look at it. You need a minimum EROI of 3 at the wellhead to drive a truck. Now, you need to have an EROI of 5 to put anything in the truck, like grains. And that includes the depreciation for the truck. Further, you need an EROI of 7 to support the families of the truck driver, the oil worker, and the farmer. And if you want education for them, you need 8 or 9. And if you want health care for them, you need 10 or 11.
We need a minimum of 7-9 EROI to sustain the comfortable functioning of modern industrial society. In this article, we will analyze the historical EROI trends of our primary energy source, i.e., fossil fuels; examine the EROI of low-carbon alternatives in sustaining our material standard of living; and explore the global energy workload involved in mitigating the climate crisis.
Calculation methodology and limitations of EROI as a metric
The supply chain of any energy source typically includes four stages: extraction, processing, transportation, and end use. To calculate the EROI at any of these stages, the output energy delivered (after accounting for conversion losses) is divided by the total energy inputs.
EROI is generally calculated either at the point of extraction (ST) or the point of use (POU). EROIST > EROIPOU because energy is consumed in processing and transporting fuel down the supply chain. In the interview referenced earlier, Charles Hall was referring to the EROI of oil at the point of extraction.
When comparing the EROI of different energy sources, it’s essential to use the same system boundary to ensure an apples-to-apples comparison—this due diligence has been followed in this article’s analysis. EROI values can vary widely for the same energy source, depending on technology, geography, operational practices, and other factors. Some studies report EROI as a range rather than a single value. Taking these variabilities into account, including this article’s focus on the qualitative linkage between EROI and broader macro factors, recent aggregated estimates at the point of extraction have been used for the analysis.
Using EROI as a metric comes with limitations and has faced criticism within the scientific community. First, it lacks a ‘time factor’—EROI does not account for energy payback time, a key parameter in evaluating renewable energy options. Second, EROI alone is insufficient without additional techno-economic modelling to comprehensively assess energy sources or systems. Third, as Schalk Cloete from Norway's SINTEF research organization notes, “The true EROI of a technology is just too complex to quantify accurately, given the challenges with assessing full lifecycle energy inputs, converting between different energy sources, and accounting for effects on the rest of the energy system.”
Despite its limitations, EROI has its value in a quick first-cut analysis of energy options for our transition, which we will explore in this article.
The golden era of high EROI
Between 1940 and 1990, coal and oil served as the world’s primary energy sources, largely because they offered a high average EROIST of 30–60. Combined with factors like low population levels, effective governance, and a culture that encouraged wealth creation, this high EROI played a key role in the rapid development of Western economies.
It was a time when, beyond basic sustenance, abundant discretionary energy supported entertainment, the arts, and leisure. This enabled Western societies and oil-rich countries in the Middle East to enjoy a significantly higher material standard of living. It also sustained the growth of the middle class across large populations, particularly in China and India.
All good times come to an end
At the turn of the 21st century—an era that marked a significant shift in global political and business dynamics—there was also a global decline in the EROI of fossil fuels. This was accelerated by rising per capita energy consumption in developed nations and a population surge in developing countries seeking a similar lifestyle.
Today, most oil extraction sites around the world yield an EROI of 15–20, except for a few locations in the Middle East where oil can still be extracted through surface exploration. For any non-renewable energy source, EROI naturally declines over time, as we must go farther and drill deeper to extract the same amount of oil. Moving offshore—away from the 'sweet spots' of oil formations—requires hydraulic fracturing (fracking) and horizontal drilling, which are highly energy-intensive processes. This lowers the EROI further and can eventually make extraction uneconomical.
Although oil and gas exploration in the Middle East has not yet shifted offshore, it is reasonable to conclude that these nations are proactively diversifying into tourism, sports, and service-based industries in anticipation of declining oil EROI.
EROI evaluation of our energy choices
As we transition to renewables and other low-carbon energy sources, maximizing their EROI is just as important (if not more so) than reducing their emission intensities. Figure 1 presents a plot of emission intensity versus EROI for all energy sources used in global production in 2020. The size of the bubbles on the graph represents their relative production volumes for that year. Detailed data sources for the graph are provided in the Appendix.
As we can see in Figure 1, it would be ideal for the bulk of our energy production to fall in the lower right-hand quadrant of the graph as this represents the best of both worlds: high EROI and low emission intensity. Unfortunately, that is not the case with current global energy production, which remains dominated by emission-intensive fossil fuels.
