WO2019161420A1

WO2019161420A1 - A methodology and system for optimising and improving the cost and performance of direct air carbon dioxide capture systems, thereby contributing to a lessening of the threat of catastrophic climate change

Info

Publication number: WO2019161420A1
Application number: PCT/US2019/023716
Authority: WO
Inventors: Peter Eisenberger
Original assignee: Eisenberger, Peter And Chichilnisky, Graciela, Jointly
Priority date: 2018-01-26
Filing date: 2019-03-22
Publication date: 2019-08-22

Abstract

A method of cyclically removing carbon dioxide from carbon dioxide-carrying ambient air by directing a flow of carbon dioxide-carrying air through a carbon dioxide capture contactor structure having channels, said contactor structure supporting sorbent which is capable of binding carbon dioxide, to remove carbon dioxide from the air by binding carbon dioxide to the sorbent, thereafter causing said carbon dioxide capture contactor structure to be exposed to regeneration, whereby saturated steam of a temperature not greater than about 120 degrees C is directed at the carbon dioxide bound to the sorbent, thereby facilitating separation of the carbon dioxide from the sorbent and regenerating the sorbent, then withdrawing the carbon dioxide that has been unbound from the sorbent together with any remaining steam, and selectively thereafter repeatedly exposing said carbon dioxide capture structure to a flow of carbon dioxide-carrying ambient air, thereby enabling the regenerated sorbent to be again used to bind carbon dioxide, to remove carbon dioxide from the newer flow of carbon dioxide-carrying ambient air, said carbon dioxide capture contactor structure having or being formed with substantially parallel channels of a predetermined geometry that provides particularly favorable operating conditions, said predetermined geometry facilitating laminar flow whereby the proportionality of the pressure drop with respect to the velocity of carbon dioxide-carrying air moving through said channels is substantially linear as said velocity is increased.

Description

Title: A METHODOLOGY AND SYSTEM FOR OPTIMIZING AND IMPROVING THE COST AND PERFORMANCE OF DIRECT AIR CARBON DIOXIDE CAPTURE SYSTEMS, THEREBY CONTRIBUTING TO A LESSENING OF THE THREAT OF CATASTROPHIC CLIMATE CHANGE

The application claims the benefit or priority pursuant to 35 U.S.C. 119(e) from a U.S.

Provisional Patent Application having Application No. 61/330,108 filed April 30, 2010; from a U.S. Provisional Patent Application having Application No. 61/351,216 filed June 3, 2010 and from a U.S. Provisional Patent Application having Application No. 61/443,061 filed February 15, 2011, and from copending U.S. Application No. 13/098,370, filed on April 29, 2011.

This application incorporates by reference and claims priority from U.S. Application No. 15/898,531 dated February 17, 2018, which is a divisional of U.S. Application No.

14/587,716 filed December 31, 2014 as well as U.S. Application No. 15693250, dated August 31, 2017, U.S. Provisional Patent Application 61/922,338 filed December 31, 2013, the text of which is fully incorporated by reference herein as if repeated below.

BACKGROUND

[0001] The present invention relates to a methodology and systems for optimizing and improving the cost and performance of direct air carbon dioxide capture systems, in order to lessen the threat of catastrophic climate change.

[0002] There is much attention currently focused on trying to achieve three somewhat conflicting energy related objectives: 1) provide affordable energy for economic development; 2) achieve energy security; and 3) avoid the destructive climate change caused by global warming. However, there is no feasible way to avoid using fossil fuels during the rest of this century if we are to have the energy needed for economic prosperity and avoid energy shortfalls that could lead to conflict.

[0003] It is mostly undisputed by scientists that an increase in the amount of so-called greenhouse gases like carbon dioxide (methane and water vapor are the other major greenhouse gases) will increase the average temperature of the planet. [0004] It is also clear that there is no solution that only reduces the ongoing human contributions to carbon dioxide emissions that can successfully remove the risk of climate change. Removing additional CO₂ from the atmosphere is also necessary. With air extraction and the capability to increase or decrease the amount of carbon dioxide in the atmosphere, one can in principle compensate for other greenhouse gases like methane (both naturally occurring and from human activity) that can increase their concentrations and cause climate change.

[0005] Until the recent inventions by the present applicant, it was the generally accepted belief among experts in the field that it was not economically feasible to capture carbon dioxide directly from the atmosphere because of the low concentration of that compound, in order to at least slow down the increase of so-called 'greenhouse' gases in the atmosphere. It was subsequently shown by the co-pending, commonly owned, prior applications that it was in fact practical and efficient to carry out such CO₂ reductions under specified conditions.

[0006] It was shown that under ambient conditions CO₂ can be efficiently extracted from the air, at ambient conditions, using a suitable regeneratable sorbent system and a low temperature stripping or regeneration process, and that such a process can be expanded to remove C02.from mixtures of effluent gases mixed with a major amount of ambient air, so as to not only remove the CO₂ from flue gas but to remove additional CO₂ from the atmosphere so as to achieve a net reduction in CO₂ in the atmosphere at lower cost and higher efficiency.

SUMMARY OF THE PRESENT INVENTION

[0007] The present invention provides an optimized lower cost direct air carbon dioxide capture (DAC) system and apparatus for removing carbon dioxide from the atmosphere,. This is accomplished according to the present invention by utilizing a method of cyclically removing carbon dioxide from carbon dioxide-carrying ambient air by directing a flow of carbon dioxide-carrying air through a porous carbon dioxide capture contactor structure having channels extending through the thickness of the contactor structure. The contactor structure supporting a sorbent on the porous interior surfaces of the channels to remove carbon dioxide from the ambient air passing through the flow channels of the contactor structure. The sorbent is capable of binding carbon dioxide, to remove carbon dioxide from the ambient air. The carbon dioxide-loaded capture contactor structure is next caused to be exposed to regeneration, whereby saturated steam of a temperature not greater than about 120 degrees C is directed at the capture structure , which causes the carbon dioxide bound to the sorbent to separate and thereby regenerate the sorbent bound to the capture structure. The separated carbon dioxide that has been unbound from the sorbent together with any remaining steam, is withdrawn and the carbon dioxide capture structure is again exposed to a flow of carbon dioxide-carrying ambient air, thereby enabling the regenerated sorbent to be again used to bind carbon dioxide, to remove carbon dioxide from the newer flow of carbon dioxide-carrying ambient air. The carbon dioxide capture contactor structure has or is formed with substantially parallel flow channels extending between the major surfaces of the capture structure of a predetermined geometry that provides particularly favorable operating conditions. The predetermined geometry, as defined below, facilitates laminar flow through the flow channels, whereby the proportionality of the pressure drop with respect to the velocity of carbon dioxide-carrying air moving through said flow channels is substantially linear as said velocity is increased.

[0008] In accordance with the present invention, The flow channels extend through the structure, from the downstream surface through to the upstream surface of the capture structure, so as to allow flow into the channel at the upstream surface and exiting from the channel at the downstream surface. The length of the channels, preferably, is in the range of 0.1 - 0.45 m and preferably between 0.1 and 0.2 m. The capture structure is thus limited to that thickness whereas the major surfaces can be as large as 50 sq. ft., or even greater. [0009] Each channel opening can have an effective diameter in the range of 1.75 mm to 3.5 mm with the most preferred diameter being between 2 and 3 mm. The shape of the opening is not limited, and can be round, circular or oval, or polygonal, but preferably a regular shape and most preferably hexagonal, or rectangular or square, or circular. The distance separating each channel, i.e. the wall thickness between the channels, is preferably in the range of 0.15 - 0.65 mm, with a most preferred range between 0.3 and 0.4 mm.

[0010] As explained above, it is desired that the flow through the flow channels is laminar and it has been found that an air velocity entering the flow channels should be in the range of between 2.56 m/s with a preferred range between 4 and 6 m/sec. The inner surface of the flow channels, regardless of their cross-sectional shape, are highly porous to provide a reservoir for the sorbent material, which is supported within the pores so as to readily make contact with the carbon dioxide in the lemon are airflow through the channels.

[0011] The preferred porous contactor material is an extruded ceramic substrate such as an extruded porous silica or alumina, corrugated fiberglass sheets, corrugated metal sheets, structured carbon and engineered plastics. All of these materials should be formed to be highly porous and the important requirement is that the interior surfaces of all of the channel be porous as that provides the necessary reservoirs for the sorbent material to contact the carbon dioxide in the air, as the laminar airflow passes through the channels. The primary limit on the number of channels through each of these types of material is based upon the requirement for maintain the structural integrity of the contactor and thus limits not only the thinness of the walls between adjacent channels but also the distance of the 1st and last channels from the outer edges of the contactor monolith. One further point is one of the most preferred materials is a porous silica monolith wherein all of the surfaces are coated with alumina, so as to achieve the greater porosity attainable with alumina, while maintaining the structural integrity of the monolithic possible with the underlying silica skeleton.

[0012] The sorbent substrate can comprise a single monolithic monolith or, e.g., a plurality of the CELCOR^® cellular, ceramic substrates, stacked as bricks, or a single substrate, having the type of pancake shape described above in connection with FIG. 6 (i.e. front surface area much greater than thickness), and the CO₂ laden air is directed through the cells of the sorbent structure. It is also contemplated that the sorbent structure can be formed by embedding the sorbent material in the, e.g., alumina, coating en the walls of the CELCOR^® cellular, ceramic structure to form a monolithic sorbent structure. Each brick has a thickness equal to the desired final thickness of the stacked monolith substrate, and the stack forms the two major surfaces. Each brik has the formed parallel hannels between the two surfaces forming the major stack surfaces.

[0013] It is increasingly recognized that the removal of carbon dioxide from the atmosphere is needed to avoid the threat of catastrophic climate change. Renewables and carbon emission reductions will no longer suffice. The present invention provides lower costs when DAC is done at scale, less than $50 per tonne, than any other costs claimed by other known to the undersigned existing DAC technologies.

[0014] The following is presented as a relatively simple heuristic analysis that attempts to explain why the DAC costs associated with the present invention are lower. The intent is not to in any way disparage the sincere efforts of others in DAC, and it can be said that the DAC efforts of all may have contributed to the increased acceptance of DAC as a viable technology. The objective is to demonstrate that a lower cost path to DAC exists utilizing the present invention. In so doing it is hoped that future innovations will be catalyzed that will push DAC costs to lower than $25 per tonne. Low cost renewables and low cost CO₂ from the air will initiate the Renewable Energy and Materials Economy (REME) which will stimulate economic growth while removing carbon from the air and sequestering it in, for example, materials of construction. This will be very important for reaching the gigatonne scale of carbon dioxide removal (CDR) needed to address the treat of climate change.

[0015] In collecting CO₂ from the air, one of the challenges relates to its low concentration. This means one has to move about 3000 tonnes of air for each tonne of CO₂ collected. This in turn means that the path to lowest cost DAC must maximize the rate of CO₂ contacting the collection surface, have kinetics fast enough to capture all the CO2 that contacts the surface so that it can be efficiently captured, and to do it in a way that minimizes the resistance to air flow so that the energy required to move the air is minimized.

