Carbon capture is set to play a key role in decarbonizing industrial sites across the world. But how far have we gotten and where are we headed? To understand the pace and scale of growth in different segments and regions, you’ll have to stay on top of carbon capture project development.
With an overwhelming and growing amount of data available, staying on top is easier said than done. This data-driven overview presents the state of carbon capture in spring of 2024. We base our overview on data available in CaptureMap, and have deliberately added as little interpretation as possible on top of the data to stay close to the facts. We prefer to leave the interpretation to you.
This article is the second of a three-part blog series, where we will tackle the following topics:
- New projects – read here
- Capture capacity across segments – this article
- Projects pipeline and infrastructure – coming soon!
This second article is about how the capture capacity is developing across segments and emission sizes. Here we look into the distribution of capture capacity and how it strongly correlates with the CO2 emissions distribution of the facilities in our database. We also dive into the unsung heroes of capture projects, often forgotten in a CCUS context, but relevant in proving industry experience with capture.
While a fair share of capture projects will announce their intended capture capacity, many do not. We have therefore taken a first stab at estimating what the real global capacity under development might be. Finally, let’s not forget about biogenic CO2, where we rely on CaptureMap’s well-developed overview of biogenic CO2 emissions and connect that to the development of capture projects.
Contents
PS: We recommend reading the appendix to be clear on what we mean when we refer to different terms.
Most capture plants are relatively small, just like most CO2 emitters
Despite the potential for economy of scale, most installed capture plants are relatively small in terms of capacity, with half of the installed capacity below 200 000 tonnes CO2 per year. There are however large variations between plants, and on average, capture plants (both operating and in project) have a capacity of ca. 800 000 tonnes CO2 per year. This value has been relatively stable across project announcements since 2020.
This “long-tail” distribution of capture plants capacities (many small, few large) also reflects the distribution of CO2 emitters, with many emitters with emissions around or below 200 000 tonnes CO2 per year, and fewer emitters over a million tonnes per year.
Note: the data above is based on facility reporting of CO2 emissions in North America, Europe and Taiwan. Different countries and regions have different thresholds for reporting, and many sites below 10 000 tonnes CO2 per year are missing from the overview.
There is actually more carbon capture already operational than often reported
When talking about carbon capture, the same operating projects are often mentioned, like the two Norwegian capture facilities at Sleipner and Snøhvit/Hammerfest (gas processing), the Boundary Dam Power Station in Canada (post-combustion capture from coal power plant), or the Petra Nova Plant in Texas, USA (post-combustion from natural gas power plant that ran for a while, stopped and then started again). Those are large scale capture and sequestration projects which deserve attention to exemplify the functioning CCS projects already in place. However, they take the attention away from the hundreds of plants already operational around the world that capture CO2 from various industrial facilities and sell it for utilization or EOR.
A quick look at the US EPA data highlights for Supply of CO2 illustrates the point. It shows that the USA alone counted 117 industrial suppliers of CO2, capturing 18.1 million tonnes of CO2 per year in 2022. These are ethanol plants, natural gas processing plants, ammonia factories, refineries, but also paper mills, breweries, distilleries and others.
Source: US EPA data highlights for Supply of CO2
The market seems to go away from rigid source-to-sink large CCS projects, to increasingly flexible value chains, sometimes combining sequestration and utilization, and sharing transport and storage infrastructure. As a consequence, the lines get blurred between the conventional industrial capture of CO2 for industrial purposes, and the capture of CO2 for emissions reductions purposes. The same facilities that today capture the CO2 and sell it to the food and beverage industries, may tomorrow decide to sequester the CO2 if the business case for doing so is better.
We therefore map as many as possible of the existing capture projects, and categorize the fate of their CO2 according to what they disclose. The result? When others report 40 to 50 operational capture plants around the world, we can document more than 230 plants already capturing CO2, and the count keeps on growing the more we look.
Most project overviews probably also underestimate the total capture capacities in project
Not all project announcements are equal in terms of details provided by the companies involved. Sometimes key elements of projects are not communicated publicly, making it difficult to build a complete overview of the market at any moment in time.
