Analysis of Equilibrium-Based TSA Processes for Direct Capture of

Article pubs.acs.org/IECR

Analysis of Equilibrium-Based TSA Processes for Direct Capture of CO2 from Air Ambarish R. Kulkarni and David S. Sholl* School of Chemical and Biomolecular Engineering, Georgia Institute of Technology, Atlanta, Georgia 30332-0100, United States S Supporting Information *

ABSTRACT: Direct capture of CO2 from air is a concept that, if successfully implemented, could lead to capture of CO2 from disperse sources. We have developed process models to consider the viability of adsorption-based air capture technologies. Our models focus on using an amino-modified silica adsorbent, TRI-PE-MCM-41, and a structured monolithic contactor unit. We have studied several different temperature swing adsorption processes using the purity of CO2 and annual product throughput as metrics for comparing process performance. This analysis identifies some of the operational parameters, adsorbent characteristics, and other factors that have a significant effect on the performance of the process. Using the total energy requirement of the process and available sources of energy, such as low pressure steam and electricity, we carry out an economic analysis to obtain a net operating cost for air capture of CO2. We identify a process with a daily throughput of ∼1.1 t CO2 at 88.5% purity using standard shipping container sized air capture units. The total energy required (6745 MJ/t CO2) is dominated by the parasitic lossessensible heat requirements of the contactor (40%) and the adsorbent (28%) and not by the mechanical energy associated with air flow (∼5%). On the basis of our analysis of factors such as source of electricity, availability of low pressure steam, and geographic location, the net operating cost of capture is estimated to be ∼$100/t CO2. These cost estimates do not include capital expenses necessary to construct or maintain the air capture units. Potential strategies for further reducing the energy and monetary cost of these processes are identified. Our analysis supports continued work to establish the technological and economic feasibility of adsorption-based air capture.

1. INTRODUCTION The growing levels of CO2 in the atmosphere and their possible detrimental effect on the global climate have made carbon management technologies one of the most widely researched areas of recent times. The current CO2 level in the atmosphere is 379 ppm.1 One benchmark technology for reducing CO2 emissions is using aqueous solutions of amines to capture CO2 from postcombustion flue gas. Other strategies available for capture of CO2 from flue gas using absorption and adsorption have recently been reviewed.2 Point sources like large coal-fired power plants typically account only for about one-third of the anthropogenic CO2 released to the atmosphere.3 Much of the remaining two-thirds is due to transportation, small power plants, and chemical industries. Under the plausible assumption that fossil fuels are going to continue to be the predominant source of global energy in the near future, no commercial carbon capture technology currently exists that can be used to offset CO2 emissions from this two-thirds of total emissions. If deep reductions in global CO2 emissions are to be achieved, a broad range of technology and policy options need to be explored. Direct capture of CO2 from air, which we will refer to below as air capture, is one technology that has the potential for capturing CO2 emissions from all possible sources. Air capture aims to make use of the concepts and technologies developed for CO2 capture from flue gas capture and apply them to capture CO2 from ultradilute concentrations in air. The concentration of CO2 is ∼250 times less in air than in flue gas. The theoretical minimum energy required for air capture, however, is only 3.4 times that for flue gas capture.4 There has been far less work performed on air capture than on flue gas capture, but the © 2012 American Chemical Society

economic and technical feasibility of air capture has been intensely debated.5−7 Jones has recently reviewed the development to date of air capture technologies.3 The first discussions of air capture as a technology to address rising atmospheric CO2 levels was presented by Lackner in 1999 and 2001.8,9 In 2004, a process was suggested that used sodium hydroxide solution to absorb CO2 from air and convert it into soluble sodium carbonate.10 The sodium carbonate was precipitated by using quicklime that was then calcined to recover CO2. Not surprisingly, the calcination step in this process is the most energy intensive step, as high temperatures (>900 °C) were required to break the strong carbonate bond.11,12 Work has since been performed with the aim of decreasing the energy requirement by using different sources of thermal energy,13−18 changing the contactor configuration,19 and using alternative regeneration chemistry.20 A recent report on the feasibility of air capture by the American Physical Society (APS) focused exclusively on this sodium hydroxide based process, and estimated the cost of using this technology to be close to $600/t CO2.7 It is perhaps tempting to summarize the APS report as concluding that air capture is economically infeasible. This summary, however, is incorrect; the report showed only that a particular process for air capture based on sodium hydroxide is highly unattractive. Similar conclusions were drawn by Gebald et Received: Revised: Accepted: Published: 8631

March 15, 2012 May 22, 2012 June 1, 2012 June 1, 2012 dx.doi.org/10.1021/ie300691c | Ind. Eng. Chem. Res. 2012, 51, 8631−8645

Industrial & Engineering Chemistry Research

Article

al.21 An alternative to the sodium hydroxide process is to move away from solution processes and focus on adsorption of CO2 from air by using solid sorbents. Any adsorbent for which the heat of adsorption of CO2 is moderate compared to the heat of carbonate formation in CaCO3 will sidestep the severe energy penalty associated with calcination in the sodium hydroxide process. There are compelling reasons to consider whether other processes exist that have more reasonable performance. Based on an empirical analysis of processes similar to air capture, a recent report by House et al.5 concluded that the cost of air capture systems exceeds ∼$1000/t CO2. The authors analyzed three large-scale trace gas removal systems and developed a Sherwood plot, which was then used to calculate the cost of air capture based on ambient CO2 concentration.5 As we will show below, we feel that constraints on the processes considered by House et al.5 led to cost estimates that are considerably too high. We reach this conclusion by analyzing specific adsorption-based processes based on known materials used in configurations that are experimentally achievable. The idea of using an adsorption process for capture of CO2 from air was first suggested by Lackner and Brennan22 in 2009. Their adsorbent was an ion-exchange resin functionalized with quaternary ammonium ligands. The uptake of CO2 depended strongly on the partial pressure of water in air because of the possibility of forming both carbonate and bicarbonate products. A process was suggested based on humidity swing where CO2 adsorption is carried out in nearly dry conditions and recovery is accomplished by exposing the system to moisture. A limitation of this initial report was that the details of an actual process or a device that would be used were not described by the authors. The possible drawbacks of a humidity swing process are the necessity of operation in an arid climate and the low adsorption capacity of the adsorbent. In a more recent work, Wang et al.23 have reported the CO2 adsorption isotherms in dry and humid conditions for the ion-exchange resin at different temperatures. They also hypothesized a mechanism for the uptake of CO2 based on absorption by the mobile hydroxide and carbonate counterions present in the system. Other adsorbents have also been demonstrated that are candidates for use in adsorption-based air capture. One example of such an adsorbent is the hyperbranched amino silica materials described by Jones and co-workers.24,25 In 2010, another class of amine functionalized silica, denoted TRI-PE-MCM-41, was reported to have a large selectivity for CO2 over N2 at air capture conditions.26 The heat of adsorption for CO2 ranges between 50 and 118 kJ/mol for these amine functionalized materials.21 Previous research relevant to adsorption-based air capture has primarily focused on increasing the amine loading of the adsorbent to improve CO2 adsorption capacity under ultradilute conditions.26−30 These and other studies have reported high initial uptake rates for CO2,28,31 enhancement of CO2 adsorption at low values of relative humidity,26−28,32−34 and good stability of the adsorbent after multiple regeneration cycles.21,28,33,35,36 Other amine-based sorbents have also been reported with significant CO2 uptake at concentrations relevant to ambient air.37 Wurzbacher et al.35 have demonstrated lab scale extraction of CO2 from dry and humid air by using a temperature vacuum swing adsorption process. A packed bed of diamine-functionalized silica gel beads was used to study the effect of parameters such as adsorption time, desorption temperature, and desorption pressure on the cyclic adsorption capacity. The authors reported

extraction of CO2 with purities >97% at an energy requirement of 440 kJ/mol CO2 for desorption at 95 °C. They also predicted the energy requirements to decrease to 166 kJ/mol CO2 if the working capacity can be increased from 0.2 to 2 mmol/g adsorbent. Recently, the same group has reported a higher cyclic capacity of 0.69 mmol CO2/g using an amine-functionalized nanofibrillated cellulose as the adsorbent.21 They employed 2 h adsorption/1 h desorption cycles and found the adsorbent to be stable for multiple cycles in the presence of humidity. Stuckert et al.38 have compared zeolites and amine-grafted silica of air capture. Using a temperature vacuum swing adsorption cycle, purities >90% were obtained during desorption. They conclude that zeolites have faster uptake rates that amine functionalized SBA-15, but are not suitable for operation under humid conditions.38 Although the studies listed above have reported useful physical data for considering adsorbents for air capture, no study has yet examined a process-level description of using these materials. Modeling of air capture at this level is critical to determining whether specific processes are worth additional development and identifying the factors that limit the performance of a material or process. The key contribution of the APS report7 on the sodium hydroxide process was to approach air capture from this process level. The aim of this paper is to describe process-level models of adsorption-based air capture of CO2 with a specific adsorbent, TRI-PE-MCM-41. We examine a number of related processes based on temperature swing adsorption (TSA), using product purity and throughput as a metric for comparison. For a specific contactor configuration, we calculate the mechanical and thermal energy requirement for each individual step of the air capture process and suggest possible strategies for reducing the energy requirements. We examine what kinds of cyclic processes can be used to enhance the purity and volume of CO2 that is captured, and provide initial operating cost estimates for these processes. Our results show that the estimated operating cost of capturing CO2 from air using adsorption processes to be ∼$100/t CO2 in terms of net CO2 capture. This preliminary study indicates that more detailed analysis of the technological and economic feasibility of adsorption-based air capture processes is worthwhile. The paper is organized as follows. The process model and the assumptions involved are described in Section 2, which is followed by the methodology used for estimating energy requirements in Section 3. Section 4 presents the results of the model and the economic analysis, and suggests potential avenues to reduce the energy and monetary cost of adsorption-based air capture.