Fossil fuels still maintain an edge over renewables due to their relatively higher EROI range (15–30), but this advantage is steadily declining each year. Oil production is expected to peak globally within this decade, which may further slow the transition to renewables. Biofuels are not encouraging for the energy transition either, given their limited global EROI range of 2–4. Wind and solar, by contrast, are already delivering consistent EROIs of 10–15. Efficiency breakthroughs, such as the use of perovskite cells in solar wafers and improved wind turbine designs, are expected to enhance these figures further.
Notably, wind turbine manufacturer Vestas has claimed that some of its latest models are delivering EROIs in the range of 32–41. This would further accelerate price parity between renewable electricity and coal, especially as coal's EROI also continues to decline.
The intermittency in power generation from renewables needs to be balanced with battery storage to ensure a reliable and continuous supply. However, this step partially offsets EROI gains from efficiency improvements, due to the additional energy required to build and maintain the storage infrastructure.
As we move forward, EROI modelling of energy grids incorporating renewables and long-duration stationary storage will become increasingly complex, and a critical area of research.
Although hydrogen and nuclear energy sources are in the desired quadrant on the graph, both options are largely efficient only on paper. Hydrogen is logistically not scalable across the world and is insufficient to meet global energy demand. This leaves us with nuclear fission and fusion as the primary candidates for achieving both high EROI and low emission intensity.
Nuclear fission faces regulatory hurdles, historical safety concerns, and requires significantly more time and capital to get a plant operational compared to deploying solar or wind farms. The cost of developing nuclear energy remains high when measured by the dollars-per-unit-electricity-generated ($/MWh). The latest estimate places the cost of nuclear fission production at around $95/MWh, compared to $30–40/MWh for land-based solar and wind farms.
Small Modular Reactors (SMRs) offer some hope of overcoming the challenges in scaling nuclear fission, but their commercial deployment is unlikely to materialize before the climate crisis reaches an inflection point. Nuclear fusion, still regarded as the holy grail of energy, has seen recent laboratory breakthroughs notably, achieving net positive ignition (outputting more energy than input). However, major engineering challenges remain before we can scale to a commercial fusion reactor.
It's worth noting that the energy requirements involved in scaling fusion from lab-scale to pilot-scale to commercial deployment are highly non-linear. This introduces a significant risk: that the fossil fuel energy investment required to build fusion infrastructure may ultimately need to be redirected toward more immediate mitigation and adaptation strategies in response to climate change.
ESOI of fuel cells
Fuel cells such as lithium-ion and hydrogen batteries deserve mention here, as they are increasingly becoming part of the modern energy mix. I excluded them from Figure 1 because it uses the metric ESOI (Energy Stored on Energy Invested), and reliable ESOI data for these technologies is still emerging due to their relatively early stage of real-world deployment. Preliminary estimates put the ESOI of lithium-ion and green hydrogen fuel cells at around 10 and 6, respectively (see Appendix).
While they cannot fully replace fossil fuels, they can serve as partial substitutes. The ESOI of long-duration storage technologies is still being developed and will be critical to reducing grid intermittency.
If you're wondering about the ESOI of lead-acid batteries, commonly used in home inverters for power backup, it’s just 2. That explains why inverters often last only a few hours, not days (especially in Asia); the low ESOI is the culprit.
Lithium-ion battery adoption in two- and four-wheeler electric vehicles (EVs) is growing steadily. However, their ESOI remains too low for use in maritime or air transport, unless there is a major breakthrough in battery chemistry to significantly increase energy density.
Moreover, the rise of the EV industry is expected to put increasing pressure on the mining of critical metals required for battery production. As ore grades decline and we dig deeper, energy and processing costs will increase super-linearly. This trend will lower the EROI of metal mining and, in turn, place long-term upward pressure on the cost of EVs.
Probable impact of plateauing of global EROI
Since nuclear energy is unlikely to scale fast enough to meaningfully impact the current climate crisis, we must come to terms with the fact that our current energy sources (those with EROI or ESOI around 10-15) are sufficient for stability but inadequate to sustain the immense industrial rate of growth experienced in the last two decades, especially in China and India.