As a matter of fact, we are missing figures for capture capacities for 1/3rd of the projects in EPC stage, and half of the projects in feasibility. When in the early feasibility stage, project developers sometimes don’t know yet the future capture capacity that may be installed on their facility. It is therefore logical that that information is sometimes missing. However, capture capacities are often not being disclosed even when in operation. This is not limited to our database, and is a common issue amongst all organizations providing market data about carbon capture projects.
The consequence? We think that carbon capture capacities are being underestimated by half compared to what is actually being considered for development, built or already in operation. Our data available for 460 projects with known capture capacity sum up to 439 million tonnes CO2 per year. Extrapolated to all the 950 projects (excluding those on hold or inactive) in our database, the total capacity could reach 912 million tonnes CO2 per year, given that projects with data are representative to those without data (uncertain, but possible). This is good news, with the important caveat that not all capacity will necessarily be built or operated to its full extent.
There are still plenty of facilities to decarbonise in the world
All carbon capture projects in CaptureMap are precisely mapped and matched with industrial facilities with CO2 emissions. This allows evaluating the share of emitting facilities that have carbon capture projects in place, for regions with near-full coverage on emissions (North America and Europe).
The highest share of facilities with projects are found for industrial activities where CO2 is readily available in high concentration, lowering the capture costs:
- ethanol and food-grade alcohol production (biogenic CO2 from fermentation),
- ammonia, methanol, ethylene oxide and hydrogen (high concentration CO2 from production processes),
- gas treatment facilities in the oil and gas sector (CO2 removal from natural gas),
- refineries (often from on-site hydrogen production).
Note: the figure above only shows industrial segments that have carbon capture projects in place, and do not represent a comprehensive overview of all industrial activities possible.
In total, our overview of emitters includes 23 349 facilities in the world at the time of this writing. While we’re fairly confident we cover the largest emitters well, we know that there’s a forest of smaller emitters (the long-tail distribution is back) that fly under our radar. Based on our estimates, we believe that the actual total number of large CO2 emitters out there is closer to 60 000. This means that facilities with capture projects cover ca. 3.5% of all facilities in CaptureMap, and probably only 1.4% of all relevant facilities in the world. In other words, still plenty of facilities relevant for decarbonisation solutions!
Large amounts of biogenic CO2 can be captured
CaptureMap includes data on both fossil and biogenic emissions (CO2 from the combustion of biomass). By connecting the capture projects to this CO2 emissions data, it is possible to estimate the share of project capacities that may capture biogenic CO2.
Biogenic CO2 is a tricky landscape, so there are quite a few limitations on the analysis below. First of all, ethanol and biogas (renewable natural gas) assets are left behind from the overview below, due to incomplete emissions reporting on these types of assets. Large volumes of biogenic CO2 may be captured from these assets. Read more about biogenic emissions in our biogenic CO2 breakdown blog and our ethanol CCUS blog. Second, the figures below only represent retrofit of carbon capture on existing facilities, not entirely new facilities. Third, we base our figures on capture plants with available data on capture capacities (half of all the plants in the database). We are therefore underestimating the total amount of CO2 potentially captured by ongoing projects.
Provided the limitations above, our estimate shows that documented capture projects could capture at least 22 million tonnes of biogenic CO2. To put things in perspective, our overview in CaptureMap for North America and Europe includes 386 million tonnes of biogenic CO2 released every year.
The largest volumes of biogenic CO2 covered by carbon capture projects and plants in operation are within:
- Biomass-fired power plants. Some of these plants were originally coal-fired and are gradually converted into biomass. One example is the Drax power plant in the UK.
- Waste-to-energy plants typically include 50% biogenic content due to the biomass present in some of the waste fractions being incinerated.
- Cement plants in Europe often use a large share of biomass and bio-based waste for energy purposes. Heidelberg Materials Brevik, in Norway, is an example of carbon capture on a cement plant with some share of biogenic CO2.
