1. Introduction
A significant reduction in greenhouse gas emissions will be necessary in the coming decades to enable the global community to avoid the most dangerous consequences of man-made global warming. In particular, this means keeping the global temperature rise well below two degrees Celsius above pre-industrial levels [
1]. In the Paris Agreement of the 21st United Nations Conference of the Parties (COP 21) in 2015, the international community agreed to make efforts to limit the rise in temperature to 1.5 degrees Celsius by 2100 [
2]. Such a restrictive goal requires a comprehensive and, above all, rapid transformation in almost all sectors, with a substantial reduction in greenhouse gas emissions by 2030, i.e., within just over ten years [
3].
Such ambitious goals can only be achieved with a mix of different technologies. A number of these technologies are known today. Many of them have already been developed for the most part and are ready for the market. Others can achieve market penetration if general conditions in the energy industry are suitable. Furthermore, there are a few core requirements for the further development of technologies, for the provision of complementary technologies (e.g., grids and storage facilities) that enable market introduction or penetration, and for the narrowing of technology gaps.
These ideas are reflected in Germany’s 7th Federal Energy Research Program (EFP), entitled “Innovations for the Energy Transition” and adopted in 2018 [
4]. In the 40 years of its existence, the EFP has been repeatedly adapted to new societal challenges. To prepare for the 7th EFP, the “Technologies for the Energy Transition” research project [
5] analyzed a large number of technologies with regard to their development status, their potential for growth, and their possible research needs. To do so, the project carried out a systematic comparative multi-dimensional analysis. It assessed these technologies with regard to their abilities to meet the requirements of the energy industry, as well as climate policy, their capacity to help leading German companies continue to expand, and their likelihood of securing technological options by promoting technology on a broad front. The analysis employed a total of 12 different criteria (see
Figure 1). Due to the nature of the requirements for this analysis, the criteria were neither quantified nor weighted, and the different technologies were not ranked.
Direct Air Capture (DAC), sometimes also known as Direct Air Carbon Capture (DACC), was identified as one of the technologies for which there was still a large technology gap. In contrast to CO2 capture from power plants or industrial processes that target large point sources, DAC technologies are capable of directly capturing carbon dioxide from ambient air and thus from distributed sources. The use of such systems is theoretically conceivable at any location worldwide, and thus also offers regions without concentrated CO2 sources the potential to contribute to capturing CO2. However, this also presumes that the demand for electricity or heat—depending on the DAC process—is met by renewable sources, and that sufficient space is available. In addition, some processes require considerable amounts of water.
DAC could make a significant contribution to achieving the objective of reducing greenhouse gas (GHG) emissions through two primary applications. First, by providing climate-neutral carbon dioxide for synthesis processes, DAC could function as an enabler for the production of electricity-based fuels (Power-to-Liquids (PtL)), gases (Power-to-Gas (PtG)), and chemicals (Power-to-Chemicals (PtC)). Second, DAC could act as a Negative Emission Technology (NET), by capturing large amounts of carbon dioxide from the atmosphere in combination with subsequent long-term geological storage (Direct Air Carbon Capture and Storage (DACCS)). On the one hand, if determined and purposeful action is taken and a massive reduction of energy consumption through the implementation of various energy efficiency measures is achieved, it may be possible to meet the climate targets without NET [
6,
7,
8]. On the other hand, the majority of the Intergovernmental Panel on Climate Change (IPCC) scenarios see the use of NETs as unavoidable to achieve the 2 °C target, and especially the 1.5 °C target [
1,
3]. Their use cannot be limited to a single technology, but all NETs have their own role and potential [
9,
10]. Since both technology options are still largely ignored in energy policy, and the use of NETs has not yet been widely discussed in public, DAC was included in the multi-dimensional analysis.