At a societal level, stagnant or declining EROI means that a growing share of energy output and economic activity must be devoted to simply acquiring the energy needed to run the economy. This leaves less discretionary energy for “non-essential” consumption that are primarily driving the GDP in modern capitalist economies. Economic degrowth would gradually become inevitable. A higher cost of living will define the low-EROI phase. Developmental projects will slow as contractors become more cautious about rising energy costs.
Moreover, extreme climate events, floods, heatwaves, wildfires, and crop failures, will further delay such projects. To make matters worse, a lower-EROI world will struggle with slower damage control and longer recovery periods, resulting in a highly constrained and reactive approach to development.
Impact of lower EROI on job creation
Another less obvious impact of the lower EROI of renewables compared to fossil fuels is a reduction in the number of skilled jobs and lower wages. Renewable energy companies do not have high profit margins (due to lower EROI) and thus may not be able to pay the labor force as well as traditional oil and gas companies. Higher EROI enables greater division of labor; hence, job reductions are likely in the renewables industry. Overall, the transition to a ‘green’ economy from an oil-and-gas-based economy is not a like-for-like replacement from a labor benefit perspective.
Impact of carbon removal projects on lowering industrial EROI
The climate crisis is worsening with each passing year, and so is the responsibility to remove both historical and current carbon emissions. A number of pilot projects—such as direct air capture, enhanced rock weathering, mineralisation, and ocean sequestration—are being tested across the world to remove carbon. These projects are broadly classified as carbon capture and storage (CCS) or carbon capture and utilization (CCU) technologies.
As industrial decarbonisation regulations tighten globally, coal gasifiers, oil upgraders and refineries, and steam methane reformers (used for hydrogen fuel) will increasingly need to be paired with CCS or CCU systems to reduce emissions. This, in turn, will lower the EROI of these industrial units, as a portion of the usable output energy must be diverted to build and maintain CCS or CCU setups.
However, the cost of financing such industrial systems remains high, exceeding $100 per tonne of CO₂ removed and we need to capture and store CO₂ at the scale of billions of tonnes per year by 2030. For context, we are currently capturing and storing only millions of tonnes annually.
This reflects a staggering three orders of magnitude shortfall—a 1,000x deficit—in carbon removal, underscoring the massive scale of the climate crisis and the energy and financial burden involved in supporting carbon removal operations.
If we do not voluntarily allocate our industrial fossil fuel energy to CCS or CCU efforts now, we will be forced to do so later—at a much higher cost—when extreme climate events cause even greater damage.
Thus, it is the need of the hour to prioritize voluntary climate mitigation now, rather than face forced and costlier mitigation and adaptation later as illustrated in Figure 2 below.
We are looking at a future where global EROI is constrained simultaneously from the depletion of concentrated energy sources and the massive engineering workload to bridge the emission reduction deficit gap.
EROI of extreme energy projects
Our insistence on maintaining current standards of living and consumption is pushing us to explore extreme energy projects like asteroid mining and deep-sea mining. Academic research on these ventures is still in its infancy, and reliable EROI data is not yet available for inclusion in this article. However, I strongly suspect that the EROI will not be favorable for scaling such projects, especially considering the potential harm to marine ecosystems in the case of deep-sea mining. Fortunately, the International Seabed Authority recently decided to place all deep-sea mining projects from various countries on hold until the negative impacts are fully understood.
Conclusion
The reality is that fossil fuels will continue to remain a major part of the global energy mix for the foreseeable future in order to sustain our current standard of living and meet several competing global energy demands.
Why? First, fossil fuels are still necessary to develop and deploy renewable energy infrastructure, to manufacture and transport solar panels and wind turbines, build factories for battery storage, and mine critical and rare earth metals for electric vehicles. Second, they are essential for building and scaling industrial CCS and CCU setups needed for large-scale carbon removal.
Third, with the rise of artificial Intelligence (AI), massive computing power will be required to operate cloud data servers—demand that will likely be met, at least in part, by coal-powered thermal plants worldwide. Fourth, fossil fuels will also be consumed in ongoing global conflicts, such as the Russia–Ukraine and Israel–Gaza wars, as well as in arguably non-essential ventures like Mars exploration.
Given these energy-intensive workloads, we cannot phase out fossil fuels immediately, despite their high emission intensity. At the same time, continued dependence on them jeopardizes our 2030 emission targets and worsens the frequency and intensity of extreme climate events around the world.
