7

The Role of Carbon Dioxide Removal in Deep Decarbonization

The sixth session of the workshop encouraged attendees to imagine a future economy with massive deployment of carbon dioxide removal solutions. Speakers discussed the natural solutions for sequestering carbon including forestry and other land use options, as well as engineering solutions such as direct air capture (DAC). Giana Amador (moderator) described her organization, Carbon180, a nonprofit that works at the interface of business, research, and policy to champion carbon removal solutions. Amador introduced the two speakers: Jennifer Wilcox (Worcester Polytechnic Institute) and Steve Hamburg (Environmental Defense Fund).

DIRECT AIR CAPTURE

Jennifer Wilcox, Worcester Polytechnic Institute

Jennifer Wilcox opened with an overview schematic of the primary available negative emissions technologies, shown in Figure 7.1. Wilcox’s presentation focused on direct air capture (DAC)—a technology which uses chemicals to remove carbon dioxide directly from the atmosphere. A few companies are currently operating DAC facilities, including Climeworks, which has removed 2-3 kilotons of carbon dioxide from the atmosphere to date. DAC is challenging and expensive, as carbon dioxide is relatively dilute in the atmosphere (ca. 400 parts per million), but advantages of DAC include its potential to be a full scale negative emissions technology, its potential to offset emissions that are difficult to avoid (e.g., heavy industry and some transport sector emissions), and its lack of

need for arable land. Disadvantages of DAC include its high energy input and large land footprint requirements. Wilcox stated that DAC should not replace emissions reduction efforts, but can be used as part of an overall carbon dioxide mitigation approach to offset some emissions.

Wilcox presented thermodynamic calculations of the minimum work required for separation, finding that energy requirements scale with dilution such that three times more energy and 300 times more contactor area is required to perform DAC versus capturing carbon dioxide directly from combustion exhaust at a power plant. Moreover, high purity carbon dioxide (>98 percent) is required for efficient transport as a liquid.

Wilcox described the process of scrubbing carbon dioxide from a point source, such as power plants operating with CCS. Amine scrubbing is performed by bubbling a gas exhaust stream through an absorption tower packed with solvent, into which the carbon dioxide is deposited into the solvent stream exiting the bottom of the tower. Power plants can balance their capture requirements and costs by optimizing the size of absorption towers deployed. Wilcox noted that DAC contactors have very different footprints from absorption towers. While a coal-fired power plant with CCS can capture 1.4 million tons of carbon dioxide per year with a 115-meter-tall absorption tower, a 20 meter by 8 meter by 200-meter-long DAC contactor would only capture 100 thousand tons of carbon dioxide per year.¹ Additionally, DAC contactors are expensive, and identifying the best cost reduction opportunities—such as improving the packing materials for solvent contactors or increasing the lifetime of materials for solid sorbent contactors—will require experience in research, development, and deployment at demonstration and commercial plants.

In addition to large footprints, DAC contactors require significant energy inputs to power the fans which draw air through the device, and for heat required to regenerate the solid sorbent or liquid solvent for reuse, the latter of which dominates the total energy cost. As such, DAC sites likely require on-site power generation, and the net CDR will be dictated by the choice of fuel, where choosing coal would result in zero net CDR and choosing natural gas would offset CDR by half. 300-500 MW of power are required to capture 1 million tons of carbon dioxide per year through DAC.

Wilcox stated that the capital expense to operations expense ratio is higher for solid sorbent DAC than for liquid solvent DAC due to the high cost of the contactor array, specifically the cost of the micro and mesoporous material that are embedded on and within the honeycomb

___________________

¹ National Academies of Sciences, Engineering, and Medicine, 2019, Negative Emissions Technologies and Reliable Sequestration: A Research Agenda, The National Academies Press, Washington, DC, https://doi.org/10.17226/25259.

framework. She noted that the dominating capital cost for liquid solvent DAC is the oxy-fired furnaces deployed to produce the heat used to regenerate their solvents, while the dominating capital cost for solid sorbent DAC is the contactor array. Driving down costs of DAC will require technological advancement and regulation, subsidies, or taxes on carbon, she said. Though DAC is not a comprehensive solution to climate change, it is an option we can implement as part of a broader net emissions reduction strategy. Investing $20 billion in cost reductions to get DAC down to $100 per ton of carbon dioxide removed (from today’s benchmark of $600 per ton) would mean building 200 “synthetic forests” each capturing 1 million tons of carbon dioxide per year, equivalent to 5 percent of our annual emissions in the United States.