[0016] The following description attempts to explain that to achieve the lowest cost for all DAC technologies of differing sorbents, and all processes for collecting the captured CO₂, is to have the air move parallel to the contacting surface in the lowest flow resistance laminar flow regime and contacting the surface via diffusion perpendicular to the flow. That same regime also enables by simple geometry the highest contacting surface area per volume which, in turn, reduces the volume of contactor needed to capture the incoming CO₂ efficiently. The same geometry of parallel surfaces will also be shown to have fast kinetics for the capture process if the walls between the surfaces are thin and porous. Fast kinetics are essential to be able to capture a relatively high rate of CO₂ contacting the surface efficiently.

[0017] Another challenge for achieving low cost DAC is to minimize the energy used to collect the captured CO₂ and to minimize the cost of that energy and to have fast enough collection kinetics so that the time to collect is relatively small compared to the time to capture the CO₂. It will also be shown below that the same design of parallel surfaces separated by thin walls effective for capture also enables an innovative low temperature CO₂ process with fast kinetics for collecting the captured CO₂. The lower the temperature of removal of the captured CO₂ from the contactor means that less heat is needed to raise the temperature of the contactor. This together with the fact that the cost of heat in general decreases as the temperature decreases means that the process according to the present invention uses less energy and less costly energy.

[0018] The impact, on reducing DAC costs per tonne of CO₂ collected, of:

1 lowest resistance to flow

2 highest surface area per volume

3 high rate of contacting by diffusion

4 fast kinetics

5 low temperature collection

will be analyzed and confirmed by direct comparison of the cost of the present invention's technology that satisfies 1-5 above with published data by other DAC technologies.

Simple Cost Model

The CAPEX for Direct Air Capture Technology:

[0019] The CAPEX for DAC is primarily determined by the amount of tonnes of CO₂ captured per frontal area of the capture device per year, TPY, divided into the capital costs of the capture technology per frontal area, CC.

Where TPY = velocity of air (V) x density of CO₂ x percent of incident CO₂ captured x one year

[0020] In DAC there is no need to capture 90% or more of the incident CO₂ as is the case for flue gas capture. Since air is free and effectively infinitely available, capture efficiency becomes a variable to minimize the cost A percent of capture of the incident CO₂ of around 50-70% minimizes the capital expenditure, CAPEX, because of the well-known decrease in the rate of CO₂ capture with increased depth in the contactor due to the exponential decrease of CO₂ concentration in the contactor with depth. So the percent captured will not be a significant differentiation between different DAC approaches for this analysis since most established DAC processes report capture efficiencies in the range of 50-70%.

[0021] Similarly, since the system and apparatus according to the present invention can be mass produced and involves abundant and low cost materials, its contactor cost per volume of contactor can be comparable. However, its CC can be lower because its high surface area per volume results in less length of its contactor compared to other contactors. All DAC approaches will benefit from learning by doing. However, the learning rate utilizing the present invention will be faster in time because its low cost will enable a greater accessible market, increasing its rate of installed capacity. This has already occurred because many experts in industrial gas have chosen the present technology after evaluating other approaches.

[0022] It is argued that this fast rate of learning with time due to more effort per time is very important for reaching the gigatonne scale needed in a timely fashion to address the threat of climate change. How one depreciates the CC determines the contribution of the CAPEX to the cost per tonne— a uniform approach for all technologies is assumed so it also is not a variable for comparing CAPEX costs.

The OPEX for Direct Air Capture Technology:

[0023] Ignoring maintenance costs which in a mature technology is taken to be a small fraction of CC, and therefore relatively uniform across all DAC technologies, OPEX is determined by the energy used per tonne of CO₂ captured, EPT, times the cost per unit of energy, CE.

[0024] EFT has two main components, namely: the energy to move the air during capture and the energy to collect the captured CO₂. Both EPT and CE can vary with the approach used and it will be argued below both can in principle be as low as possible in the approach to DAC of the present invention.

[0025] In general, DAC costs are the lowest when V is the highest and when CC, EPT and CE are the lowest and set forth below is an explanation as to why the technology of the present invention has low values for all those variables.

Analysis

Contacting the CO₂ in the Air:

[0026] The parallel channel contactor embodiment of the present invention has the air moving parallel to the surface in the lowest resistance to flow laminar regime. Its contactor has the same parallel channel geometry, enabling it to process large volumes with little resistance. Its properties are adjusted to optimize the DAC process. The basic properties of such a contactor can be specified by three geometrical parameters

• Size of the channel S

• Thickness of the walls W

• Length of the channels L

One can specify the properties relevant to DAC:

1. The surface area per volume SAv, where

where if S is on the order of millimeters and is a small fraction of S, SAv is thousands of square meters per cubic meter.

2. The mass transfer rate K of the CO₂ to the walls of the channels which is based upon diffusion, so depends upon 1/S (i.e. a smaller S increases the rate of CO₂ hitting the surface) for the case of CO₂ in air:

K = 0.059 /S [m/sec] where S is expressed in mm

(http://www.mermopedia.corn/content/940/)

3. The 1/e decay length DC of the CO₂ as it passes through the channels is:

or

4. The pressure drop PD in the laminar regime

where μ is viscosity of air (1.98 x 10^-5 kg (m sec)) and is also based upon diffusion, in this case the molecules colliding with the walls.

(http://www.mhtlab.uwaterioo.ca/pdf_papers/mhtl05-5.pdf)

5. The rate of CO₂ removal per year per area, TPY, is determined by the rate of CO₂ incident on the contacting device per year per area:

where p is the density of CO₂ in air. The fraction of the incident CO₂ captured η is

Thus

[0027] When the air moves in laminar flow parallel to the surface there is a distinctive mechanism responsible for the resistance to flow at the surface boundary that is related to the viscosity of the air and by a linear dependence on velocity (see Eq. 4 above). If the air strikes a surface other than parallel to it will lose forward momentum when it hits the surface resulting in an effective pressure drop that depends upon a power higher than two on the velocity (non- laminar flow).

[0028] The distinction in performance between the two regimes is increased as the velocity of the air increases, as indicated by difference in the velocity exponent (1 vs. >2). Thus, a first design principle for low cost DAC is to maximize the surface area contacted, SAv, so that the CO₂ can be captured efficiently, and to do so in a way that minimizes the resistance to air flow, pressure drop PD. The pressure drop is at a minimum when there is laminar flow of the air parallel to the surface it is contacting and SAv by simple geometry is also high if S is small. In this way there is only resistance at the interface between the moving air and the parallel walls which minimizes the resistance per surface area. [0029] It is the fundamental property of diffusion that causes viscosity, which in turn determines the pressure drop in laminar flow regime and also determines that the rate of CO₂ transport to the walls, thus the mass transport coefficient It is in turn the relationship between the viscosity and the mass transport that enables capture of CO₂ at high efficiency on a parallel contactor surface even at high velocity.

[0030] Non laminar flow with frontal contact is very inefficient because the resistance is high, resulting from turbulent flow. Also by simple geometry if the tilt of the surface times its length is thicker than the wall thickness between the channels, the SAv will be reduced. Since according to the present invention DAC S is about 2 millimeters, it has a very high SAv compared to other known types of contactors. It also has a high K due to the low S (Eq. 2). A high and SAv results in a DC that can achieve the target η of 50-70% even for the high V of Sm/sec enabled by the low PD in the laminar regime. This has both the advantage of making the contactor small in length (low CC) and also of distributing the CO₂ impinging on the surface over a large area. Conversely when the SAv and the mass transfer rate K are low as in the case of structured packing, one needs contactor lengths on the order of 10 meters even for the relatively low V of 1-2 m/sec to achieve the target η.

[0031] The below measurement data exhibits how the pressure drop PD varies with velocity V for the GT contactor length of 0.15 m. Note that as predicted in the analysis above (Eq.4), the PD-V relationship is relatively linear under the conditions of the present invention, demonstrating the expected laminar flow regime.

Chart A - Pressure drop vs. Velocity measurements for present invention monolith contactors at high and low loadings (i.e. bigger and smaller wall thicknesses). Note at 5 m/s the pressure drop is < 150 Pa.

[0032] Under these flow conditions (V = 5 m/s, L = 0.15m, PD < 150 Pa) the monolith contactors of the present invention have also been shown to satisfy the η > 50% criterion, as shown in the breakthrough curve data in Chart2 below. In 15 minutes at 5 m/s, 40 L of CO₂ at 400 ppm (0.04%) are processed across the frontal area of 36 in² (0.0225 m²), and the monoliths of the present invention accumulate more than 20 L adsorbed in that duration.

Chart B- CO₂ breakthrough curve and adsorption uptake curve for a monolith of the present invention at V = 5 m/s. Note the cumulative uptake > 20 L / monolith at 15 minutes.

[0033] By adopting the simple principle that one cannot capture more CO₂ than is entering the device, the maximal possible TPY per PD is achieved with a monolith parallel channel contactor because it enables the highest capturing rate at the lowest resistance to air flow - which in turn enables one to capture the higher rate of CO₂ incident because of the higher velocity. It also has a much smaller volume of contactor L because of the high SAv and K thus reducing the CC.

Kinetics

[0034] Once the CO₂ has contacted the capturing material the next physical constraint is that the CO₂ has to be captured as fast as it impinges on the surface otherwise there will be concentration polarization of CO₂ at the wall surface and thus reduce the rate of diffusion. Therefore, the rate of capture will be slower than the rate of contacting and result in less TPY and thus increased CAPEX. The transport inside the contacting walls must be designed to ensure that the removal rate is faster than the contacting rate, which is a generic challenge that all DAC needs to address. The geometry enabled by the parallel channel monolith contactors is very favorable for fast kinetics because of its high surface area and very thin walls (-tenths of a millimeter). The thin walls in the contactor make mass transfer by diffusion and heat transfer by thermal conductivity very fast because they scale with W² and , respectively. The above data shows that there is minimal slowdown in the contactor of the present invention since the amount captured is close to what is predicted by the equation for TPY, 22 liters, which is how much CO₂ contacts the surface.

[0035] Thus in the contactor geometry of the present invention the air moves with low resistance laminar flow parallel to the channel surface, which maximizes TPY per PD. At the same time, it maximizes the surface area per volume SAv, which enables low contactor volume L, which reduces CC. It also spreads out the CO₂ capture over a large surface reducing the rate one needs for effectively immobilizing the CO₂. The thin walls enable fast CO₂ transport to the sorbent sites embedded in the walls.

[0036] By use of the monolith contactor technology of the present invention, a different learning curve whose limiting CAPEX cost is lower than alternatives.

Selective Sorbents;

[0037] All DAC technologies can benefit from improvements in sorbent performance and reduced sorbent costs. The reduction on the requirement of the capture sites per volume of wall material makes it easier for the design of the present invention to embed a sorbent in the walls that will both provide the capacity and kinetics needed to capture all the CO₂ contacting the surface of the contactor which is required by the higher TPY.

