Portal de Periódicos da CAPES

Negative Emissions Technologies: The Tradeoffs of Air-Capture Economics

2020; Elsevier BV; Volume: 4; Issue: 3 Linguagem: Inglês

10.1016/j.joule.2020.02.007

2542-4785

Matthew D. Eisaman,

Climate Change Policy and Economics

Matthew Eisaman is an assistant professor in the Electrical & Computer Engineering Department at Stony Brook University. From 2008 to 2011, Eisaman was an applied physicist at Xerox PARC. He received an AB in physics from Princeton in 2000 and a PhD in physics from Harvard in 2006, and he was an NRC postdoc at NIST from 2006 to 2008. He holds 17 patents and has coauthored 29 papers with over 3,700 citations. Eisaman invented indirect ocean capture in 2010–2011 while at PARC. From 2014 to 2016, Eisaman led Project Foghorn at X (formerly Google X), which used IOC to make carbon-neutral fuel from seawater. Matthew Eisaman is an assistant professor in the Electrical & Computer Engineering Department at Stony Brook University. From 2008 to 2011, Eisaman was an applied physicist at Xerox PARC. He received an AB in physics from Princeton in 2000 and a PhD in physics from Harvard in 2006, and he was an NRC postdoc at NIST from 2006 to 2008. He holds 17 patents and has coauthored 29 papers with over 3,700 citations. Eisaman invented indirect ocean capture in 2010–2011 while at PARC. From 2014 to 2016, Eisaman led Project Foghorn at X (formerly Google X), which used IOC to make carbon-neutral fuel from seawater. Climate models predict that significant deployment of negative emissions technologies (NETs) will be needed to avoid warming beyond 1.5°C.1Smith P. Davis S.J. Creutzig F. Fuss S. Minx J. Gabrielle B. Kato E. Jackson R.B. Cowie A. Kriegler E. et al.Biophysical and economic limits to negative CO2 emissions.Nat. Clim. Chang. 2016; 6: 42-50Crossref Scopus (783) Google Scholar Many NETs have been proposed, including land-based and ocean-based approaches.2National Academies of Sciences, Engineering, and MedicineNegative Emissions Technologies and Reliable Sequestration: A Research Agenda.. The National Academies Press, 2019Google Scholar,3GESAMPHigh level review of a wide range of proposed marine geoengineering techniques.in: GESAMP Reports and Studies no. 98. GESAMP, 2019http://www.gesamp.org/publications/high-level-review-of-a-wide-range-of-proposed-marine-geoengineering-techniquesGoogle Scholar Avoiding catastrophic warming and ocean acidification will require the development and deployment of a wide range of technologies,4Bindoff N.L. Cheung W.W.L. Kairo J.G. Arístegui J. Guinder V.A. Hallberg R. Hilmi N. Jiao N. Karim M.S. Levin L. et al.Changing Ocean, Marine Ecosystems, and Dependent Communities.in: Pörtner H.-O. Roberts D.C. Masson-Delmotte V. Zhai P. Tignor M. Poloczanska E. Mintenbeck K. Alegría A. Nicolai M. Okem A. IPCC Special Report on the Ocean and Cryosphere in a Changing Climate. 2019https://www.ipcc.ch/srocc/chapter/chapter-5/Google Scholar with the economically optimal NET depending on local conditions.2National Academies of Sciences, Engineering, and MedicineNegative Emissions Technologies and Reliable Sequestration: A Research Agenda.. The National Academies Press, 2019Google Scholar,3GESAMPHigh level review of a wide range of proposed marine geoengineering techniques.in: GESAMP Reports and Studies no. 98. GESAMP, 2019http://www.gesamp.org/publications/high-level-review-of-a-wide-range-of-proposed-marine-geoengineering-techniquesGoogle Scholar CO2 capture from the atmosphere is certain to play a significant role. While direct air capture (DAC)5Keith D.W. Holmes G. St. Angelo D. Heidel K. A Process for Capturing CO2 from the Atmosphere.Joule. 2018; 2: 1573-1594Abstract Full Text Full Text PDF Scopus (604) Google Scholar, 6Azarabadi H. Lackner K.S. A sorbent-focused techno-economic analysis of direct air capture.Appl. Energy. 