The Lowdown on CCS: Part 1, Carbon Capture

For a topic that‘s so central and so critical to global efforts to rein in carbon emissions, there’s an astonishing amount of confusion about Carbon Capture and Storage (CCS). 

Or maybe not so astonishing. There’s reason for the confusion. Public policy toward CCS stands to impact, positively or negatively, the fortunes of multiple competing interest groups. Those who would benefit from CCS tend to support it. They have no difficulty finding arguments to rationalize support. Those who would be hurt tend to oppose it. They likewise have no difficulty finding arguments to rationalize opposition. Both do their best to enroll others in the narratives that support their respective sides. The result is noisy. Welcome to Politics and Human Nature 101.

There are those of us who just want to understand the facts as best we can. We care about decarbonization and mitigation of climate change, and we want to know what makes sense. This is my humble attempt to sketch a map of the CCS territory for those so minded. 

CO₂ capture

It’s a major error to talk about CCS (or CCUS, as it’s also known) as if it were all one thing. There are many variations, both for capture and for utilization or sequestration. Each variation has its own cost-benefit profile, its own set of technical and scale considerations, its own future prospects. 

On the capture side, there are at least five distinct categories to consider. The first and most iconic -- the one most widely studied and talked about -- is processing to scrub CO₂ from the flue gases of power plants and furnaces.

Flue gas scrubbing

In the typical approach to flue gas scrubbing, the flue gases from fuel combustion are bubbled through a chamber of a sorbent solution. The sorbent captures a large percentage of the CO₂ content from the flue gases before being sent on to the next step. In the next step, the charged sorbent solution is heated to release the captured CO₂. The released CO₂ is compressed and piped away for use or sequestration. The regenerated sorbent is then cooled and recycled back to the scrubbing chamber.

The benchmark technology for flue gas scrubbing -- the one to which competing processes are invariably compared -- was adopted from the natural gas industry. Raw natural gas, as it comes from the wellhead, often includes a high content of CO₂ along with methane and assorted light hydrocarbons. Most of the CO₂ must be scrubbed from the gas to achieve uniform quality before the gas can be fed to the distribution network. That need for scrubbing has existed for as long as there has been a national gas distribution network. The technology that the industry settled on was the thermal swing adsorption process described above, using an aqueous solution of the organic compound monoethanolamine (MEA) as the sorbent.

Thermal swing adsorption using MEA is not especially energy efficient. For extracting CO₂ from raw natural gas, that doesn’t greatly matter. The relatively small amount of CO₂ to be removed and the high value added by its removal make the energy cost of scrubbing “small potatoes”. But using the process for flue gas scrubbing in a coal-fired power plant is another matter. Both the amount of CO2 and the volume of gas from which it must be scrubbed are many times larger. The energy needed to drive the scrubbing process and compress the captured CO₂ for transport are parasitic drains on power plant output. Total drain falls in the range of 25 to 30% of overall power output.

It’s theoretically possible to do much better. Applied R&D to identify better materials or more efficient cycling arrangements has been ongoing for many years. Alternative sorbents have been developed that are better suited to particular operating conditions or have other cost advantages. In terms of energy efficiency, however, no alternative using thermal swing adsorption has yet proved decisively more efficient than the benchmark process with MEA. 

Recently, researchers from DOE’s Pacific Northwest National Laboratory, in collaboration with Fluor Corporation and EPRI, published a study of a new liquid sorbent candidate. The organic liquid solvent, EEMPA, captures and releases CO₂ in a manner similar to MEA. However, it has low water content and requires 17% less energy to drive regeneration. A 17% reduction in parasitic energy drain on the power plant (per ton of captured CO₂) lowers the amount of extra fuel needed per MWh of net output. The compound result is a 19% reduction in the specific energy cost for carbon capture. 

If the EEMPA process can be commercialized, it may become the new benchmark for flue gas scrubbing. Its energy requirements, however, are still far above the theoretical minimum energy needed for flue gas capture. It may be as good as it’s feasible for any thermal swing adsorption process to get, but thermal swing adsorption isn’t the only game in town.

A group of researchers from MIT recently announced a new and very different approach to capturing CO₂. Instead of thermal swing, it uses an electrochemical approach. A small change in voltage to an electrochemical changes the state of a chemical within a cell. It goes from a state with a strong binding potential for CO₂ to a state in which it has zero binding potential. Not only is the approach far more energy efficient than the thermal swing method, but the change in CO₂ binding energy is large enough for the electro-swing approach to work with very low CO₂ concentrations in the input stream. That may allow it to serve for direct air capture as well as flue gas scrubbing.

