This article argues that resource and logistical constraints weighing on low-carbon energy and CO2 capture technologies are likely to pave the way for geo-engineering solutions such as Stratospheric Aerosol Injection (SAI), which demand negligible land, material, and energy inputs. This “climate transition without carbon transition”, though technically feasible, is far from being that simple, raising a whole new set of environmental risks as well as geopolitical, institutional, and ethical issues.
Our carbon budget is rapidly draining
Let us first set one fact straight: there are ample resources to grossly overshoot the 2-degree target. The IPCC’s latest Group 1 report updates our carbon budget at 900Gt of CO2 to keep an 83% chance of achieving the +2C objective. That is only 25 years of emissions at the current pace.
2000 GW of coal power capacity is operating worldwide (double the capacity of year 2000), representing ~19% of human Greenhouse Gas (GHG) emissions in CO2 equivalent. A further 500 GW is being built or planned. A load factor of 54%, a CO2 emission factor of 1000 tons / GWh, and plant lifetimes of 30 years imply that these 500 GW alone would consume about 8% of our remaining +2C budget (and 25% of our +1.5C budget).
There are abundant accessible coal reserves left and they keep being burnt. Not to mention oil (conventional oil production is in a slight decline that may accelerate, but unconventional oil reserves could be plentiful), natural gas, remaining deforestation potential, and methane emissions from livestock, melting permafrost or industrial leakage.
Additional oil-powered SUVs, trucks, ships, and planes are being delivered. Gas-heated homes, steel and chemical plants are being built. Amazon Forest concessions keep being signed. These are among the signs of our energy and agricultural system’s inertia and increasing “carbon lock-in” of our infrastructure, equipment, and consumption habits. The energy system, unfortunately, changes much, much slower than the IT system.
What can this lead to? The latest IPCC report suggests that under Business-As-Usual, we would likely reach +3C warming as soon as ~2060, and +4C by 2100. Whether this spells tough adaptation, regional disasters, perpetual wars everywhere, or civilisational collapse is open to interpretation. But quantified estimates of increasing frequency and magnitude of heatwaves, floods and droughts, and concerns about climate-induced chronic stress on soils, forests, and marine wildlife through ocean warming, acidification, stratification and deoxygenation would at least suggest a dangerous world, which future generations would probably prefer that we do not choose for them (not to mention other ecological crises, whether the climate changes or not).
How did we get in this mess?
Since 1990, CO2 emissions have stagnated in the US (they have been declining since 2007, mostly through the low-hanging fruit of replacing coal power with gas) and were reduced by 25% in the EU-27. Consumption of imported goods however has mostly offset gains, perhaps even increased overall emissions of high-income nations. Such poor performance is mainly due to climate denialism, industrial greenwashing, consumer indifference, and political cowardice self-justified by successive cycles of (unmet) technological promises. A socio-political process which shows little sign of change, with updated promises of Carbon Dioxide Removal from the air (CDR) and Solar Radiation Management (SRM).
In the meantime, CO2 emissions in Asia have exploded. Greater energy consumption has lifted out of poverty billions of Middle Easterners, Indians, and Chinese sitting on gigantic and easy-to-use coal, oil or natural gas reserves. Hundreds of millions remain in extreme poverty, billions rightfully aspire to middle-class standards of living.
Africa’s emissions have remained negligible, with unknown prospects as to access to fossil fuel resources.
Can low-carbon technologies alone avoid catastrophe?
A review of available or most-researched low-carbon energy and CDR technologies leads to some doubts as to their potential.
The number of nuclear power reactors operating worldwide has stagnated at around 430 since 1990. Nuclear power production has grown by 1% a year, with ups and downs, and today covers 4% of global energy demand, a declining share. Upper bound estimates by the IAEA lead to 20% of electricity share by 2050 (just electricity, not energy); constructions in the pipeline are however not aligned with that trajectory.
This has nothing to do with resource availability. If the whole world was to deploy nuclear energy at France’s scale and speed in the 1980s, temporary industrial bottlenecks and tensions on uranium-235 supply may arise, but research on thorium-based nuclear power could solve the issue for centuries, perhaps millennia. The main problem with nuclear power is that is it not being deployed: CAPEX and financing costs are high, large projects are suffering delays, logistical intricacies and budget slippages, and governments are timorous to launch such complex ventures that engage their people for millennia.
