Image: Turbine maintenance at Drax Power Station. Copyright: Drax
Table of Contents
When the United Nations surveyed 1.2 million people around the world in 2021, the results were clear. The three most popular policies to combat climate change were conserving forests, switching from fossil fuel power to solar and wind, and changing our methods of farming to emit less greenhouse gas.
It might be surprising then, to learn that most of the scenarios modelled by the Intergovernmental Panel on Climate Change (IPCC) – the UN body and world authority on climate policy –include burning hundreds of millions of hectares of trees and plants, offsetting the emissions from farming, and continued fossil fuel emissions until at least 2050.
The complex reasons for this include the ‘net zero’ accounting approach to climate policy, the type of models the IPCC uses, and a group of technologies known as carbon capture and storage, or CCS.
The first article in this series explained what CCS is, how it works, and what projects have happened so far. It described how CCS can be used to remove the carbon dioxide from smokestacks on power plants and industrial facilities, and then inject it underground.
This article explores how climate modelling anticipates the future of CCS. It also explains more about another group of CCS technologies, classed as creating ‘negative emissions’, rather than just reducing the pollution of a particular facility.
Some IPCC pathways have CCS capturing and storing over a thousand gigatonnes of CO2 in total by 2100. To put that in perspective, it is nearly 29 times larger than the entire planet’s carbon dioxide emissions in 2021. This would be an unparalleled technological feat, marking an historic era of change such as the Industrial Revolution or the Computer Age, involving thousands of facilities around the planet.
Twenty years ago, the CCS industry was focused on coal. This was partly because coal power was the most viable application of CCS at the time: capturing carbon in a smokestack is cheaper and simpler than smaller applications (e.g. vehicle exhausts), and had already been proven possible. Coal sector emissions were greater than all other smokestack emissions combined, making it the obvious target.
Coal was also targeted because climate discourse and diplomacy focused on energy and power at the turn of the millennium, and wind and solar were less established than they are now. In the Third Assessment Report in 2001, the IPCC laid out three possible pathways to lowering greenhouse gas emissions: exploiting “unconventional” oil and gas resources (e.g. tar sands, shale oil, or coal bed methane), developing non-fossil energy sources, or developing “fossil energy technology with carbon capture and storage”. The race to define 21st century energy and power generation was on.
CCS seemed to lose this race. Over the following decade, huge progress was made in renewable electricity technologies and extraction of unconventional oil & gas resources expanded, while numerous CCS projects proved unsuccessful. Today, no traditional coal power stations are operational using CCS, and only two are in development globally, while wind and solar power is rapidly growing worldwide. Whether the apparent failure of coal-power CCS can be blamed on political, commercial, or technological reasons is a point of debate, but the result is that the world began moving away from coal use before CCS was able to save it.
Many IPCC pathways to keep global heating below 1.5°C do still include fossil fuel CCS, but with gas, not coal. On average, they show 13% of global energy in 2050 being produced by gas, only possible by combining it with CCS. Largely though, the CCS conversation – both politically and scientifically – has now shifted to capturing or offsetting ‘residual’ or ‘hard to abate’ emissions, and the need to create ‘negative’ emissions.
Residual emissions refer to industries or sectors that would otherwise be difficult to decarbonise. While we can electrify many pollutant sectors with renewables, such as heating or road vehicles, some industries are considered reliant on greenhouse gas emissions, at least in the short term. These can include, depending on whom and where you ask: aviation, shipping, agriculture, chemical manufacturing, and steel, cement, and iron production.
Chemical, steel, cement and iron production emit through factory smokestacks, making them eligible for on-site carbon capture. This is often termed ‘industrial CCS’, and is in many ways the least contentious form of CCS. While there are other theoretical ways to decarbonise some of these industries, most agree that CCS is the best approach.
Dr. Howard Herzog is a pioneer of CCS research and a Senior Research Engineer at MIT’s Energy Initiative. He says that some industrial carbon capture such as with cement production is actually “easier” technologically than capturing carbon from power stations. This is because of the higher partial pressure – which is related to the concentration of CO2 in the flue gases. A higher proportion of CO2 in the flue gas, a higher partial pressure, making it easier to capture.