Pipeline-ready CO₂ capture

There are some applications that produce CO₂ but don’t require flue gas scrubbing in order to capture it. They yield a CO₂ waste stream sufficiently pure that it can just be compressed and transported for use or sequestration. One major example is biological production of ethanol. Another is processing of raw natural gas, landfill gas, or biogas to reduce its CO₂ content. 

There are many other processes that don’t yield a pipeline-ready CO₂ stream as commonly implemented, but that could be redesigned to do so at little cost. Perhaps the most important example is the reforming of natural gas (principally methane) to produce “blue” hydrogen. 

The usual way to produce hydrogen from natural gas is by steam methane reforming (SMR). The most common implementation of SMR yields a flue gas of CO₂ mixed with atmospheric nitrogen. If capture of CO₂ is not required, the flue gas is vented. If a requirement to capture CO₂ is later imposed, it has to be done by retrofitting a flue gas scrubber onto the plant. That can be costly. It’s the basis for the assertion often heard that production of “blue” hydrogen (SMR with CCS) is significantly more expensive than “gray” hydrogen (SMR without CCS) and that only about 80% of the CO₂ can be captured. 

While that assertion is more or less accurate for retrofitted CO₂ capture, it doesn’t apply to new SMR facilities designed from the ground up for carbon capture. One simple alternative, already implemented in some plants, involves the way heat is provided to drive the initial endothermic reaction of the SMR process. That reaction is:

CH₄ + H₂O + heat => CO + 3H₂

The usual way to supply heat is through combustion in air of a portion of the natural gas feedstock. That produces a flue gas stream with a lot of nitrogen mixed in with the CO₂ from combustion. It’s the simplest and easiest approach when CO₂ capture isn’t required. But if capture is required and a ready source of pure O₂ happens to be available, there’s an easy alternative. Combustion of fuel in air is simply replaced by oxyfuel combustion. 

In oxyfuel combustion, pure oxygen is mixed with recirculated CO₂. Fuel is then burned in the O₂ - CO₂ mix, instead of the O₂ - N₂ mix of normal air. The resulting exhaust gas stream is just CO₂ and water vapor. After condensing out the water vapor, the exhaust stream is CO₂ of sufficient purity to be “pipeline ready”. The cost of CO₂ capture in this case is essentially zero, and the capture is total. 

If a ready source of pure O2 is not available, there are still alternatives that beat retrofitted flue gas scrubbing. The Norwegion research organization SINTEF has studied a variation on SMR that they term Gas Switching Reforming (GSR). Like SMR with oxyfuel combustion, it produces a waste stream of nearly pure CO₂ that requires minimal processing to make it pipeline-ready. The way in which that result is achieved is novel, but the bottom line is a process that captures 96% of carbon emissions at an energy cost only 0.3% above “plain vanilla” SMR without carbon capture.

It’s also possible to build a hybrid system that combines water electrolysis and gas reforming. The “waste” stream of oxygen from electrolysis, normally vented to the atmosphere, is instead used to implement oxy-fuel combustion for CO₂ capture from gas reforming. The oxygen from electrolytic production of one kg of hydrogen is enough to enable capture of pipeline-ready CO₂ from about three kg of hydrogen produced by SMR.

Direct Air Capture (DAC)

The ultimate option for carbon capture is to scrub it directly from air. That avoids the need for pipelines to transport captured CO2 from the point of emission and capture to the point of use or sequestration. It doesn’t matter whether emission sources are concentrated or diffuse.

Because of the low concentration of CO₂ in air, over 2000 tons of air must be scrubbed for every ton of CO₂ captured. That makes the economics of DAC challenging. But it has been done, and experience from pilot projects provides a basis for cost projections. Researchers from Carbon Engineering, one of the companies pioneering DAC, published a detailed economic study in June 2018 in the energy journal Joule. A report on the study was published that same month in Nature. The study projected costs for the Carbon Engineering approach in the range of $94 to $232 per ton. The range depends on the projected cost of renewable electricity and certain other economic assumptions. 