Wilcox considered various DAC plant configurations, comparing a DAC contactor array powered by natural gas with CCS versus one powered by solar electricity to produced hydrogen for use in a hydrogen-fired kiln, as shown in Figure 7.2. Considerations for energy storage, energy efficiency, and land footprint of various configurations are important. Wilcox suggested a few cases where power plants and energy companies might benefit economically by deploying DAC using the U.S. 45Q tax credits. For the thermal requirements of DAC, geothermal plants can utilize their “waste” heat, nuclear plants can use 5 percent of their slipstream of steam, and natural gas plants can avoid burning stranded natural gas as flare gas. Wilcox presented a map of the United States with opportunities for DAC related geological storage, and by combining these benefits, she suggested that we can lower the cost of DAC to around $200 per ton of carbon dioxide removed. Ultimately, the primary strategy for climate change mitigation should be reducing carbon sources, but that negative emissions technologies like DAC can aid in our decarbonization efforts, she concluded.

ROLE OF FORESTS/LAND USE CHANGE IN DEEP DECARBONIZATION

Steve Hamburg, Environmental Defense Fund

Steve Hamburg spoke of the natural role that forests and other land play as a carbon sink. Though humans have manipulated the carbon sink capacity of natural systems through deforestation, building, and agricultural practices, we can still seize control of our use of natural systems to help decarbonize the economy. Hamburg suggested we think about decarbonization by looking at the whole carbon cycle to avoid spurious implications. While land use changes are currently contributing to higher carbon dioxide emissions, there is also a large offsetting sink of carbon associated with terrestrial and ocean storage, representing together a

massive transfer of carbon by natural processes of photosynthesis and respiration. In other words, there is simultaneously large amounts of carbon being emitted into and recovered from the atmosphere, and human activities are perturbing this natural system. The primary sources of carbon dioxide emissions are fossil fuel burning (87 percent) and land use changes (13 percent), while only 44 percent of those emissions remain in

the atmosphere, 29 percent are stored in the terrestrial biosphere, 22 percent are stored in oceans, and 5 percent are not offset and contribute to rising atmospheric carbon dioxide levels. If the Earth’s carbon sink weakens, then global warming will represent a bigger problem for society.

Over the last century, global emissions have shifted from coming primarily from land use changes to coming primarily from fossil fuel burning. As fossil fuel use has increased, the storage capacity of the land sink and ocean sink has increased, but these trends may not continue. Hamburg cited a recent report that suggested there is 0.9 billion hectares of potential reforestation areas globally, but Hamburg does not believe that this level of increase is practical. He mentioned that with respect to land use changes designed for decarbonization, it is easy to mix up what could happen with what is likely to happen, even with human intervention.

Hamburg noted that anthropogenic carbon dioxide emissions is only responsible for about half of the net radiative forcing we are seeing today, with anthropogenic methane contributing another 25 percent. He suggested that we think of decarbonization in terms of net radiative forcing, not just carbon dioxide emissions, as the methane budget is a complex, secondary carbon cycle which needs consideration.

Regarding reforestation in North America, Hamburg suggested that there is not a lot of potentially available land for such projects. The idea that we simply hire workers to replant trees in deforested areas is naive, as forests are already very good at natural regeneration, and replanting forests is difficult; only 8 percent of U.S. forests were generated by replanting and these forests tend to sequester less carbon than natural forests.

Hamburg presented three issues that need to be addressed when considering reforestation and land use change projects:

1. Benefits matter. You need an accurate baseline of the carbon capacity of an area to properly assess potential benefits.
2. Time matters. The carbon dynamics of natural systems play out over long time periods, sometimes over centuries, and thus one must consider how quickly a project will make an impact.
3. Spatial scale matters. The carbon accounting will be very different for projects on a particular field, a woodshed, a region, or the entire globe, due to differing local variability levels of carbon sink capacity.

Hamburg presented a research study that tracked the carbon capacity of a landed area (Grafton County, New Hampshire) over a 200-year period, revealing large changes in carbon flux due to land use changes connected to economic and societal trends. Deforestation and forest regrowth is a regional phenomenon that should be measured as such. Hamburg showed

picture evidence that deforested areas can undergo natural reforestation, but mentioned that 15 years post-harvest the ecosystem carbon is about the same as preharvest.

Hamburg mentioned that while we should pursue increasing soil carbon, we must consider system dynamics and the possibility that altering the carbon cycle in the soil also affects the nitrogen cycle and could lead to offsetting emissions increases for nitrous oxide. These cycles are complex systems that require robust understanding before we act. Regarding biomass, the challenge is ensuring that bioenergy is climate beneficial. Hamburg urged the workshop to consider system level impacts to using biomass, including albedo changes, bioenergy net carbon emissions, methane net emissions and uptake, and land use changes like fragmentation.