2019; 250: 959-975Crossref Scopus (70) Google Scholar, 7Beuttler C. Charles L. Wurzbacher J. The Role of Direct Air Capture in Mitigation of Anthropogenic Greenhouse Gas Emissions.Front. Clim. 2019; 1: 10Crossref Scopus (71) Google Scholar has received the most attention among atmospheric CO2 capture approaches, indirect ocean capture (IOC—the indirect removal of atmospheric CO2 via extraction of dissolved inorganic carbon from seawater, followed by air-seawater re-equilibration) has been prototyped, and its performance and techno-economics have been analyzed.8Eisaman M.D. Parajuly K. Tuganov A. Eldershaw C. Chang N. Littau K. CO2 extraction from seawater using bipolar membrane electrodialysis.Energy Environ. Sci. 2012; 5: 7346Crossref Scopus (98) Google Scholar, 9de Lannoy C.-F. Eisaman M.D. Jose A. Karnitz S.D. DeVaul R.W. Hannun K. Rivest J.L.B. Indirect ocean capture of atmospheric CO2: Part I. Prototype of a negative emissions technology.International Journal of Greenhouse Gas Control. 2018; 70: 243-253Crossref Scopus (37) Google Scholar, 10Eisaman M.D. Rivest J.L.B. Karnitz S.D. de Lannoy C.-F. Jose A. DeVaul R.W. Hannun K. Indirect ocean capture of atmospheric CO2: Part II. Understanding the cost of negative emissions.International Journal of Greenhouse Gas Control. 2018; 70: 254-261Crossref Scopus (32) Google Scholar Below I explore how the relative costs of IOC and DAC vary as a function of time- and site-specific conditions such as the cost of energy, cost of capital, and pumping requirements. Fundamentally, IOC and DAC are equivalent: they both involve sorption of CO2 from the air, by the ocean in IOC or by a sorbent in DAC, followed by an input of energy to desorb CO2 into a concentrated stream (Figure 1). Since the CO2 reservoirs in the surface ocean and air equilibrate relatively quickly, the theoretical minimum energy to remove CO2 from the air via IOC is equal to that of DAC. DAC systems are typically engineered to create a large interfacial area between the air and sorbent material. DAC sorbents may be liquids5Keith D.W. Holmes G. St. Angelo D. Heidel K. A Process for Capturing CO2 from the Atmosphere.Joule. 2018; 2: 1573-1594Abstract Full Text Full Text PDF Scopus (604) Google Scholar or high-surface-area solid materials,6Azarabadi H. Lackner K.S. A sorbent-focused techno-economic analysis of direct air capture.Appl. Energy. 2019; 250: 959-975Crossref Scopus (70) Google Scholar whereas in IOC the ocean surface serves as a natural sorbent. The naturally occurring ocean-air equilibrium is one of the primary advantages of IOC, as (1) it avoids manufacturing the air contactor, resulting primarily in savings of capital expenditures (capex); and (2) it avoids the need to contact air with the sorbent, resulting primarily in savings of operational expenditures (opex). An additional advantage of IOC is that there is 147 times more CO2 in a certain volume of seawater, primarily as bicarbonate ions HCO3−, than in the same volume of air. Therefore, assuming equal CO2 capture fractions, DAC must process a volume of air 147 times larger than the volume of seawater IOC must process to capture the same amount of CO2. Finally, removal of CO2 from seawater can more directly and immediately mitigate ocean acidification in relatively protected areas of the ocean, though this advantage diminishes if the seawater is allowed to re-equilibrate with the atmosphere and mix with untreated seawater. DAC, in turn, has many potential advantages over IOC, for example: (1) DAC sorbents typically have higher CO2 loadings than seawater, potentially lowering the capex for regeneration; (2) water pumps are more expensive than air fans; and (3) the density of seawater is higher than the density of air, meaning that pumping seawater uses more energy than pumping the same volume of air. In calculating the cost C of air capture, the