To the authors’ knowledge, no analysis with a comparable comprehensive scope has been published before, either in general or with a focus on German energy research needs. When assessments of DAC systems have been carried out, this has been done in two ways. First, techno-economic assessment studies have evaluated future technical development (for example, the energy and water needs, the efficiency, or process specifications), combined with economic calculations (such as the potential CO
2 savings costs or the levelized cost of captured CO
2). The most comprehensive of such evaluations was published by Fasihi et al. after completing the recommendations for the EFP [
11]. The authors reviewed all known DAC technologies, estimated their possible technical development until 2050, and conducted a dynamic cost analysis using the learning curve method in order to compare the different technology options in the long run. In addition, they briefly discussed the land use needs and water consumption of DAC.
Second, assessments of DAC potential have been conducted within integrated assessment models (IAM). One group of climate scientists assessed the global CO
2 removal potential, cost, and relevant side effects of DAC and other NETs on a global level by way of a refereed literature review [
9,
10,
12]. However, this analysis neither considered the implementation level as intended in this article, nor did it result in a multi-dimensional analysis. Before that, Smith et al. had compared different requirements and impacts of NETs on a global level (land use, water, and energy demand and cost) [
13].
Against this background, the intention of this article is to explore the possible role of DAC for the German Energy Research Program by means of a multi-dimensional analysis and to show the research needs that DAC still demands. Where possible, the results of the reference study [
5] are updated with regard to studies published afterwards.
The remainder of the article is organized as follows:
Section 2 describes the methodologies applied in the multi-dimensional analysis and gives a rough overview on the status of DAC; the outcome of each analysis step is given in
Section 3; and after discussing the results from an integrative perspective in
Section 4, some final conclusions are drawn in
Section 5.
3. Results
3.1. Criterion 1: Lead Time
Mainly for financial reasons, existing facilities are still far from large-scale implementation, and will probably rely on market incentive programs to mature into competitive products. Therefore, under conservative assumptions, it seems unlikely that DAC will be commercially available on a large-scale before 2030. In a recent MIT Technology Review, Bill Gates ranked DAC as one of ten breakthrough technologies that would have (commercial) availability in five to ten years [
52]. This has been indirectly confirmed by manufacturers who advertise their systems as essentially market-ready (i.e., TRL of 9), but who would need R&D programs to reduce their costs from the current
$600/t CO
2 to a more competitive
$100/t CO
2 within the next ten years [
25,
53]. Technical learning through R&D support would complement cost reduction through volume effects and economies of scale.
3.2. Criterion 2: R&D Risks
3.2.1. Stage of Development
As depicted in
Table 1, TRL indications differ significantly, depending on the literature sources and the given technological concept. According to Napp et al., all NETs, including DAC (but excluding BECCS), are at a very early development stage, with TRLs of 1–3 [
54]. Other studies differentiate among different sorbents, and assign TRLs of 3–5 to hydroxide solutions, which are used for DAC
noTemp and DAC
highTemp, while assigning TRLs of 2–4 to amine-based DAC
lowTemp systems [
55]. Some more recent studies go further and classify DAC
lowTemp with a TRL of 5–6 [
56] or 6 [
57], primarily justified by the existence of several prototypes far exceeding lab-scale application. Accordingly, Climeworks ranks its own DAC
lowTemp technology at TRL 9, given current costs [
25].
3.2.2. Technical and Economic R&D Risks
As depicted in
Table 2, technical risk is assumed to be rather low, because between Climeworks and Global Thermostat, two of the three larger companies under review have already begun offering their technologies on the market. The third large company, Carbon Engineering, has tried to limit its technical risks by aligning the design of its large-scale process with the availability of commercial hardware [
22]. Economic R&D risk is perceived as being high in general, since measures that induce cost reductions, such as technical learning, upscaling, and mass production, all depend on adequate sales markets; these will only emerge under appropriate carbon control mechanisms. Even more contingencies arise in cases of using DAC within synthesis processes. On the one hand, the cost of CO
2 capture is only one cost component among several. On the other hand, these processes themselves are reliant on market incentive programs.
The results of both criteria justify the inclusion of DAC in the energy research program. In particular, technical learning stimulated by R&D funding would significantly reduce the lead times and economic risk.