2. MODEL DESCRIPTION 2.1. Temperature Driving Forces. One approach we examine below to reduce the external energy required for air capture is to use ambient diurnal heating and cooling as a driving force for TSA. To examine the effect of this driving force at a range of locations, six US locations were chosen. In most cases, these locations were chosen based on their proximity to ongoing geologic sequestration experiments.39 The locations and a summary of the temperature data for each location from the National Climatic Data Center (NCDC)40 is presented in Table S1 (Supporting Information). The performance of an air capture technology in such a scenario will depend on the climate at the given location, particularly the temperature and relative humidity. For example, the temperature driving force available is smaller at a humid coastal location than an arid desert, which 8632

dx.doi.org/10.1021/ie300691c | Ind. Eng. Chem. Res. 2012, 51, 8631−8645


Article

previous gas phase is replaced by the fresh ambient air at the low temperature and the above steps are repeated. Because the system is allowed to come to equilibrium with air at the start of every cycle (Step 1), the product of a given cycle is independent of the previous cycle. This process is also considered when steam is available to control the desorption temperature, T2. When steam-assisted desorption is considered, we assume that saturated low pressure steam at 1.4 bar and 110 °C is available. The energy released by the condensation of steam will cause the temperature of the monolith and the adsorbent to increase. Depending on the amount of steam used, the heat transfer to the monolith will also cause some cooling of the condensed steam and reduce the available temperature driving force. To account for this effect, the desorption temperature, T2, was fixed at 90 °C for our calculations. Process B. A possible drawback of process A is that a significant amount of CO2 may be left in the adsorbent after equilibration in Step 3, as only a part of the gas phase is withdrawn as the product without affecting the adsorbed phase. This drawback can be overcome by using a purge gas as a carrier for desorbing CO2 (and N2). In process B, we assume low pressure steam is available as a purge gas that provides both a thermal and a concentration driving force. During the flow of steam, a low CO2 partial pressure is maintained in the gas phase, which leads to continued CO2 desorption. We will show later that the performance of this process is significantly better than that of process A. A schematic of process B is presented in Figure S2 (Supporting Information). The first two steps of this process are exactly the same as process A. The remaining steps are described below. Step 3: When a fraction of the gas phase has been removed as waste at the lower temperature, steam is introduced to desorb the CO2 and N2. It is assumed that steam at 110 °C is passed through the adsorbent continuously until a fixed fraction (usually 90%) of the adsorbed phase components is recovered as product. The continuous flow of steam results in almost zero CO2 partial pressure in the gas phase, allowing for high recoveries. It may be possible that small amounts of CO2 (close to those of ambient air) are present in steam, depending on its source. Contrary to the earlier process A, the continuous flow of steam ensures that the CO2 partial pressure at the surface does not increase because of desorption, but instead remains constant at the low inlet concentration. Step 4: The water gas and product mixture is cooled sufficiently to condense out the water and the process is then repeated. After the separation of water and CO2 at ambient pressures, purified CO2 is obtained as the product from the process. This gaseous product will be saturated with water vapor, as is the case for CO2 exiting most large-scale CO2 capture technologies such as aqueous amine absorption. 2.3. Contactor. To make even approximate estimates of the cost for operating the processes of interest, we must specify the form of the contactor containing the adsorbent. One approach is to use monoliths consisting of numerous parallel channels that are coated with the adsorbent.42 The length of the monolith must be sufficiently long to adsorb a reasonable fraction of CO2 from air. On the other hand, the mechanical costs associated with flowing significant quantities of air require that the pressure drop across the contactor be as low as possible. On the basis of a preliminary analysis (see the Supporting Information) of various contactor parameters (coating thickness, channel diameter, monolith wall thickness and contractor length), a monolith configuration was chosen and the parameters are listed in Table

may lead to a significantly different performance of the same technology at these two locations. By choosing six locations from different parts of the US, we have attempted to address a part of the overall effect of climate. We have accounted for differences in the process performance for various locations and also for the changes during a year for the same location. A second temperature driving force we examine below is to use steam to provide heat during desorption. A significant amount of low pressure steam at 105−120 °C is available as waste heat from chemical and manufacturing facilities and can potentially be utilized for useful applications.3,41 We therefore considered TSA cycles that use this low-grade steam during desorption. In both this kind of cycle and processes that involve only diurnal heating, it is possible to develop TSA processes that incur no financial costs for the temperature driving force. In estimating operating costs for TSA processes using steam desorption below, we compare scenarios in which steam is available for free from waste heat sources and where steam must be generated specifically for the air capture process. The calculations for different locations were performed for an entire calendar year and the results were scaled by using 330 operating days/year as a basis. The daily CO2 throughput was obtained by considering 365 days in a year. 2.2. Adsorption Processes. All the adsorption processes considered below are modifications of a simple equilibrium based TSA cycle. Briefly, adsorption of ambient air occurs at a lower temperature, T1, and ambient pressure and desorption and product recovery occurs at a higher temperature, T2, and a slightly lower pressure. As the heat of adsorption and the uptake of CO2 are higher than that of N2, the product contains more CO2 than ambient air. A dominant factor in the performance of the cycle is the magnitude of the temperature swing, T2 − T1. Higher temperature swings yield better performance both in terms of CO2 purity in the product and the total moles of product. Below, we consider two distinct cycles denoted processes A and B. Process A. Figure S1 in the Supporting Information shows a schematic of process A, which relies on diurnal temperature variation as a driving force. The process is initiated with the system at equilibrium at the lowest temperature of the first day (January 1 unless otherwise mentioned). This is shown as stage 0 at the start of the cycle, where the system consists of a solid adsorbent phase that is in equilibrium with the gas phase at temperature Teq. The process then proceeds through the following steps. Step 1: The gas phase from the previous stage is discarded and fresh air at temperature T1 is flowed through the contactor (using wind or fans, if required) until the adsorbent is in equilibrium with air and the contactor is then closed. Step 2: A predetermined fraction (90% unless otherwise specified) of the gas phase, α, is removed while the temperature is kept constant at T1. We assume that the gas phase is removed rapidly such that the composition of the adsorbed phase does not change. The adsorbent and remaining gas phase now contain more CO2 than ambient air. Step 3: After the ambient temperature increases to T2, another prespecified fraction (90% unless otherwise specified) of the gas phase, β, is withdrawn as the product while the adsorbed phase remains unchanged. The equilibration and product withdrawal steps can be repeated to increase the product yield as explained later. Step 4: The system is allowed to equilibrate until the low temperature of the next day is achieved. For the next cycle, the 8633



Article

has been reported for fused silica impregnated with polyethylenimine at air capture conditions.29 For process A using steam desorption, it is assumed that only the thermal energy of steam is transferred into the adsorbent monolith and there is no direct contact of the adsorbent with the flowing steam. When steam is used as a purge in process B, we assume that the presence of water does not have an effect on the adsorption isotherm of CO2. In this case, steam maintains a low partial pressure in the gas phase during desorption and provides a thermal driving force for desorption. When steam comes in contact with the monolith, the heat of condensation of steam is utilized for desorption of the adsorbates. A portion of the available thermal energy is also used to raise the temperature of the adsorbent and the monolith to the desorption temperature. Additionally, stability of functionalized silica adsorbents in steam has been shown to depend on the nature of the silica support and the operating conditions.49 Our work implicitly assumes that the presence of low pressure steam does not cause significant degradation of the adsorbent. As the processes we consider use ambient temperatures as all or part of the driving force, the cycle half times of the individual cycles are of the order of a few hours. On this time scale, heat and mass transfer restrictions are assumed to be negligible. We assume that the time scale for attaining local equilibrium is much smaller than the half-cycle time of the adsorption cycle. This appears to be reasonable when compared to earlier studies that have shown that the adsorption of CO2 on amines is relatively fast, especially in the presence of water.27,29,31,47 Similarly, we assumed that the gas phase can be removed rapidly such that the composition of the adsorbed phase remains unaffected. These assumptions imply that three distinct time scales exist, namely (in increasing order), withdrawal of the gas phase by external pumps, achievement of local equilibrium with the gas phase, and the halfcycle time of the process.

1. We assume cordierite as the material of construction, as it is a common choice for ceramic monoliths.40,43 Our choice of Table 1. Details of the Monolith Structure Used in This Study parameters 3

physical dimensions (m ) channel density (cells per in.2) channel internal dimensions (mm) density of monolith material (cordierite), ρm (kg/m3) monolith wall thicknessa (mils) internal surface area (m2/m3) adsorbent coating thickness (μm) adsorbent density, ρa (kg/m3) adsorbent loading (kg/m3) empty volume (%) pressure drop (Pa) Ma/Mma

value 2.24 × 2.24 × 0.5 100 2.47 × 2.47 2600 2.86 1530 100 1000 146.8 79.6 100 1

a

Chosen to ensure equal mass of the adsorbent and bulk monolith. Ma: mass of adsorbent. Mm: mass of monolith.

