CCUS is an important technological option for reducing CO₂ emissions in the energy sector and will be essential to achieving the goal of net-zero emissions. As discussed in Chapter 1, CCUS can play four critical roles in the transition to net zero: tackling emissions from existing energy assets; as a solution for sectors where emissions are hard to abate; as a platform for clean hydrogen production; and removing carbon from the atmosphere to balance emissions that cannot be directly abated or avoided. These roles are evident in the IEA Sustainable Development Scenario, in which global CO₂ emissions from the energy sector fall to zero on a net basis by 2070 worldwide compared with the Stated Policies Scenario, which takes into account current national energy- and climate-related policy commitments. The contribution of CCUS to emissions reductions grows over time as the technology progresses, costs fall and other cheaper abatement options are exhausted.

Around 60% of the CO₂ captured in the period to 2070 in the Sustainable Development Scenario is from fossil fuel use; the rest is from industrial processes, bioenergy and DAC. This reflects difficulties in eliminating CO₂ emissions in certain industry sectors (and hence the need for carbon removal options), the high share of process emissions in some industries (in particular cement), the scope for capturing CO₂ in the production of biofuels for transport; and the increasing role for DAC in providing CO₂ as feedstock for producing synthetic aviation fuels as well as for removing CO₂ from the atmosphere. Fossil fuels are still the source of about half of CO₂ captured in 2070, with around one-third of this from low-carbon hydrogen production from natural gas.

When net-zero emissions are reached in 2070 in the Sustainable Development Scenario, 9.5 Gt CO₂ is captured and stored and another 0.9 Gt is captured and used. The power sector accounts for around 40% of the captured CO₂ in 2070, with almost half of it linked to bioenergy. Around one quarter of the CO₂ captured is in heavy industries, where emissions are hard to abate in other ways. Almost 30% is in the fuel transformation sector to produce hydrogen and ammonia from fossil fuels as well as CO₂ captured from biofuel plants, which is used to make synthetic fuels or stored for carbon removal. A further 7% comes from DAC, again, as a carbon-neutral source of CO₂ for fuel and feedstock production or to create negative emissions (DAC with CO₂ storage, or DACS). While only around 8% of total captured CO₂ is used or “recycled”, it plays an important role in supporting the decarbonisation of the transport and industry sectors through the production of transport fuels and as a feedstock for the chemical industry.2

The contribution of CCUS to reducing global energy sector CO₂ emissions in the Sustainable Development Scenario evolves over the projection period, with three distinct periods. In the first phase to around 2030, the focus is on capturing emissions from existing power plants and factories. In the power and industry sectors, over 85% of all CO₂ emissions captured in this decade are from plants retrofitted with CO₂ capture equipment: coal-fired power units (and, to a lesser extent, gas-fired power units); chemical plants (mainly fertilisers), cement factories, and iron- and steelworks. Some low-cost CO₂ capture opportunities in hydrogen and bioethanol production are also developed, building on the current portfolio of projects. Total capture reaches 840 Mt in 2030. Cumulatively to 2030, CCUS contributes around 4% of the overall emissions reductions in the Sustainable Development Scenario relative to the Stated Policies Scenario.

During the second phase, from 2030 to 2050, the contribution of CCUS to cumulative emissions reductions grows to 12% relative to the Stated Policies Scenario. CCUS deployment expands most rapidly in the cement, steel and chemicals sectors, which together account for around one third of the total growth in global CO₂ capture during that period. In power generation, the focus shifts to natural gas-fired stations, which help to integrate variable renewable energy sources (mainly solar and wind) in some regions by providing short-term flexibility and to balance seasonal variations in electricity demand or renewable generation, for which batteries are less well suited. Hydrogen production from fossil fuels (primarily natural gas) is responsible for a fifth of the overall growth in CO₂ capture in the 2030-50 time frame, driven by increasing hydrogen demand in long-distance transport modes (such as trucks and shipping). BECCS also expands significantly, accounting for around 15% of the growth in CO₂ capture over that period. By 2050, around half of BECCS capacity is in the power sector and the remainder primarily in producing alternative low-carbon fuels, in particular biofuels. BECCS benefits from economies of scale and cost reductions through technological advances and learning-by-doing, which are generally highest at the early stages of the adoption of a technology.

In the last phase from 2050 to 2070, the amount of CO₂ captured jumps by 85%, as carbon removal and the use of CO₂ accelerate. Around 45% of the growth during this period comes from BECCS and 15% from DAC, while capture from natural gas dominates the increase in CO₂ capture from fossil fuels, driven by the production of hydrogen and electricity in regions with low-cost gas resources. In 2070, about 35% of all CO₂ emissions captured are from bioenergy or DAC, most of which are stored, generating negative emissions to balance all remaining emissions from transport, industry and buildings so as to achieve a net-zero emissions energy system. Around one fifth of all the CO₂ captured from bioenergy or directly from the air is used in combination with clean hydrogen to produce synthetic hydrocarbon fuels, notably for use in aviation, where synthetic fuels meet 40% of aviation fuel demand. The scale-up of carbon removal in the Sustainable Development Scenario implies an average of around 50 BECCS and 5 DACS plants of 1 Mt/year being added each year from 2020 to 2070. By 2070, 800 Mtoe (33 EJ), or more than a quarter of global primary bioenergy use, is linked to BECCS, with almost half of the bioenergy in the power and fuel transformation sectors being used in plants equipped with capture facilities. The deployment of these carbon removal technologies is constrained by their cost-competitiveness with other mitigation measures and (potentially) access to suitable storage, with BECCS also constrained by the availability of sustainable bioenergy and DAC by the availability of low-cost electricity and heat (see the section on carbon removal below).