3.3. Criterion 3: Market Potential
Table 3 shows the scenario results for a possible carbon-neutral CO
2 demand caused by PtL and PtC production, both in Germany and globally, according to the assumptions set out in the definition of the criterion. In the case of Germany, an almost complete substitution of conventional chemical production by PtC would result in a CO
2 demand of 55 Mt CO
2/year by 2050, which is approximately twice as much as that needed for PtL (however, transportation is mainly covered by other decarbonization strategies in the assumed scenario, which makes this assessment rather conservative).
An almost complete worldwide substitution of (increasing) conventional chemical production by PtC would result in a CO2 demand of 4500 Mt CO2/year by 2050. These figures are approximately 100 times higher than those in the German scenario and are determined by vigorous growth in demand, especially in China, India, and the U.S. According to the scenario assumption in the traffic sector, PtL production would result in a relatively small demand of 400 Mt CO2/year in 2050.
The second aspect to consider is the necessary effort to globally achieve negative emissions. As
Figure 2 illustrates, the global demand for DAC, as modeled by Chen and Tavoni [
48] and Marcucci et al. [
49], far exceeds those quantities required by PtL and PtC, which are assumed to remain constant from 2050 onwards. In comparison, the more moderate global demand derived from Rogelj et al. [
50], by assuming that 50% of the stated negative emissions may be achieved by DAC, is in the same order of magnitude as the demand for PtC.
Figure 3 shows the proportion of CO
2 required for negative emissions for which Germany might be responsible, in comparison to the demand of 80 Mt/year that is required by PtL and PtC domestically.
3.4. Criterion 5: Contribution to Energy and Resource Efficiency (Energy, Land, and Water Demand)
3.4.1. Energy Demand
In theory, the capture of CO
2 by DAC would require two to four times as much energy as the capture of exhaust gases from a power plant [
14,
19], which is relatively low in view of a 250–300 times lower concentration of CO
2 in the air. On the other hand, ideal consumption in reality is often considerably exceeded, so that much worse values can be expected than in the case of CCS. APS has stated 0.124 MWh/t CO
2 as the thermodynamic minimum required to capture 50% of the CO
2 present in the ambient air at a concentration of 500 ppm, supplemented by 0.06 MWh/t CO
2 for CO
2 compression [
19]. In practice, however, these minimum values are never reached.
Table 4 depicts specific energy-related parameters of the various types of DAC processes. The figures are not directly comparable because they were determined for specific system configurations. Nevertheless, they show that future large-scale commercial plants can still be expected to deliver significant energy savings. Lackner’s process has by far the lowest energy requirement, but is still only a theoretical concept. Compared to Carbon Engineering’s system, the Climeworks system can cover its heat requirement from waste heat due to the low temperature level, which makes it predisposed for use in PtX processes. The same is true for the newly developed system from ZSW, while no data is available for Global Thermostat. Overall, it is evident that development from initially power-intensive (DAC
noTemp) to thermal (DAC
highTemp and DAC
lowTemp) processes has occurred.
3.4.2. Land Use
Table 5 summarizes the land use parameters of the various DAC processes (no data is available for DAC
noTemp). The figures regarding system dimensions vary by several magnitudes. According to APS, a DAC
highTemp facility capable of capturing 6 Mt CO
2/year (the emissions of a 1 GW coal-fired power plant) would require approximately 9 km
2, i.e., 1.5 km
2/(Mt CO
2·year) [
19]. Since this figure does not include any area set aside for energy generation facilities [
13], high-temperature heat delivering plants, such as natural gas or concentrated solar power plants (the latter including large areas for concentrating tubes or mirrors) must be accounted for. Carbon Engineering has indicated dimensions of 8 m × 200 m for capturing 0.1 Mt CO
2/year, i.e., 0.0016 km
2/(Mt CO
2·year). However, the company itself uses those dimensions only to refer to the packings, and concedes that actual land use would be significantly higher [
59].