contactor configuration is practically realizable as cordierite monoliths with similar wall thicknesses (2 mils/3 mils) and higher cell density (900/600 cpsi) are commercially available.40 The surface area and volumes associated with a single monolith block40 are too small to have a significant removal rate of CO2 and larger units comprised of many individual monoliths will be necessary for practical applications. For ease of manufacturing and transportation, we assume that the dimensions of a commercial air capture unit will be smaller than those of a standard shipping container (2.4 × 2.4 × 12.0 m3). Specifically, each unit is designed to consist of 20 contactormodules, each 2.24 × 2.24 × 0.5 m3 with a combined volume of 50 m3 per unit. Each module has a length of 0.5 m and an exposed surface area of 5 m2, and the flow of air occurs over the smallest dimension of the module. All the results presented later are with respect to a 50 m3 air capture unit. 2.4. Adsorbent. Due to the dilute concentrations of CO2 in air, an ideal adsorbent needs to have high adsorption capacity and high selectivity for CO2 at ambient conditions. The adsorbent used in our calculations in an amine functionalized silica, TRIPE-MCM-41, developed by Sayari and co-workers.44,45 The adsorption of CO2 is explained by two independent adsorption mechanismschemisorption by the amine groups and physisorption by the bare silica surface.46 The experimental adsorption isotherm for CO246 and N247 is modeled by the Toth equation, eq S1−S5 (Supporting Information), with the parameters listed in Table S4 (Supporting Information). 2.5. Assumptions. To make our calculations feasible using physical data that are currently available, a number of assumptions are required. An important assumption is that the air entering the process contains only N2 and CO2 under dry conditions. This assumption is driven by the lack of available isotherm data for humid gases for amine-functionalized silica adsorbents. We are not advocating that air be dried before entering an air capture unit, as this would likely be economically infeasible and technically disadvantageous. Nevertheless, studying models based on dry air is a reasonable initial approach because the presence of water vapor typically enhances the CO2 selectivity and capacity of amine-functionalized adsorbents.26,27,32−35,48 For TRI-PE-MCM-41, the CO2 uptake has been shown to be higher in the presence of humidity for 5% CO2/N2 mixtures.48 More recently, enhancement of CO2 uptake

3. ESTIMATION OF ENERGY REQUIREMENTS To judge how viable the above processes would be in relation to other approaches, it is critical to examine the costs associated with their operation. In this section, we outline the methodology used in estimating the energy requirements for these processes. These estimates are not intended to provide the fidelity that could be achieved in a detailed economic analysis; rather, they provide guidelines for deciding whether future detailed analysis is warranted. The concentration of CO2 in air is ∼400 ppm, which is equivalent to 16.6 mmol CO2/m3 of air. Because of this dilute concentration, a large quantity of air must be flowed through the contactor to adsorb a significant amount of CO2. The pressure drop across a contactor depends on the average velocity of air and on the physical parameters of the contactor. Higher velocities lead to a higher gas side mass transfer coefficient but a smaller residence time for the monolith, changing the equilibration time during adsorption. Higher velocities also cause a larger pressure drop and increase the energy requirement for blowers. Accurate calculation of the equilibration time and associated breakthrough curves requires models of significant complexity. These calculations are well established but a significant hurdle in performing calculations of this kind is the lack of experimental data for various rate processes. Instead of attempting detailed calculations of this type, we use the approach suggested by the recent APS report.7 The authors defined a characteristic length of the adsorbent, L0, which defines the capture fraction, αcapture, for the system. The capture fraction 8634



Article

⎡ P ⎛ P ⎞⎤ P W = −PextV ⎢ 1 − 2 + ln⎜ 2 ⎟⎥ ⎢⎣ Pext Pext ⎝ P1 ⎠⎥⎦

is defined as the fraction of CO2 that is adsorbed by the contactor from air during the adsorption stage. The characteristic length is defined as 2

L0 = ηUd /D

(1)

αcapture = 1 − exp( −L /L0)

(2)

The energy costs described above yield a CO2-enriched product exiting the process at a relatively low pressure. Finally, it is useful to also consider the additional compression cost that would be associated with compressing the product to CO2 pipeline ready pressures (specifically, 14 MPa). The thermodynamic analysis presented by House et al.51 was used to estimate this contribution. Considered together, the mechanical energy calculations include the energy required for flowing air through the monolith (Ef), for removal of waste and products by application of partial vacuum (Ev) and for compression of the product stream (Ec). For all mechanical energy requirements (blower, vacuum pump, and compressor), an overall efficiency of 80% was assumed. The calculation of the thermal energy component is more straightforward. The minimum energy required for regeneration of the adsorbent is calculated by using the heat of adsorption of CO2 and N2. This quantity depends only on the adsorbent characteristics and not on the capture process. For the diurnal process, the thermal energy for desorption is provided by the ambient surroundings and does not require external energy inputs. The actual thermal energy for desorption of CO2 will depend on the relative contribution of physisorption and chemisorption component. An upper bound for thermal energy is obtained by using heat of chemisorption of CO2. If pure CO2 is obtained as the product, the minimum energy required for desorption is ∼1530 MJ/t CO2 for the chosen adsorbent. This by itself is higher than the reported thermal energy input of ∼1100 MJ/t CO2 for the humidity swing air capture process of Lackner et al.22 When steam desorption is used, it is unlikely that all the steam supplied will be used only for desorption. Though it is undesirable, a part of this energy will be used for the sensible heating requirements of the bulk monolith and will depend on the material of construction employed. The energy associated with the sensible heat requirements for the adsorbent, Eta, and the monolith, Etm, can then be calculated as,

where U is the average velocity in the monolith channels, L is the length over which the air flow occurs, D is the gas phase diffusivity of CO2, and η is a parameter whose value depends the rate of removal of CO2 at the surface. For the ideal case of instantaneous adsorption at the wall, η is 0.068.7 Allowing for the possible limitations associated with the dynamics of the system, we assumed a value of η that is 25 times larger than the ideal case. The parameters used in the energy calculations are listed in Table 2. Table 2. Physical Parameters of the System Used in This Study

a

parameters

value

kinematic viscosity, νa (m2/s) diffusivity, D (m2/s)a equivalent diameter, db (mm) density of air, ρa (kg/m3) heat capacity of adsorbent, Cpa (J/g °C) heat capacity of monolith material, Cpm (J/g °C) ηa Schmidt number, Sc

1.5 × 10−5 2.1 × 10−5 2.27 1.2 1 1.4 1.7 0.71

Values obtained from ref 7. bFor 100 μm adsorbent coating.

The mechanical energy for air flow depends directly on the pressure drop across the system. The velocity through each monolith channel was calculated by assuming laminar flow as50 U=

Pdropd 2 32ρLν

(4)

(3)

where Pdrop is the pressure drop across the length L. The power required is the product of the volumetric flow rate and the pressure drop. The total energy required for air flow will depend on the duration of time for which air flows through each contactor, tads. This time is estimated by calculating the average CO2 uptake required by the adsorbent to achieve equilibrium with air for every adsorption−desorption cycle. The second component of mechanical energy that needs to be included is the electrical energy required for withdrawal of gases from the monolith with use of a vacuum pump. In process A we assumed that 90% of the gas phase of the monolith is evacuated, in step 2 as waste and in step 3 as the product, by application of a partial vacuum. A bound can be placed on the cost of the partial vacuum based on the energy required for compression of gas from the vacuum to ambient pressure. The minimal amount of energy is required when an isothermal reversible process is used and is given by W = nRT ln(P1/P2), where P1 and P2 are the initial and final pressures, respectively. Because of the time scales associated with the vacuum pump, we assumed an irreversible process for our calculations. The energy required for reducing the pressure inside the monolith from P1 to P2 and irreversibly compressing the withdrawn gases to fixed external pressure, Pext (1 bar), is described in the Supporting Information and is given by

Eta = MaCpaΔT

(5)

and

Etm = M mCpmΔT

(6)

where Ma and Mm are the masses and Cpa and Cpm are the specific heats of the adsorbent and the monolith, respectively. Unlike the thermal energy for desorption, the sensible heat requirements for the monolith represent a parasitic loss, as they do not scale favorably with the throughput of CO2 obtained as the product. During the regeneration step of process B, it is not necessary to raise the temperature of the entire monolith to the desorption temperature; the only requirement for steam flow being that a sufficient amount of CO2 has been desorbed. Practically, the amount of the thermal energy utilized as sensible heat will depend on the relative rate of desorption of CO2 and the thermal conductivity of the monolith material. By considering the entire mass of the monolith for the sensible heat, we obtain an upper bound for the parasitic losses associated with the process. Our energy calculations assume that the pressure at the inlet of the process is close to ambient conditions, implying that the energy required for the pressurization of air is smaller than Ef. For processes that require higher inlet pressures for operation, the 8635



Article

scenario examined, the purity of CO2 obtained at the end of the process and the total amount of product gas obtained per unit (50 m3 volume) per year (tons CO2 unit−1 year−1) were used as metrics for the process performance. 4.1. Process A. The performance of process A at various locations is shown in Table 3 for the contactor configuration defined earlier. Using the diurnal temperature swing, only about 900−1000 kg of the gas product is obtained at purities of less than 0.5% for all the location. This corresponds to only 2−4 kg/ annually of CO2 captured every 50 m3 volume of the air capture unit. There is some correlation between the purity and the quantity of the product stream. To determine the effect of temperature on the performance of the process, similar calculations were performed at fixed maximum and minimum temperatures. As expected, higher temperature differences resulted in better performance in terms of both purity and product yield. This trend can also be seen in Table 3 as the location in Texas has the highest average daily temperature difference. The performance of the process can be increased further by using an external source of energy such as low pressure steam. We assume that low quality steam at 110 °C is available and that a cross-flow configuration is used where there is no direct contact between steam and the adsorbent. Allowing for some cooling of condensed water due to transfer of heat to the monolith, a lower temperature of 90 °C is used for desorption. Not surprisingly, the performance of this process is greatly improved relative to that of the pure diurnal cycles, with annual product purities of 9−20% and capture of 1−3 t CO2 unit−1 year−1. Because larger temperature swings give improved performance, the location with the lowest average daily temperatures (UT) has the best performance. The mean working capacity at the UT location is 3.2 × 10−4 mmol/g for diurnal temperature swing and 0.03 mmol/g for steam desorption. Such low working capacities result in low throughputs for the process. Using equations described in Section 3, a pressure drop of 100 Pa results in a velocity of 1.78 m/s and an adsorption time of the order of a few minutes. The low yields can be improved by repeating the adsorption and steam desorption step multiple times a day. To estimate the outcome from this approach, the daily maximum and minimum temperature data were linearly interpolated to obtain approximate hourly temperature data for each location. Calculations were performed for 1, 2, 3, 4, 6, 8, and 12 cycles per day. For instance, when 12 cycles are use on a daily basis at UT, the product throughput can be increased to 4.8 t CO2/unit/year. Due to the low adsorption time, the number of possible adsorption−desorption cycles per day will be limited by the rate of heating and cooling the contactor.