[0038] The binding sites in the porous structure are determined by the amount and dispersion of the amines throughout the porous structure. There are three generally known classes of supported amine sorbents which have been used for the present situation. The presently preferred Class 1 adsorbents are based on porous supports impregnated with monomelic or polymeric amines (Figure 12). The amine species are thus physically loaded onto or into the pores of the support structure. This class of sorbents is described in the technical literature, for example in Xu, X.C., et al., Preparation and characterization of novel CO2 "molecular basket" adsorbents based on polymer-modified mesoporous molecular sieve MCM-41. Microporous Mesoporous Mat., 2003. 62(1-2): p. 29-45 and Xu, X.C., et al., Influence of moisture on CO2 separation from gas mixture by a nanoporous adsorbent based on polyethylenimine-modified molecular sieve MCM-41. Ind. Eng. Chem. Res., 2005. 44(21): p. 8113-8119 and Xu, X.C., et al., Novel polyethylenimine-modified mesoporous molecular sieve of MCM-41 type as high-capacity adsorbent or CO2 capture. Energy Fuels, 2002. 16(6): p. 1463-1469. Class 2 adsorbents are based on amines that are covalently linked to the solid support. Methods of forming such Class 2 adsorbents in the porous structure of the present invention are known to the art. This has most often been achieved by binding amines to the ceramic monolith porous walls, e.g., silica oxides or alumina oxides, via the use of silane chemistry, or via preparation of polymeric supports with amine-containing side chains.

[0039] Class 3 adsorbents are based on porous supports upon which aminopolymers are polymerized in-situ, starting from an amine-containing monomer. This Class 3 type was described for use as adsorbents for CO₂ capture by Hicks, J.C., et al., Designing adsorbents for CO2 capture from effluent gas-hyperbranched aminosilicas capable, of capturing CO2 reversibfy. J. Am. Chem. Soc., 2008. 130(10): p. 2902-2903.and by Drese, J.H., et al., SynthesisStructure-Property Relationships for Hyperbranched Aminosilica CO₂ Adsorbents. Adv. Funct. Mater., 2009. 19(23): p. 3821-3832. Each of these adsorbent classes can be used for CO₂ capture and steam-regeneration studies.

[0040] A highly preferred sorbent structure is one in which the primary amine is incorporated into the monolith structure itself requiring only one step to make it. Such a specific embodiment can be made from plastic/polymers, which can survive because of the mild conditions utilized in the system of the present invention. The monolith can be a composite include inorganic non polymeric materials- such a composite would have properties in terms of strength, porosity, stability that could be useful.

[0041] The following procedures can be followed to provide amine sorbent supported on commercial particulate silica supplied by the PQ Corporation (PQ-9023) or on mesocellular foam. For the preparation of all the adsorbents, the silica substrate was first dried under vacuum at 100 °C for 24 hrs. to remove absorbed water on the surface before use. Λ commercial particulate silica supplied by the PQ Corporation (PQ-9023) and a lab-synthesized mesocellular foam were used as supports. The commercial silica is characterized by a surface area of 303 nrVg, an average pore volume of 1.64 cc/g. and an average pore diameter of 60 nm. The mesocellular foam was prepared following a literature methodology, Wystrach, V.P., D.W. Kaiser, and F.C. Schaefer, PREPARATION OF ETHYLENIMINE AND TRIETHYLENEMELAMINE J. Am. Chem. Soc., 1955. 77(22): p. 5915-5918. Specifically, in a typical synthesis, 16 g of Pluronic P123 EO-PO-EO triblock copolymer (Sigma-Aldrich) was used as template agent and dissolved in 260 g Dl-water with 47.1 g concentrated HC1. Then 16 g oftrimethylbenzene (TMB, 97 %, Aldrich) was added at 40 °C and stirred for 2 hrs before 34.6 g tetraethyl orthosilicate (98 %, Aldrich) was added to the solution. The solution was kept at 40 °C for 20 hrs before 184 mg NH₄F (in 20 mL water) was added. The mixture is later aged at 100 °C for another 24 hrs. The resulting silica was filtered, washed with water, dried in oven, and calcined at 550 °C in air for 6 hr to remove the organic template before further use. The mesocellular foam silica is characterized by a surface area of 615 m²/g, an average pore volume of 2.64 cc/g and average window and cell diameters of 12 nm and 50 nm.

[0042] Generally, for a Class 1 sorbent, the amine compound may be applied to the porous substrate structure by physical impregnation from the liquid or vapor phases. The amine compound can diffuse into the pores of the substrate structure. In this embodiment the pore volume becomes the critical parameter determining loading and pores 5-15 nm being preferable but the conclusion of wanting as thin walls as possible and thus as high a porosity as possible that is also physically strong enough so that the monolith is structurally strong. As an example of the preparation of the Class 1 adsorbent, 18 kg low molecule-weight poly(ethylenimine) (PEI, MN ~ 600, Mw ~ 800, Aldrich) and 90 L methanol (99.8%, Aldrich) were mixed first for 1 hr. Subsequently, 30 kg of amorphous particulate silica (PQ Corporation, PD-09023) [or a suitable substrate (175 in²) of the CELCOR^® monolith] was added and the liquid stirred for an additional 12 hrs. The methanol solvent was later removed by rotavap, and the resulting supported adsorbent ("PQ-ΡΕI") was further dried under vacuum at 75 °C overnight before using.

[0043] For preparation of the Class 2 adsorbent, 90 L anhydrous toluene (99.5%, Aldrich) and 3 kg of particulate silica (PQ Corporation), or a suitable monolith substrate (e.g., a brick of the CELCOR^® monolith having a front surface area of 36in², and a pore surface area of 175 in²) was mixed in a pressure vessel for 1 hr, then 30 kg of 3 -aminopropyltrimethoxy silane (APTMS, Aldrich) was added into the mixture. The mixture was kept under vigorous stirring for 24 hrs at room temperature. The resulting supported adsorbent (PQ-Mono) was recovered by filtration, washed with toluene and acetone, and then dried overnight, under vacuum, at 75°C.

Collection of Captured CO₂:

[0044] A generic DAC technology needs to remove the captured CO₂ from the sorbent and collect it for use. This is an energy intensive process because it takes energy to remove the CO₂ that is strongly bound to the sorbent and to heat up the sorbent/contactor. There are many known ways to remove the CO₂:

1. Heating and effectively evaporating the CO₂ from the sorbent

2. Pumping it off by creating a vacuum over the sorbent

3. Using another gas to effectively sweep the CO₂ off the sorbent by creating a low CO₂ partial pressure

There are also combinations of the above that are used in other separation technologies.

[0045] Here as described above there are two performance parameters and one constraint to minimize energy costs. The two parameters to minimize are the collection energy amount EFT and the cost of the energy used CE. Not all energy costs are the same. The cost of electricity per joule is greater than the cost of heat per joule, and the cost of high temperature heat is more than low temperature (below 100°C) heat. The constraint is that the collection step has to be fast compared to the adsorption step or the contactor will spend less of the year adsorbing and thus reduce TPY and increase the CAPEX.

[0046] The undersigned is a named inventor in patents teaching a low energy, low temperature temperature-vacuum swing + sweep gas combination where the low temperature (less than 100°C to as low as 70°C) steam heats the contactor, evaporating and stripping the CO₂ from the contactor, and sweeping to achieve a low partial gas pressure of CO₂ in the contactor during collection. It also satisfies the timing constraint with collection lOx faster than capture. This is not only because of the contactor design which maximizes heat transfer (direct contact/thin walls) but because the process of the present invention uses the latent heat of steam condensation on the contactor walls (more than 5x the sensible heat of 100°C steam) to deliver the heat at a very fast rate for relatively little mass of steam. Data illustrating the CO₂ collection rate from the regeneration process of the present invention is shown below in Chart3.

[0047] The key to being able to use low temperature heat is that before it condenses downstream in the monolith the steam acts as sweep gas creating a low partial pressure of CO₂ in the channels of the contactor. This shifts the Langmuir equilibrium to favoring the gas phase for the adsorbed CO₂ at lower temperatures. Currently this is to about 85°C. The undersigned has shown that one can shift the temperature to as low as 70°C, and by altering the steam and pumping rates it is believed that it can be shifted even lower. This is a less energy intensive way than pumping to create low absolute pressures.

[0048] In many cases the upstream power production or downstream CO₂ conversion processes have unused low temperature heat that can be used to cause the release or stripping of the CO₂ from the contactor sorbent. Often that energy is otherwise rejected (e.g. power plant cooling towers). The undersigned is a named inventor in patents which teach this process and/or waste energy to remove the CO₂— heat that has already been paid for by generation of the electricity to power our technology or the production of a useful product using the CO₂ we produce. This cogeneration is enabled because the source of the CO₂ is the air which is available anywhere, meaning the DAC system can be installed directly adjacent to the coupled process. This use of low temperature steam provided by cogeneration of our CO₂ with the production of our electricity and/or the conversion of the CO₂ into a product allows for the lowest potential cost for our useful heat energy possible, as low as zero. The availability and the desirability of using the cogenerated heat has been confirmed by many potential users of the CO₂ produced.

Cost Comparisons with three Other DAC Technologies:

[0049] Cost is important since it determines the commercial feasibility of the technology and as described above is critical to achieving the gigatonne scale needed. Here a primary purpose is to show that the factors discussed below do in fact result in lower costs. a. Carbon Engineering (CE)

[0050] An assessment of the Carbon Engineering technology has been published in http://rstaToyalsocietypublishing.org/content/370/1974/4380:

[0051] Carbon Engineering uses more energy because of the larger delta T between adsorption and collection driven by the higher heat of reaction of their sorbent. They necessarily therefore use higher temperature heat in their regeneration process which makes the OPEX costs higher compared to an equivalent amount of low temperature heat.

[0052] The real difference in cost as described above is in the difference in the performance of their contactors compared to the technology of the present invention (hereinafter sometimes "GT"):

[0053] Thus as expected the GT monolith contactor has higher surface area per volume, faster mass transfer to the contacting surface, a higher velocity and less volume of contactor all which result in lower costs.

[0054] This is reflected in the much higher Capex cost per tonne compared to that enjoyed by GT. For Carbon Engineering the capital cost per frontal CA from the table above is $3700, with a velocity V of 1.5 m/sec and a capture efficiency of 0.75. For GT the CA is $1500 per frontal area and the velocity V is 5 m/sec and a capture efficiency of 0.55. This gives ratio of CE Capex per tonne captured to GT of (3700/1500) (5/1.5) (.55/75) = 6. So GT has both a lower OPEX and much lower CAPEX than Carbon Engineering. It is worth noting that even though the GT length is about 60 times shorter its CC is only about 3 times less. With future mass production of the GT contactor and learning by doing the costs of the GT contactor are expected to drop. b. Infinitree

[0055] Infinitree technology can only concentrate CO₂ to 5% and can only work where there is not a lot of humidity so on its own it is not a scalable technology to address climate change. c Climeworks

[0056] Climeworks uses a technology that has some similarities with GT in using low temperature heat, however its contactor has significant direct contact that also limits its velocity of air. Detailed information regarding the Climeworks process parameters are not fully known at this point.

From the Climeworks Website

Technical Specifications:

htlp://www.dimeworks.com/capture^rocess.html [00S7J Climeworks' thermal energy requirement is 50% higher than that of the GT system (1500 kWh vs 1000 kWh). The Climeworks kinetics is slower and so that they only capture 300 metric tonnes per year per 40 foot container, whereas the GT system produces 2000 tonnes per year per 40 foot container. Assuming similar monolith capital costs, GT is a factor of about 6 less in Capex cost per tonne.