The Sustainable Development Scenario reaches net-zero emissions from the energy sector within five decades on the back of ambitious technological change and optimised innovation systems comparable to the fastest and most successful clean energy technology innovation success stories in history. The Faster Innovation Case explores the opportunity to accelerate this transition to bring the global energy system to net-zero emissions 20 years earlier, by 2050. This variant of the Sustainable Development Scenario considers a more rapid deployment of new technologies, and innovative techniques to enable additional carbon removal, for example by expanding sustainable biomass supply.

Captured CO₂ can be either permanently stored deep underground in geological formations or used in a variety of ways, including for EOR or as a raw material in the production of fuels, chemicals or building materials.4 More than 90% of all the CO₂ captured over 2020-70 in the Sustainable Development Scenario is stored, with 80% of the stored CO₂ coming from fossil sources and industrial processes and 20% from bioenergy and DAC. Of the CO₂ that is used, around 95% is used as feedstock for synthetic fuel production, while the remainder is used in the chemicals sector.5 This represents a major change in how CO₂ is used. Today, the majority of CO₂ captured and used is for EOR – where nearly all of the CO₂ is permanently stored – or in the chemicals sector where the CO₂ is captured and used within the same process to produce fertiliser and ultimately released into the atmosphere.6 In the period to 2030, CO₂ use for synthetic fuel production is scaled up, building on projects already underway such as the planned Norsk-e Fuel project in Norway (see the section Long-distance transport below). This shift increases CO₂ use by around three-quarters relative to today, albeit from a small base, and paves the way for greater deployment of synthetic fuels in aviation in the longer term.

The contribution of CO₂ use to reaching net-zero emissions depends in large part on the source of the CO₂. By 2070, all synthetic fuel production uses CO₂ sourced from bioenergy or DAC so that burning these fuels is carbon-neutral (using CO₂ captured from fossil fuel sources would still result in emissions). In the preceding period, some of the CO₂ used come from fossil fuels or from industrial plants, which contributes to CO₂ reductions by reducing reliance on the direct use of fossil fuels in the transport and industry sectors (see Chapter 3 for a discussion of the climate benefits of CO₂ use). 

One of the defining challenges for global energy transitions is how to reduce CO₂ emissions from the existing stock of energy-consuming assets – vehicles, factories, public and residential buildings, and infrastructure. Some of these assets, notably power stations and industrial plants, are built to last for decades, effectively locking in their emissions unless they are modified in some way to emit less or are retired early. Retrofitting CO₂ capture facilities to existing plants and storing the CO₂ underground is one way of addressing this challenge.

Existing industrial and power plants, if they continue to operate as they do now through to the end of their normal operating lifetimes, would generate over 600 Gt of CO₂ emissions worldwide between now and 2070 – around 17 years’ worth of current global emissions. Continued operation of the existing transport fleet and building stock would increase cumulative locked-in emissions by a further 150 Gt. Emissions of that magnitude would exhaust the majority of the remaining CO₂ budget in the Sustainable Development Scenario through to 2070. In other words, it would permit hardly any new energy-consuming assets of any description to be brought into use ever again.

The power sector is the main source of emissions from existing assets, accounting for 410 Gt worldwide to 2070 – 80% of which is from coal plants. China alone contributes almost half of global cumulative emissions from existing power plants, and other emerging economies most of the rest, mainly due to their younger fleets. Most of the investment in those assets occurred over the past two decades, when their economies were growing most rapidly. The average age of coal plants is less than 20 years in most Asian countries and just 13 years in China; in Europe, it is 35 years and in the United States around 40 years. Of the 2 100 GW of coal-fired capacity in operation worldwide today and 167 GW under construction, around 1 440 GW could still be operating in 2050 – 900 GW of it in China alone.

Gas-fired power plants are generally younger, averaging less than 20 years in all major countries with the exception of Japan, the Russian Federation (hereafter “Russia”) and the United States, since gas was introduced as a fuel for power generation in many countries only after the 1990s. Because of their shorter technical lifetime, 350 GW of the around 1 800 GW of gas power plants in operation today and 110 GW under construction could still be operational in 2050.

Industry, particularly heavy industry sectors, is the other major contributor to emissions from existing assets. Of the nearly 200 Gt of cumulative CO₂ emissions from existing industrial assets, the steel and cement sectors each account for around 30% and the chemicals sector for around 15%. As with the power sector, China is the main contributor, due to its dominance as an industrial producer and the relatively young age of its industrial plants. The country accounts for nearly 60% of the capacity used to make iron from iron ore, the most energy-intensive step of primary steel production, for just over half of the world’s kiln capacity for making cement, and for around 30% of total production capacity for ammonia, methanol and high-value chemicals (HVCs) combined.

The majority of China‘s industrial capacity is at the lower end of the age range for each type of assets, averaging between 10 and 15 years, compared with a typical lifetime of 30 years for chemical plants and 40 years for steel and cement plants. The phenomenal growth over the last two decades in China’s output of steel – more than sevenfold – and cement – nearly fourfold – bears testimony to the relatively short time frame over which most of the country’s steel works and cement plants were added. In contrast, the chemical sector is characterised by a more even distribution of capacity both regionally and across different age ranges.

There are three options for cutting locked-in emissions in the power generation and industrial sectors:

• Investing in modifications to existing industrial and power equipment to either use less carbon-intensive fuels or improve energy efficiency
• Retiring plants before the end of their normal operating lifetimes, or making less use of them (e.g. by repurposing fossil fuel power plants to operate at peak-load rather than base-load)
• Retrofitting CO₂ capture facilities and storing or using the CO₂.