energy required for air pressurization may be significant. Process B involves continuous flow of steam through the monolith, causing some water to remain condensed in the channels. In reality, the cost of product recovery will also involve the mechanical energy associated with the flow of this two-phase mixture of steam and condensed water. This aspect has not been included in the energy estimates that follow. We have not accounted for energy requirements or possibilities of heat recovery in the separation of the resulting CO2−steam mixture via condensation, under the assumption that this process will typically be straightforward by using cooling from ambient conditions. Because of these assumptions, our energy estimates for the total energy requirements of the process are lower bounds. For each of the scenarios examined, it is useful to compare the energy required for the real process to the theoretical minimum work required for the same separation by a completely reversible process. The process outlined in Figure 1 is an isobaric and

Figure 1. Schematic of the air capture process used for the calculation of minimum work.

isothermal separation of ambient air into two streams, with stream 3 being the purified product. Assuming ideal gas mixtures, the minimum work required per mole of CO2 captured is ⎡n Wmin = − RT ⎢ 1 (y1CO2 ln y1CO2 + y1N2 ln y1N2 ) ⎣ n3 n2 CO2 − (y2 ln y2CO2 + y2 N2 ln y2 N2 ) n3 ⎤ − (y3CO2 ln y3CO2 + y3 N2 ln y3 N2 )⎥ ⎦

(7)

where ni is the total number of moles and yi the compositions of the various streams. For a given temperature, the minimum work is only a function of the purity of the product obtained, y3CO2 and the process throughput, n3/n1.

4. RESULTS Initially, calculations were performed for a single contactormodule with the adsorbent coating thickness of 100 μm. In each

Table 3. Process A Performance Using Diurnal Desorption and Steam Desorption at Different Locations diurnal desorptiona

steam desorptionb c

location

purity CO2 (%)

CO2 throughput

product throughput

Tdiff (°C)

purity CO2 (%)

CO2 throughput

product throughput

FL AL UT GA WA TX

0.28 0.34 0.33 0.32 0.28 0.40

2.3 3.1 3.4 2.9 2.8 3.8

818 911 1019 922 978 948

11.1 12.9 13.8 12.4 11.6 15.1

9.5 14.4 19.9 15.0 18.2 15.7

0.99 1.83 3.13 1.95 2.70 2.10

11 13 16 13 15 13

a

Throughput units for diurnal desorption in kg/unit/year. bThroughput units for steam desorption in t/unit/year. cTdiff is the annual average difference between the maximum and the minimum temperature at a given location. 8636



Article

kgCO2 are predicted for one desorption step per day. Comparison of these results to those for process A shows that the flow of steam and the resulting concentration driving forces cause dramatic improvements in the process. This is a result of continuous removal of desorbed CO2 from the system by the purge (steam), which drives further desorption of CO2 from the adsorbent into the gas phase. For a pressure drop of 100 Pa across the contactor, the characteristic length, L0, is calculated as 74.2 cm. Using eq 2, this gives an overall capture fraction of 49%. Table S6 (Supporting Information) shows that the average adsorption time, tads, is less than 2 h for the six chosen locations. As earlier, the performance of the process can be improved by repeating both adsorption and desorption steps, multiple times a day. For air capture conditions, faster desorption half cycles have been reported for the TSA process with use of air at 85 °C as the purge,29 though the oxidative stability of materials at these conditions is unclear.52 The combination of both temperature and concentration driving force for desorption would ensure that the time required for the desorption step will be significantly lower than tads. Thus, the minimum cycle time used is set at 4 h (i.e., six cycles per day) to ensure that the total time for each case is at least two times tads. Table S7 (Supporting Information) shows the results of multiple cycle calculations for UT. For each case, the product purity remains almost constant at approximately 88.5%. The total amount of CO2 recovered, however, increases almost linearly with the number of daily cycles. Allowing for considerable margin of error in the calculation of adsorption times and possibly slower uptake rates at lower temperatures, a total cycle time of 6 h is used. For four cycles per day, this version of the process produces >400 t CO2 per unit per year with an average purity of 88.5%. We now consider the energy associated with operating process B. When four cycles per day are used, 1.1 t CO2/day is obtained at a purity of 88.5%. The energy requirements for this case are shown in Figure 3. The total energy required is 6328 MJ/t CO2

The energy requirements of process A are presented in Table S5 (Supporting Information). Although the idea of diurnal desorption is superficially attractive, electrical energy needs to be supplied externally. Our calculations show that these external energy requirements are very high. When using steam for desorption, for example, the energy required for flow of air and for desorption is 0.5 MJ/t CO2 and 3.1 GJ/t CO2, respectively. The sensible heat requirements for the adsorbent and monolith with this approach are estimated to be 80.9 GJ/t CO2 and 71.1 GJ/t CO2, respectively, and these quantities do not scale with the quantity of CO2 obtained. These numbers are an order of magnitude higher than the energy obtained from combustion of natural gas (9 GJ/t CO2).5 Implementation of a process that requires such high energy requirements will not result in a net capture of CO2. Thus, this process is rejected and further analysis is not discussed. 4.2. Process B. We now turn to process B (shown in Figure S2, Supporting Information), where we will show that significantly improved performance is possible. As described above, the key difference between process B and process A is the removal of desorbed CO2 (and N2) by a purge stream of steam that prevents the increase of the CO2 partial pressure in the gas phase. That is, in process B, desorption of CO2 does not occur in a closed constant volume system, which prevents the attainment of equilibrium during the desorption step. Figure 2a shows the annual variations of temperature for two locations, FL and UT. Calculations assessing process B were

Figure 2. (a) Variations of ambient temperature and (b) working capacity for UT (black) and FL (red). The desorption temperature is fixed at 90 °C for all locations.

performed for each day, with adsorption being carried out at the lower temperature and desorption using purge steam. Figure 2b shows a strong correlation between the adsorption temperature and the working capacity of the process. The average working capacity, 0.87 mmol/g for FL to 1.12 mmol/g for UT, is far higher than in process A. The better performance in UT is because the average adsorption temperature at UT (5.5 °C) is almost 10 °C lower than that at FL (16.3 °C). This also implies that the uptake rate of CO2 may be slower in UT than FL. Table S6 (Supporting Information) shows the average purity and throughputs obtained for each location when process B is used. Average purities of ∼88% and daily recoveries of 250−320

Figure 3. Energy requirements of process B (four cycles per day) at the UT location. Energy required for air flow, Ef, is calculated for a pressure drop of 100 Pa across the contactor.

and is dominated by the thermal energy requirements (Et + Eta + Etm) of the system. As noted earlier, this thermal energy is provided by the condensation of steam inside the monolith channels. The energy required for desorption, Et, is 1580 MJ/t CO2, which compares favorably with the theoretical minimum for 100% CO2. This energy scales with the amount of CO2 8637



Article

presented above, four cycles per day using process B in UT, is used as a reference scenario for the remainder of this work. Monetary Cost. In the previous section we have described a plausible process that can capture ∼1.2 tons of CO2/day at 88.5% purity, and requires 5962 MJ/t CO2 from steam and 784 MJ/t CO2 of mechanical energy from electricity. Depending on the energy source used, we must account for CO2 that would have been emitted back to the atmosphere for powering the process. In this section we aim to estimate the monetary cost for net CO2 capture. We only estimate operational costs, and have not attempted to estimate capital expenditures and maintenance costs. Instead of using an average cost of electric power, we follow a methodology similar to House et al.5 In their analysis, the authors calculated the cost of air capture for various sources of electricity, accounting for differences in the cost and CO2 emission factors among these sources. Assuming electricity as the only source of energy, the authors predicted air capture costs >$1000/t CO2 without the energy required for compression of capture CO2 to pipeline specifications. For process B, the total electricity required is only 5.4% of the total energy requirements. When the energy for compression is included, this number increases to a little over 11%. The US Energy Information Administration has reported the CO2 emission factor for electric power generation of 0.676 t CO2/ MWh.53 The same study suggested CO2 emission factors to be used for steam generation of 88.18 kg CO2/MMBtu,54 and for calculation of avoided emissions from energy sales as 79.71 kg CO2/MMBtu.54 These data are summarized in Table 4 using