[0058] In the Climeworks commercial plant which captured 900 tonnes per year they said it cost $3.6 million, yielding a CAPEX of $4000 per tonne. For GT it is approximately $2 million for 4000 tonnes per year, giving a ratio of CAPEX per tonne of CO₂ of 8.

Comparing Technologies

[0059] GT technology has a distinctive patentable approach that minimizes the CAPEX and the OPEX based upon geometry and kinetic principles. The prediction based upon those simple considerations is confirmed by GT data and the available comparisons. What should be clear is that GT is on a distinctive learning curve producing economically viable DAC CO₂ with a significantly lower cost limit than other known technologies. Again, to be clear, there is no intent to be negative about other DAC approaches; intent of the undersigned is to use the comparison to support the position that a low cost under $50 per tonne DAC using GT technology is reasonable. In fact if one took the published projections for longer term costs for Carbon Engineering and Climeworks of $100 -$150, and used the ratios calculated above based upon simple principles, then future under $25 dollar per tonne for a mature GT technology is plausible.

[0060] It is estimated that the limiting cost that GT technology could reach at the gigatonne scale (actually by the ten million tonnes per year installed capacity scale) is not the 50 dollars per tonne, but is believed to be under 25 dollars per tonne. The CAPEX is reduced - even today monoliths made in China cost about 80% less than what was used for the $50 dollar per tonne cost. Also as renewable energy becomes less costly than the $0.07/kWh assumed for the $50 case, the OPEX will also be reduced.

[0061] At $50 per tonne of CO₂, the carbon cost is equivalent to $23 dollar per barrel oil. This will enable the production of cheaper carbon materials using CO₂ instead of oil as the feedstock, such as plastics or fuels. Thermoplastic polymers will be made directly from CO₂ from the air. Carbon black will be produced for tires or carbon fibers for building materials cheaper than their current fossil carbon based processes with C4 Composites. In this way, gigatonnes of CO₂ can be sequestered as one converts to a REME.

[0062] The present invention provides further new and useful systems and methods for removing carbon dioxide from a mass of carbon dioxide laden air, at higher efficiencies and lower overall costs including lower capital expenses ("CAPEX") and lower operating expenses ("OPEX").

[0063] In one embodiment of the present invention, a novel process and system has been developed utilizing assemblies of a plurality of monoliths, or beds, that are combined with a single regeneration box, in a ratio dependent upon the ratio of the speed of adsorption compared to the speed of regeneration of the sorbent. In preferred embodiments, the monoliths are supported on a closed loop track, preferably forming a closed curve; upon which the monoliths are moved along the track, in succession, while being exposed to a moving stream of ambient air or a mixture of gases comprising a major proportion of ambient air. At one location along the track, the rotation is halted and one of the monoliths is moved into a sealed box for processing to strip CO₂ from the sorbent to regenerate the sorbent. When the sorbent is regenerated, the monoliths are rotated around the track until the next monolith is in position to enter the regeneration box, when the rotation of all of the monoliths is next halted. The moving stream of ambient air continues to flow into the monolith channels that are not being regenerated.

[0064] Each monolith is formed of a porous substrate having on its surfaces carbon dioxide adsorbing amine sites, preferably with a high proportion of primary amines. As the monoliths move along the track, they adsorb CO₂ from the moving gas streams passing through the channels until each monolith reaches the sealable regeneration box. Once sealed within the box, the sorbent is treated to cause the CO₂ to be stripped from the sorbent, regenerating the sorbent. The stripped CO₂ is removed from the box and captured. The monolith with the regenerated sorbent then moves out of the sealed box and moves along the track with the other monolith to adsorb more CO₂, until the next monolith is rotated into position to be moved into the regeneration box. At the stripping/regeneration location, the monolith can be moved into a box located above or below the grade of the track, or the box can be located so that the monolith moves into the box at the same grade level as the track, forming a seal with the monolith. These several alternatives are further defined below and diagrammed in the accompanying drawings.

[0065] In the instances where the regeneration box is below or above grade, the system must include a sub-system for raising or lowering the monolith. In systems where the regeneration box is on grade with the tracks, a more complex sealing arrangement will be required, for providing a seal along the sides as well as along the edge surfaces. CO₂ Adsorption and Removal Process

[0066] The basic premise of this process is that CO₂ is adsorbed from the atmosphere by passing air or a mixture of air and effluent gas, through a sorbent bed, preferably at or close to ambient conditions. Once the CO₂ has been adsorbed by the sorbent, the CO₂ has to be collected, and the sorbent regenerated. The latter step is performed by heating the sorbent with steam in the sealed containment box to release the CO₂ and regenerate the sorbent The CO₂ is collected from the box, and the sorbent is then available to re-adsorb CO₂ from the atmosphere. The only primary limitation on the process is that the sorbent can be de-activated if exposed to air if it is at a "too high" temperature. Thus the sorbent may have to be cooled before the monolith leaves the box and is returned to the air stream.

[0067] Generally, a longer time is required for adsorption of CO₂ from ambient air than for the release of the CO₂ in the regeneration step. With the current generation of sorbent this difference will require an adsorption period approximately ten times greater for the adsorption step compared with that required for CO₂ release and sorbent regeneration, when treating ambient air. Thus a system with ten monoliths and a single regeneration unit has been adopted as the current basis for an individual rotating system. If the performance of the sorbent is improved over time, this ratio of adsorption time to desorption time, and thus the number of monoliths, required in a system, should be reduced. In particular, if a higher loading embodiment of the sorbent is used a one hour adsorption time would be viable, thus requiring one regeneration box to serve only five monoliths. In addition the relative treatment times will vary with the concentration of CO₂ in the gas mixture treated, such that the higher the CO₂ content, the shorter the adsorption time relative to the regeneration time, e.g., by mixing a combustion effluent ("flue gas") with the ambient air through a gas blender. [0068] The chemical and physical activity within the monoliths, both during the adsorption cycle and the regeneration cycle in the sealed box, is substantially the same as is described in prior copending applications Nos. 13/886,207 and 13/925,679 and 15/898531. The disclosures of those copending applications are incorporated by reference herein as if repeated in full, as modified by the new disclosure presented herein. In the system according to the present invention, each rotating system provides one seal able regeneration box for each group of rotating monoliths, the number of monoliths being dependent upon the relative times to achieve the desired adsorption and the desired regeneration. In addition, it has been found that greater efficiencies and lower costs are achieved by spatially relating and temporally operating two of the moving systems in a suitable relationship to allow the regeneration boxes for the two rotating monolith systems to interact, such that each is preheated by the remaining heat in the other as a result of the prior regeneration in the other; this also efficiently cools down the regenerated monolith before it is returned to its adsorption cycle on the rotating track.

[0069] This interaction between the regeneration boxes is achieved in accordance with this invention, by lowering the pressure of the first box system so that the steam and water remaining in the first box evaporate after the release of CO₂, and the system cools to the saturation temperature of the steam at its lowered partial pressure. Furthermore, as described below, the heat released in this process is used to pre-heat the second sorbent bed and thus provides approximately 50% sensible heat recovery, with a beneficial impact on energy and water use. This concept can be used even if an oxygen resistant sorbent is utilized. The sensitivity of the sorbent to oxygen de-activation at higher temperatures is being addressed during the development process and it is anticipated that its performance will be improved over time.

[0070] As discussed above, the sorbent bed is preferably cooled before it is exposed to air so as to avoid de-activation by the oxygen in the air. This cooling is achieved by lowering the system pressure and thus lowering the steam saturation temperature. This has been shown to be effective in eliminating the sorbent deactivation issue as it lowers the temperature of the system. There is thus a significant amount of energy removed from the bed that is cooled during the de-pressurization step. A fresh bed that has finished its CO₂ adsorption step has to be heated to release the CO₂ and regenerate the sorbent This heat could be provided solely by the atmospheric pressure steam, but this is an additional operating cost. In order to minimize this operating cost, a two-bed design concept has been developed. In this concept the heat that is removed from the box that is being cooled by reducing the system pressure, and thus the steam saturation temperature, is used to partially pre-heat a second box containing a bed that has finished adsorbing CO₂ from the air and which is to be heated to start the CO₂ removal and sorbent regeneration step. Thus the steam usage is reduced by using heat from the cooling of the first box to increase the temperature of the second box. The remaining heat duty for the second box is achieved by adding steam, preferably at atmospheric pressure. This process is repeated for the other rotating monoliths in each of the two boxes and improves the thermal efficiency of the system.

[0071] These and other features of this invention are described in, or are apparent from, the following detailed description, and the accompanying drawings.

BRIEF DESCRIPTION OF THE FIGURES AND EXHIBITS

[0072] FIG. 1 is a diagrammatic top view of a mutually interactive pair of rotating multi- monolith systems for removing carbon dioxide from the atmosphere according to an exemplary embodiment of this invention;

[0073] FIG. 2 is a diagrammatic elevation view of the rotating multi-monolith system of FIG. 1 for removing carbon dioxide from the atmosphere according to an exemplary embodiment of this invention;

[0074] FIG. 3 is a diagrammatic top view of an alternative mutually interactive pair of rotating multi-monolith systems for removing carbon dioxide from the atmosphere according to another exemplary embodiment of this invention;

[0075] FIG. 4 is a diagrammatic elevation view of the rotating multi-monolith system of FIG. 3 for removing carbon dioxide from the atmosphere according to that exemplary embodiment of this invention;

[0076] FIGS. S and 5 A-H are schematic illustrations of a vertical offset version of a pair of regenerating chambers for removing carbon dioxide from the monolith medium of FIGS. 1 through 4, utilizing a vertical motion system or elevator to move the monolith between the rotating track level, upper air contact position (where the air movement is aided by a mechanical blower) and the vertically offset regeneration chamber position;

[0077] FIG. 6 is a top plan [schematic elevation] view of the regeneration chambers and monoliths on adjacent monolith systems showing the piping system arrangement for each chamber and between the chambers;

[0078] FIGS. 7A and B are schematic elevation views showing fans which are stationary and which rotate with each monolith, respectively;

[0079] FIG. 8A is a diagrammatic side elevation view of a Design for Dual Induced Axial Fans and Plenums of FIGS. 7 A, B;

[0080] FIG. 8B is a diagrammatic front elevation view of a Design for Dual Induced Axial Fans and Plenums of FIGS. 7 A, B; [0081] FIG. 9 is a diagrammatic cut-away elevation view of the Design for Dual Induced Axial Fans and Plenums of FIG. 8B, along lines 9-9;

[0082] FIGS. 10A, 10B and 10C depict the Design of Seal Systems on the monoliths, depending on the location of the regeneration position, where the Angles and Dimensions are Exaggerated for Explanation Purposes;

[0083] FIG. 11 is a diagrammatic top view of a mutually interactive pair of rotating multi- monolith systems for removing carbon dioxide from the atmosphere according to another exemplary embodiment of this invention; and

[0084] FIG. 12 is a diagrammatic elevation view of the mutually interactive pair of rotating multi-monolith system, taken along lines 11-11 of FIG. 11, for removing carbon dioxide from the atmosphere.

[0085] FIGS. 13(a)-(c) are schematic diagrams for a suitable porous substrate interior surface of a channel, showing the supported amine adsorbent in the pores of the substrate; and

[0086] Figs.14 (a) and (b) diagrammatically depict a portion of the front or rear major surfaces of the monoliths useful in this invention, showing the parallel channel openings and the thin walls separating the openings for the laminar flow of ambient air or mixtures of ambient air and a minor portion of an effluent gas containing a higher percentage of CO₂.