This could make some industrial CCS cheaper than power CCS too.
“The reason cement has higher partial pressure is because of all the CO2 coming off of the limestone. So we’re driving CO2 off of the feedstock, and that’s adding to the CO2 in the flue gas. So that’s why it has higher partial pressure, but you have got to be careful on what the impurities are in your stream. There’s a lot of factors.”
Aviation, shipping, and agriculture are more tricky: without smokestacks, emissions are more likely to be offset using negative emissions technologies (NETs). For CCS, this kind of carbon removal involves either bioenergy with CCS (BECCS), or direct air CCS (DACCS). Rather than minimising the pollution of one particular source, such as a power plant, NETs promise to remove CO2 from the atmosphere itself. Under this strategy, hard-to-abate sectors continue to emit greenhouse gases, while equivalent levels of carbon dioxide are removed from the atmosphere elsewhere.
Dr. Steve Smith, Executive Director of both The Oxford Net Zero Initiative and CO2RE, a research hub focussed on greenhouse gas removals, says that offsetting agriculture using CCS is particularly interesting, “because actually, a lot of the residual emissions aren’t CO2 emissions, they’re nitrous oxide emissions and methane emissions.” These three gases all have different impacts on global heating, with different ‘lifetimes’ in our atmosphere before being broken down, and different abilities to trap heat. While “nitrous oxide is quite a long lived gas,” methane only impacts our atmosphere for around a fifth of the time that carbon dioxide does – but in its first 20 years in the atmosphere, it has more than 80 times the warming power of CO2.
Unsurprisingly, Smith said that offsetting methane or nitrous oxide by removing carbon dioxide requires “a lot of gory details” in the carbon accounting. But there are bigger challenges ahead with NETs than managing the books.
The challenges ahead with BECCS & DACCS
Speaking to the UK’s Environmental Audit Committee in November, Dr. Smith opened a session by describing the “very wide range” of negative emissions technologies (NETs) available. The caveat? “Essentially, all of these are at zero or close to zero levels of deployment”.
Not all NETs include CCS. Alternatives include processes that store carbon in a charcoal-like substance called biochar by heating plant matter but not burning it, or that change the alkalinity of the ocean with powered silicate or carbonate rock, known as enhanced weathering. Bioenergy with CCS is used far more than these methods in IPCC modelling – and direct air capture is its up-and-coming competitor, with major investments from super-corporations like Microsoft, Shopify and Stripe.
There are various forms of BECCS, but typically it is similar to coal-plant CCS, with the power plant burning wood or plant matter instead of coal. Burning biomass for power like this is considered a renewable and carbon neutral process, despite it often emitting more CO2 than coal in the short term. This is because trees can be replanted and regrown, in theory re-sequestering the CO2 released from burning. Another explanation of BECCS’ carbon neutral label is that bioenergy emissions are not counted in the energy sector under IPCC rules – whoever cuts the trees down records them instead. For the bioenergy company, or its government, if the biomass is bought overseas, no emissions need be recorded in the carbon accounting. Once they add CCS, they can record negative emissions instead.
DACCS’ negative emissions are less complicated to understand, and less controversial: it works similarly to the CCS processes in power plant, but captures CO2 directly from the air itself, rather than in a smokestack. Large, outdoor fans blow air through a solvent or membrane which can separate and capture the CO2 from other gases.
According to the International Energy Agency (IEA), there are now 19 direct air capture projects operating globally, capturing more than 0.01 Mt CO2/year. In order to be on track for the IEA’s Net Zero Emissions by 2050 Scenario, we need to be capturing and storing 8500 times this amount by 2030 through direct air capture alone. There are plans for DAC plants that can capture much larger quantities – Carbon Engineering are developing one in the US that hopes to capture 1 Mt CO2/year alone – but even with this, upscaling by a magnitude of 8500 in eight years seems improbable. And the IEA says that before we can begin this level of deployment, “several more large-scale demonstrations to refine the technology” are required first.
Experts interviewed for this article, however, described issues around cost and logistics, but not tech: “Yeah, it exists, people are doing it,” said Smith. “It’s very small scale, and it’s very expensive. It’s much more expensive currently than doing pretty much any other abatement we would do.”