desorbed and can (in principle) be reduced on a per t CO2 basis by using an adsorbent with a lower heat of adsorption if other properties of the adsorbent are unchanged. For the chosen contactor configuration, 40.4% (28.9%) of the total energy is associated with the sensible energy requirement of the monolith (adsorbent). Together, these parasitic losses are responsible for approximately 70% of the total energy requirements. In contrast, the electrical energy associated with flowing air through the contactor is ∼5.4% of the total energy at a pressure drop of 100 Pa. Thus, even though the low concentration of CO2 in air requires the flow of a considerable amount of air through the contactor, the energy required for this does not necessarily dominate the energy cost. Instead, the most critical energy cost is incurred by regeneration of the adsorbent, specifically in cyclic heating of the bulk monolith that does not directly contribute to the CO2 throughput. The sensitivity of the various process parameters and possible options for reducing the parasitic losses are explored later in Section 4.3. At this stage it is useful to consider the second law efficiency of the overall process. From eq 7, the minimum work calculated for an air capture process with 49% CO2 capture and final product purity of 88.5% (at 1 bar) is 466 MJ/t CO2 (or 20.5 kJ/molCO2). When compared to the total energy requirement, the overall second law efficiency of the process is 7.4%. Our analysis of energy requirement is based on a simplistic model and gives a reasonable lower bound for the process; a detailed analysis will predict higher energy requirements and lower second law efficiencies. Nonetheless, we will show later in Section 4.3 that it is possible to improve the thermodynamic efficiency of the process (∼11%) by using a better contactor configuration and higher working capacities. It is important to compare our results with the analysis of House et al.,5 which considered design of air capture processes based on existing gas purification methods. House et al. predicted the efficiency of air capture to be less than 5%. The thermodynamics and the optimal second law efficiency for the industrial processes considered by the authors (flue gas capture, N2/O2 separation, ethanol distillation) are constrained by downstream process requirements such as flow rate, conversion, removal fraction, etc. Such constraints are not directly applicable to air capture, so the optimal second law efficiencies are not determined by identical process economics. Our thermodynamic analysis is based in part on the observation that adsorption is a spontaneous process (it requires no energy input). After this spontaneous adsorption step, the system is no longer dilute with regard to CO2, and a higher thermodynamic efficiency for the energy intensive desorption step is possible than is seen in the analysis of House et al. Both these factors are addressed again in Section 5. As the throughput obtained by this process is of the order of hundreds of kilograms of CO2 per day per unit, it is useful calculating the energy, Ec, required for compressing the product CO2 to sequestration ready pressures. A compression energy requirement of ∼13 kJ/mol gas,51 is used for this calculation. Using the value of purity obtained earlier and assuming no further concentration of the CO2 stream, the energy for compression, 417 MJ/t CO2, is comparable with the energy for air flow (344 MJ/t CO2) and represents ∼55% of the overall electricity cost. It is important to note that compression costs are also present in any process that removes CO2 from concentrated point sources such as power plant flue gases. The overall energy estimates for operating process B for various locations are presented in Table S8 (Supporting Information). The analysis

Table 4. Comparison of Emission Factors for Steam and Electricity Generation type of energy

reported valuea

steam generation

88.1854

avoided emissions from sale of steam electric power generation (US average)

79.7154

a

0.67653

unit kg CO2/ MMBtu kg CO2/ MMBtu t CO2/MWh

value (MJ/t CO2) 11964.8 13236.2 5325.4

Valid for reporting years 2003 and later.

consistent units. A key observation here is that for every ton of CO2 released, twice as much energy can be obtained from using steam than from using electricity. Equivalently, for the same amount of energy used, the emissions caused by using steam are almost a factor of 2 lower than those from using electricity. It also should be noted that the suggested emission factor for avoided emissions due to purchased steam is more favorable than that for steam generation. Table S9 (Supporting Information) summarizes the dollar cost and the CO2 intensities for different sources of electricity analyzed by House et al.5 and includes corresponding values for using steam. The monetary costs for electricity have been updated from a recent report by the US Energy Information Administration55 and differ slightly from those of House et al.5 An important observation is that the cost of steam, $15.2/ MWh,56 is significantly lower than that of electricity. We have considered four different cases for calculation of cost based on the availability of steam, which are described as Scenarios I−IV as follows. Scenario I (no steam): All the energy required for the process is provided by electricity. Even though it is impractical to assume 8638



Article

that the energy for desorption and sensible heating of the monolith is provided by electricity, this analysis is consistent with the method of House et al.5 In this case, we assume no loss of electrical power while supplying the thermal energy to the system. Scenario II (dedicated steam): This is a more realistic scenario in which all the steam required by the process is obtained by production of low-pressure (LP) steam dedicated for the air capture process. The cost of steam and the emission factor is obtained from Table S9 (Supporting Information). The energy for the blowers and pump is provided by electricity. Scenario III (purchased steam): In this scenario we assume that a source of LP steam is present in the vicinity of the air capture process and this steam is purchased and used for performing the air capture process. The cost of steam and the CO2 capture fractions are again considered in the calculation of net CO2 capture cost. Scenario IV (waste steam): This scenario is very similar to Scenario III in the sense that all the steam requirements of the process are provided by the same nearby source. The difference is that we assume that unless the air capture plant were to use the LP steam, it would be entirely wasted at the source. Thus, because of the presence of the air capture plant all the CO2 emissions caused due to the generation of LP steam are avoided. For these calculations, the CO2 emission factor for steam was set to zero to replicate the avoided emissions. Realistically, the analysis of the actual process will be a combination of Scenarios II, III, and IV depending on the extent of availability of steam from other sources. The range of values defined by these scenarios should be a reasonable estimate of the monetary cost. The estimation of monetary costs for process B with four cycles per day in UT is shown in Table 5. The values for Scenario I are significantly higher than the other cases and range from $160 to $500/t CO2-captured. This is not surprising since (1) the CO2 emission factor is unfavorable for electricity relative to steam and (2) the cost of electricity is higher than that for steam. For Scenario I the optimum source of electricity becomes a tradeoff between having the most favorable emission factor and having the lowest cost per kWh. The minimum cost is obtained by using hydroelectric power. This analysis is consistent with previous work,5 but is unrealistic for processes based on adsorptionregeneration as it is undesirable to provide thermal energy via electricity. The estimated costs for Scenarios II and III are similar as they only differ in the value of CO2 emission factor used for steam. For most of the cases, the cost of the process when steam must be generated is ∼10% higher than that for Scenario III. Recalling that the net cost of capture is favored by low cost electricity and a favorable CO2 emission factor, two trends can be observed. First, carbon free sources of electricity such as solar and off-shore wind are more expensive ($130 to $170 MJ/t CO2, Scenario III) because of the high cost of generating this electricity. Similarly, cheap sources of electricity such as coal (without CCS) are more expensive (∼$140/t CO2) because of high CO2 emissions from these sources. Second, the minimum cost is obtained with electricity sources that have both zero CO2 emissions and a reasonable cost (advanced nuclear = $90.9, wind = $84.2, and hydro = $80.0/t CO2), or cheap conventional sources with lower emissions than coal such as natural gas (∼$85/t CO2). When compression costs are not included these costs reduce to $60 to $70/t CO2 for natural gas and $70 to $90/t CO2 for coal based electricity. Depending on the source of electricity, decreases of

Table 5. Cost of Reference Process B (Four Cycles Per Day at UT) for Different Scenarios monetary cost ($/t CO2-net)a electricity source coal conventional coal IGCC IGCC with CCS natural gas conventional combined cycle advanced combined cycle advanced CC with CCS other advanced nuclear wind windoffshore solar PV solar thermal biomass hydro

Scenario I: no steamc

Scenario II: dedicated steam

Scenario III: purchased steam

Scenario IV: waste steam

N/Ab

162.8

138.2

58.6

N/Ab 408.1

161.3 120.2

138.5 108.3

60.9 57.3

494.3

95.9

85.5

43.3

471.9

94.3

84.1

42.6

196.8

92.5

83.8

45.4

213.4

100.0

90.9

50.0

181.7 455.7 394.8 584.2 337.1 161.9

92.7 156.4 142.2 186.3 108.9 88.0

84.2 142.1 129.2 169.3 98.1 80.0

46.3 78.1 71.0 93.0 51.9 44.0

a Electricity requirement = 783.6 MJ/t CO2, steam requirement = 5961.6 MJ/t CO2. bThe net amount of CO2 captured is negative and results in net emission of CO2. cSimilar to the analysis of House et al.

25% to 45% are obtained when the cost of compression is not added. Predicted costs for other sources of electricity are reported in Table S10 (Supporting Information). For practical purposes, Scenario IV is exactly the same as Scenario III. The subtle difference between the two scenarios is the ownership of emitted CO2 released due to the generation of low pressure steam used in the process. Assuming that there is a source of LP steam that will otherwise be wasted, the values reported for Scenario IV should be interpreted as the monetary cost that will be incurred to prevent wasteful emission of a ton of CO2 to the atmosphere. Unlike the previous scenarios, this will depend strongly on the CO2 emission portfolio of the relevant industry as the responsibility of the relevant CO2 emission still lies with the industry. This cost is again the lowest for natural gas based sources and is ∼$40/t CO2 (Table 5). An interesting observation is associated with the implementation of carbon capture and sequestration (CCS) processes for conventional sources like coal and natural gas. When CCS is implemented for natural gas, the estimated price of electricity increases but the net cost for air capture decreases slightly (from $84.1 to $83.8/t CO2). For coal the effect is more dramatic as the cost reduces from ∼$138/t CO2 (for IGCC) to $108/t CO2 when CCS is implemented. In this case, implementation of CCS processes causes an increase in the electricity cost, but at the same time reduces the CO2 emissions from 0.9 to 0.2 t CO2e/ MWh. We emphasize that air capture processes are not intended to be an alternative for point source capture. Instead of being a competitive technology, air capture is intended toward being a complementary technology that captures CO2 that is inaccessible to other conventional methods. 8639



Article

At this stage it is useful to analyze the cost of air capture at the different locations listed in Table S1 (Supporting Information). Figure 4 compares the cost of air capture process for four cycles

Figure 5. Energy contributions (in %) for 1.5, 2.86, and 4 mil wall thickness. The total energy requirements are 5536, 6745, and 7774 MJ/t CO2, respectively.

correspond to adsorbent to monolith ratios of 1.9, 1.0, and 0.71 kg-adsorbent/kg-monolith, respectively. By reducing the wall thickness, the total energy decreases from 7774 MJ/t CO2 (4 mil) to 5536 MJ/t CO2 for 1.5 mil and the contribution of Etm decreases from ∼47% to less than 25%, respectively. The decrease in energy requirements affects the cost required for capture. Figure 6 shows the monetary cost for different

Figure 4. Cost of process B (Scenario II) for four cycles per day at different locations assuming natural gas advanced combined cycle as the source of electricity.