MORE DETAILED DESCRIPTION OF AN ASPECT OF THE INVENTION

[0087] A conceptual design for one embodiment of a system to perform the present invention operations is shown in Figures 1 and 2. A slight variation on the concept is shown in Figures 3 and 4. The overall conceptual design is discussed above, and a detailed discussion of the operation and the ancillary equipment that will be required is set out below.

[0088] In this embodiment, there are ten "monoliths" located in a decagon arrangement and which are located on a circular track. There are two circular/decagon assemblies associated with each process unit and they interact with each other (see Figures 1-4). Air is passed through the Channels extending between the 2 major surfaces of the monoliths by induced draft fans located on the inner sides of the monoliths. At one location one monolith is in a position adjacent to a single sealable chamber box, into which each monolith is inserted, as shown by vertically moving the bed out from the track, for processing (i.e. where they are heated to a temperature of not greater than 130 C, and more preferably not above 120 C, and most preferably not above 100° C, preferably with precise heat steam to release the CO₂ from the sorbent and regenerate the sorbent). Alternatively, the box can be on grade. In this embodiment, the adsorption time for adsorbing CO₂ by the monolith is ten times as long as sorbent regeneration time.

[0089] It should be understood that although the use of porous monoliths is preferred, it is feasible to use stationary beds of porous particulate, or granular, material supported within a frame, in place of the monolith. In both cases the porous substrate supports an amine sorbent for CO₂, when the bed has the same surface area as the monolith for supporting the adsorbent.

Mechanical Requirements

[0090] FIGS. 1-4, 11 and 12 show the basic operational concepts of the system. There are ten "monoliths" 21, 22 located in each decagon assembly arrangement and which are movably supported on a circular track 31, 33. There are two circular/decagon assemblies A, B associated with each process unit and they interact with each other. Air is passed through each of the monoliths 21, 22 by induced draft fans 23, 26, located radially interiorly of each of the decagon assemblies, and inducing a flow of air out of the inner circumferential surface of each monolith, and up away from the system. At one location along each track 31, 33, one monolith 21, 22 is adjacent to a sealable regeneration box 25, 27 into which one monoliths 21, 22 is inserted for regeneration processing after having completed one rotation around the track.

[0091] Thus, as shown in FIGS. 1 and 2, a first Bed 21 is rotated into position beneath the regeneration box 25 and then moved vertically upwardly into the box 25 for processing; or if the box 27 is located below grade, FIG. 4, the bed 22, is then moved vertically downwardly into the box 127 for processing; or if on grade, assembly is rotated to move the Bed 21, 22 out of the box 27, so that Bed 21 , 22 is in position when movement along the track is halted for all of the monoliths. When the Bed 21 has been regenerated it is moved back onto the track and the bed assembly is rotated, so that the next Bed 21 -2, 22-2 is in position. Bed 22 is then moved into the box for processing and then returned to the ring. This process is repeated continually. The two ring assemblies operate together, although the monoliths for each decagon are moved in and out of their boxes at slightly different times, as explained below, to allow for the passage of heat, e.g., between Box 25 and Box 27, when regeneration in one is completed to provide for preheating of the other box. This saves heat at the beginning of the regeneration and reduces cost of cooling the bed after regeneration.

[0092] Three locations for the regneration boxes 25, 27 are presented. In FIGS. 1 and 2, the regeneration boxes 25, 27 are placed above the rotating bed assemblies (at nominal grade) and the monoliths are moved vertically up into the boxes for regeneration. The only elevated structure is that required for the boxes, which are located above the rotating monoliths on a cantilevered structure.

[0093] In Figures 3 and 4 the regeneration boxes 125, 127 are located below grade and under the rotating bed assemblies. The boxes would be located in a single excavation with adequate access for maintenance and process piping. The monoliths 121, 122 are moved vertically downwardly into the boxes.

[0094] In Figures 11 and 12 the regeneration boxes 321, 327 are located on grade with the rotating bed assemblies. The boxes would be located with adequate access for maintenance and process piping also on grade. Suitable mutually sealing surfaces would be located on the regeneration box and on each bed, so that as the bed rotates into position in the box, the box 322, 327 is sealed. [0095] In all cases ancillary equipment (such as pumps, control systems, etc.) would preferably be located at grade within the circumference of the track supporting the rotating bed assemblies 29, 39. The regeneration boxes could be located in different levels, in particular situations without departing from the concept of this invention.

[0096] These designs, compared to prior disclosed apparatus in the prior art, would:

• Minimize structural steel;

• Place all major equipment at grade level apart from the regeneration boxes which are only acting as containment vessels;

• Ensure that there is no interference with air flow to the monoliths, where the boxes are at different levels from the track;

• Only require one or no vertical movement equipment for the monoliths, for insertion into the single box for each group of, e.g., 10, monoliths;

• Minimize or eliminate the time required for bed movements in and out of box, especially when the boxes are on grade;

• Allow all piping to be in fixed positions; and

• Allow the two regeneration boxes to be adjacent to each other with minimum clearance to permit the heat exchange desirable for increased efficiency.

[0097] The mechanical operations, with necessary machinery and power, that are required include:

• Rotation of the two sets of bed assemblies around a circular track on a support structure

• Precise locating elements to precisely locate the position where the monoliths are to be stopped so as to ensure the tree movement of the monoliths into and out of the regeneration box

• Removal of the bed from the bed assembly on the track, insertion of the bed into the regeneration box, removal of the bed from the regeneration box and reinsertion of the bed into its position on the track assembly. All of these movements occurring in a vertical direction, or alternatively as part of the horizontal rotational movement on the track. The monoliths and regeneration boxes are designed so that, for vertically movable monoliths there is an air-tight seal between the top or bottom of each monolith and the support structure of the box. Examples of some conceptual designs for such seals are shown in Figure 10.

[0098] In all cases, referring to FIGS. 1-6, a Bed 21-1 (Ring A) is rotated into position and then moved up or downwardly into the Box 25 for processing. The pressure in Box 25 (containing Bed 21-1, Ring A) is reduced using, e.g., a vacuum pump 230, to less than 0.2 BarA. The Box 25 is heated with steam at atmospheric pressure through line 235 and CO₂ is generated from Bed 21-1 and removed through the outlet piping 237 f om the Box 25 for the CO₂ and condensate which is separated on a condenser 240 (FIG. 5A). Bed 22-1 (Ring B) is then placed in Box 27 (Ring B) while Box 25 is being processed, as above (FIG. 5B). The steam supply to Box 25 is stopped and the outlet piping for the CO₂ and condensate isolated. Box 25 and Box 27 are connected by opening valve 126 in connecting piping 125 (FIG. 5C).

[0099] The pressure in Box 27 is lowered using a vacuum pump 330 associated with Box 27. This lowers the system pressure in both boxes and draws the steam and inerts remaining in Box 25 through Box 27 and then to the vacuum pump. This cools Box 25 (and thus Bed 21- 1 Ring A) to a lower temperature (i.e. the saturation temperature at the partial pressure of the steam in the box) and reduces the potential for oxygen deactivation of the sorbent when the Bed 21-1 is placed back in the air stream. This process also pre-heats Box 27 (and thus Bed 22-1 Ring B) from ambient temperature up to the saturation temperature at the partial pressure of the steam in the box 250. Thus energy has been recovered and the amount of atmospheric pressure steam required to heat the second Box 27 (and Bed 22-1 Ring B) is reduced (FIG. 5D). As the vacuum pump 330 lowers pressure in the Boxes 25 and 27, the first Box 25 is reduced in temperature (from 100° C. approx. to some intermediate temperature) and the second Box 27 is increased in temperature (from ambient to the same intermediate temperature). CO₂ and inerts are removed from the system by the vacuum pump 330.

[00100] The valve between the first Box 25 and the second Box 27 is closed and the boxes isolated from each other. Bed 21-1 Ring A is now cooled below the temperature where oxygen deactivation of the sorbent is of concern when the bed is placed back in the air stream. The second Box 27 and Bed 22-1, Ring B, have been preheated and thus the amount of steam required for heating the Box and Bed is reduced (FIG. SE). Bed 21-1 Ring A is then raised back into the bed assembly. The Ring A bed assembly is rotated by one bed and Bed 21-2 Ring A is then inserted into Box 25, where it is ready for preheating. Box 27 is heated with atmospheric steam and the stripped CO₂ is collected (FIG. SF).

[00101] When the second Box 27 (containing Bed 22-1 Ring B) has been fully regenerated the steam supply to Box B is isolated and the piping for the CO₂ and condensate is isolated using valves 241, 242. The valving 126 between the first Box 25 and the second Box 27 is opened and the pressure in the Boxes 25, 27 is reduced using the vacuum pump 230 system for Box 25. The temperature of the second Box 27 (and thus Bed 22-1 , Ring B) is reduced (see 5 above). The temperature of the first Box 25 (containing Bed 21-2, Ring A) is increased (see 5 above) (FIG. 5G). The vacuum pump 230 lowers pressure in Boxes 25, 27. Box 25 is reduced in temperature (from 100°C approx. to some intermediate temperature). Box 27 is increased in temperature, (from ambient to the same intermediate temperature). CO₂ and inerts are removed from the system by the vacuum pump 230. Bed 22-1, Ring B, is raised back into the ring assembly and the assembly rotated one bed. Bed 22-2, Ring B, is then inserted into Box 27. Box 25 (containing Bed 21-2 Ring A) is heated with atmospheric steam to release the CO₂ and regenerate the sorbent (FIG. 5H). The pre-heating of Box 27 then occurs as described above. The process is repeated for all of the beds as the Decagons are rotated many times.

Design Parameters

[00102] The current basis for the design of the system is as follows:

Weight of individual monolith to be moved: 1,500 - 10,000 lbs. (including support structure)

Approximate size of bed: Width - 5-6 meters

Height - 9-10 meters

Depth - 0.15-1 meter

[00103] It should be noted that the bed dimensions could be adjusted depending upon the particular conditions at the geographic location of each pair of systems, and the desired, or attainable, processing parameters. [00104] For a system including 10 monoliths in each of the Decagon rings, the outer dimensions of a preferred circular/decagon structure would be about 15-17 meters, preferably about 16.5 meters. The monolith support structures could be individually driven, for example by an electric motor and drive wheel along the track, or the support structures could be secured to a specific location along the track and a single large motor used to drive the track and all of the structures around the closed loop. In either case, the regeneration box is placed at one location and all of the structures can stop their movement when one of the support structures is so placed as to be moved into the regeneration box. The economics of a single drive motor or engine, or multiple drive motors or engines, will depend on many factors, such as location and whether the driving will be accomplished by an electrical motor or by some fuel driven engine. The nature of the driving units is not itself a feature of this invention, and are all well-known to persons skilled in the art. Examples of suitable engines include internal or external combustion engines or gas pressure driven engines, for example operating using the Stilling engine cycle, or process steam engines or hydraulic or pneumatic engines.

[00105] When a regeneration box is located above the track level, the top will be about 20 meters above the grade of the track, and when the regeneration box is located below the grade of the track, the top of the box will be immediately below the track grade. A box on grade will only be minimally above the tops of the monoliths, so as to accommodate the monolith wholly within the box during regeneration.