To determine if deep decarbonization decreases net radiative forcing, one needs to consider the full range of impacts of a particular action, including direct emissions changes, changes to land use, and societal and consumer changes influenced by those decisions. Hamburg pointed out that the CO2e metric is an oversimplification of a complex issue, and suggested instead using metrics which reveal the time implications of different actions in the short term versus the long term. Hamburg believes that the land use/forestry issue is a goldilocks problem—we need to find the right balance of action, in which we implement some measures to increase the carbon storage capacity of the terrestrial biosphere, but we avoid perverse outcomes which may come from acting too drastically and upsetting the balance within these interconnected planetary systems.

DISCUSSION

Before breaking for the workshop lunch, Amador told the workshop participants that society will likely need to be deploying carbon removal solutions on a scale of 10 gigatons of carbon dioxide per year by 2050. Amador urged participants to ponder what actions must be taken today to allow us to scale-up and reach that target within 30 years.

Upon returning from lunch, the discussion session commenced, with Roger Aines (Lawrence Livermore National Laboratory) filling in for Jennifer Wilcox on the panel. A participant asked Hamburg to comment on the difficulty of measuring the carbon capacity of land as well as land use changes. She wondered what actions we can take to increase certainty of measurement in the sector. Hamburg mentioned that he is less concerned about measurement uncertainty as he is about failing to take a systems approach and consider collateral outcomes of our action. He stated that we need accuracy in land use measurements at a regional scale, but not precision. Due to the enormous heterogeneity in nitrous oxide emissions for individual farms, estimations of very small systems

are often inaccurate and/or imprecise, but as you scale up to larger land areas you can achieve high accuracy. Hamburg offered a more important question: what is the scale of the measurement needed to ensure accuracy? He said that these assessments must be regional, not project-based. Further, Hamburg is interested in identifying the boundary conditions where certain measurement methods become less accurate.

Hamburg was asked about the possibility of using ocean biology to increase the carbon sink capacity of the oceans. He said that oceans are important, but like all ecosystems, we need to be careful in how we manipulate them due to unintended perverse outcomes.

A participant asked how climate change has affected land use and forestry practices and ecosystems, and whether it has created any positive feedback loops. Hamburg stated that climate change is affecting species and dynamics in these ecosystems, but he had not yet seen evidence of it affecting the carbon cycle.

A participant asked what the best use is for seaweed biomass in fighting climate change and ocean acidification. Aines stated that depending on your carbon price, the field traditionally thought that best option was to burn the biomass for energy and deploy CCS. Aines went on, that for higher carbon prices (over $100 per ton), it may be reasonable to simply bury the biomass. Hamburg agreed with Aines, but added that while we have multiple potential uses of the biomass (e.g., feedstock, process heat, or electricity) and we are still using fossil fuels, we might as well use the biomass as fuel as long as it does not require deforesting large areas. Hamburg worries also about potential perverse outcomes associated with burying biomass, such as ecological disruption. Aines pointed out an interesting potential use for biomass: as feedstock for fishotrope reactions to produce hydrogen fuel, resulting in a full negative-emissions use with nearly the same energy harvest as combustion of the biomass would produce.

A participant asked, if we capture, transport and sequester carbon dioxide on the scale of 10 billion tons per year by 2050, what is the business case for these activities? Will CCS be a publicly-funded utility like some trash collection services, or does it require a carbon price? Aines stated that the negative emission business case is the trash collection model, and that reaching our 2050 targets will cost the world around 1 percent of global GDP, similar to what we spend on trash management today. Amador added that many carbon tech products have strong business cases today, but to encourage carbon sequestration on the scale required to meet our 2050 targets we will need policy incentives and markets created by the federal government.

The panel was asked: where does the United States stand compared to other countries in managing forests and where are the best practices

found? Additionally, are we cutting more forests down now than historically? Hamburg mentioned that the United States actually has a very large forest carbon sink, offsetting about 15 percent of U.S. fossil fuel emissions. The eastern deciduous forest was cut down wholesale during the 18th and 19th centuries, but was then largely allowed to naturally regenerate resulting in a growing carbon sink that persists today. Hamburg noted that we may soon saturate our available forest sink capacity, and that some projections show that the amount of forest area in the United States will decrease moving forward. The most important factor in maintaining forested area is keeping disturbances out, such as grazing livestock.

A participant asked, is it widely true across different geographies and different forest types that natural regeneration is preferable to replanting forests—what about areas where trees cannot regrow? Hamburg clarified his position: it’s not that one should never plant new trees, it’s just that one should not plant most of the time because natural generation is remarkably robust. In degraded systems with dense plants that prevent natural regeneration, one may want to plant new trees.

The panel was asked: if it is not economically feasible to bury biomass underground, is there a role for very decay-resistant tree species to serve as useful building materials rather than fuel? Hamburg said wood is very underutilized in modern U.S. construction. He mentioned that replacing steel fence posts with wood would generate a greenhouse gas reduction benefit while providing an adequate level of structural stability.