According to UNEP, 2 500 billion dollars have been invested in renewable energy over the past decade, mainly in solar and wind. This has enabled a “quantum leap” of solar and wind from a near-zero share of global energy supply to a negligible share: 2% of primary energy for wind and 1% for solar. Germany has already spent 300-400 billion euros in its energy transition; in 2019, solar and wind covered 5% of the country’s total energy supply. Solar and wind power are now very cheap, though the cost-effectiveness of managing intermittency in an energy system dominated by renewables remains to be demonstrated (and energy storage through artificial water reservoirs raises a host of environmental issues).
It is physically easier to go from 0% to 3% than from 3% to something like 50%. Next steps could see energy and material limits to the deployment of renewables. Energy return on investment (EROI) is the amount of usable energy delivered from an energy source versus the amount of energy used to get that energy resource. EROI calculation faces many methodological challenges, and solar and wind power are subject to vastly divergent estimates, from most optimistic researchers suggesting an EROI higher than fossil fuels, to pessimistic ones suggesting EROI too low for sustaining our civilization. The Fraunhofer Institute estimates solar PV EROI in Europe of over 20. Conversely, Capellan-Perez and De Castro estimate an “extended EROI” of about 1.8 for solar PV and 2.9 for onshore wind. While waiting for researchers to converge (or for reality to provide answers) it seems for now difficult to conclude on the energetic viability of a system dominated by renewables.
The data shows less uncertainty on the quantities of copper, lithium, nickel or cobalt necessary for large-scale deployment of energy production and distribution systems based on diluted and variable sources such as the sun and the wind, with no natural storage function. The energy transition has barely started that the IEA is already alerting on potential tensions on critical minerals. A study by IFP Energies Nouvelles estimates that 89% of known copper resources will have been extracted from mines by 2050 in a 2-degree scenario.
The problem in the short term is not the abundance of minerals in the Earth (or in asteroids), but the metal industry’s ability to increase production fast enough to provide for the energy transition in a cost-effective way. Production is very concentrated in a few countries, some of which are vulnerable to water scarcity. The mining industry in Chili is multiplying projects for building desalination plants (a technology which is quite energy-intensive) due to freshwater shortages.
Another matter is the exponentially increasing energy intensity of metal production as ore grade decreases (which it does, as historical megatrends show).
These are of course no excuse for rejecting renewable energy. There remains potential for technological innovation and lifestyle adaptations. And given the situation, it would be insane for humanity not to even try to harness solar and wind energy.
We are nevertheless in a puzzling world where we need ever more energy to get energy (EROI is decreasing).Ever more energy to get water (pumping deeper, desalinating). Ever more metals to get energy. Ever more energy to get metals. And so far, more energy means more CO2. Whether this is all irreversible, and how things could eventually unfold is unclear. But these are trends worth keeping an eye on.
Carbon capture, utilization and storage (CCUS) is another story of unmet technological promise. The (theoretical) potential is to capture up to a good third of global CO2 emissions (ones from concentrated sources such as coal power, gas power, cement and steel plants). Despite hopes over the last 20 years (and excuses for pursuing Business-As-Usual), only 21 projects were operating in 2020. Ironically, CCUS’ main client is the oil industry, CO2 being a useful raw material for Enhanced Oil Recovery.
Another 30 CCUS facilities have been announced since 2017, which is little sign of exponential growth. Though 15-30% of the power plant’s energy is needed to capture its CO2, CCUS’ slow uptake has for now little to do with resource constraints: retrofitting plants is so costly that one might as well build an equipped plant from scratch, projects are long and costly (trillions were printed to cover for Covid-19 costs, but it is apparently not worth investing into a viable climate and possibly avoiding hundreds of millions, perhaps billions of deaths), the Chinese and Indians rapidly need coal power plants to support their economic growth (which is legitimate), and people have other cats to feed (which, literally speaking, is among causes of marine wildlife devastation.
Good news though is that only 5% of electric plants are responsible for 73% of power sector CO2 emissions, making the necessary effort less intimidating (and excuses less valid).
The case of “green” and “blue”hydrogen is not discussed herein. This will be relevant when water electrolysis is deployed at industrial scale using low-carbon energy input, or when natural gas reforming is combined with CCUS.