[00106] Where the regeneration box is not on grade, the elevator system for moving the monolith into and out of the regeneration box should be able to accomplish the movement into and out of the box during a period within the range of 30 seconds to 120 seconds, and preferably between 30 and 45 seconds. The shorter the time period, the greater the flexibility in the process parameters that are available for the process. It is recognized that there are certain inherent mechanical limitations in moving the massive monoliths. One advantage where the regeneration box is on grade, is that vertical movement is not needed, as the monolith merely rotates into the box, as part of its rotational movement, and seals; thus avoiding the vertical movement, the loss of time and the additional capital cost of the elevators. In each case, the two edges of the bed are solid and form seals with the edges of the regeneration box.

Operational and Design Details [00107] This section is divided into the following sub-sections:

• Section i - Description of the overall system design and the use of the carburetor system for energy recovery

• Section ii - Process description including simplified PFD and description of major items of equipment

• Section iii - Conceptual mechanical design

• Section iv - Issues that have to be examined in more detail to arrive at a final optimized design

Discussion i. CO₂ Adsorption and Removal Process

[00108] In the process of this invention, CO₂ is adsorbed from the atmosphere by passing air, or mixtures of air and effluent gases, through a sorbent bed, suitable sorbents preferably include amines, and preferably polyamines with at least a major proportion of the amine groups on the sorbent being primary amines. Once the CO₂ has been adsorbed by the sorbent it is stripped from the sorbent and collected, while the sorbent is regenerated. This step is performed by heating the sorbent with steam in a sealed containment, or regeneration, box. This releases the CO₂ and regenerates the sorbent. The CO₂ is collected and the sorbent is then available to re-adsorb CO₂ from the atmosphere. A limiting parameter on the process is that the sorbent can be de-activated if exposed to air at too elevated a temperature. Thus, usually the sorbent has to be cooled before it is returned to contacting the air stream. This is achieved, in accordance with the present invention, by lowering the pressure of the system so that the steam and water remaining in the regeneration box after the release of CO₂ evaporate, thus cooling the system to the saturation temperature of the steam at its new lowered partial pressure. Furthermore, as described below, the heat released in this process is used to pre-heat a CO₂- loaded sorbent bed, so as to provide approximately 50% sensible heat recovery, with a beneficial impact on energy and water use. This concept is useful even if an oxygen resistant sorbent is utilized to further lengthen the effective life of the sorbent and of the monolith substrate. [00109] Generally, a longer time is required for adsorption of CO₂ from the air by the sorbent, than is required for the release of the CO₂ in the regeneration step. With the current generation of sorbent this difference will require an adsorption period approximately ten times greater for the adsorption step compared with that required for CO₂ release and sorbent regeneration. Thus a system with ten monoliths and a single regeneration unit has been adopted as the current basis. If a sorbent is operating in a system where it will have an adsorption period only approximately five times greater for the adsorption step compared with that required for CO₂ release and sorbent regeneration, the number of monoliths required in a system, for each regeneration box, could be reduced, e.g., to one regeneration box to serve 5 monoliths. This also depends upon the concentration of CO₂ in the gas mixture being treated, and the desorption period for any particular sorbent

[00110] As discussed above, the regenerated sorbent bed is preferably cooled before it is exposed to air so as to avoid potential de-activation by the oxygen in the air. In accordance with this invention, this cooling is achieved by lowering the system pressure in the regeneration box, after regeneration has occurred, thus lowering the steam saturation temperature. According to this invention, this is accomplished in a way that a significant amount of energy removed from the regenerated monolith during the de-pressurization step, is transferred to a second bed containing CC -loaded sorbent prior to its desorption step, thus providing some of the energy to heat the second bed to release the CO₂ and regenerate the sorbent. This heat transfer from one regeneration box to a second reduces the operating cost of providing solely fresh steam to heat the monolith bed. The remaining heat duty for the second box is achieved by adding atmospheric steam, but less is required thus saving costs. This process is repeated for alternate monoliths in each of the two boxes and improves the overall thermal efficiency of the system. This concept is shown in Figures 1 through 6, 1 land 12.

[00111] In the preferred embodiment as shown in these drawings, there are ten "monoliths" located in a decagon arrangement and which are located on a circular track. There are two circular/decagon assemblies associated with each process unit and they interact with each other (see Figure 1 and Figures SA-SH). Air is passed through the monoliths by induced draft fans preferably located opposite the radially inner surfaces of the monoliths. At one location the monoliths are adjacent to a box into which the monoliths are inserted, as shown by vertically moving the bed out from the track, for processing (i.e. where they are heated with steam to release the CO₂ from the sorbent and regenerate the sorbent). Alternatively, the box can be on grade, so that the monolith merely moves along the track into the regeneration box 1 or moves outwardly from the track, into a box, and on grade. The latter method reduces the energy used in moving the bed, while allowing the two regeneration boxes to be located adjacent, closer to each other.

[00112] The basic operational steps for the systems of Figures 1-4 and 11-12 as defined above would thus be:

1. Bed 21-1 (Ring A) after making one full rotation, is rotated into position and then moved, e.g., vertically into the Box 25 for processing, FIGS. 1-4 and S.

2. Box 25 (containing Bed 21-1 (Ring A)) is heated with steam at atmospheric pressure and CO₂ generated is removed, FIG. 5A-H.

3. Bed 22-1 (Ring B) is placed in Box 27 while Box 25 is being processed to regenerate the sorbent.

4. The steam supply to Box 25 is stopped and the outlet piping for the CO₂ and condensate isolated. Box 25 and Box 27 are connected by opening valves in connecting piping 125.

5. The pressure in Box 27 is lowered using a vacuum pump 330 associated with Box 27. This lowers the system pressure in both boxes and draws the steam and inerts remaining in the regenerated Box 25 into the other Box 27 and then to the vacuum pump 330. This cools the regenerated Box 25 (and thus Bed 21-1 Ring

A) to a lower temperature (i.e. the saturation temperature at the partial pressure of the steam in the box) and reduces the potential for oxygen deactivation of the sorbent when it is placed back in the air stream. This process also heats Box 27 (and thus Bed 22-1 Ring B) from its temperature after adsorption up to the saturation temperature at the partial pressure of the steam in the box 27. Thus energy has been recovered from the regenerated Box 25, and the amount of atmospheric pressure steam required to heat Box 27 (and thus Bed 22-1 Ring

B) is reduced. 6. The valve 125 between the two Boxes 25, 27 is closed and the boxes isolated from each other. Bed 21-1, Ring A is now cooled below the temperature where oxygen deactivation of the sorbent is of concern when the bed is placed back in the air stream. The second Box 27 and Bed 22-1 Ring B have been preheated and thus the amount of steam required for heating the Box and Bed is reduced.

7. Bed 21-1 Ring A is then vertically moved back onto the Decagon track assembly. Box 27 is heated with atmospheric steam and the CO₂ is collected. The Ring A bed assembly is rotated by one bed and Bed 21-2 Ring A is then inserted into the regeneration Box 25, where it is ready for preheating. FIG. 5H.

8. When Box 27 (containing Bed 22-1 Ring B) has been fully regenerated the steam supply to Box 27 is isolated and the piping 337 for the CO₂ and condensate is closed using valves. The valving between the Box 25 and the regenerated Box 27 is opened and the pressure in Boxes 27, 25 is reduced using the vacuum pump 230 for Box 25. The temperature of Box 27 (and thus Bed 22-1 Ring B) is reduced (see 5 above). The temperature of Box 25 (containing Bed 21-2 Ring A) is increased (see 5 above).

9. Bed 22- 1 Ring B is raised back into the bed assembly and the assembly rotated one bed. Bed 22-2 Ring B is then inserted into Box 27. Box 25 (containing Bed 21-2 Ring A) is heated with atmospheric steam to release the CO₂ and regenerate the sorbent.

[00113] It is understood that reference to a "bed" includes both a monolithic substrate as well as an enclosed particulate bed held within the same size volume.

[00114] This process is repeated continually and the two ring track assemblies operate together, although the monoliths for each decagon are moved in and out of their boxes at slightly different times, so that the heat from cooling the earlier regenerated box preheats the later box when the later monolith is in place.

[00115] In Figures 1 and 2 the boxes are placed above the rotating bed assemblies (which are located at nominal grade) and the monoliths are moved up into the boxes. The only elevated structure is that required for the boxes, which are located above the rotating monoliths on a cantilevered structure.

[00116] In Figures 3 and 4 the boxes are located below grade and under the rotating bed assemblies. The boxes would be located in a single excavation with adequate access for maintenance and process piping.

[00117] In Figures 11 and 12, the boxes are located on grade, preferably over the track so that no additional vertical movement at the machinery is necessary. Alternatively, the regeneration box on grade can be located outwardly from the Decagons, and moved radially from the track.

[00118] In either case ancillary equipment (such as pumps, control systems, etc. - see section 2) would be located at grade radially inside of the rotating bed assemblies. ii. Process Equipment and Controls

[00119] Figure 6 shows the general design from the proposed system:

• There are two decagons of monoliths in a single system. Thus a single system contains 20 (twenty) monoliths.

• There are nine fan installations for each decagon (there is no set of fans at the location where the monoliths are inserted into the boxes). At present it is preferred that there will be two vertically arranged axial fans associated with each bed of the size described above, i.e., a height of 10 meters and a width of 5 meters. Thus for a single system there will be 2x 18 = 36 axial fans. However, the selection of the number and size of fans depends upon many factors.

• The nine fans per decagon each remain stationary (i.e. they will not rotate with the beds). Preferably a sealing system such as walls with a flexible end seal is provided with each fan, to minimize bypassing of the air around the monoliths. It is understood that the monoliths do not move continuously, but rather stop as one bed reaches the regeneration box location, and then restarts as that bed leaves the regeneration box. The stationary fans are located so that when a bed enters a regeneration box, each bed is located opposite to and sealed with a fan installation. Alternatively, the fans can be attached to the rotating bed structure and be fixed with the beds. In that case the number of fans would increase to 2 x 20 = 40 axial fans per single system. (See Section 3).

• There are two regeneration boxes 25, 27 in a single double track ring system; each box serves one of the decagons.

• The size of the monoliths is not standardized. As an initial estimate it should be assumed that each bed is S meters wide x 10 meters tall by 1 meter deep. This initial size can be modified based upon economic analysis and other factors.

• Only the major valving is shown in Figure 6 and additional valving, instrumentation, piping and controls are required for safe commercial operation, which are well known to the ail.