For the world to reach net-zero emissions by 2070 or earlier, a combination of the three will be required. Their relative contribution will vary by country depending on their economic viability, social acceptability and implications for energy security. At the level of an individual plant, the least-cost option, in terms of the cost per tonne of CO₂ emissions avoided, depends on the age and technological characteristics of the assets as well as on the market conditions and regulatory framework. In practice, plant modifications and repurposing may be limited by the specific plant characteristics and, in the case of industry, by non-combustion processes. For example, CCUS is effectively the only option for achieving significant reductions in emissions from cement production short of closing the plant, due to the large amount of process emissions and the need for high-temperature heat, which cannot be provided easily and cheaply by non‑fossil energy.

In many cases, early retirement of assets before full repayment of capital costs is an expensive option for plant owners and governments, particularly in emerging economies with younger assets. Retrofitting these assets with CCUS to allow continued operation can provide plant owners with an asset protection strategy and may prove cheaper than early retirement, depending on the size of any carbon penalty and other policy incentives.

From a broader economic perspective, retrofitting CCUS generally makes most sense for power plants and industrial facilities that are young, efficient and located near places with opportunities to use or store CO₂, including for EOR, and where alternative generation or technological options are limited. Other technical features that have to be considered when assessing whether a retrofit is likely to make commercial or economic sense are capacity, availability of on-site space for carbon capture equipment, load factor, plant type, proximity to CO₂ transport infrastructure and confidence in the long-term availability of CO₂ storage capacity. In advanced economies, where industrial capacity is generally older, there is greater potential for early retirement, as the economic losses involved are typically lower. In emerging economies with younger assets, the emphasis is likely to be more on retrofitting plants with more energy-efficient and CCUS technologies.

Retrofitting with CCUS plays a major role in reducing emissions from coal and gas-fired power assets in the Sustainable Development Scenario. Around 190 GW of coal-fired capacity, mainly in China, and 160 GW of gas-fired capacity is retrofitted with CCUS. Globally, retrofits on existing plants account for around a third of all the CO₂ captured from power plants over the period 2020-70, and account for 16% of emissions reductions from existing plants relative to the Stated Policies Scenario. Some existing coal power plants are also repurposed to provide reserve capacity to the power system, thus generating smaller amounts of electricity and CO₂ emissions, and some plants co-fire coal with biomass, also reducing CO₂ emissions. Despite these measures, early retirements of some power plants are unavoidable: around 600 GW of the existing global coal capacity of 2 100 GW today are retired globally earlier than in the Stated Policies Scenario.

CCUS retrofits also play an important role in addressing emissions from existing assets in heavy industry. They account for nearly 90% of CO₂ captured in heavy industry sectors by 2030 and 55% of the cumulative capture volume to 2070 in the Sustainable Development Scenario. Post‑combustion capture technologies are generally more suited to retrofitting than oxy-fuelling and pre-combustion technologies as there is less need for fundamental overhauls in combustion equipment. As with power stations, the regional deployment of retrofits in heavy industry is primarily determined by the age of existing assets, as well as future growth in production: if existing capacity levels are largely sufficient to meet local demand, such as in Europe and North America, retrofits may be an attractive option. Conversely, in countries such as India where production capacities are set to grow strongly, the share of retrofits is lower, given the relatively high investment in new assets. Worldwide, CO₂ capture from retrofits in heavy industry declines rapidly after 2060 in the Sustainable Development Scenario, as the bulk of existing capacity today will have come to the end of its lifetime.

No part of the energy system can avoid the need to reduce emissions, including those sectors where it is particularly difficult or expensive, if the world is to reach net-zero emissions. The main sectors in which emissions are hard to abate are heavy industry, notably iron and steel, cement and chemicals, and the three modes of long-distance transport – trucking, shipping and aviation.

CCUS – alongside electrification, bioenergy and hydrogen – is a major component of the portfolio of technology options to deliver deep emissions reductions in the hard-to-abate sectors in the Sustainable Development Scenario. Improvements in the performance of existing technologies, material efficiency in heavy industry and measures to conserve energy in transport, by avoiding journeys and shifting between modes, can deliver substantial emissions reductions in the near‑term. But for the energy sector as a whole to reach net-zero emissions in the longer‑term, technologies that significantly reduce the emissions intensity of producing a tonne of material or of moving passengers and freight around the world are required.

Heavy industry and long-distance transport together emit around 10 Gt of CO₂ today, or around 30% of total emissions from the energy system, including industrial process emissions. In the Sustainable Development Scenario, emissions from these sectors decline by almost 90% to around 1.5 Gt in 2070. Achieving these reductions requires widespread adoption of near-zero emissions production routes in heavy industry and the substantial replacement of fossil fuels with low‑carbon alternatives in long-distance transport. CCUS plays a critical role in both sectors. By 2070, around 2.7 GtCO₂ is captured in the steel, cement and chemicals sectors, and around 5 mb/d of synthetic fuels are consumed in the aviation sector using around 0.8 Gt of captured CO₂.

Technology options and costs

In heavy industry, CCUS can be applied directly to production facilities to manage industrial process and energy-related CO₂ emissions, through both retrofits as well as the construction of new plants with integrated CO₂ capture facilities. Industry produces large quantities of bulk materials for sale in highly competitive global market places; margins tend to be slim and energy costs account for a large share of overall production costs. In this context, technology costs, along with local regulatory contexts and infrastructure constraints, will be critical in determining the eventual deployment of CCUS alongside other emissions abatement options.