The precautionary and proactive approach to addressing this problem is to reduce energy consumption–particularly among the higher economic strata–at least until the low-carbon energy landscape (renewables and storage) reaches a sufficient threshold of coverage to meet competing global energy demands. By doing this we also give developing countries much needed buffer time (through minimizing average global temperature rise) to develop their climate adaptation solutions across the globe. The masses remain highly resistant to this proposal. In an ideal fair world, a detailed energy accounting and planning would be done by an organization like UN to balance the energy expenditure for global wellbeing of masses and the energy required for carbon removal projects. All countries would then adhere to that global policy. Unfortunately, we are finding hard to reach such consensus and finalize effective policies like for example significant loss and damage fund for developing countries at global events like COP.
By doing this, we also give developing countries the much-needed buffer time—by minimizing the average global temperature rise—to develop their climate adaptation solutions. However, the masses remain highly resistant to this proposal.
In an ideal world, detailed energy accounting and planning would be conducted by an organisation like the United Nations (UN) to balance global energy expenditure between the well-being of the masses and the energy required for carbon removal projects. All countries would then adhere to this global policy. Unfortunately, we are finding it difficult to reach such consensus and finalise effective policies. A glaring example of this is the delay and underfunding of the Loss and Damage Fund for developing countries at global events like COP (Conference of the Parties), held annually to address climate change preparedness.
In light of the plateauing global EROI, many current and near-future lifestyle choices of the global rich come under scrutiny. The extravagant lifestyles of Ultra-High Net Worth Individuals (UHNIs)–including private jets, space tourism, and luxury yachts–represent an excessive and inequitable use of energy resources, especially in the context of global scarcity. This is particularly unjust when people in developing countries are still striving to achieve the development benchmarks outlined in the UN Sustainable Development Goals. Wealth inequality feeds into energy access inequality in a reinforcing loop.
Some aspects of our current global ‘green economy’ are doing more harm than good by distracting us from the fundamental constraints of EROI. Carbon offset projects, which are designed to avoid future emissions and traded as credits in the voluntary carbon market, are one such major distraction. These projects can, at best, serve as a tertiary solution given the enormous energy workload required for large-scale carbon removal.
Other distractions arise from misleading labels like ‘sustainable aviation fuels’ and ‘carbon neutrality,’ which create the illusion that we can maintain our current lifestyles while new technologies and free markets solve the climate crisis. Behavioral change–both from consumers and through regulatory interventions by policymakers–remains underappreciated and disincentivized.
Ultimately, global EROI constraints are masked by this form of ‘green capitalism’ and only become visible to the public through rising energy prices. Governments often intervene with fossil fuel subsidies to keep these prices in check, but in doing so, we merely postpone confronting the inevitability of declining global EROI and the deepening climate crisis.
PS: If you liked this article, you would enjoy browsing the comic on 'Energy slaves' by Australian comics artist Stuart Mcmillen.
This article has been edited by Anjaly Raj.
Appendix
Source description and data for Figure 1: The energy production data (in TJ) for Figure 1 is taken from IEA energy statistics data browser except for geothermal, which is taken from Global Geothermal alliance 2021 report. Although it would be more relevant to analyze 2023 global energy production data, IEA publicly available data was only till 2020. Also, global fossil fuel usage has gone up in the last two years along with renewable installation, so relative bubble sizes would remain almost unchanged for 2023.
The emission intensity data is taken for the entire lifecycle of energy production and is validated from Electricity Maps, a reliable open-source database. EROIST data of fossil fuels is interpreted from 2021 research paper in Journal of Applied Energy. A study on biofuels in Ecuador was used to get EROI range of commonly used biofuels. EROI of geothermal was accessed from a 2012 technical report by US Department of Energy. It is expected that geothermal's EROI would be improved slightly by now from that report. EROI of wind and solar was taken from 2020 research paper in energies, published by MDPI. EROI of hydro was referenced from 2019 research paper in Energy strategy reviews. EROI of coal and nuclear was referenced from 2015 Forbes article. All the data for Figure 1 is shown in the table below:
Data for fuel cells:
ESOI values of Li cell is taken from a presentation and ESOI of hydrogen cell is taken from a research paper.
Really appreciate your monetary support (links below) if you enjoyed reading this post and found it insightful.
If you are interested to be a guest writer on this platform, do check out this page.
This work is licensed under CC BY-SA 4.0