Consumption levels show little sign of reduction
In the meantime, it could have been conceivable to implement some degree of “energy sobriety” and set an alternative, appealing and socially just prosperity example for developing countries. Instead, we find ourselves in an uncomfortable situation with an even greater effort to make, in an even shorter period, with even more hypothetical technological promises. We are not quite in a situation where we have the luxury of rejecting any viable solution, whether it be nuclear, renewables, carbon capture, energy efficiency, or sobriety. Beggars cannot be choosers.
A review of the scientific literature suggests that “energy sobriety” is not some eco-fascist delirium promoted by Pol Pot nostalgics, but a lever for solving this century’s complex Energy-Climate equation. In its 2018 special report (Global warming of 1.5C) the IPCC uses the term “lifestyles” 56 times, the term “behavior” (in terms of consumption behavior) over 100 times. The report for instance states:
“Demand-side measures are key elements of 1.5°C pathways. Lifestyle choices lowering energy demand and the land- and GHG-intensity of food consumption can further support achievement of 1.5°C pathways (high confidence).”
What exactly is not clear in that statement and in the rest of the report?
Among most ambitious researchers are IIASA’s authors of the LED scenario (Low Energy Demand). They propose a 40% reduction of energy demand by 2050 while providing decent living standards to everyone. This is possible by deploying readily available low-carbon technologies (no need to wait for miracles) and implementing adequate and socially equitable demand reduction measures.
This, however, does not quite seem to be how decision-makers see things.
Are there enough natural resources to capture CO2 from the air?
Since the early 2010s, a scientific consensus has formed around the necessity of capturing at least some CO2 already present in the air (Carbon Dioxide Removal, CDR). Given our carbon budgets and current levels of emissions, this is indeed mathematically inexorable. One equally inexorable fact is that the more CO2 we emit that can be avoided with accessible and reasonable means, the more must be captured with means that are currently inaccessible and, as we are about to see, ecologically unreasonable.
The general public is little familiar with CDR, which mainly draws attention from fossil fuel and agribusiness players seeing opportunities for new markets and for scoring billions in “R&D” subsidies. They perceive CO2 as “just another waste management problem”.
A comprehensive review of CDR techniques is outside the scope of this article. One which appears desirable is soil organic carbon sequestration through adoption of modern agroecological methods such as permanent soil cover, reduced or no-tillage, introduction of temporary pastures and green manure in crop rotations, increase of permanent pasture productivity, agroforestry, controlled composting, and plantation of hedgerows. The scientific literature however suggests that the CO2 sequestration potential is uncertain, and at best about 10-15 years of global CO2 emissions… if all farmers in the world fully adopt such methods. This of course does not mean it should not be done, especially as the agroecological transition presents numerous co-benefits for biodiversity, food security, fertilizer and pesticide reduction, and soil resilience against droughts and floods.
Similarly, other nature-based CDR solutions such as reforestation and mangrove restoration present potential for drawing amounts of CO2 going from a few years to 10-15 years of emissions, among other ecological co-benefits.
But the viability of industrial CDR techniques is for now elusive. “Negative Emissions Technologies” most simulated in Integrated Assessment Models include BECCS (Bioenergy with Carbon Capture and Storage, a remake of the competition between biofuels and human nutrition, except this time carbon is captured when converting biomass into energy), and DACC (Direct Air Carbon Capture, some sort of atmospheric vacuum cleaner fixing CO2 with purpose-made sorbents).
Resource consumption orders of magnitude are mind-boggling. Realmonte et. al estimate that the capture of 1.5GT of CO2 per year through DACC (hence 4% of current emissions) would require 300EJ primary energy input, that is half of current global energy supply. Colossal quantities of energy are indeed required for producing and regenerating CO2 sorbents such as sodium hydroxide. Estimates across other studies confirm that hundreds of exajoules would be required yearly for capturing a few % of CO2 emissions. Whether it is possible to discover resource-efficient CO2 sorbents is presently uncertain, but not to be excluded.
BECCS is gradually being used by industry players for “carbon neutrality” allegations, but its impacts on food security and biodiversity could be disastrous. Hippman et al. assess that a decarbonation scenario heavily depending on BECCS would require the equivalent of one third of global cropland as soon as 2050, to capture 15-20% of CO2 emissions. To give an idea, such a “gain” would be cancelled out with about 12 years of CO2 emissions growth at the current pace. Other estimates confirm that millions of square kilometers would be required to capture minor quantities of CO2.