It is generally difficult to reduce industrial process emissions, which are inherent to the chemical reactions involved in producing certain bulk materials, without CO_2 capture. The production of clinker – the key active ingredient in Portland cement – is the prime example here. Process emissions account for around two-thirds of the emissions in a cement kiln. Even if the kiln in which it is produced were to be electrified or fuelled with bioenergy, these emissions would persist. Industrial process emissions amounted to 2.5 GtCO₂ in 2019, of which the cement sector accounted for 63% (the chemicals and steel sectors account for more than half of the rest). There are no alternatives to CCUS at comparable levels of technology maturity that can support deep emissions cuts in this sector. Alternative binding agents could one day constitute an alternative to the use of Portland cement, which produces around 520 kgCO₂ of process emissions per tonne of clinker. Alternative binding materials that could lead to substantial reductions in process emissions (e.g. magnesium oxide derived from magnesium silicates) are still in the research and development (R&D) phase today.

There are also limited alternatives to CCUS for now for reducing emissions in steel and chemicals production. CCUS concepts in the steel and chemicals sectors also tend to be at higher levels of technology maturity than their hydrogen-based alternatives. The hydrogen-based direct reduced iron (DRI) route for making steel, which reduces emissions substantially, could emerge as an economically viable alternative to CCUS-equipped facilities, but probably only in regions with access to very low-cost renewable electricity for hydrogen production via water electrolysis. Based on current estimates of the levelised costs of production for commercial-scale plants, producing one tonne of steel via CCUS-equipped DRI and innovative smelting reduction processes is typically 8-9% more expensive than today’s main commercial production routes, but the hydrogen-based DRI route typically raises costs by around 35-70%. The story is similar in the chemicals sector. Electrolytic hydrogen as a feedstock for ammonia and methanol production could become an important alternative to CCUS, but in most regions today, it is more expensive than applying CCUS to existing or new plants. The cost of CCUS-equipped ammonia and methanol production is typically around 20-40% higher than is that of their unabated counterparts, while the cost of electrolytic hydrogen routes is 50-115% higher.

CCUS accounts for nearly 40% of the total cumulative reduction in global CO₂ emissions in the steel, cement and chemicals sectors combined in the Sustainable Development Scenario relative to the Stated Policies Scenario. CO₂ emissions captured in industry – including those utilised in the commercial production of urea – increase by a factor of four in the period to 2030 (to 0.5 Gt) and by almost a factor of 25 to 2070 (to 2.7 Gt) in the Sustainable Development Scenario. Industrial CCUS applications account for around 54% of total CO₂ capture in the energy system by 2030 and 26% by 2070. The amount of CO₂ captured cumulatively is largest in the cement industry (43 Gt), followed by the chemicals (18 Gt) and steel industries (16 Gt). By 2070, almost 90% of the CO₂ generated in cement production is captured, about 75% in the steel sector and just under 80% in the chemicals sector. 

The role of CCUS in reducing emissions in each sub-sector grows rapidly over time to 2050, after which its contribution begins to plateau. By 2070, CCUS accounts for 61% of annual emissions reductions in the cement sector, 31% in the steel sector and 33% in the chemicals sector. Cement production accounts for the largest share of the process CO₂ emissions captured. CCUS is also deployed in the pulp and paper sector, but on a much smaller scale. Around 18 Mt is captured annually in this sector by 2070 in the Sustainable Development Scenario, or 1% of all the CO₂ captured in industry. CCUS in that sector is mainly applied to steam boilers, some of which are fed by biomass (e.g. bark and other pulp and papermaking residues), resulting in some CO₂ removal (-3.5 Mt in 2070).

The pace of CCUS deployment in industry is daunting, emphasising the need to get the ball rolling as quickly as possible. By 2030, around 450 MtCO₂ is captured per year in industry worldwide, mainly in the cement sector, in the Sustainable Development Scenario. Assuming an average capture rate of 0.5 MtCO₂ per year for each facility, this implies a need for close to one CCUS-equipped cement facility coming online per week between now and then. The rate accelerates to almost 6 per month on average in the period 2030‑70. Much of this capacity is retrofitted to existing plants or those currently under construction. This deployment hinges on a matching expansion of CO₂ transport and storage infrastructure. 

Technology options

CCUS is an option for effecting deep emissions reductions in long-distance transport, including heavy-duty trucking, shipping and aviation. Together, these three sub‑sectors make up nearly half of global annual energy use (1 192 Mtoe) for transport and related CO₂ emissions (3.6 Gt). They are among the most difficult to decarbonise. Electrification of trucks and the direct use of hydrogen and ammonia in ships are among the main alternatives to the use of biofuels (the supply of which is constrained by the availability of land to grow crops for energy purposes) and synthetic fuels. By contrast, electrification of long-distance air travel is a nascent technology, constrained by limits on the gravimetric energy density of batteries. An all-electric passenger commercial aircraft capable of operating over ranges of 750‑1 100 kilometres, for instance, would require battery cells with densities of 800 Wh/kg – more than three-times the current performance of lithium-ion (Li-ion) batteries (Schäfer et al., 2019).

CCUS can contribute to the decarbonisation potential of long-distance transport as a source of CO₂ for synthetic hydrocarbon fuels. Captured CO₂ can be used to convert low-carbon hydrogen into synthetic hydrocarbon fuels (diesel, gasoline and kerosene) that are easier to store, transport and use, but with potentially lower lifecycle CO₂ emissions than conventional fossil fuels (see Chapter 3). However, the production of synthetic hydrocarbons is energy-intensive and requires large amounts of hydrogen, making them relatively expensive. As CO₂ emissions constraints increase over time, the feedstock CO₂ must increasingly be sourced from biomass or the air (DAC).