opex arises from moving air (DAC) or seawater (IOC) and the energy to regenerate CO2 from the capture solution, while the capex depends on the type and size of the required equipment. An expression for the cost of captured CO2 applicable to both DAC and IOC systems is:C=β+cp[ρQghηR+Eregen],(Equation 1) where the first term represents capex and the second and third terms represent opex. C is the cost of captured CO2 in $/tCO2; tCO2 is metric ton, i.e., tonne, of CO2; R is the rate of CO2 capture (tCO2 y−1); Q is the volumetric flow rate of seawater (IOC) or air (DAC) required to achieve a CO2 capture rate of R assuming a CO2 capture fraction of 1; η is the fraction of CO2 captured; g=9.81 m s−2 is the acceleration due to gravity. ρQgh/ηR is the energy per tCO2 to pump a fluid of density ρ across a pressure head h and flow rate Q/η resulting in CO2 capture rate R. β represents the amortized capital cost of captured CO2 with units of $/tCO2. cp is the cost of energy ($/kWh) and Eregen is the energy (kWh/tCO2) to regenerate CO2 from the sorbent. Equation 1 aims to capture the most important costs but ignores less significant yet non-negligible costs such as make-up chemicals and replacement membranes. How do we understand the relative tradeoffs, advantages, and disadvantages of IOC and DAC from Equation 1? Since seawater contains 147 times as much CO2 (as dissolved inorganic carbon mostly in the form of HCO3− ions) as the same volume of air contains as CO2 gas, the volumetric flow rate of seawater required by IOC is 147 times smaller than the flow rate of air required by DAC, assuming equal CO2 capture fractions for IOC and DAC. That is, QIOC=QDAC147≡QDACfv, where fv is defined as the volumetric CO2 concentration factor of seawater relative to air. At the same time, the density of seawater is higher than that of air by a factor of: ρIOC=ρseawater=1,025 kg m−3 =854ρair=854ρDAC≡fρ⋅ρDAC, where fρ is defined as the density ratio of seawater to air. Specific versions of Equation 1 for IOC and DAC can be written separately:CIOC=βIOC+cpIOC[(fρ⋅ρDAC)(QDAC/fv)ghIOCηIOCR+EregenIOC]=βIOC+cpIOC[(fρ/fv)⋅ρDACQDACghIOCηIOCR+ fr⋅EregenDAC],(Equation 2) andCDAC=βDAC+cpDAC[ρDACQDACghDACηDACR+EregenDAC],(Equation 3) where fr is the ratio of IOC and DAC regeneration energies and the differences between Equations 2 and 3 have been highlighted in bold. For the opex term representing the cost of pumping, the higher density of seawater relative to air (by a factor of fρ=854) overwhelms the volumetric CO2 concentration factor ratio of fv=147, resulting in fρ/fv = 5.8. However, since the pressure head h over which air and seawater are pumped in DAC and IOC may differ, as may the CO2 capture fraction η, the ratio of the IOC opex pumping energy to the DAC opex pumping energy is (ηDAC/ηIOC)⋅(fρ/fv)⋅(hIOC/hDAC). The energy costs cp may not be the same for IOC and DAC since IOC primarily uses electricity while DAC may use natural gas,5Keith D.W. Holmes G. St. Angelo D. Heidel K. A Process for Capturing CO2 from the Atmosphere.Joule. 2018; 2: 1573-1594Abstract Full Text Full Text PDF Scopus (604) Google Scholar waste heat,6Azarabadi H. Lackner K.S. A sorbent-focused techno-economic analysis of direct air capture.Appl. Energy. 2019; 250: 959-975Crossref Scopus (70) Google Scholar,11Realmonte G. Drouet L. Gambhir A. Glynn J. Hawkes A. Köberle A.C. Tavoni M. An inter-model assessment of the role of direct air capture in deep mitigation pathways.Nat. Commun. 2019; 10: 3277Crossref PubMed Scopus (167) Google Scholar heat pumps,7Beuttler C. Charles L. Wurzbacher J. The Role of Direct Air Capture in Mitigation of Anthropogenic Greenhouse Gas Emissions.Front. Clim. 2019; 1: 10Crossref Scopus (71) Google Scholar or electricity, depending on the approach. Moreover, integration of low-temperature DAC with synthetic fuel production allows process intensification that can further reduce energy costs. From Equations 2 and 3, the ratio of IOC cost to DAC can be written as:CIOCCDAC=βIOC+cpIOC[((fρ/fv)ρDACQDACghIOC)/(ηIOCR)+ fr⋅EregenDAC]βDAC+cpDAC[(ρDACQDACghDAC)/(ηDACR)+EregenDAC].