In the case of DAC
lowTemp, Climeworks has stated that it only requires 90 m
2 for a DAC-18 type system, which would be capable of capturing 0.9 kt CO
2/year, i.e., resulting in an area of 0.1 km
2/(Mt CO
2·year), mainly consisting of clearance between facilities [
25,
26]. Again, energy generation plants have been excluded from these figures. If they couple their system with synthesis plants, only the necessary electricity demand would have to be taken into account, since waste heat would be available. In the case of DACCS, accounting for additional heat supply is relevant since waste heat delivery facilities would not necessarily be available. If, for example, photovoltaic-operated heat pumps were used, the necessary area would increase to approximately 2 km
2/(Mt CO
2,a) at optimal sites [
25,
26]. Lackner states 30 m
2 for capturing 365 t CO
2/year, i.e., 0.08 km
2/(Mt CO
2·year) [
15]. Global Thermostat has indicated 20–500 t CO
2/(a·m
2), which is 0.05–0.002 km
2/(Mt CO
2·year) [
31].
3.4.3. Water Demand
In their comparison of various NETs for DAC, Smith et al. calculated a water consumption of 19.9–30 m
3/t CO
2 for DAC [
13]. The range of values is not comprehensible, since their Supplementary Information referred to amine solutions on the one hand, and the referenced source says nothing about this on the other. Stolaroff et al. have stated that for a DAC
highTemp process working with aqueous NaOH solutions, water losses due to evaporation would be 20 mol H
2O/mol CO
2 for a certain parameter setting [
60], which would thus correspond to 8.18 m
3/t CO
2. Even if these values could be reduced by process improvements, significant water losses would have to be expected. Carbon Engineering has declared a net water consumption of 4.7 m
3/t CO
2 for its aqueous KOH solution pilot plant. This ratio fluctuates with temperature, ambient conditions, and solution molarity [
23].
Climeworks, on the other hand, produces approximately 1 m
3 water/t CO
2, since the TSA process does not consume any water, while, at the same time, the vapor taken in with the ambient air can be used [
25,
26]. In this way, the Climeworks system can capture CO
2 at a single location, produce hydrogen from the water obtained, and generate synthetic products from syngas, while the waste heat is in turn used for air capture. The first such cogeneration project has already been successfully demonstrated for the production of kerosene (using solar thermochemical water splitting via CSP instead of electrolysis) [
24]. No figures were available for Global Thermostat.
3.5. Criterion 6: Costs
Table 6 depicts both current and estimated future costs of the various types of DAC processes, as stated by the manufacturers. Carbon Engineering recently reported a fall in cost expectations from
$600/t CO
2 to a (levelized)
$94–
$130/t CO
2 for providing CO
2 for fuel synthesis and to
$113–
$232/t CO
2 including compression of the CO
2 to 15 MPa for storage. In so doing, they considered various forecasts of future electricity prices (
$30–
$60/MWh) and a natural gas price of
$12.6/MWh, but omitted the assumed cost of water [
23]. It should be noted that their work refers to a conceptual design, modeled with Aspen, which has not yet been built. Climeworks has stated costs of
$600/t CO
2 for actual operation and has aimed to reach
$200/t CO
2 by 2020 as a result of upscaling and mass production, and
$100/t CO
2 by 2025–2030 as a result of further R&D, in each case without mentioning the underlying energy prices [
25,
53]. Global Thermostat has even asserted costs of only
$50/t CO
2 [
53], which seems fairly unrealistic, given the lack of data and detailed explanations. Further cost reductions depend on how long the amine surfaces last [
61]. Lackner’s design had the same cost targets, but it is in a very early development stage. No cost data was given for the ZSW’s newly developed process [
37].
Older cost statements in scientific literature spanned even more widely, ranging from
$25/t CO
2 to
$1000/t CO
2. However, both upper and lower bounds referred to either very early or vague assessments (see, e.g., [
21,
62]). No data was available for the early electricity-based processes developed by ZSW and PARC. Furthermore, none of the papers assessed used learning rates to estimate future cost developments. The only models that considered endogenous learning were climate models exploring DAC as NET between 2060 and 2100, each without specifying the corresponding learning rates [
48,
49]. Climeworks cited learning rates of 20% according to their cost forecasts [
25], which seems high but not unrealistic, given similar developments of other low-carbon technologies [
63].