Given the high cost of producing synthetic fuels, their long-term use at scale is likely to be mostly limited to long-distance aviation, where practically no other alternatives to conventional oil-based fuels and biofuels exist, and where the higher cost are likely to be more easily absorbed. The estimated levelised cost of synthetic fuels is around two to seven times that of kerosene produced from crude oil (at a price of USD 50/bbl) and bio kerosene is around 1.5-4 times that of kerosene produced from crude oil (at a price of USD 50/bbl). The cost of capturing the CO₂ needed to make synthetic fuels is a major component of the total cost of making those fuels. CO₂, generated in the production of bioethanol is expected to be the cheapest source of biogenic CO₂, at around 15-30/tCO₂. CO₂ captured from the atmosphere in a DAC facility is projected to cost in the region of USD 135-345/t, though future costs are highly uncertain since this family of technologies is at a comparatively early stage of development. Electricity accounts for around 30-80% of the cost of synthetic fuel production, based on a future renewable energy electricity price of USD 20-60/MWh. 

CO₂ for synthetic fuels in the Sustainable Development Scenario

CCUS contributes indirectly to emissions reductions in all three long-distance transport sub-sectors in the Sustainable Development Scenario. While conventional biofuels expand most rapidly in the near term, synthetic hydrocarbon fuels and BTL with CCUS start to make inroads in the late 2020s. By 2070, biofuels make up 17% (418 Mtoe) of the total fuel mix and synthetic hydrocarbon fuels make up 10% (254 Mtoe).

Synthetic hydrocarbon fuels make the largest contribution in aviation, accounting for almost all synthetic fuel use and 40% of the total demand for kerosene in 2070 (biofuels account for 35%). In the Sustainable Development Scenario, synthetic hydrocarbon fuels play a modest role in the trucking sector over the period 2030 to 2060, with the sector transitioning to other low-carbon fuels thereafter. By 2070, the production of synthetic hydrocarbon fuels across all sub sectors requires around 120 Mt (350 Mtoe) of electrolytic hydrogen, 830 Mt of CO₂, and 5 500 TWh of electricity – around 8% of all the electricity produced worldwide in 2070. As the CO₂ is sourced from the atmosphere (55% of all CO₂ used in 2070) or captured at biomass power or biofuel production plants (45%), the aviation fuel produced is carbon-neutral.

CCUS can play an important role in facilitating the production of low-carbon hydrogen for use across the energy system. Hydrogen is a low-carbon fuel or feedstock that can be used without direct emissions of air pollutants or GHGs. It offers a way to decarbonise a range of energy sectors, in particular where direct electrification is difficult, including long-haul transport, chemicals, iron and steel production, and power and heat generation (see the previous section). CCUS can help decarbonise hydrogen production in two key ways:

• By reducing emissions from existing hydrogen plants: Around 75 Mt H₂ of hydrogen is currently produced each year for industrial use,7 almost entirely from natural gas (76%) and coal (23%), with the remainder from oil and electricity. This is associated with more than 800 MtCO₂, corresponding to the combined total energy sector CO₂ emissions of Indonesia and the United Kingdom (IEA, 2019b). Unabated production of hydrogen from fossil fuels results in emissions of 9 tCO₂/t H₂ in the case of natural gas and 20 tCO₂/t H₂ in the case of coal. Seven projects based on the generation of hydrogen from fossil fuels with CCUS are in operation today8 with a combined annual production just over 0.4 MtH₂, capturing close to 6 MtCO₂. Of the seven projects, four are at oil refineries and three at fertiliser plants. There is significant potential to expand CCUS retrofitting to reduce emissions from existing facilities and enable these facilities to continue operations sustainably. Capturing CO₂ from hydrogen production is a relatively low-cost CCUS application, and existing facilities are often concentrated in coastal industrial zones, with potential to share CO₂ transport and storage infrastructure with other industrial facilities.
• By providing a least-cost pathway to scale up new hydrogen production: Gas- and coal-based hydrogen production with CCUS is currently less expensive than using renewable energy for water electrolysis in most regions and will remain so where both CO₂ storage and low-cost fossil fuels are available.

Hydrogen production from natural gas using reforming processes and from coal using gasification are well-established technologies. In the case of natural gas, steam methane reforming (SMR) is the leading production route today, with part of the natural gas (30-40%) used as fuel to produce steam, giving rise to a “diluted” CO₂ stream, while the rest of it is split with the help of the steam into hydrogen and more concentrated “process” CO₂. The concentration of the CO₂ in the output streams affects capture costs. Capturing CO₂ from the concentrated “process” stream costs around USD 50/t_, leading to overall emission reductions of 60%. CO₂ can also be captured from the more diluted gas stream. This can boost the level of overall emissions reduction to 90% or more, but it also increases costs to around USD 80/tCO₂ in merchant hydrogen plants. Several SMR CCUS projects are currently at the feasibility study stage with ambitions to be operational by 2030, in particular in densely industrialised zones. These include the H-Vision project, which aims to retrofit CO₂ capture to up to 0.6 MtH₂/yr for industrial use in Rotterdam, the Netherlands (PoR, 2019), and the Magnum Project in the Netherlands, which could create demand for 0.2 MtH₂/yr for each of the three gas power plant units converted to hydrogen (NIB, 2018).

Auto‑thermal reforming (ATR) is an alternative technology in which the required heat is produced in the process itself. This means that all the CO₂ is produced inside the reactor, which allows for higher CO₂ recovery rates than can be achieved with SMR. ATR can also be cheaper than SMR because the emissions are more concentrated. A large share of global ammonia and methanol production already uses ATR technology, and two new projects in the United Kingdom – HyNet and H21 – plan to use that technology, too (Northern Gas Networks, 2018; HyNet, 2020).