(Equation 4) Defining cpDACρDACQDACghDAC/(ηDACRβDAC)≡Rpump as the dimensionless ratio of DAC opex pumping costs to DAC capex and cpDACEregenDAC/βDAC≡Rregen as the dimensionless ratio of DAC opex regeneration costs to DAC capex, Equation 4 becomes:CIOCCDAC=(βIOC/βDAC)+(ηDAC/ηIOC)⋅(fρ/fv)⋅(cpIOC/cpDAC)⋅(hIOC/hDAC)⋅Rpump+(cpIOC/cpDAC)⋅fr⋅Rregen1+Rpump+Rregen.(Equation 5) IOC's volumetric concentration advantage of fv=147 relative to DAC means that IOC can be more compact with cheaper capex, perhaps deployed on floating platforms located near offshore wind farms. However, by mass there is fρ/fv=5.8 times less CO2 in a tonne of seawater than in a tonne of air. This means that IOC may have larger opex resulting from seawater pumping, as encapsulated by (ηDAC/ηIOC)⋅(fρ/fv)⋅(cpIOC/cpDAC)⋅(hIOC/hDAC) in Equation 5. For IOC to be cheaper than DAC, the pump head loss hIOC must be minimized. To find typical parameter ranges for IOC and DAC, recent techno-economic analyses are used. R is assumed to be 1 MtCO2/y, with $60≤βDAC≤$200/tCO2 for high-temperature DAC5Keith D.W. Holmes G. St. Angelo D. Heidel K. A Process for Capturing CO2 from the Atmosphere.Joule. 2018; 2: 1573-1594Abstract Full Text Full Text PDF Scopus (604) Google Scholar and βDAC as low as $10/tCO2 for low-temperature approaches,7Beuttler C. Charles L. Wurzbacher J. The Role of Direct Air Capture in Mitigation of Anthropogenic Greenhouse Gas Emissions.Front. Clim. 2019; 1: 10Crossref Scopus (71) Google Scholar yielding an overall range of $10≤βDAC≤$200/tCO2. cpDAC=$3.69/mmbtu=$12.6/MWh when natural gas is used and EregenDAC=5.5GJ/tCO2=1.53MWh/tCO2.5Keith D.W. Holmes G. St. Angelo D. Heidel K. A Process for Capturing CO2 from the Atmosphere.Joule. 2018; 2: 1573-1594Abstract Full Text Full Text PDF Scopus (604) Google Scholar The measured regeneration energy for IOC is 2.49 MWh/tCO2,9de Lannoy C.-F. Eisaman M.D. Jose A. Karnitz S.D. DeVaul R.W. Hannun K. Rivest J.L.B. Indirect ocean capture of atmospheric CO2: Part I. Prototype of a negative emissions technology.International Journal of Greenhouse Gas Control. 2018; 70: 243-253Crossref Scopus (37) Google Scholar suggesting fr=2.491.53=1.63. In this case: cpDAC⋅EregenDAC=$19.28/tCO2, resulting in:Rregen=cpDAC⋅EregenDACβDAC=1.928−0.096. The DAC head loss is hDAC = (67.9 Pa)/(1.2 kg m−3 ∙ 9.81 m s−2) = 5.8 m, and the DAC CO2 capture fraction is ηDAC=0.745.5Keith D.W. Holmes G. St. Angelo D. Heidel K. A Process for Capturing CO2 from the Atmosphere.Joule. 2018; 2: 1573-1594Abstract Full Text Full Text PDF Scopus (604) Google Scholar For IOC, the head loss typical of intake piping for desalination plants is hIOC = 30 m – 100 m for a 1500 m pipe, so a typical range for hIOC/hDAC is 5<hIOC/hDAC<20. Since IOC uses electricity, cpIOC is $20/MWh−$60/MWh, so DAC powered by natural gas yields 2<cpIOC/cpDAC<6, while cpIOC/cpDAC=1 for DAC powered by electricity. The DAC opex due to pumping is then:cpDACρDACQDACghDACηDACR=$12MWh−1⋅1.2kgm−3⋅3.95⋅104m3s−10.745⋅9.81ms−2⋅5.8m⋅11MtCO2y−1=$0.38 (tCO2)−1, where 408 ppm CO2 concentration is used to calculate QDAC. For $10≤βDAC≤$200/tCO2, this results in Rpump=0.038−0.002. Figure 2 plots the IOC to DAC cost ratio CIOC/CDAC from Equation 5 as a function of βDAC and hIOC/hDAC for multiple values of cpIOC/cpDAC, with plotted ranges of $10/tCO2≤βDAC≤$200/tCO2, 0 ≤hIOC/hDAC≤20, and 1≤cpIOC/cpDAC≤4. In Figure 2, βIOC/βDAC=0.145=0.9⋅(1/fv)0.6+0.1=0.9⋅(1/147)0.6+0.1 represents the case where capex scales with equipment volume according to the "0.6 rule" and 90% of the capex depends on the required volume of air or seawater. Given $10/tCO2≤βDAC≤$200/tCO2, this corresponds to $1.45/tCO2≤βIOC≤$29/tCO2. Previous techno-economic analysis of a land-based IOC plant yielded βIOC=$162/tCO2, with $48/tCO2 from electrochemical equipment and $114/tCO2 from balance of plant, including the desorption of CO2 gas from solid CaCO3 and production of 99% pure CO2 (dry basis).10Eisaman M.D. Rivest J.L.B. Karnitz S.D. de Lannoy C.