Since this study was prepared in an effort to assess the need for DAC’s inclusion in a research program, standardizing the specific company data and making rigorous estimates of future costs were outside the scope of this paper. After completion of the study, a comprehensive techno-economic evaluation was published [
11] in which the authors calculated the levelized cost of DAC (LCOD). The cost figures projected for 2050 resulted in
$60–
$79/t CO
2 for DAC
highTemp and
$42–
$60/t CO
2 for DAC
lowTemp (using a conversion factor of €0.90/
$1 as of July 2019) (each without compression of CO
2). It depended on various general conditions and learning rates, as well as electricity costs of
$55.6/MWh, low-temperature heat (<100 °C), costs of
$22.2/MWh, and high-temperature heat (900 °C) costs of
$27.8/MWh. These figures show that, depending on the assumptions made about scenarios and parameters, the vague figures supplied by companies could fall even further in the long term.
3.6. Criterion 7 (Domestic Value Added)
Currently, no calculations according to the criteria description are possible, as there is no German company developing DAC processes. As described above, the research center ZSW has developed its own DAC
lowTerm process. In addition, Climeworks and its German subsidiary are involved in various research projects in Germany. Additional research work might also be underway at universities. Nevertheless, if one considers the strong position of the German chemical industry and the mechanical engineering industry, it can be assumed that German companies would also enter a DAC market as soon as the demand for the provision of climate-neutral CO
2 for PtX applications becomes more specific. In the foreseeable future, however, CO
2 from industrial plants will probably be used for this purpose. During this time, corresponding production capacities for DAC could be developed. The discussion on negative emissions has so far only been conducted among climate scientists and, due to the long-term nature of its implementation, is unlikely to be the focus of industry for the time being. Even without the ability to calculate the specific contribution to value added, the market potential to be expected estimated in
Figure 3 justifies the inclusion of DAC in the German energy research program solely on the basis of German demand.
3.7. Criterion 8 (Status-Quo and Trends of R&D in International Comparison)
Currently, there are no known German manufacturers of DAC technologies. To the authors’ knowledge, the public R&D budget has only been made available for two research projects, each by the Federal Ministry of Education and Research: The CORAL Project (“CO2 Raw Material from Air”) received funding with a duration of 2016–2019 (the resulting process, from ZSW, was one of the aforementioned DAClowTemp systems). Within the ongoing “Power-to-X” Copernicus project, the “A2: Low Temperature Co-Electrolysis” research cluster is developing an integrated plant for the production of liquid fuels. A number of additional research activities, e.g., on absorption processes, are likely to be supported by R&D funding, but cannot be directly attributed to the development of DAC plants.
Regarding the research output, the patent search yielded no results for Germany. Worldwide, 20 patent applications have been filed: four each in the USA and Canada, two in China, and one each in Croatia and Mexico (as of March 2018). In addition, three European (EP) and three World Intellectual Property Organization (WIPO) patents were identified.
Finally, the publication analysis revealed 290 refereed papers from 2006 to July 2019; of these, papers that contained no direct reference to DAC, or where the main focus was clearly not on DAC (e.g., in the case of other meanings of DAC, such as “diamond-anvil cell” in medicine; other meanings of “air capture”, such as in wind energy, trees, or space travel), were manually excluded. In total, 57.5% of the remaining 167 papers presented results from basic technical research, while around 13% dealt with technical–economic valuation and reviews. Only 17% considered overall issues, such as DAC in the context of energy systems, political issues, or applications for PtX, while 12.5% discussed DAC in the context of NETs (
Figure 4).
The linear interpolation shows a steady increase, particularly in the case of basic technical research. One noteworthy aspect is the strong increase in both crosscutting issues and issues dealing with DAC in the context of NET from 2018 onwards, which can be expected to reach a similar magnitude in 2019. Only 11 out of the 167 papers (6.6%) are from German authors (2 from 2019 and 9 from 2018), of which 3 papers include overarching topics and 6 papers are focused on DAC in the context of NET.