[00120] During regeneration and CO₂ release f om a bed, steam at atmospheric pressure and a temperature of 100°C - 120°C is supplied directly to the regeneration Box 25, 27 containing the bed. The effect of the steam is to heat the bed and the box and release CO₂ and produce condensate. The condensate is removed to a collection system. The CO₂ is removed from the box, together with some steam and inerts, by the action of the CO₂ Blower 225, 227. The exhaust stream from the box is passed through a heat exchanger (condenser) 240 where the stream is cooled and further condensate is produced, which is sent to the condensate collection system 291. Finally the product CO₂ is sent via line 229 to storage and compression or can be used directly in another process, such as algae growth, without compression. The compression of the CO₂ is not included in the scope of this process description. Preferably, the air is at least partially withdrawn from the regeneration box 25, 27, after it is sealed with the bed, before the steam flow is started, especially where the CO₂ is to be compressed. Preferably, the pressure in the sealed regeneration box is reduced to not greater than 0.2 Bar A before feeding the steam and stripping the CO₂. It is preferred that as much of the non-condensibles from air be removed as feasible, in order to reduce the cost of compression. [00121] It is desirable to reduce the amount of water in the CO₂ exhaust stream after the condenser, as the more water present the higher will be the compression costs associated with storing the CO₂ product; more condensate will have to be removed in the inter-stage coolers of the compressors if not removed upstream. The amount of steam left in the exhaust stream sent to storage will be a function of the lowest temperature of coolant that is available and the size of the condenser that is installed. Determination of these values in any particular case is based upon an economic assessment of the relative costs of compression (capital and operating), coolant temperature (e.g. whether to use ambient air, cooling water or a refrigerant) and capital cost of the heat exchanger.

[00122] If correctly designed, the condenser should also be able to separate the liquid and vapor streams. However, under certain circumstances, a knock-out drum or similar type unit may be required to separate the liquid and vapor streams before the vapor stream is passed to the CO₂ Blower 225, 227.

[00123] The CO₂ Blower 225, 227 could be a liquid ring pump. If that type of unit is selected then it will be able to handle liquid condensate in the incoming feed and the condensate will be eliminated from the liquid ring system and sent to condensate storage. If a liquid ring type pump unit is not used then additional steps may be required to ensure that the vapor stream entering the blower does not contain a significant amount of liquid. Therefore, the selection of the type of unit used for the CO₂ Blower may have an impact on the design of the upstream equipment.

[00124] When the regeneration step is completed, all valving is closed and thus both boxes are isolated. In order to next cool the box and bed that have just finished the CO₂ release and sorbent regeneration step and pre-heat the other box and bed, which are at ambient temperature the following steps occur:

• The isolation valve 126 between the boxes is opened

• The vacuum pump 230, 330 associated with the bed at ambient conditions is turned on

• The effect of the vacuum pump is to draw the steam (initially at, e.g., atmospheric pressure and approximately 100°C) from the box that has finished CO₂ production and bed regeneration (the "hot" box), into the box at ambient temperature. The lower pressure will cool the hot regenerated box and regenerated bed to a temperature substantially below the initial temperature after regeneration, i.e., approximately 100°C, due to the reduction in partial pressure of the steam which reduces the saturation temperature of the steam. As the vapor and steam are drawn from the "hot" box and bed this stream will start to heat the second box and bed (initially at ambient temperature) due to condensation of the steam on the walls of the box and inside the channels of the sorbent bed. As the vacuum pump operation continues, the pressure in both boxes decreases and reaches a final pressure (approximately 0.2 Bar A in the current example). At this point both boxes and their monoliths will be at approximately the same temperature (approximately 60°C in the currently example). Thus the "hot" bed has been cooled to a temperature where, when it is returned to the air stream for further CO₂ adsorption, the sorbent will not be deactivated to any significant extent by the presence of oxygen in the air. Simultaneously, the bed at ambient temperature has been provided with a significant proportion of the heat needed to raise its temperature to approximately 100°C for the CO₂ stripping from, and regeneration of, the sorbent. The final pressure to which the combined boxes will be brought is determined by the temperature restrictions on the sorbent in the presence of oxygen.

• Once the defined pressure level in both boxes 25, 27 is reached the vacuum pump 230, 330 is stopped, the isolation valve 126 between the boxes is closed and the regeneration bed is returned to atmospheric pressure.

• The cooled bed is returned to the ring track assembly, which assembly rotates until the next bed is moved into position to enter the box, and the rotation then stops.

• The second box and bed in the second box 25, 27 that were pre-heated to approximately 60°C, is in the meantime supplied with atmospheric pressure steam and heated to 100°C for CO₂ removal and sorbent regeneration. The CO_¾ steam and inerts are removed by the CO₂ Vacuum Blower 225, 227 associated with that Box. (See text above and FIG. 6). • The process is then repeated continually, to alternatingly regenerate Boxes 25, 27.

[00125] It is possible that only a single CO₂ Blower and a single CO₂ Vacuum Pump could be used for each pair of regeneration boxes, a separate blower and pump for each box, or a central system, i.e. a single CO₂ Vacuum Pump 230, 330 and a single CO₂ Blower 225, 227 could be used to serve multiple system pairs.

[00126] Figures 1 and 2 show the conceptual mechanical design where there are two decagons in each system and where the beds are raised into or from the boxes which are located above the circular track system and supported by a cantilevered structural steel structure. Figures 3 and 4 show a similar concept except that the boxes are located below grade in a single excavation and the boxes are lowered into the boxes. It is also feasible to have the box on grade, and merely rotate each bed into a sealed relationship with the box, as the ring rotates and then stops when the bed is sealed in the regeneration box.

[00127] Figure 7A shows the conceptual design of the fan support system for the induced draft axial fans. Vertical walls 38 extending from each edge of the beds to a location radially inwardly of the fans (only one such wall is shown in FIG. 7A) along with a surface seal 136 where the walls contact the edge of the beds, plus top and bottom surfaces 36, 37 shown in cross-section, extending between the vertical walls, will prevent air from bypassing around the beds 21, 22, with the fans 26 remaining in a fixed position. Preferably, each of the walls 38 and top 36 and bottom 37 surfaces are provided with an elastomer bumper 136 that would not contact the front of the bed 22 but which would press against the edges of the bed when the bed 21 was fully rotated into the air capture position.

[00128] Figure 7B shows a conceptual design where the fans 326 are rotated with their associated monoliths 21. This would require the fan support structures to be part of the ring rotation system and would increase the power required for rotating the monoliths, particularly the initial torque required to start the rotation. This option would allow the bypassing of air around the bed to be eliminated as the seals would be permanent and would not have to move. [00129] Figures 8A, B and 9 show a conceptual arrangement of the fans 326 and plenums 425 that could be employed to ensure even distribution of the air across the monoliths using two fans per bed, when the beds are 10 meters tall.

[00130] The mechanical operations that will be required of the positioning system to ensure that the monoliths will be moved into and out of the boxes precisely include:

• Rotation of the two sets of bed assemblies around a circular track on a support structure.

• Precise location of the position where the monoliths are to be stopped so as to ensure the free movement of the monoliths into and out of the regeneration boxes, and into and out of the sealable relationships with the air guidance walls and seals, when the fans are stationary.

• Removal of the bed from the bed assembly, insertion of the bed into the regeneration box, removal of the bed from the box and re-insertion of the bed onto the circular track assembly, where the bed is to be vertically moved. When the regeneration box is on grade, removal of the bed would not be necessary.

[00131] The monoliths are to be designed so that there is an air-tight seal between the monoliths and the internals of the box, and between the bed and the fan support structure when in the positions where air is passed through the bed. FIGS. 10A, 10B and 10C show conceptual designs for a side by side tapered seal system that will seal the bed in either the upper and lower regeneration box (Fig. 10A) positions of a regeneration box (Fig. 10B). Fig. 10C depicts an elevation side view.

[00132] Two seal systems are installed side by side on each bed frame, each matched with a channel 150 in a regeneration box. One channel is in the box and the other channel is in the ring assembly where the bed is located for CO₂ removal from the air stream.

[00133] Each of the channels 150 into which the seals will pass is also tapered. When inserted upwards the seal used is narrow at the top - relative to the channel which is wide at the bottom relative to the seal. This results in a tolerance for the seal to be inserted into the channel in which it will slide and seal. The channel into which the seal slides is also tapered to match the taper of the seal. As the bed is raised the gap between the channel and the seal narrows. This both gradually centers the bed in the correct location and also gradually decreases the gap between the seal and the channel . When fully raised the seal and the channel are the same width from top to bottom, the seal is tight against the channel, producing the seal, and the bed is located in exactly the correct position.

[00134] When inserted downwards, the other seal is used which is narrow at the bottom, which allows a tolerance for the seal to be inserted into the tapered channel (which is wide relative to the seal) and has the same taper as the seal) in the lower position within which it will slide and seal. As for the seal operation in the upward direction, the gap between the seal and the tapered channel will decrease as the bed moves into position, centering the bed and producing the required seal. In addition, there is also a seal focused between the bottom of the bed and the bottom of the regeneration box above the track and the top of the bed and the top of the regeneration box when the box is below the track as in FIGS. 3 and 4. When the regeneration box is on grade as in FIGS. 11-12, the edges or sides of the bed for the seal.

[00135] When designing the elevator system for vertical movement of the bed, either up or down, the approximate time period desired for bed vertical movement, for monoliths weighing about 10,000 lbs, and having the dimensions 5 ms x 10ms x lm, between the track and the box - is 30 seconds to 120 seconds. The shorter this time period, the greater the flexibility in the process parameters that is available for the development of the process. It is for this reason that a regeneration box on grade holds some advantages.

4.1 Sorbent Properties and Bed Thickness

[00136] It should be understood that the specific dimensions and other numerical parameters set out above are based upon the use of the now conventional Polyethyleneamine ("PEA") as the sorbent. As improved sorbents are realized, that adsorb more quickly and/or are less susceptible to the effects of oxygen at elevated temperatures, for example, dimensions and temperatures of operation, as well as the number of beds per regeneration box and the speed of the beds around the track can change.

[00137] At present the pressure drop through the sorbent bed (which is usually a porous silica or alumina substrate with PEI present on its surfaces) is preferably limited to 1 inch H₂O and, given the current structure of the sorbent bed and the superficial air velocity used for the design (2.5 m/s in the free duct) results in a defined depth (in the direction of air flow) for the bed. This, in turn, affects the depth of the box. The assumed pressure drop, bed porosity, flow channel size, superficial air velocity can all be modified with changes in the sorbent and/or the substrate, so that in conjunction with the sorbent performance, that can result in a different prefered bed depth. One improved system is achieved by using a substrate formed from an alumina-coated silica with a primary amine polymer, such as a poly(allyl)amine, or one of its derivatives, coated on its surfaces.

4.2 Minimum Design Pressure - Regeneration Boxes

[00138] The most significant effect of the minimum design pressure selected will be on the cost of the boxes used for heating the sorbent monoliths. The minimum design pressure is selected based upon achieving a steam saturation temperature (at the steam partial pressure in the box at the minimum design pressure) such that the bed is cooled below the temperature at which significant deactivation of the sorbent occurs when it is exposed to oxygen in the air stream. The lower the pressure the thicker the plates and heavier the stiffening structures required for the box. Utilizing a primary polyamine, such as poly(allyl)amine, as now generally available, preferably the current minimum design pressure of 0.2 Bar A the box is required to be a large, heavy and expensive item of equipment even with a bed size of approximately 3m x 5m x lm. In a commercial unit it would be desirable to have a larger bed. However, as the bed size is increased the weight and cost of the box will increase in a power relationship (not linearly) with the dimensions of the box. In addition, a higher minimum design pressure would allow a greater amount of heat recovery, as the "cold" box could be heated to a higher temperature and less atmospheric steam would be required. Thus, being able to use a higher minimum design pressure (i.e. greater than 0.2 Bar A) would bring significant advantages, if a sorbent is used that would not be deactivated at the higher temperature.