-F. Jose A. DeVaul R.W. Hannun K. Indirect ocean capture of atmospheric CO2: Part II. Understanding the cost of negative emissions.International Journal of Greenhouse Gas Control. 2018; 70: 254-261Crossref Scopus (32) Google Scholar In contrast, this paper intentionally explores low-capex implementations of IOC such as ocean-based, in situ precipitation of CaCO3 without conversion to CO2 gas, which can drastically reduce βIOC. Figure 2 shows that IOC can be cheaper than DAC for sufficiently low IOC head loss hIOC/hDAC and that the region where IOC is cheaper than DAC grows as the IOC-to-DAC energy cost ratio cpIOC/cpDAC decreases (moving from rightmost to leftmost column). Finally, for fixed cpIOC/cpDAC and hIOC/hDAC, IOC becomes relatively cheaper than DAC as the capital cost βDAC increases from left to right along the x axis due to the CO2-per-volume advantage that decreases IOC capex relative to DAC. Figure 3 explores the effect of increasing βIOC/βDAC for a fixed energy-cost ratio of cpIOC/cpDAC=1. While IOC can be cheaper than DAC for cpIOC/cpDAC=1 and 0.25≤βIOC/βDAC≤0.75, for cpIOC/cpDAC=1 and βIOC/βDAC=1, no combination of βDAC and hIOC/hDAC values can make IOC less expensive than DAC. This analysis highlights conditions where IOC is economically favored relative to DAC (CIOC/CDAC 1). For example, the pumping of seawater can be avoided by performing IOC directly in the ocean, leading small values of hIOC/hDAC and CIOC/CDAC<1. In contrast, land-based implementations likely favor DAC over IOC. In one imagined scenario, IOC equipment could sit on a ship or floating platform, taking advantage of wind energy from a nearby offshore wind farm, although the capacity factor would be an important consideration. This implementation would have multiple, competing effects on capex, decreasing it by eliminating certain equipment like holding tanks and piping, but increasing it due to the marine environment. In situ IOC could be performed in a volume of ocean from which marine life had been excluded by an expandable net. Within this volume of ocean, the cyclic pH shifts required for IOC could be performed directly with no need for seawater pumping. Before the net is removed and marine life is allowed to re-enter, the ocean would be returned to a targeted alkalinity and pH by (1) CO2 invasion from the air into this CO2-depleted, alkalinized region, and (2) mixing with surrounding, untreated seawater. In this way, IOC could be locally deployed to help protect and restore sensitive or valuable marine areas from acidification. Further research is required to understand the biological consequences of implementing this process or related concepts. In addition, the analysis considered here does not include storage costs. It is possible that IOC's capability for onsite conversion of the captured CO2 to enhanced ocean alkalinity12Rau G.H. CO2 mitigation via capture and chemical conversion in seawater.Environ. Sci. Technol. 2011; 45: 1088-1092Crossref PubMed Scopus (71) Google Scholar may be less expensive, less risky, and more environmentally beneficial than the geological storage suggested for onshore DAC. Looking into the future, the deployment of atmospheric CO2 capture technologies around the globe will likely feature multiple implementations of both DAC and IOC, with process specifics determined by local conditions. M.D.E. is a consultant for X Development, an advisor to Prometheus Fuels, and an inventor on multiple patents related to the subject of this work.