3.8. Criterion 9: Societal Acceptance
To the authors’ knowledge, no studies or public discussions on the social acceptance of DAC systems are known at present. There have been critical reports and analyses on NET in general [
47,
64,
65,
66,
67], but there are no acceptance studies available here either. With regard to the use of DACCS, synergies with the general issue of CO
2 storage may be conceivable, but all corresponding acceptance studies refer to CCS for power plants and industry. For the case of Germany, one study demonstrated, for example, that the storage of CO
2 from industrial plants had a somewhat higher acceptance than CO
2 from power plants [
68]. If it is a matter of achieving negative emissions or the release of climate-neutral CO
2 for PtX processes, a completely different perception could emerge in society. Therefore, there is a considerable need for research on decisive factors regarding the acceptance of DAC and possible obstacles to its implementation, as well as to communication and participation processes with various groups of actors regarding DAC and its background.
3.9. Criterion 10: Path Dependency
Little or no path dependency is expected for using DAC within synthesis processes, as DAC facilities can be constructed on site with synthesis plants. In contrast, the risk of path dependency for the use of DAC as an NET is expected to be very high due to the extensive need for land, energy, and infrastructure. If DAC as an NET is to be implemented, it will be necessary to develop an overall concept that would be sustainable in the long term. Specifically, this would have to include not only the CO2 storage sites to be used and the infrastructure required, but also the provision of the energy required.
3.10. Criterion 11: Dependence on Infrastructure
Depending on the underlying concept, DAC facilities need thermal and/or electrical energy and thus rely on the commensurate infrastructure. Thermal options could include waste heat (according to the process), heat pumps, or high-temperature heat generation (e.g., natural gas, CSP); electricity can be provided either by public supply or via direct coupling with generation plants. Moreover, DAC facilities also require a significant amount of land (though this amount varies depending on the project), as well as a still partly unspecified need for water (see
Section 3.4). Moreover, if on-site processing or storage of the captured CO
2 in the case of DACCS is not an option, additional transport infrastructure (e.g., pipelines) would also be required. Ideally, plants for (renewable) power generation, water supply, and synthesis processes, as well as CO
2 capture, would be located at the same site, so that only the synthesis products would have to be transported.
3.11. Criterion 12: Systems Compatibility
Given the respective scaling, the energy demands of DAC facilities in Germany would require adaptions in renewable energy generation. As
Table 7 shows, for PtL and PtC purposes alone, satisfying the demand for climate-neutral CO
2 (which was roughly and conservatively estimated above, see
Figure 3) would require 121.5 TWh
th of low-temperature heat and 40.5 TWh
el of electricity (based on
Table 4). This calculation was done using the example of a Climeworks’ DAC-36” system, as this is the only DAC
lowTemp system for which sufficient data is currently available. This meant using a target heat consumption of 1.5 MWh
th/t CO
2. While this level of heat consumption may be covered by waste heat from the necessary synthesis plants, the electricity demand (41 MWh) would represent an increase of 7.8% compared to the 527 TWh
el required in 2018 in Germany [
69]. About twice as much heat and electricity would be needed if the estimated German share of negative emissions derived from the moderate scenario of Rogelj et al. [
50] would have to be managed in Germany. In this case, additional heat sources would have to be taken into account. The feedback loops from the additional load on the entire energy system, including its spatial distribution, should therefore be analyzed in advance.
Land use is also a crucial parameter in a densely populated country such as Germany. Without taking into account the land consumption for electricity production, the sample DAC plants alone may require an area of 8.1 km
2 in the case of PtX processes and 16.9 km
2 for negative emissions (based on
Table 5), which equals 0.9% and 1.8% of the area of Berlin, respectively. Since this estimate contains some uncertainties, these values should be considered an initial estimate. On the one hand, only the area of the respective individual plants was taken into account, while for combined plant, parks in practice areas between the rows of plant towers would also have to be added. On the other hand, a large part of the considered area is used for control technology, which in the case of individual plants (e.g., the “DAC-36”), accounts for around 60% of the total area, but would no longer be significant in the case of large plants [
25].