- Pulp production facilities often have high biogenic content in their emissions, and represent a large potential for capturing biogenic CO2.
Conclusion: Lots to learn about the details of capture capacity
This blog post aimed to take a deeper dive into how the capture capacity is being developed across emission sites. We’ve shown how the “long-tail” distribution of capture plants (many small, few large) also reflects the distribution of the CO2 emitters. There are more capture projects out there than are often talked about, and while most of them today go to EOR or utilization purposes, they might be eligible for storage in the time to come – so we think it makes sense to include them.
When it comes to total capture capacity, it’s a surprisingly big blind spot in the industry. We don’t publish capture capacities in CaptureMap unless we can source the figures. That said, we did a little mental math to ballpark estimate what the capture capacity pipeline could look like based on extrapolation from what we do know. And it’s nearing a gigatonne of CO2/year.
Even with that, there’s lots more potential in many industries, so we wish all the actors across the CCUS value chains best of luck and encourage you to roll up your sleeves. We in Endrava also have a job to do to get a hold of even more asset-level CO2 data, a challenging and meaningful endeavor.
Finally, about biogenic CO2. Capture capacity is building, but the volumes are small compared to what’s out there. Given their role in negative emission projects (carbon dioxide removals or CDRs) as well as e-fuel projects, we expect these volumes to come up significantly in the time to come.
Appendix – Where does our carbon capture project data come from and what does it contain?
Our ambition with CaptureMap is to be the world’s most accurate overview of large CO2 emitters and their associated carbon capture projects. We can’t do that without laser focus on data quality. And it’s this elusive chase of data perfection that drives us and that makes our users come to us. We’re the first to tell you that our dataset is still not perfect. Yet, from our users we also hear that we collectively know more than anyone else, and that our advantage is growing.
Historic CO2 data in CaptureMap is made from a combination of different public databases that we clean, harmonize and consolidate. It’s already quite challenging, but doable. See a blog here about one of the databases that makes us smile and also pull our hair out.
For capture projects we needed a different approach. That’s in part because the landscape changes so quickly, but also because the publicly available databases lack the granularity and update frequency we’re looking for. Therefore, we decided to develop an in-house database of capture projects using Python sprinkled with a bit of artificial intelligence (AI). We won’t divulge all the secrets of our recipe, but in short, it’s a tech stack that cuts about 90% of the processing time compared to manual approaches. And it scales incredibly well.
Our main data source for capture projects are press releases and news articles about the projects, all available in the public domain. This means that confidential projects are not included. We tie information about every capture project back to a specific facility, existing or future, on the map. This means that we do not include projects for clusters or hubs of emitters when the exact facilities are not specified. This strict mapping framework allows us to avoid double-counting, and makes it possible to tie our precise CO2 emission data and activity segments to capture projects.
In CaptureMap and in this document, a capture project consists of a capture unit being developed or already operating at a specific industrial facility. We focus on the capture of CO2 from combustion, process or fermentation emissions, and the database does not include Direct Air Capture (DAC).
We tag each capture project with relevant data about its engineering stage, the planned capture capacity, companies involved in the project, the transport mode and fate of CO2, and other relevant fields.
In terms of engineering stages, projects are categorized within:
- Feasibility: from early concept through FEED studies, until the final investment decision
- EPC: engineering, procurement and construction, from the final investment decision to the commissioning
- Operation: when the plant is operational
- Hold: for projects that are temporarily set on hold but not abandoned
- Inactive: for decommissioned facilities or abandoned projects.
When an industrial facility has several different projects for several capture units in different phases, we distinguish between each of these phases, and each one will be counted as a project. We’ve come across facilities that have one capture unit in operation, building the next one and a third one in feasibility, so the distinction is more relevant than you’d think.
We chose deliberately to include pilot and demonstration capture projects in our database, as they tend to be an indicator that the industrial actor hosting those is seriously considering CCUS as a decarbonisation solution. These projects represent about 13% of all the projects in the database at the time of this writing, and we’ll also make a separate post about these soon.