This high cost is largely for the same reason that MIT’s Dr. Herzog stated CCS for cement production was relatively easy: CO2 concentration. Typical flue gas in a natural gas or oil plant might contain 10-13% CO2. The air typically is closer to 0.04%, according to Herzog, making every tonne of CO2 captured significantly more costly.
Direct air capture needs a lot of space, which becomes a problem when envisaging large scale roll out. As Dr. Herzog explained, “you have to process a lot of air. And the air can only go through these machines at a certain speed. So Carbon Engineering is designed for 1 million tons per year, which means they need a cross sectional area of 46,000 square meters. That’s 10 meters high – a three storey building – and almost five kilometres long.”
“Wind takes land too but you can do other uses. The question is, could you really do other uses with these fans blowing all the time? And you’ve got these chemicals, these absorbers, over say, one or two square kilometres, then you have to bring everything to a central place to collect the CO2 and to regenerate the solvent. They’re going to have issues trying to implement this.”
Herzog compares direct air capture with wind power, but the two will likely come as a package. Carbon Engineering plan to build wind farms solely to power the DACCS process. According to their Vice-President Dr. Amy Ruddock, “for a megatonne-scale direct air capture plant, you are talking about a typical-scale wind farm to support it” . A 2019 study that explored modelling that relied on DACCS to keep global heating under 2C estimated the technology would need 300 exajoules of energy per year by 2100. That is well over half of the entire world’s energy demand today – just to power direct air capture.
There are similar issues with BECCS. Even aside from questions about whether the process really does create negative emissions – there are studies that show otherwise – burning biomass at scale will require huge areas of land to grow and harvest trees or crops. One estimate showed that in some IPCC scenarios BECCS would require space more than twice the size of India. This is not only logistically difficult, but raises issues around the human impacts of land use change and food security, as well as risks around biodiversity and deforestation. Bioenergy without CCS has already driven unsustainable logging in protected forests across the US, EU, and Russia.
Supply chains increase emissions but are not yet properly accounted for. The world’s flagship BECCS project, Drax’s plant in England, is not yet storing carbon. It is the UK’s largest single emitter of greenhouse gases, but records no emissions in official accounts. As the biomass is sourced from overseas, the UK government does not even record the carbon loss from cutting the trees down, nor guarantee new trees are regrown.
Overshoot and the net zero balancing act
Read more about risks with the net zero framework:
Rethinking net zero: why Holly Jean Buck’s ‘Why Net Zero is Not Enough’ is not enough
BECCS facilities are considered to have ‘carbon debt’ until new trees or crops have grown to replace the burned ones – a process that can last over a century. Carbon debt is in keeping with a general rule of net zero modelling: what is happening today is not the priority, as long as everything balances out in the end.
In 2015, the Paris Agreement set the deadline of the year 2100 to keep global heating to 2, or ideally, 1.5°C above preindustrial levels. In 2018, the IPCC’s Special Report on Global Warming of 1.5°C established net zero emissions by 2050 as the carbon accounting benchmark for achieving this temperature target.
Yet both of these targets – emissions and temperature – rely on this balancing act approach. The carbon debt from BECCS is just one example of how carbon accounting can be distorted to present something as carbon neutral or negative today, despite present-day emissions not being sequestered until a distant future. Carbon offset markets are now selling carbon credits for future projects, so you can offset emissions by paying for a BECCS or DACCS company to promise they will remove those emissions a decade later. It is conceivable – likely even – that we will hear success stories of net zero targets being met in coming years, when in reality the carbon removal or offsetting has been paid for but has not actually happened.
In 81 out of 90 of the scenarios to limit global heating to 1.5°C in the IPCC’s 2018 Special Report, global heating does in fact rise higher than 1.5°C temporarily, before being brought back down by colossal amounts of carbon dioxide removal in the second half of the 21st century. These are commonly described as ‘overshoot’ models. On average, most IPCC scenarios use carbon removal to offset residual emissions until 2050. Then, in the latter half of the century, negative emissions become the priority, so much that the cumulative volume of removed CO2 is over double the amount emitted by hard to abate sectors.