Negative interdependence (i.e., competition) with short- and medium-term climate policy goals might arise, particularly for the use of DAC as an NET, because the prospect of drastically reducing emissions after 2050 might diminish the sense of urgency for active climate policy within the coming decades. As a consequence, additional emissions might accrue (overshooting) [
67,
70].
4. Discussion
The previous sections indicate that the future role of Direct Air Capture will be affected by a variety of uncertainty factors. The technology is still in an early stage of development and has yet to prove its large-scale technical feasibility, as well as its economic viability. Currently, there is neither consensus within the literature itself, nor between literature and manufacturers, on the potential for technical efficiency, as well as cost decreases. Furthermore, comprehensive technology assessments (e.g., in regard to coupling with low- or high-temperature synthesis facilities), which would be necessary to fully evaluate the applications outlined above, are lacking. Moreover, additional applications are conceivable for DAC that have not been examined in this study, e.g., as part of power-to-gas for heating or as a strategy for dealing with an intermittent energy supply from renewables.
The literature under review shows that there is a particular need for technical research in the development of CO
2 sorbents and processes for the clusters under consideration. Even though adsorption/desorption processes have proven to be the most promising with regard to applications for the electricity-based production of fuels and chemicals, since they can utilize the low-temperature heat of synthesis plants [
11], other processes should also be further developed in the spirit of open-ended research. In the case of high-temperature DAC, this could be, for example, the use of renewable energies instead of natural gas, which could include unprecedented investigations of the coupling of DAC with CSP plants, thus leveraging both their high-temperature heat and the electricity generated.
Several authors have given an overview of specific technical research needs [
14,
71,
72,
73]. Processes already developed to market maturity must be prepared for large-scale use and mass production, for example. This requires the minimization of energy requirements, plant scaling and, in particular, process integration, so that they can be optimally used for the production of electricity-based products—both for fuels and basic chemicals. Carbon Engineering has performed theoretical calculations to understand how to achieve this process with as little risk as possible [
23]. Minx et al. have pointed out that a sense of urgency about the upscaling and diffusion of NET and thus also DAC technologies has not emerged, either among scientists or policymakers. They argued that the period from 2030 to 2050 would be essential for the massive expansion of NET [
9]. The Innovation for Cool Earth Forum (ICEF) has summarized the most important innovation steps required in a roadmap for the next 20 years [
74].
In addition to detailed cost analyses, Minx et al. also pointed to the need for comprehensive life-cycle assessments to assess the various environmental impacts of these chemical-based processes [
9]. Williamson noted, in particular, the lack of safety and environmental impact assessments for NET technologies, but determined that DAC could still be the most environmentally friendly of these [
46].
In light of the uncertainties associated with NET in general and with DAC in particular, climate scientists have been calling for integrated assessments of these technologies. Along these lines, Smith et al. pointed out which driving and braking forces should be considered in the evaluation of NET, which would also be relevant for DAC to a large extent [
13]. Fuss et al. have described four criteria for a major new transdisciplinary research agenda regarding the management of NET: technologies and their sustainability impacts; carbon cycle uncertainties; policy and technology mix regarding cost, risk, and timing; and institutional and policy challenges [
67]. To a certain extent, these also play a role in the use of DAC for PtX processes. This would also include critically scrutinizing existing potential assessments for DAC and/or NET, making more detailed estimates of the possible need for DAC for PtX and for achieving negative emissions, and also analyzing possible alternative options and the measures that would have to be taken to implement them. Although these demand-side issues should be urgently considered, a detailed investigation by Nemet et al. noted that 83% of the scientific work on NET has been concerned with the supply side (59 percentage points of which are R&D alone, supplemented by work on demonstrations and scaleup) [
11]. DAC itself has an even larger share. This confirms our results regarding Criterion 8, where we found around 68.5% of the articles dealing with technical research, illustrating an urgent need for studies on technology (impact) assessment.
Last but not least, systems-level analyses are needed to analyze the interaction between the various plants in Germany and abroad. On the one hand, this concerns the potential demand for power-based fuels and chemicals and the timelines in which they could replace fossil-based substances. On the other hand, questions of the geographic location of the individual process steps have to be answered—from the location of DAC plants for CO2 separation and CO2 transport to the location of synthesis plants and the supply of the required electricity, heat, and water.