Other options for using natural gas to produce hydrogen exist, but are still at a laboratory or demonstration stage. In an alternative SMR design, natural gas would still be required as feedstock, but the necessary steam could be produced by alternative sources, such as electricity or concentrated solar energy, thus eliminating the diluted CO₂ stream from heat generation in conventional SMR designs. Methane pyrolysis (or splitting) is another emerging technology. It involves splitting methane at high temperatures, for example in a plasma generated by electricity, to produce hydrogen and solid carbon, but no CO₂. The resulting carbon can be potentially used as feedstock in the chemical, steel or aluminium industry, providing another revenue stream besides the hydrogen (Daloz et al., 2019). In the United States, Monolith Materials operates a pilot methane pyrolysis plant in California and a commercial demonstration plant in Nebraska. In Australia, the 100 t H₂/yr Hazer Commercial Demonstration Plant, which will use biogas to produce hydrogen and graphite, is under construction (FuelCellsWorks, 2020).

Coal gasification is a mature technology used today mainly in the chemical industry for the production of ammonia, in particular in China. Coal gasification can be combined with CCUS, though there are technical challenges. In particular, few technologies exist that produce both high-purity hydrogen and CO₂ that is pure enough for other uses or storage, since gas separation technologies focus on either hydrogen removal or CO₂ removal. The choice and design of the capture technology therefore depends on what the hydrogen is going to be used for, as well as on production costs. In Australia, the planned Hydrogen Energy Supply Chain Latrobe Valley project is seeking to produce hydrogen from lignite using gasification, with the CO₂ being transported and stored via the CarbonNet project.

Producing hydrogen from fossil fuels with CCUS will likely remain the cheapest low-carbon route in regions with low-cost domestic coal and natural gas and available CO₂ storage, such as the Middle East, North Africa, Russia and the United States. The cost of producing hydrogen that way is currently in the range of USD 1.2/kg H₂ to USD 2.6/kg H₂, depending on local gas and coal prices. This cost is not projected to change significantly in the coming decades. The future economics of the technology and competing options depend on factors that will continue to vary regionally, including prices for fossil fuels, electricity and carbon. The cost of electrolytic hydrogen is expected to come down substantially in the long term, driven by cost reductions from scaling up the deployment of electrolysers and their manufacturing capacities as well as due to declining costs for electricity from renewables. Water electrolysis could become a competitive option for low-carbon hydrogen production in regions with abundant renewable energy resources, including Northern Africa and most of Australia, with costs projected to range from USD 1.3/kg to USD 3.3/kg of hydrogen by mid‑century.

Low-carbon hydrogen plays a key role in decarbonising transport, industry, buildings and power generation in the Sustainable Development Scenario, with global hydrogen demand increasing seven-fold to 520 Mt by 2070. Hydrogen is used in a wide range of new applications as an alternative to current fuels and raw materials, including as a transport fuel for cars, trucks and ships, as an input for chemicals and steel making, to produce heat in buildings and industry, and for energy storage to balance the variability of renewables in the power sector. The direct use of hydrogen in transport, buildings, industry, and power generation accounts for two-thirds of hydrogen demand in 2070, while nearly a quarter is used to produce synthetic hydrocarbon fuels and 10% is converted into ammonia as a fuel for the shipping sector. Ammonia produced from natural gas with CCS covers more than a third of fuel needs in the shipping sector in 2070.

Overall, the production of hydrogen with CCUS and its use leads to cumulative CO₂ reductions of around Gt by 2070, or 3.5% of the cumulative emissions reductions in the Sustainable Development Scenario relative to the Stated Policies Scenario. Production reaches 18 Mt (52 Mtoe) worldwide in 2030 in the Sustainable Development Scenario, meeting 20% of global hydrogen needs, compared with 7.8 Mt9 (22 Mtoe) in 2020. Both existing and new hydrogen plants are equipped with CO₂ capture, including in some of the main industrial clusters in ports of the North Sea, the US Gulf Coast and southeast China.

The shares of water electrolysis and fossil fuels with CCS in total low-carbon hydrogen supply is roughly equal up to 2030, but moves slightly in favour of water electrolysis over time, reflecting expected cost reductions for electrolysers and renewable energy generation. By 2070, low-carbon hydrogen production from fossil fuels with CCUS accounts for 40% of global hydrogen production or around 210 Mt (600 Mtoe) – nearly 500 times more than the total hydrogen capacity with CCUS in operation today. Around 1.9 GtCO₂ is captured and stored from hydrogen production in that year, representing around 18% of all CO₂ being captured globally in 2070.

Carbon removal technologies involve extracting CO₂ from the atmosphere, directly or indirectly (via the absorption of CO₂ in biomass), and permanently storing it. The main attraction of carbon removal technologies is their potential to offset residual emissions from sectors where emissions are hard to abate, to achieve net-zero emissions across the energy sector. While some CO₂ could be stored in products (e.g. concrete), geological storage will undoubtedly be needed to achieve large-scale carbon removal with these technologies.

Carbon removal is also often seen as a way of producing net-negative emissions in the second half of the century to counterbalance excessive emissions earlier on. This feature of many climate scenarios however should not be interpreted as an alternative to cutting emissions today or a reason to delay action. In the Sustainable Development Scenario, carbon removal technologies are part of the portfolio of technologies and approaches to cut emissions in the near term and in the future, helping a faster transition to net-zero emissions. From a policy perspective, support for carbon removal technologies can additionally serve as a means to hedge against the risk of delay or even failure in the development and deployment of other CO₂ abatement technologies across the energy sector: technology development and deployment tends to be a non-linear process in which delays can occur for many different reasons (IEA, 2020c). 