IPCC scenarios are integrated assessment models: rather than proposing the best pathways on purely scientific and climatic terms, they also involve societal, economic, and arguably political considerations, as Professor Wim Carton discussed on a recent episode of our ELCI podcast. As the UN-backed Center for International Climate Research (CICERO) wrote in 2021, these models are “useful tools for financial institutions”, but should not be misunderstood as presenting “the most feasible or desirable” pathways.
Direct air capture’s current high cost has minimised its role in most IPCC models, even though it has fewer issues than BECCS around environmental and human impact. Like coal-CCS and renewables, it could be that direct air capture becomes cheaper, and plans change. A 2021 study in Frontiers in Climate also suggested policymakers might be wrong about BECCS’ cost, using systems modelling to demonstrate that DACCS would actually be cheaper at scale. Yet for now, the IPCC’s two main competitors for the late 21st century carbon removal surge are BECCS and afforestation (planting new forests).
The 1.5°C Special Report suggests that the use of BECCS and afforestation is largely interchangeable. Unless offset markets make afforestation particularly lucrative, commercial interests are likely to back BECCS in this fight. Unlike planting trees, BECCS requires corporations to run the facilities, various markets to supply biomass, transport, and technologies, it produces energy, needs a larger workforce, and can make use of old fossil fuel power plants. It might seem counter-intuitive, but the same factors that make BECCS more logistically difficult than afforestation make it more commercially and politically friendly. If it is unprofitable on its own, governments will subsidise it.
Experts are increasingly sounding warning bells about these overshoot models. Numerous studies in Nature and Nature Climate Change in critiqued the concept of temperature overshoot. The studies found that overshoot would:
- Cause more variable and extreme weather conditions such as flooding, wildfires and droughts, which in turn will harm food security, biodiversity, ecosystems, and living standards.
- Cost a lot more than avoiding overshoot. Even without the economic impact of climate change, so just looking at the transition costs, the authors found that global GDP would be 2% higher in scenarios avoiding temperature overshoot. This directly challenges assumptions in modelling that overshoot models save money by allowing more time to transition.
- Increase the chances of crossing irreversible climate tipping points, such as ice sheet melts.
The IPCC’s Working Group Two acknowledged many of these issues itself in its February 2022 report. According to Hoesung Lee, chair of the IPCC, this was the “first time” the IPCC had analysed the impact of a temporary overshoot “in very high detail”. “There will be some impacts that will be irreversible even if the temperature will return to 1.5°C at the end of the century. That is one of the major, major findings of this report,” he said on publication. It remains to be seen whether this scepticism will affect Working Group Three’s revised pathways, which are set to be published in early April.
Other studies have demonstrated the risks involved with relying on unproven future technologies to reverse temperature overshoots. A 2017 study in Nature Geoscience warned that overshoot models would “would turn into potential sources of political flexibility”, removing accountability around decarbonisation. This would likely continue even after the temperature overshoot has happened, they wrote:
In other words, unless our political and geopolitical processes radically change within the next two decades, overshoot models are likely to lead to a permanent overshoot – global heating will not return to below 1.5°C.
There are risks with relying on new technologies, too: what if negative emissions technologies do not work as well as expected, or what if it is too difficult to deploy them at scale? A 2020 study in Climatic Change found that not only could these technologies fail to reverse temperature overshoot, they could actually cause “an additional temperature rise of up to 1.4 °C.” This followed a similar study in Nature Communications that looked at direct air capture as an alternative to BECCS in IPCC modelling. The authors found that “the risk of assuming that DACCS can be deployed at scale, and finding it to be subsequently unavailable, leads to a global temperature overshoot of up to 0.8 °C.”
What are the alternatives?
The IPCC does provide scenarios without overshoot, but these mostly rely on even greater levels of CCS before 2050. Given that 78% of large-scale CCS projects have failed to date, and there are only 132 commercial CCS projects currently being developed, it increasingly unlikely that enough CCS facilities will be developed in time, or that they would meet their own targets. Most of the 132 projects are years away from being operational, while we would need roughly this many facilities to come online every year, from now, to meet these pathways.