[00139] The above parameters assume that a certain constant loading (of sorbent groups, e.g., primary amine groups) is achieved. In addition, the velocity of the air coming in was assumed constant in the comments above. It must be understood that the pressure drop per tonne of CO₂ capture increases as the velocity of the air flow increases, which increases the cost of the electricity to move the air, to the extent natural forces, such as the wind, are not sufficient to achieve the desired airflow. The cost of the whole process other than the electricity cost decreases as the airflow velocity increases. Thus, the air velocity choice is a compromise between capital cost, which is reduced as the airflow velocity increases, and operating costs, that increase as the airflow velocity increases. It is preferred to operate with an incoming airflow in the range of 2-4m/sec. The relative costs will vary depending upon the local conditions at each plant site, e.g. is there a dependable prevailing wind present or not and the local cost of electricity.

[00140] In general as one increases the loading one also wants high amine efficiency as defined by the fraction of amine sites present that are available to bind the CO₂. This is the reason for preferring primary amines and also for adjusting the loading so as to minimize pore blockage. Experimental results indicate that the optimum loading that balances amine efficiency with increased loading is between 40 - 60 % by volume organic amine content relative to the porous substrate/skeleton to which it is attached or into whose pores it is deposited.

[00141] Stripping of the CO₂ from the sorbent is equally practical at relatively low temperatures. Thus, this invention contemplates the use of other sorbents having the desirable properties of the primary amines with respect to the air capture of CO₂; such sorbents would be used in the invention of the process described in this application.

[00142] The primary amines work effectively at air capture (from atmospheric air) concentrations under ambient conditions. The loading of CO₂ depends strongly upon the ratio of the heat of reaction/ (boltzmann constant) T (temperature); the heat of reaction difference between primary and secondary amines, as shown above, can cause a factor of about 100 times difference in loading, following the well-known langmuir isotherm equation. The amine groups are preferably supported upon a highly porous skeleton, which has a high affinity to the amines or upon which, or in which, the amines can be deposited.

[00143] Alternatively, the amine groups may be part of a polymer that itself forms the highly porous skeleton structure. A highly porous alumina structure is very effective when used as the skeleton to support the amines. This ceramic skeleton has a pore volume and surface to achieve high loadings of amines in mmoles of amine nitrogen sites per gram of porous material substrate. Another embodiment of a preferred such skeleton support material has 230 cells per cubic inch with a thickness of six inches. Another structure that can be used is based upon a silica porous material known as cordierite and is manufactured and sold by Corning under the trademark CELCOR. CELCOR product is made with straight macro channels extending through the monolith, and the interior walls of the channels are coated with a coating of porous material, such as alumina, into the pores of which the amine can be attached or deposited(and which preferably is adherent to the amine compounds). This can be achieved by utilizing a monolith contactor skeleton that is made out of a primary amine- based polymer itself, but is also at least partially achieved by forming the structure of the monolith of alumina. Although alumina does not form as structurally durable a structure as does cordierite, for the conditions met at the ambient temperature of the air capture or the relatively low temperatures at which the CO₂ adsorbed on the amines at ambient temperatures can be stripped off, the structural strength of alumina is adequate.

[00144] When using the alumina coated CELCOR cordierite, or any monolith structure provided with channels passing the full thickness of the monolith, the length of the contactor in the direction of air flow, for a fixed pressure drop and fixed laminar air flow, and with a fixed void fraction, scales like the area of the individual square channel openings in the CELCOR monolith; and the cycle time, as determined by the sorbent becoming_saturated with CO₂ or to some fixed level of CO₂ sorption, scales with the same factor. The void fraction is the ratio of open input area to total input area of the front face of the monolith, facing the air flow. Preferably, the void fraction of the monolith is between 0.7 and 0.9, i.e., between 70% and 90% open channels.

4.3 Box Materials of Construction

[00145] When the regeneration box is constructed of carbon steel and stainless steel, it results in a structure that is heavy and expensive. Other construction materials include, for example, carbon fiber (or other man-made material), which would allow for savings in cost, as well as in weight.

4.4 Air Distribution Into and Out of Monoliths

[00146] It is essential that the air flow across the monoliths be as uniform as possible. The use of induced draft axial fans with suitably designed plenums to guide the air flow are useful in this context, and are used, for example, with petro-chemical air cooler installations. [00147] A second issue associated with the air distribution involves the velocity of the air passing out of the circle of monoliths in the decagon system. Depending upon the ratio of the height of the bed to its width, the air velocity in the plume of air rising out of the circular opening formed by the tops of the monoliths may be high, and should be considered in the design of the fan plenums.

4.5 Use of a Single Outlet Plenum with the Potential for Energy Recovery

[00148] It is understood that if the size of the monoliths were to be reduced there is the potential to use a single very large axial fan installed horizontally in the circular opening at the top of the monoliths. This would draw air through the monoliths and then move all of the air vertically out of the assembly. There would be a plenum above the fan to guide the air and prevent re-circulation. In addition, the outlet plenum could be designed to achieve some energy recovery by using a small constriction and then an expansion, as is done in cooling towers with a similar fan and plenum arrangement. If the amount of air to be moved becomes too large then this option would not be practical.

4.6 Use of Central CO₂ Blower and Condensing System and Amount of Condensing Required Prior to CO₂ Blower

[00149] In the current design there is a condenser 240 upstream of the CO₂ Blower 225. This removes water and reduces the vapor load on the blower. Alternatively, a single central condensing system can be used; that would process all of the CO₂ product streams from all of the units in multiple system pairs. This would reduce the complexity of the systems and reduce costs. However, the penalty for this would be that each CO₂ Blower would have to be designed to handle a wet vapor stream with a higher fiowrate. Bach system should be evaluated to determine the most economic option.

4.7 Use of Central CO₂ Vacuum Pump

[00150] During the de-pressurizing of the system and transferring heat from the "hot" regeneration box to the "cold" regeneration box, a CO₂ Vacuum Pump 230 is used. In the preferred design shown, a vacuum pump is associated with each regeneration box. Under certain circumstances one CO₂ Vacuum Pump can serve for both of the boxes in the two-ring system. In addition, a single large CO₂ Vacuum Pump serving multiple systems can be used. Reducing the number of vacuum pumps should reduce the capital cost associated with the system.

[00151] Preferably, the use of a liquid ring type pump would appear to be advantageous as any condensate produced will be contained in the liquid ring system and more readily removed.

4.8 Bed Removal/Sorbent Replacement

[00152] The sorbent monoliths will have to be serviced during the life of the process. This would involve maintenance activities on the bed movement systems (both rotational and vertical), replacement of the sorbent and maintenance, etc. These activities might be performed with the monoliths in position or they may require that the monoliths be removed from the assembly. Removal of the monoliths is achieved by installing a second lift system which could then move the monoliths out from the track for access. Alternatively, the monoliths could be designed to be removed using a crane. Other options are available.

[00153] With the foregoing disclosure in mind, it is believed that various other ways of operating multiple bed systems for removing carbon dioxide from a gaseous mixture, in accordance with the principles of this application, will become apparent to those skilled in the art, including the use of many conventional steps and components that are or shall become well-known and would be useful in carrying out the present invention without themselves being a part of the invention. The scope of this invention is to be determined only in accordance with the scope of the following claims.

Claims

What is claimed is:

1. A methodology for improving the performance characteristics of systems that utilize direct air capture of carbon dioxide from carbon dioxide-carrying air, comprising: directing a flow of carbon dioxide-carrying air through a carbon dioxide capture contactor structure having channels, said contactor structure supporting sorbent capable of binding carbon dioxide, to remove carbon dioxide from the air by binding carbon dioxide to the sorbent, causing said carbon dioxide capture contactor structure to be exposed to

regeneration, whereby substantially saturated steam of a predetermined temperature is directed at said carbon dioxide bound to the sorbent, thereby facilitating separation of the carbon dioxide from the sorbent and regenerating the sorbent, withdrawing the carbon dioxide that has been unbound from the sorbent together, and selectively thereafter again exposing said carbon dioxide capture structure to a flow of carbon dioxide-carrying air, thereby enabling the regenerated sorbent to be again used to bind carbon dioxide, to remove carbon dioxide from the newer flow of carbon dioxide-carrying air, said carbon dioxide capture contactor structure being formed with parallel channels of a predetermined geometry, utilizing said predetermined geometry to cause areas of laminar substantially non-turbulent flow of the carbon dioxide-carrying air, exhibiting characteristics whereby the proportionality of pressure drop with respect to the velocity of portions of carbon dioxide-carrying air moving through said channels is substantially linear as said velocity is increased.

2. The methodology of claim 1, wherein said channels are substantially parallel.

3. The methodology of claim 1 , wherein said contactor material includes one or more of the following: extruded ceramic, corrugated fiberglass sheets, corrugated metal sheets, extruded alumina, structured carbon and engineered plastics.

4. The methodology of claim 1, wherein incoming velocity is increased to Sm/sec while being dominated by relatively lower resistance laminar flow for a size S of an order of approximately 2 millimeters, and the wall thickness of the contactor W is relatively small W/S <l/4 (higher ratios raise the pressure drop), and where W is the size of the sum of the contactor substrate thickness and a coating, said substrate being of a predetermined thin dimension.

5. The methodology of claim 1, wherein there is a tradeoff between wall thickness and opening size S, with better performance with smaller S and thus smaller W, results in a higher surface area to volume SAv which will yield a higher carbon dioxide capture efficiency, whereby S and W yield 100-150 cpsi is best.

6. The methodology of claim S, wherein for smaller W the kinetics are improved, thereby improving the overall process.

7. The methodology of claim 1 , wherein as one increases the velocity to capture more C02 per year it is best to decrease the cycle time for capture and regeneration to one half their previous values, to 15 minutes and 1.5 minutes, and where higher velocities are used one is able to reduce those times further or when low cost heat is available one is able to increase TPY by going to even shorter times for both at 5 m/sec.

8. A method of cyclically removing carbon dioxide from carbon dioxide-carrying ambient air, comprising: directing a flow of carbon dioxide-carrying air through a carbon dioxide capture contactor structure having channels, said contactor structure supporting sorbent capable of binding carbon dioxide, to remove carbon dioxide from the air by binding carbon dioxide to the sorbent, causing said carbon dioxide capture contactor structure to be exposed to regeneration, whereby saturated steam of a temperature not greater than about 120 degrees C is directed at said carbon dioxide bound to the sorbent, thereby facilitating separation of the carbon dioxide from the sorbent and regenerating the sorbent, withdrawing the carbon dioxide that has been unbound from the sorbent together with any remaining steam, and selectively thereafter again exposing said carbon dioxide capture structure to a flow of carbon dioxide-carrying ambient air, thereby enabling the regenerated sorbent to be again used to bind carbon dioxide, to remove carbon dioxide from the newer flow of carbon dioxide-carrying ambient air, said carbon dioxide capture contactor structure having substantially parallel channels of a predetermined geometry that provides particularly favorable operating conditions, utilizing said predetermined geometry to cause laminar flow having characteristics whereby the proportionality of pressure drop with respect to the velocity of carbon dioxide-carrying air moving through said channels is substantially linear as said velocity is increased.