Temporal development also plays an important role, as large point sources with fossil CO
2 could be used at first, until infrastructure for the use of climate-neutral CO
2 is built. In turn, completing DAC facilities for PtX processes would serve as a path toward a possible massive deployment of DAC for the sake of negative emissions. Land consumption also must play a central role in the analysis of these issues, especially in densely populated industrialized countries, such as Germany. Finally, further research questions concern the output of the required plants and the establishment of corresponding mass production. Only then would it be possible to make estimates for the expected domestic added value of German companies (Criterion 7). If, for example, 1% of global CO
2 emissions (350 Mt/year) were to be separated by DAC alone, this would require around 200,000 of Climeworks’ DAC-36 plants (
Table 7). For 50% of global emissions, 10 million plants would be required, which in terms of plant volume, would correspond to almost 14.5% of one year of current global passenger car production. In order to prepare for such an enormous technology diffusion, several hundred NET facilities may have to be installed annually between 2030 and 2050 [
9].
Nemet et al. have pointed out that “the thousands to millions of actors that potentially need to adopt these technologies for them to achieve planetary scale” should also be considered (e.g., with regard to research on public acceptance) [
12]. All these questions of technology assessment raised here require interdisciplinary and transdisciplinary studies at an early stage to be prepared for possible developments.
5. Conclusions
The previous sections show that it is currently impossible to determine either the future role of DAC or the possible role of the German industry, based on the current level of knowledge. The results of this multi-dimensional evaluation justify the inclusion of research on DAC technology in the German energy research program:
Although many DAC prototypes and demonstration facilities exist, several development and upscaling steps will be necessary to achieve a Technology Readiness Level of 9, related to a large-scale commercial use, within the next ten years. This means the need for further technological development in order to reduce costs from the current $600/t CO2 to less than $100/t CO2 and thereby reduce the consumption of energy, water, and land during operation;
While the technical risk is assumed to be rather low, the economic R&D risk is assessed as high;
The evaluation of the R&D landscape shows only a small contribution of Germany in an international comparison in terms of R&D budgets (very low), scientific publications (6.6%), and patents (0);
A large market potential for the use of DAC, not only in Germany, but also worldwide, may be expected to make a major contribution to value creation, which could lead to the expansion of the leading position of German industry;
Even though no explicit research on DAC has taken place in Germany so far, German companies from the chemical and mechanical engineering industries are well positioned to advance the research and development of DAC processes.
In view of the considerable importance that DAC could receive in the future considering climate protection requirements, the role of DAC is likely to increase significantly. There is therefore also a great need for research worldwide, both with regard to technical development and in particular with regard to the development of deployment strategies and roadmaps for setting up DAC.
Apart from the research needs described above, there are major uncertainties and challenges in this field of technology:
On the one hand, estimates of market potential are very uncertain (both for PtX applications and even more for use as NET). At the moment, it is not possible to predict in which sectors PtX applications will be needed in the medium to long term, nor whether NET will be needed at all, and if so, what role DAC could play;
On the other hand, the need to reach a carbon-neutral economy by 2050 does not leave much time for either long-term research or the deployment of DAC in the case that this technology would prove to be an important part of the transition in the energy, transport, and chemical sector;
A potentially massive deployment leads to the question of building up sufficient production capacities so that they are available in time. As the above comparison with automobile production has shown, the sheer volume of systems alone is likely to be a considerable challenge. However, not only the production structures, but also the substantial demand for energy and land in densely populated countries, could become a considerable challenge;
Last but not least, the possible application of DAC technologies has not yet been discussed in politics or in public. In particular, both socio-political and community acceptance could be critical and delay possible implementation by years.
Having all these issues in mind, a strong research focus on integrated assessments and system-level analyses is recommended. This should be accompanied by technical research as outlined above, so that the technologies are ready for use when needed. Systems analytical studies could also provide valuable information for technology development in order to identify possible risks and problems during subsequent implementation at an early stage and allow them to be taken into account in the technology development process.