The no-overshoot 1.5°C scenario with the least amount of BECCS and DACCS has carbon removal in the low hundreds of gigatonnes of CO2 over the 21st century, rather than closer to a thousand GtCO2. This is only considered possible alongside radical behavioural changes happening very quickly. People and companies would need to buy less, use less energy, travel less, and eat less red meat, dairy, and imported food in the next few years. The report also notes that the more that models include these changes, the lower the “number of trade-offs with respect to sustainable development”.
For the modelling behind the report, only one out of 21 teams explored scenarios without any CCS, whether industrial, power plant CCS, or negative emissions technologies. This no-CCS model was the German Global Energy System Model – but even before the 1.5°C Special Report was published, this model was changed to include BECCS. The singular reason given for this change is that it would allow the team to “better compare their results” with other models. Despite the different types of scientists across the IPCC modelling groups – including one team from Shell, the oil company that is alone responsible for nearly 2% of global greenhouse gas emissions since 1988 – there is clearly a trend towards uniformity with these models. On a recent ELCI podcast, Professor Wim Carton blamed this on a lack of geographical and political diversity in the modelling teams: scientists from the global south might be less dismissive of climate justice and land-use issues around BECCS and DACCS.
For this reason, real alternatives to the tech-heavy and cost-optimised IPCC models may need to come from elsewhere. New IPCC pathways are set to be published next week, but are not expected to scale down the emphasis on NETs. In small numbers and at national levels, climate modelling and 2050 scenarios that aim for ‘true’, ‘real’ or ‘absolute’ zero rather than net zero, and which do not rely on CCS, do exist. UK FIRES, a research programme across various universities, produced a report for the UK in 2019, for example. There is no global equivalent, although various campaigns exist that promote total fossil fuel phase-outs and commitments to end emissions rather than offsetting.
Listen to our podcast with Wim Carton:
Is climate modelling undermined by economics and ideology?
Rather than accepting IPCC models as definitive pathways, we should ask whom they are really made for, and made by. If UN-stamped publications like CICERO’s acknowledge that integrated assessment models are primarily useful for financial institutions, why is it that climate journalists, researchers and politicians treat them otherwise?
Climate mitigation represents the first time in history the entire planet has needed to work together towards a single goal. It is natural, and positive in many ways, that the world looks to the UN to lead. In science fiction, the existential crisis that leads to global cooperation or governance often marks the beginning of a utopian future. We do need centralised and global planning around climate change, but we should be very cautious before basing our plans on models catered to certain commercial interests in the Global North. Hopefully, it is not yet too late to hit our temperature targets without overshooting them. It is vital that we know the most effective and ethical ways of doing this, not just the cheapest.
 One of the primary sources for this series is Dr. Herzog’s 2018 book, Carbon Capture, published by MIT Press.
 For example, HYBRIT‘s iron and steel.
 More specifically, she described the energy requirements as “80 to 370 kWh per tonne of CO2 of electrical energy and about 5 GJ of thermal energy needed per tonne.” UK Environmental Audit Committee, Oral evidence: Technological Innovations and Climate Change: Negative Emissions Technologies, HC 738.
 For example, see Yassa, S., & Greene, N. (2021) A Bad Biomass Bet: Why the Leading Approach to Biomass Energy with Carbon Capture and Storage Isn’t Carbon Negative. NRDC.
 E.g. see Dart, T., & Milman, O. (2018). ‘The dirty little secret behind ‘clean energy’ wood pellets’, The Guardian; Kuresoo, S. (2021). ‘EU renewable energy policy subsidizes surge in logging of Estonia’s protected areas’. Economy, Land and Climate Insight; or read Earthsight’s investigations into Russian wood exports, or listen to ELCI’s podcast with their director Sam Lawson here.
 Ritchie, P.D.L., Clarke, J.J., Cox, P.M. et al. (2021). ‘Overshooting tipping point thresholds in a changing climate’. Nature 592, 517–523.
 They used an average scale-up rate of 1.5GtCO2/yr.
Bertie Harrison-Broninski is an Assistant Editor at Land and Climate Review, and a researcher studying bioenergy and BECCS policy for Culmer Raphael. He is also a freelance investigative journalist, with recent work including Al Jazeera’s Degrees of Abuse series and John Sweeney’s book on Ghislaine Maxwell. Find him on twitter @bertrandhb.