per day at different locations assuming generation of dedicated steam (Scenario II) and advanced combined cycle as the source of electricity. The cost in UT, $94.3/t CO2, is approximately 15% smaller than the cost in FL ($113.1/t CO2). As explained earlier, this is a direct result of higher working capacities in UT. For a commercialized air capture process, electricity will typically be obtained from the grid. It may appear that the discussion of the source of electrical power is moot as any operator would always pay a power cost that is weighted over all the sources of electricity generation. The usefulness of the above analysis lies in the fact that for all the possible sources of energy, the cost of air capture is not as high as previously expected.5 In this section, our analysis of a specific TSA process suggests the operating cost to be of the order ∼$80 to $150/t CO2 captured. It should be stressed that these estimates include only the energy of operation of each process; they do not include costs of capital expenditures for constructing and/or maintaining equipment. Keeping in mind the relatively high yields (∼1 t CO2/day) and purities (88.5%) possible, this process may form the basis for a useful approach to capturing CO2 from air. Efforts targeted toward establishing detailed process models and economic analyses for this process would be worthwhile. 4.3. Potential Improvements. From the analysis presented above, it is clear that the dominant contribution to the cost of adsorption-based air capture is associated with the sensible heat requirements of the monolith. For the reference case presented above (including compression), the thermal heat requirements are ∼38% for the monolith and ∼27% for the adsorbent. In this section we discuss two different strategies for reducing the overall cost of air capture. Of the different sources of electricity explored in the earlier section, IGCC with CCS (coal), advanced combined cycle (NG), and wind are used as examples (Scenario II). Monolith Configuration. The first approach is to reduce the mass of the monolith by keeping the amount of adsorbent constant and to change the wall thickness of the monolith, since this will directly decrease the sensible heat required for the process. Figure 5 shows the distribution of energy requirements for three different wall thicknesses, 1.5, 2.86, and 4 mil. These

Figure 6. Monetary cost of four cycles per day of process B at UT for a 1.5 mil (red), 2.86 mil (green), and 4 mil (blue) channel wall thickness of the contactor.

monolith configurations. For electricity generated from natural gas, the cost decreases from $94.3/t CO2 (reference case) to $66.2/t CO2 for the minimum thickness. For the more unfavorable scenario of a larger wall thickness, the cost increases to $132.0/t CO2. The trends for other sources of electricity are very similar, and are shown in Figure 6. These results show that there is a strong dependence on the cost of the process with the energy required for regeneration of the adsorbent. This depends on the parasitic losses incurred for the sensible heat requirements of the monolith. It is important to note that a minimum parasitic loss is determined by the structural strength of the material of construction rather than by thermodynamics. An ideal material of construction should have a low thermal mass without compromising on the structural strength required for manufacturing processes. Adsorbent Capacity. We have shown that the amount of product recovered in the process is not affected by the sensible heat requirements of the process. It is of course useful to 8640



Article

minimize the parasitic losses per unit of CO2 captured, and ideally a high ratio of the adsorbent to the monolith is desired. Practically, this ratio will be decided by the manufacturing limitations of the contactor and cannot be easily improved beyond a certain value. Another approach is to increase the throughput of CO2 obtained during each adsorption−desorption cycle. One way of achieving this is by increasing the adsorption capacity of CO2 at ambient conditions. Recently, various adsorbents with higher uptakes of CO2 than TRI-PE-MCM-41 have been reported.21,27−29 For the amine modified silica adsorbents of interest, this is possible by increasing the amine loading of the adsorbent.27,28 To explore the impact of increased adsorption capacity, we performed calculations with modifications of the adsorption isotherm used above. Specifically, the maximum chemisorption capacity of CO2, (ns0)chemi, was increased from 3.64 mmol/g in steps of 20%, which resulted in increased uptakes at ambient conditions. It should be noted that a change in ns0 does not directly imply a corresponding change in CO2 uptake, as ns0 is a theoretical capacity obtained from a mathematical fit, while the real uptake will depend both on the partial pressure and the temperature. From the data presented in Table S11 (Supporting Information) the average uptakes during adsorption increase from 1.07 mmol/g for the original case to 2.14 mmol/g for the largest change in (ns0)chemi. Even the highest uptake is reasonable as recent studies have reported values as high as 1.71 mmol/g at 420 ppm CO2 (dry) for PEI grafted fumed silica.29 The same study has reported CO2 uptake of 1.74 mmol/g (humid) by a similar adsorbent.29 This increase in the CO2 adsorption capacity implies that a larger amount of air must flow through the contactor to saturate the adsorbent, resulting in longer adsorption time. The calculated adsorption time increases from 1.7 h for the original case to a maximum of 3.3 h (Table S11, Supporting Information). The assumption of four cycles per day is still reasonable, as the adsorption times are still much smaller than the total cycle time of 6 h per cycle. For the specific case of amine-modified adsorbents, studies have shown that increasing the amine loading causes a reduction in the adsorption rates of CO2.27 In these calculations we have inherently assumed the validity of the equilibrium model at large adsorption time scales for all working capacities. From Table S11 (Supporting Information), the increase in the adsorption capacity of CO2 causes more CO2 to be present in the system relative to N2, and results in a higher purity of the product. More importantly, the increased CO2 uptakes directly result in a linear improvement of CO2 product recovery (Figure 7). The total CO2 obtained increases from 1.1 t CO2/day (at a working capacity $1000/t CO2. This conclusion was drawn from comparing the concentration factor of CO2 in air with that of large-scale industrial processes such as flue gas capture, ethanol distillation, N2/O2 separation, and trace gas removal processes such as SOx and NOx recovery. A second law efficiency of 100% will necessarily require contactor of infinite size and infinite capital cost. Trace gas removal processes need to have a certain throughput and a predefined removal fraction that is determined by the size and type of the power plant. For these processes, the second law efficiency is a trade-off between the operational costs and capital costs of electricity generation, while throughput and removal ratio act as constraints. For example, even though the reaction of NOx is thermodynamically favorable, the removal fraction will define the length of the contactor, and the flow rate will define the cross-sectional area. Lower operational costs will require a large cross-sectional area (smaller pressure drop) but will increase the capital costs. For air capture processes, there are no such restrictions on the volume of air processed or on the overall capture fraction. This is an important distinction, as the optimum capture faction for an air capture process is the one that balances the capital costs and the operational costs, solely based on minimizing $/t CO2 captured or maximizing the throughput (t CO2/year/unit). This difference between air capture of CO2 and trace contaminant removal means that the economics of processes for the two problems are not identical. It is also important to note that the analysis of House et al.5 considers air capture as a nonspontaneous one-step process, which requires energy input for concentration of CO2, and results in a high concentration factor. For the process outlined above, the separation consists of two distinct steps: adsorption (spontaneous) and desorption (nonspontaneous). A second law analysis is useful when considering processes that require external energy input: in our case, desorption. The first step of the process, adsorption of CO2 (and N2) from air, is a spontaneous, thermodynamically favorable process and no energy needs to be provided for achieving equilibrium. This consideration means that the overall second law efficiencies of the processes we considered are higher than the estimates of House et al.,5 which did not consider the spontaneous nature of adsorption. Our results are not based on detailed optimization of any process, contactor, or adsorbent parameters. It is likely that with some optimization and further analysis the operating costs could be reduced. Nevertheless, our cost estimates should be viewed as lower bounds on the total cost that would be associated with implementing these processes. In process B, the energy required for flowing steam through the adsorbent monolith during purging, any energy associated with separation of the steam/CO2 mixture in the product, and energy costs for opening/closing the monolith have not been included. The operational life of the air capture plant will exceed the lifetime of the adsorbent, requiring periodic replacement of the material. As this operational cost will depend on factors such as the market price and stability of the adsorbent, it is difficult to estimate these costs with current publically available data. These factors will increase the overall cost of operation, although none of them seem likely to make dominant contributions. Our estimates only include ongoing

based technologies has focused on materials development for CO 2 adsorption and on lab scale adsorption experiments.26−28,31−36,44−48 To determine the economic feasibility of the technology, analysis at the process simulation level is essential. In this paper, we have considered several variations of temperature swing adsorption processes based on a highly selective adsorbent that has been demonstrated previously in lab scale experiments.26,33,44,46 To our knowledge, this paper is the first report that analyzes a specific process for capture of CO2 using highly selective adsorbents. For each process, we developed preliminary process models to estimate the product purity, throughput, and operation costs. These models made it possible to vary a number of key process parameters to understand which factors have the most significant impact on process performance. We first considered an adsorption process (Process A) based solely on diurnal temperature swings. Although this concept is superficially attractive, our results indicate this process will be ineffective in practice due to the low throughputs of the product. A central drawback of this process is that the enhancement in the CO2 level of the product relative to ambient air that is possible is limited. When low quality steam is used for desorption, the performance of process A is improved. Unfortunately, the penalty associated with the energy cost of the process makes this largely unfeasible. We also considered a second process (Process B) that uses low quality steam to supply heat for desorption and also act as a purge. Because of this, the performance of this process greatly improves upon Process A. Using steam for desorption and purging, a specific process was described in which about 1.1 t CO2 is obtained per day from a unit the size of a standard shipping container at purities of ∼88.5% CO2. For this process the total energy requirement was calculated as 6745 MJ/t CO2 (including compression), of which 5962 MJ/t CO2 was thermal. Assuming that the electricity is obtained from natural gas and that dedicated steam needs to be generated for the process, the operational cost for this process is estimated to be $95/t CO2 for the reference scenario. This cost estimate includes the cost for compressing the product to sequestration-ready pressures; this step accounts for ∼25% of the total cost. It is useful to compare our predictions with the experimental data from Wurzbacher et al.,35 who used a continuous vacuum for desorption, which caused a low constant gas phase partial pressure of CO2, and obtained purities >97% CO2. Our calculations assumed the use of purge steam for desorption of CO2. In this regard, the continuous vacuum desorption is similar to steam flow in Process B. Their reported thermal energy for a packed bed,35 including both sensible and desorption energy, was 430 kJ/molCO2 (i.e., 9773 MJ/t CO2). This is higher than our estimate (5962 MJ/t CO2, thermal) as the working capacity of Process B (0.97 mmol CO2/g) is higher than the working capacity reported by the authors (∼0.2 mmol CO2/g).35 Realizing that the sensible heat for monolith contributes ∼39% of the total energy requirements and represents a parasitic loss to the process, we explored two possibilities of improving the process. The idea of decreasing the relative thermal mass of the monolith has limited applicability, since this will be determined by the structural strength of the material of construction and ease of manufacturing. Another approach is to increase the uptake of CO2, which will lead to higher working capacities. As described earlier, a working capacity of 1.93 mmol CO2/g will reduce the total energy requirement to ∼4000 MJ/t CO 2 , which corresponds to a second law efficiency of 11.8%. It is reasonable to hypothesize that a combination of these two approaches, using 8642