Carbon removal approaches can include either nature-based solutions, enhanced natural processes, or technological solutions.11 Nature-based solutions include afforestation (the repurposing of land use by growing forests where there were none before) and reforestation (re-establishing a forest where there was one in the past)12. Enhanced natural processes include land management approaches that increase the carbon content in soil through modern farming methods (for instance, by adding biochar13 or fine mineral silicate rocks) and ocean fertilisation, in which nutrients are added to the ocean to increase its capacity to absorb CO₂. BECCS and DACS are the main technological solutions available today – they are the primary route for the energy sector to contribute to carbon removal in the transition to net-zero emissions, and therefore the focus of this analysis.

While all these approaches can be complementary, technology solutions can offer advantages over nature-based solutions, including the verifiability and permanency of underground storage; the fact that they are not vulnerable to weather events; including fires that can release CO₂ stored in biomass into the atmosphere; and their much lower land area requirements. BECCS and DACS are also at a more advanced stage of deployment than some carbon removal approaches. Land management approaches and afforestation/reforestation are at the early adoption stage and their potential is limited by land needs for growing food. Other non‑technological approaches – such as enhanced weathering, which involves the dissolution of natural or artificially created minerals to remove CO₂ from the atmosphere, and ocean fertilisation/alkalinisation, which involves adding alkaline substances to seawater to enhance the ocean’s ability to absorb carbon – are only at the fundamental research stage. Thus, their carbon removal potentials, costs and environmental impact are extremely uncertain.

BECCS, DAC, land management approaches and ocean fertilisation/ alkalinisation have the highest cumulative potential; however, they all come with potential negative side effects such as land use changes, food security and biodiversity losses (BECCS, land management approaches), high CO₂ capture costs (DAC), and ocean eutrophication (ocean fertilisation/alkalinisation). DAC has the smallest land footprint among the most mature carbon removal options, while BECCS and afforestation/reforestation have similar ranges for the land footprint, which depends mainly on the source of biomass.

BECCS

BECCS involves the capture and permanent storage of CO₂ from processes where biomass is converted to energy or used to produce materials. Examples include biomass-based power plants, pulp mills for paper production, kilns for cement production and plants producing biofuels. Waste-to-energy plants may also generate negative emissions when fed with biogenic fuel. In principle, if biomass is grown sustainably and then processed into a fuel that is then burned, the technology pathway can be considered carbon-neutral; if some or all of the CO₂ released during combustion is captured and stored permanently, it is carbon negative, i.e. less CO₂ is released into the atmosphere than is removed by the crops during their growth. In practice, a life cycle assessment is needed to identify whether a specific technology and application is genuinely producing negative emissions or not, depending on the sustainability of the biomass feedstock, the scope of the application, changes in land management and use, and the timing of emissions and removals (IEA Bioenergy, 2013).

BECCS is the most mature of all the carbon removal technologies, as both bioenergy production and CCS have been separately proven at commercial scale. BECCS is already operating in the fuel transformation and power generation sectors, with different levels of maturity depending on the specific application. The most advanced BECCS projects capture CO₂ from ethanol production or biomass-based power generation, while industrial applications of BECCS are only at the prototype stage (IEA, 2020d). There are currently more than ten facilities capturing CO₂ from bioenergy production around the world. The Illinois Industrial CCS Project, with a capture capacity of 1 MtCO₂/yr, is the largest and the only project with dedicated CO₂ storage, while other projects, most of which are pilots, use the captured CO₂ for EOR or other uses. 

DAC

DAC technologies extract CO₂ directly from the atmosphere for permanent storage (carbon removal), or for use, for example, in food processing or to produce synthetic hydrocarbon fuels (where the CO₂ is ultimately re-released). Currently, technologies to capture CO₂ from the air rely either on liquid sorbents (liquid DAC [L-DAC]), using a hydroxide solution (Carbon Engineering, 2020)) or solid sorbents (solid DAC [S-DAC]), making use of a CO₂ “filter” (Climeworks, 2020) or dry, amine-based chemical sorbents (Global Thermostat, 2020).

While existing DAC technologies rely on both fuel for heat and electricity to power rotating equipment for their operation, S-DAC could operate solely on electricity, which could come from renewable power sources. On the other hand, L-DAC will most likely always need a source of heat such as natural gas in order to achieve the high operating temperature needed in the calciner (around 900°C), unless a new way of providing a low-carbon source of heat (which does not currently exist [IEA, 2019a14] becomes commercially available. If gas were used to provide the heat (as it is the case nowadays), the associated CO₂ emissions would also need to be captured and stored along with the CO₂ captured directly from the air to maximise carbon removal.

An advantage of DAC is the potential for flexibility in siting. For example, a DAC plant could be located next to a plant that needs CO₂ as a feedstock or on top of a geological storage site to reduce the need for CO₂ transport. DAC facilities can also be co-located with other CO₂ capture facilities, such as CCUS-equipped power or industrial plants, to facilitate access to existing CO₂ transport infrastructure and enabling these facilities to reach net zero or even negative emissions.

The main drawback of DAC is the low CO₂ concentration in ambient air compared with other sources of CO₂, such as industrial or power plants, which makes this technology highly energy-intensive and expensive compared with other options for carbon removal. The amount of energy needed and the share between fuel and electricity differs depending on the type of technology and whether the CO₂ needs to be compressed for transportation and storage. L-DAC for CO₂ use applications requires relatively small amounts of electricity (less than 5% of total energy needs); S-DAC for storage typically requires more (23%). Natural gas – usually the cheaper source of energy for heat-raising – is mainly used to regenerate the solvent, either at around 100°C (S-DAC) or around 900°C (L-DAC).