Hof et al. estimate that a +1.5C world with wide deployment of BECCS would be worse for biodiversity that a +4C world (which is not hospitable at all to biodiversity). Stenzel et al. estimate that irrigation needs of BECCS could raise water scarcity risks for an additional 2 billion people, a risk comparable to a +3C world.
At first sight, devastating the Earth’s habitability in the name of climate change mitigation appears to be a questionable idea.
Spraying more chemicals into the atmosphere to save the planet?
Faced with assessments such as the ones described so far, what was once deemed to be science fiction is gaining ground: a growing number of climate scientists are turning towards the controversial possibility of controlling the Earth’s thermostat though Solar Radiation Management (SRM), with pioneers such as David Keith and Frank Keutsch from Harvard leading the SCoPEx project.
If we exclude costly and uncertain SRM techniques that will only ever exist in fertile imaginations (e.g. space mirrors), those that are judged ineffective (e.g. surface brightening), and those which would at best have regional climate effects (marine cloud brightening, cirrus cloud thinning), Stratospheric Aeorosol Injection (SAI) is described in the IPCC’s special 2018 report as “the most-researched [solar geoengineering] method, with high agreement that it could limit warming to below 1.5°C.”
But the same report states:
“Solar radiation modification (SRM) measures are not included in any of the available assessed pathways. Although some SRM measures may be theoretically effective in reducing an overshoot, they face large uncertainties and knowledge gaps as well as substantial risks and institutional and social constraints to deployment related to governance, ethics, and impacts on sustainable development.”
The principle behind SAI is simple and consists of spraying aerosols into the stratosphere (e.g. sulfur dioxide, perhaps preferably calcium carbonate to avoid re-destroying the ozone layer or triggering acid rains). The aerosols’ properties reflect incoming solar radiation back into space, thereby diminishing radiation that reaches the Earth’s surface. SAI would thus have no effect on atmospheric CO2, which would for instance leave ocean acidification unsolved (but do not panic, there are also geo-engineering solutions being considered to enhance ocean alkalinity).
The longer we extend our fossil fuel addiction, the longer we need to shoot aerosols up into the stratosphere and increase dosages. If the system were to be stopped (e.g. because of unforeseen and unwanted consequences), the climate “termination shock” would be brutal, especially if we fail to reduce CO2 concentration in the meantime.
Deployment costs of SAI are negligible: a few billion dollars, perhaps 10-20 billion in case of climate inaction. Neither is there much technological challenge or resource issue. Required quantities of aerosols are small compared to what industry is already processing (ironically, sulfur is a by-product of the oil industry, mostly used for fertilizer manufacturing). A few hundred planes would be needed to permanently patrol the skies and spray aerosols, perhaps a thousand in case of climate inaction (presently there are on average 14 000 planes in the sky for commercial aviation).
Deployment time of SAI would be a few years, not half a century (or never) as in the case of the energy transition.
It would be the end of blue skies, which can be seen as a small price to pay, but also as quite a dystopian world with white-yellowish skies patrolled by drones, ads hiding the stars, and no birds.
Solastalgic considerations aside, simplistic modelling assumptions could fail to anticipate detrimental side effects such as changes in plant growth and precipitation or stratospheric heating. Another element is the moral hazard of discouraging already meagre efforts for reducing our ecological footprint. With or without climate change, soils are degrading, water stress is worsening, and biodiversity is collapsing.
SRM is unsurprisingly promoted by fossil fuel industry leaders such as Rex Tillerson (former Exxon CEO and Trump Secretary of State), who describes climate change as “an engineering problem, with engineering solutions”. But SRM would be at best a temporary palliative to gain time for fossil fuel phase-out. More generally, no technological solution makes sense as long as we fail to engage into a solid and sincere trajectory that preserves and regenerates the biosphere.
If we were to engage on such a path (perhaps not one that would achieve the 1.5C target which now seems out of reach, but at least the beginning of something) and if after 20-30 years decision-makers were to announce that despite efforts, we temporarily need to resort to SRM, this would be more credible and ethically acceptable (and the “termination shock” would be less brutal in case we stop).
But we are in nearly nothing that even remotely resembles that. Another preoccupying observation is that CDR and SRM are absent from public debate. Decisions are thus likely to be made outside democratic processes.
Energy Voices is a democratic space presenting the thoughts and opinions of leading Energy & Sustainability writers, their opinions do not necessarily represent those of illuminem.