Article

sequestration of CO2 may be easier to resolve if numerous injection sites with small individual injection rates are used rather than a single site with a large injection rate. Efficient air capture technologies could play a significant role in this scenario. Another area that appears especially interesting is the use of CO2 in growth of biomass for generating fuels or other useful products4,6 or for wastewater treatment.58 A number of studies have shown that enhancing the CO2 content in bioreactors can increase algae growth rates.58−60 It is not necessary in these applications to have large throughputs of highly concentrated CO2; even modest increases in concentration above the quantities available in ambient air at reasonable throughputs might be advantageous. Implementation of processes like the adsorption-based approaches we model in this paper for bioreactors may be one avenue where economic benefits could be realized from air capture on small or moderate scales, which will simultaneously create experience with materials and equipment that could lead to more precise design models for large-scale systems. In conclusion, we have developed process models that account for some of the issues that need to be taken into account while considering the viability of adsorption-based air capture technologies. We have identified one process that appears promising based on these initial models. Our work should motivate future experimental and modeling efforts to refine the specifications of processes like this one and define more detailed estimates of their economics.

operational costs; they do not include any information about capital costs or maintenance costs. According to the recent APS report,7 the capital cost for postcombustion capture using absorption is ∼35% of the total cost while for air capture using absorption it is ∼60% of the total cost. Due to the detailed design necessary for estimation of capital costs, similar reports for adsorption-based postcombustion CO2 capture are very limited in the literature.57 For the air capture processes described in this work, we expect that capital costs would be incurred both through fabricating and siting the adsorbent units and also in providing the low quality steam necessary for the process. In some locations, it may be possible to use waste steam from existing manufacturing processes, but this is unlikely to be universally true if air capture is to be performed on large scales. Despite all of these caveats, the estimated cost associated with generating CO2 from the process we have outlined makes a compelling case that more detailed examinations of this process and related process economics are worthwhile. The models we have discussed in this paper are based on a specific amino-modified silica adsorbent, TRI-PE-MCM-41, developed by Sayari and co-workers.26,34,46 A number of other related materials have also been reported by other groups,21,27,29,35 so opportunities exist for considering the relative performance of multiple materials in air capture processes. We focused on TRI-PE-MCM-41 because the adsorption isotherms of CO2 and N2 are available with greater levels of detail to date than other similar materials. Even for TRIPE-MCM-41, however, data are not currently available that quantify the adsorption of CO2 and N2 in the presence of humidity. This creates an important simplification in the models we have described; our models assume that adsorption occurs with dry gases. We are not advocating that air be dried as part of an air capture process, but data are not currently available to fully describe the effects of humidity during adsorption or steam during purging. We described in the Introduction why, in general, it is reasonable to expect that the CO2/N2 selectivity of amino-modifed silicas will be enhanced by the presence of humidity relative to their behavior with dry gases, so it is reasonable to expect that the conclusions from our models are not seriously undermined by not treating adsorption as occurring from a three-component CO2/N2/H2O mixture. Nevertheless, obtaining quantitative data that would allow this issue to be addressed within process models should be a high priority for future work by the experimental community on these materials. Making a determination of the relative benefits of capturing CO2 from air compared to other possible strategies for lowering net CO2 emissions (e.g., CO2 capture from large point source emitters, reducing carbon intensity of power and fuel use, improvements in energy efficiency in large-scale applications, etc.) is beyond the scope of this paper. A common way to look at air capture, and the narrow approach taken in the recent APS report,7 is to ask whether this process can “improve upon” carbon capture from concentrated point sources such as power plant flue gas. We do not expect that air capture can be more economically efficient than treatment of large point sources. An alternative way to frame this issue is to ask whether there are applications in which the geographically dispersed nature of air capture might have advantages over flue gas capture or other means of reducing net CO2 emissions. For example, a key feature of air capture relative to flue gas treatment is that it has the potential to be decentralized, so it can possibly capture CO2 close to the locations that will be used for sequestration. More speculatively, it is possible that the regulatory issues associated with geological

■

ASSOCIATED CONTENT

S Supporting Information *

Material as described in the text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail:[email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the Department of Energy’s National Energy Technology Laboratory under contract DE-FE0002438. Helpful conversations with Prof. Christopher W. Jones and Ms. Stephanie Didas of Chemical & Biomolecular Engineering at Georgia Tech are gratefully appreciated.

■

REFERENCES

(1) Solomon, S.; Qin, D.; Manning, M.; Chen, Z.; Marquis, M.; Averyt, K. B.; Tignor, M.; Miller, H. L. Summary for Policymakers: The physical science basis; Contribution of Working Group I to the Fourth assessment report of the Intergovernmental Panel on Climate Change, 2nd ed.; Cambridge University Press: Cambridge, United Kingdom and New York, NY, 2007. (2) Choi, S.; Drese, J. H.; Jones, C. W. Adsorbent Materials for Carbon Dioxide Capture from Large Anthropogenic Point Sources. ChemSusChem 2009, 2 (9), 796−854. (3) Jones, C. W. CO2 Capture from Dilute Gases as a Component of Modern Global Carbon Management. Annu. Rev. Chem. Biomol. Eng. 2011, 2 (1), 31−52. (4) Keith, D. W. Why Capture CO2 from the Atmosphere? Science 2009, 325 (5948), 1654−1655. (5) House, K. Z.; Baclig, A. C.; Ranjan, M.; van Nierop, E. A.; Wilcox, J.; Herzog, H. J. Economic and energetic analysis of capturing CO2 from ambient air. Proc. Natl. Acad. Sci. U.S.A. 2011, 108 (51), 20428−20433. 8643



Article

(29) Goeppert, A.; Czaun, M.; May, R. B.; Prakash, G. K. S.; Olah, G. A.; Narayanan, S. R. Carbon Dioxide Capture from the Air Using a Polyamine Based Regenerable Solid Adsorbent. J. Am. Chem. Soc. 2011, 133 (50), 20164−20167. (30) Didas, S. A.; Kulkarni, A. R.; Sholl, D. S.; Jones, C. W. Role of Amine structure on CO2 Adsorption from Ultra-Dilute Gas Streams such as Ambient Air. In ChemSusChem 2012, DOI: 10.1002/ cssc.201200196. (31) Serna-Guerrero, R.; Sayari, A. Modeling adsorption of CO2 on amine-functionalized mesoporous silica. 2: Kinetics and breakthrough curves. Chem. Eng. J. 2010, 161 (1−2), 182−190. (32) Sayari, A.; Belmabkhout, Y. Stabilization of Amine-Containing CO2 Adsorbents: Dramatic Effect of Water Vapor. J. Am. Chem. Soc. 2010, 132 (18), 6312−6314. (33) Serna-Guerrero, R.; Belmabkhout, Y.; Sayari, A. Triamine-grafted pore-expanded mesoporous silica for CO2 capture: Effect of moisture and adsorbent regeneration strategies. Adsorption 2010, 16 (6), 567− 575. (34) Serna-Guerrero, R.; Da’na, E.; Sayari, A. New Insights into the Interactions of CO2 with Amine-Functionalized Silica. Ind. Eng. Chem. Res. 2008, 47 (23), 9406−9412. (35) Wurzbacher, J. A.; Gebald, C.; Steinfeld, A. Separation of CO2 from air by temperature-vacuum swing adsorption using diaminefunctionalized silica gel. Energy Environ. Sci. 2011, 4 (9), 3584−3592. (36) Li, W.; Choi, S.; Drese, J. H.; Hornbostel, M.; Krishnan, G.; Eisenberger, P. M.; Jones, C. W. Steam-Stripping for Regeneration of Supported Amine-Based CO2 Adsorbents. ChemSusChem 2010, 3 (8), 899−903. (37) Choi, S.; Watanabe, T.; Bae, T.-H.; Sholl, D. S.; Jones, C. W. Modification of the Mg/DOBDC MOF with Amines to Enhance CO2 Adsorption from Ultradilute Gases. J. Phys. Chem. Lett. 2012, 3 (9), 1136−1141. (38) Stuckert, N. R.; Yang, R. T. CO2 Capture from the Atmosphere and Simultaneous Concentration Using Zeolites and Amine-Grafted SBA-15. Environ. Sci. Technol. 2011, 45 (23), 10257−10264. (39) National Oceanic and Atmosphere Administration. http://www. ncdc.noaa.gov/oa/ncdc.html (accessed March 10, 2011). (40) Corning Celcor ultrathin-wall substrates. www.corning.com/ WorkArea/downloadasset.aspx?id=6283 (accessed January 2, 2011). (41) Elimelech, M.; Phillip, W. A. The Future of Seawater Desalination: Energy, Technology, and the Environment. Science 2011, 333 (6043), 712−717. (42) Eisenberger, P. M.; Chichilnisky, G. System and method for removing carbon dioxide from an atmosphere and global thermostat using the same. U.S. Patent 0289319 A1, 2008. (43) Corning substrates product catalog. http://www.corning.com/ environmentaltechnologies/products_services/ceramic_substrates. aspx (accessed January 2, 2011). (44) Harlick, P. J. E.; Sayari, A. Applications of pore-expanded mesoporous silica. 5. Triamine grafted material with exceptional CO2 dynamic and equilibrium adsorption performance. Ind. Eng. Chem. Res. 2007, 46 (2), 446−458. (45) Harlick, P. J. E.; Sayari, A. Applications of pore-expanded mesoporous silicas. 3. Triamine silane grafting for enhanced CO2 adsorption. Ind. Eng. Chem. Res. 2006, 45 (9), 3248−3255. (46) Serna-Guerrero, R.; Belmabkhout, Y.; Sayari, A. Modeling CO2 adsorption on amine-functionalized mesoporous silica: 1. A semiempirical equilibrium model. Chem. Eng. J. 2010, 161 (1−2), 173−181. (47) Belmabkhout, Y.; Sayari, A. Effect of pore expansion and amine functionalization of mesoporous silica on CO2 adsorption over a wide range of conditions. Adsorption 2009, 15 (3), 318−328. (48) Serna-Guerrero, R.; Belmabkhout, Y.; Sayari, A. Further investigations of CO2 capture using triamine-grafted pore-expanded mesoporous silica. Chem. Eng. J. 2010, 158 (3), 513−519. (49) Li, W.; Bollini, P.; Didas, S. A.; Choi, S.; Drese, J. H.; Jones, C. W. Structural Changes of Silica Mesocellular Foam Supported AmineFunctionalized CO2 Adsorbents Upon Exposure to Steam. ACS Appl. Mater. Interfaces 2010, 2 (11), 3363−3372.