A total of 15 DAC plants are currently operating in Canada, Europe, and the United States. Most of them are small-scale pilot and demonstration plants, with the CO₂ diverted to various uses, including for the production of chemicals and fuels, beverage carbonation and in greenhouses, rather than geologically stored. Two commercial plants are currently operating in Switzerland, selling CO₂ to greenhouses and for beverage carbonation. There is only one pilot plant, in Iceland, currently storing the CO₂: the plant captures CO₂ from air and blends it with CO₂ captured from geothermal fluid before injecting it into underground basalt formations, where it is mineralised, i.e. converted into a mineral (CarbFix, 2020). In North America, both Carbon Engineering and Global Thermostat have been operating a number of pilot plants, with Carbon Engineering (in collaboration with Occidental Petroleum) currently designing what would be the world’s largest DAC facility, with a capture capacity of 1 MtCO₂ per year, for use in EOR (Carbon Engineering, 2019).

Costs of BECCS and DAC

At present, BECCS is the cheaper of the technology-based approaches for carbon removal. Generally speaking, the higher the initial concentration of CO₂ before capture, the lower the capture cost, which is why BECCS is cheaper than DAC. In the case of BECCS, capture from fuel transformation processes (such as bioethanol production from sugar or starch cane) or biomass gasification (where only pre-treatment and compression are needed to capture CO₂) are the cheapest at present, with costs ranging from about USD 15/tCO₂ to USD 30/tCO₂. Capture in biomass based power generation costs around USD 60/tonne, while BECCS applied to industrial processes has a capture cost of around USD 80/t. 

Capture costs for DAC are much higher than for BECCS capture – by a factor of between 2 and 25 – due mainly to the lower initial concentration of CO₂ compared with industrial streams. DAC costs vary according to the type of technology (solid- or liquid- based technologies) and whether the captured CO₂ needs to be compressed to high pressure for transport and storage rather than used immediately at low pressure. As the technology has yet to be demonstrated on a large scale, future costs are extremely uncertain. Cost estimates reported in the literature are wide, typically ranging from USD 100/tCO₂ to 1 000/tCO₂ (Realmonte et al., 2019). Carbon Engineering claimed that costs as low as USD 94/t to USD 232/t were achievable depending on financial assumptions, energy costs and the specific plant configuration (Keith et al., 2018).

The energy needs for a DAC plant will be a major factor in determining plant location and production costs. The choice of location needs to take into account the source of the energy needed to run the DAC plant, which will also determine if the system is carbon negative, as well as the cost of the energy. For instance, low-carbon energy sources such as solar thermal, photovoltaic (PV) and wind power generation could power DAC plants in isolated areas, though the utilisation rate of the plant (and, therefore, its economic viability) would vary according to the availability of sunshine and wind.15

Carbon removal accounts for a large and increasing share of the CO₂ captured over the projection horizon in the Sustainable Development Scenario. Bioenergy with CCS/CCU and DAC together account for 25% of all the CO₂ cumulatively captured to 2070. Of all the CO₂ captured by the two technologies, around 48 Gt, or 78%, is stored permanently and so counts as carbon removal. Captured and stored volumes reach around 2.7 Gt for BECCS and almost 0.3 Gt for DACS in 2070. In both the Stated Policies and Sustainable Development Scenarios, the availability of sustainable biomass is assumed to be limited to around 3 000 Mtoe/year (125 EJ/year), which constrains the deployment of BECCS (see below). Stronger policy incentives, including higher carbon prices, are nonetheless assumed to drive much faster growth in BECCS in the Sustainable Development Scenario.

BECCS starts to be deployed at scale from 2030, and by 2070, it has captured a cumulative total of around 45 GtCO₂ in the Sustainable Development Scenario. It is mainly installed in power generation (55%) and fuel transformation (40%), with the remainder in the cement and pulp and paper industries. By 2070 half of biomass-fired generation is associated with CCUS. When BECCS is deployed in the fuel transformation sector (where CO₂ capture is cheaper than in other sectors) around half of the carbon remains in the biofuel product, providing a carbon-neutral fuel for hard-to-abate transport modes. 

BECCS and DACS can play a decisive role in getting the global energy system to net-zero emissions. However, there remains considerable uncertainty regarding the potential contribution of these technologies in practice, notably with respect to future costs and performance, how fast they can be commercialised, public understanding and acceptance, the limits to the availability of sustainable biomass, and how quicklCO₂ transport and storage infrastructure can be developed. This underscores the need for intensive RD&D to ensure that these technologies are ready to be deployed on a large scale within the next decade given the lead times involved.

Particular concerns have been expressed regarding the land requirements associated with both BECCS and DACS. The land footprint for BECCS is estimated at between 1 000 and 17 000 km² per Mt of CO₂ removed, depending on a number of factors including location and the source for the biomass (e.g. forest and agricultural residues, and purpose-grown energy crops). The land needs for DAC are smaller, at a maximum of around 15 km² per Mt of CO₂ removed, including the space needed for solar PV panels if they are the sole source of the electricity used to run the plants.16 The 740 MtCO₂  captured by DAC in 2070 in the Sustainable Development Scenario would require approximately 10 500 km² of land if using solar PV – roughly one-third the size of Belgium. The same level of removal through afforestation would require between 0.5 and 11.5 million km², the latter being a land area bigger than the United States. An emerging DAC technology, based on electro swing adsorption (ESA-DAC)17 has potential for a smaller land footprint (Voskian and Hatton, 2019).

Water requirements for DAC are highly dependent on the chosen technology. L-DAC requires significant amounts of water while, by contrast, some S-DAC options produce water, which could be beneficial within integrated systems with water demand such as hydrogen production (Breyer, et al., 2019).