A range of $94 to $232 is well above the $50 per ton typically cited for the cost of flue gas capture from coal-fired power plants. However it’s well below what IPCC scientists had previously estimated for DAC. It’s above any currently implemented price on carbon emissions, but not beyond prices that have been mooted as fair for the uncaptured externalities of fossil fuels. Carbon Engineering’s approach, by choice, focuses on simple and well understood materials and chemistry. Energy efficiency was not high on their priority list. Various competitors are pursuing approaches that are more energy efficient. That could ultimately translate into lower costs. 

At this point, it’s hard to draw any firm conclusions about DAC. Other forms of carbon capture are more economical, and one would expect to see those implemented first. But DAC’s ability to be located in proximity to the point of CO₂ use or sequestration and its ability to capture emissions from diffuse sources are very attractive. It cannot be ruled out as a future option.

Photosynthetic capture

Capture and fixing of CO₂ by photosynthetic plants is a natural form of DAC. However, the capture mechanism and the economic issues are so different that it makes sense to consider photosynthetic capture as a distinct category.

The graph below, from Wikipedia’s article on the Keeling Curve, plots atmospheric CO₂ concentration levels over time in the northern hemisphere. The rising sawtooth shape of the curve illustrates both the powerful promise and the difficulty of photosynthetic capture as a means of reducing atmospheric CO₂ levels.

In spring and summer, plant growth in the northern hemisphere converts CO₂ from the atmosphere into biomass at a rate fully an order of magnitude faster than anthropogenic emissions add to it. But then in fall and winter, plant growth slows or stops, and consumption of biomass by oxygen-respiring lifeforms releases captured CO₂ much faster than the reduced level of plant growth removes it. The resulting seasonal variation overlays a steadily rising annual average from anthropogenic emissions.

If there were a magic way to prevent dead biomass from decaying while plant growth continued normally, then photosynthetic capture would remove CO₂ from the atmosphere very quickly. In only one or two decades, CO₂ levels would be back to pre-industrial levels. In three decades, they’d be well below pre-industrial levels, and the world would be heading into a new glacial period. Fortunately, preventing decay of all biomass is impossible. But we can perhaps tilt the balance between growth and decay in ways that will increase the global inventory of fixed carbon. That’s the intent behind regenerative agriculture and reforestation.

Hydrological capture

Like photosynthetic capture, hydrological capture is a natural form of direct air capture. And as with photosynthetic capture, the capture mechanism and profile for use are sufficiently different from other forms of DAC to justify labeling it as a distinct category. 

I use hydrological capture as a general term for two closely related processes. Both capture carbon through the dissolution of atmospheric CO₂ in water. Both are part of the hydrological cycle. One process is natural rock weathering. The other is ocean uptake of CO₂ through enhanced ocean alkalinity.

Rock weathering begins when small amounts of CO₂ dissolve in raindrops. That makes rainwater mildly acidic and able to erode certain types of exposed mineral surfaces. Silicate or sulfate groups in the exposed surfaces are replaced with bicarbonate groups. The bicarbonate forms are water soluble and carried, ultimately, to the ocean. The process is natural and always ongoing, but is normally slow. It speeds up during epochs of mountain building, when uplift and erosion result in greater exposure of fresh mineral surfaces. It can also be artificially accelerated by grinding and exposing large amounts of serpentine or other abundant alkaline minerals. 

For ocean uptake of CO₂ through added alkalinity, alkaline minerals are dissolved directly in seawater. Hydroxyl ions (OH^-) from the added alkalinity combine with hydrogen ions to lower seawater pH. The lower H⁺ concentration drives the equilibrium for dissolved CO₂ toward the right: 

CO₂(g) + H₂O(l) ⇌ HCO₃^- + H⁺

I.e., more CO₂ from the atmosphere is taken up. 

Bottom line on carbon capture

The above descriptions of approaches to carbon capture are by no means complete. They don’t cover all of the approaches and variations that are currently being pursued, and new variations are regularly being invented. But the variety that’s out there makes a key point: there’s no single best approach, and no single “cost of capture” that makes sense to talk about. Costs for carbon capture, per se, range from near zero for those applications that naturally produce a stream of pipeline-ready CO₂, to over $100 per ton of CO₂ for current methods of direct air capture.

Of course, capture alone is only half the story. Something has to be done with the captured CO₂. Ideally, it should be stored in a long-term carbon reservoir other than the atmosphere. Short of that, it must at least be used in a manner that will reduce consumption of fossil fuels. And that brings us to the other half of our CCS roadmap. Look for it in Part 2: Use and Sequestration.