(6) Keith, D. W.; Ha-Duong, M.; Stolaroff, J. K. Climate strategy with CO2 capture from the air. Clim. Change 2006, 74 (1−3), 17−45. (7) Direct Air Capture of CO2 with Chemicals. http://www.aps.org/ policy/reports/assessments/dac-biblio.cfm (accessed March 5, 2012). (8) Lackner, K. S.; Grimes, P.; Ziock, H.-J. Capturing carbon dioxide from air. In Proceedings of the First National Conference on Carbon Sequestration, Washington, DC, USA, 2001. (9) Lackner, K. S.; Ziock, H.-J.; Grimes, P. Carbon dioxide extraction from air: is it an option? In 24th International Conference on Coal Utilization & Fuel Systems, Clearwater, FL, 1999. (10) Zeman, F. S.; Lackner, K. S. Capturing Carbon Dioxide Directly From the Atmosphere. World Resour. Rev. 2004, 16 (2), 157−172. (11) Baciocchi, R.; Storti, G.; Mazzotti, M. Process design and energy requirements for the capture of carbon dioxide from air. Chem. Eng. Process. 2006, 45 (12), 1047−1058. (12) Zeman, F. Energy and material balance of CO2 capture from ambient air. Environ. Sci. Technol. 2007, 41 (21), 7558−7563. (13) Nikulshina, V.; Hirsch, D.; Mazzotti, M.; Steinfeld, A. CO2 capture from air and co-production of H2 via the Ca(OH)2-CaCO3 cycle using concentrated solar powerThermodynamic analysis. Energy 2006, 31 (12), 1715−1725. (14) Nikulshina, V.; Galvez, M. E.; Steinfeld, A. Kinetic analysis of the carbonation reactions for the capture of CO2 from air via the Ca(OH)2CaCO3-CaO solar thermochemical cycle. Chem. Eng. J. 2007, 129 (1− 3), 75−83. (15) Nikulshina, V.; Ayesa, N.; Galvez, M. E.; Steinfeld, A. Feasibility of Na-based thermochemical cycles for the capture of CO2 from air Thermodynamic and thermogravimetric analyses. Chem. Eng. J. 2008, 140 (1−3), 62−70. (16) Nikulshina, V.; Steinfeld, A. CO2 capture from air via CaOcarbonation using a solar-driven fluidized bed reactorEffect of temperature and water vapor concentration. Chem. Eng. J. 2009, 155 (3), 867−873. (17) Nikulshina, V.; Nikulshin, V. Exergy analysis of the solar thermochemical cycle for capturing CO2 from air. Int. J. Exergy 2010, 7 (2), 243−254. (18) Sherman, S. R. Nuclear Powered CO2 Capture from the Atmosphere. Environ. Prog. Sustainable Energy 2009, 28 (1), 52−59. (19) Stolaroff, J. K.; Keith, D. W.; Lowry, G. V. Carbon dioxide capture from atmospheric air using sodium hydroxide spray. Environ. Sci. Technol. 2008, 42 (8), 2728−2735. (20) Mahmoudkhani, M.; Keith, D. W. Low-energy sodium hydroxide recovery for CO2 capture from atmospheric airThermodynamic analysis. Int. J. Greenhouse Gas Control 2009, 3 (4), 376−384. (21) Gebald, C.; Wurzbacher, J. A.; Tingaut, P.; Zimmermann, T.; Steinfeld, A. Amine-Based Nanofibrillated Cellulose As Adsorbent for CO2 Capture from Air. Environ. Sci. Technol. 2011, 45 (20), 9101−9108. (22) Lackner, K. S. Capture of carbon dioxide from ambient air. Eur. Phys. J. Spec. Top. 2009, 176 (1), 93−106. (23) Wang, T.; Lackner, K. S.; Wright, A. Moisture Swing Sorbent for Carbon Dioxide Capture from Ambient Air. Environ. Sci. Technol. 2011, 45 (15), 6670−6675. (24) Hicks, J. C.; Drese, J. H.; Fauth, D. J.; Gray, M. L.; Qi, G. G.; Jones, C. W. Designing adsorbents for CO2 capture from flue gas-hyperbranched aminosilicas capable of capturing CO2 reversibly. J. Am. Chem. Soc. 2008, 130 (10), 2902−2903. (25) Choi, S.; Drese, J. H.; Eisenberger, P. M.; Jones, C. W. A new paradigm of anthropogenic CO2 reduction: Adsorptive fixation of CO2 from the ambient air as a carbon negative technology. In AIChE Annual Meeting, Nashville, TN, 2009. (26) Belmabkhout, Y.; Serna-Guerrero, R.; Sayari, A. Amine-bearing mesoporous silica for CO2 removal from dry and humid air. Chem. Eng. Sci. 2010, 65 (11), 3695−3698. (27) Choi, S.; Drese, J. H.; Eisenberger, P. M.; Jones, C. W. Application of Amine-Tethered Solid Sorbents for Direct CO2 Capture from the Ambient Air. Environ. Sci. Technol. 2011, 45 (6), 2420−2427. (28) Choi, S.; Gray, M. L.; Jones, C. W. Amine-Tethered Solid Adsorbents Coupling High Adsorption Capacity and Regenerability for CO2 Capture From Ambient Air. ChemSusChem 2011, 4 (5), 628−635. 8644



Article

(50) Bird, R. B.; Stewart, W. E.; Lightfoot, E. N., Transport Phenomena, 2nd ed.; Wiley: New York, 2010. (51) House, K. Z.; Harvey, C. F.; Aziz, M. J.; Schrag, D. P. The energy penalty of post-combustion CO2 capture & storage and its implications for retrofitting the US installed base. Energy Environ. Sci. 2009, 2 (2), 193−205. (52) Bollini, P.; Choi, S.; Drese, J. H.; Jones, C. W. Oxidative Degradation of Aminosilica Adsorbents Relevant to Postcombustion CO2 Capture. Energy Fuels 2011, 25 (5), 2416−2425. (53) U.S. Department of Energy Energy Information Administration: Electricity Emission Factors. http://www.eia.doe.gov/oiaf/1605/pdf/ Appendix%20F_r071023.pdf (accessed March 5, 2012). (54) U.S. Department of Energy Energy Information Administration: Emission Benchmarks for Purchased Steam and Chilled/Hot Water. http://www.eia.gov/oiaf/1605/pdf/Appendix%20N_20110128.pdf (accessed March 5, 2012). (55) U.S. Department of Energy Energy Information Administration: Annual Energy Outlook 2012 (Early Release Overview). http://www. eia.gov/forecasts/aeo/er/pdf/0383er(2012).pdf (accessed March 5, 2012). (56) Smith, R.; Varbanov, P. What’s the price of Steam? CEP 2005, 101 (7), 29−33. (57) Ho, M. T.; Allinson, G. W.; Wiley, D. E. Reducing the Cost of CO2 Capture from Flue Gases Using Pressure Swing Adsorption. Ind. Eng. Chem. Res. 2008, 47 (14), 4883−4890. (58) Park, J. B. K.; Craggs, R. J. Wastewater treatment and algal production in high rate algal ponds with carbon dioxide addition. Water Sci. Technol. 2010, 61 (3), 633−639. (59) Kurevija, T.; Kukulj, N. Algae biofuels as a possible alternative to environmentally doubtful conventional methods of biodiesel production: Survival and Sustainability; Springer: Berlin, Germany, 2011; pp 499− 510. (60) Benemann, J.; Pedroni, P. R.; Davison, J.; Becker, H.; Bergman, P. Technology roadmap for biofixation of CO2 and Greenhouse Gas Abatement with Microalgae. . http://www.netl.doe.gov/publications/ proceedings/03/carbon-seq/PDFs/017.pdf.

8645


Analysis of Equilibrium-Based TSA Processes for Direct Capture of

Recommend Documents