Constraints on the in situ density of CO2 within the Utsira formation from time-lapse seafloor gravity measurements

At Sleipner, CO₂ is being separated from natural gas and injected into an underground saline aquifer for environmental purposes. Uncertainty in the aquifer temperature leads to uncertainty in the in situ density of CO₂. In this study, gravity measurements were made over the injection site in 2002 and 2005 on top of 30 concrete benchmarks on the seafloor in order to constrain the in situ CO₂ density. The gravity measurements have a repeatability of 4.3 μGal for 2003 and 3.5 μGal for 2005. The resulting time-lapse uncertainty is 5.3 μGal. Unexpected benchmark motions due to local sediment scouring contribute to the uncertainty. Forward gravity models are calculated based on both 3D seismic data and reservoir simulation models. The time-lapse gravity observations best fit a high temperature forward model based on the time-lapse 3D seismics, suggesting that the average in situ CO₂ density is about to 530 kg/m³. Uncertainty in determining the average density is estimated to be ±65 kg/m³ (95% confidence), however, this does not include uncertainties in the modeling. Additional seismic surveys and future gravity measurements will put better constraints on the CO₂ density and continue to map out the CO₂ flow.     

Amine volatility in CO2 capture

Amine volatility is a key screening criterion for amines to be used in CO₂ capture. Excessive volatility may result in significant economic losses and environmental impact. It also dictates the capital cost of the water wash. This paper reports measured amine volatility in 7 m MEA (monoethanolamine), 8 m PZ (piperazine), 7 m MDEA (n-methyldiethanolamine)/2 m PZ (piperazine), 12 m EDA (ethylenediamine), and 5 m AMP (2-amino-2-methyl-1-propanol) at 40–60 °C with lean and rich loadings giving CO₂ partial pressures of 0.5 and 5 kPa at 40 °C. The amine concentrations were chosen to maximize CO₂ capture capacity at acceptable viscosity. At the lean loading condition (where volatility is of greatest interest), the amines are ranked in order of increasing volatility: 7 m MDEA/2 m PZ (6/2 ppm), 8 m PZ (8 ppm), 12 m EDA (9 ppm), 7 m MEA (31 ppm), and 5 m AMP (112 ppm). The apparent amine partial molar excess enthalpies in these systems were estimated to range from ～10 to 87 kJ/mol of amine.     

Experimental investigation of the CO2 sealing efficiency of caprocks

Using a combination of experimental (petrophysical and mineralogical) methods, the effects of high-pressure CO₂ exposure on fluid transport properties and mineralogical composition of two pelitic caprocks, a limestone and a clay-rich marl lithotype have been studied. Single and multiphase permeability tests, gas breakthrough and diffusion experiments were conducted under in situ p/T conditions on cylindrical plugs (28.5 mm diameter, 10–20 mm thickness).The capillary CO₂ sealing efficiency of the initially water-saturated sample plugs was found to decrease in repetitive gas breakthrough experiments on the same sample from 0.74 to 0.41 MPa for the limestone and from 0.64 to 0.43 MPa for the marl. Helium breakthrough experiments before and after the CO₂ tests showed a decrease in capillary threshold (snap-off) pressure from 1.81 to 0.62 MPa for the limestone.Repetitive CO₂ diffusion experiments on the marlstone revealed an increase in the effective diffusion coefficient from 7.8 × 10^?11 to 1.2 × 10^?10 m².Single-phase (water) permeability coefficients derived from steady-state permeability tests ranged between 7 and 56 nano-Darcy and showed a consistent increase after each CO₂ test cycle. Effective gas permeabilities were generally one order of magnitude lower than water permeabilities and exhibit the same trend. XRD measurements performed before and after exposure to CO₂ did not reveal any distinct change in the mineral composition for both samples. Similarly, no significant changes were observed in specific surface areas (determined by BET) and pore-size distributions (determined by mercury injection porosimetry). High-pressure CO₂ sorption experiments on powdered samples revealed significant CO₂ sorption capacities of 0.27 and 0.14 mmol/g for the marlstone and the limestone, respectively.The changes in transport parameters in the absence of detectable mineral alterations may be explained by carbonate dissolution and further precipitation along a pH profile across the sample plug which would not be subject to quantitative mineral alteration.     

Study of individual reactions of the sour compression process for the purification of oxyfuel-derived CO2

Following the feasibility study of sour compression process as a novel purification method of producing NO_x-free, SO₂-free oxyfuel-derived CO₂ using actual fluegas, in this paper, we present the study of the individual reactions taking place in the process in a controlled environment. We have previously showed that an increase of NO/NO₂ concentration in the inlet stream is beneficial for SO₂ removal as NO₂ promotes SO₂ oxidation and the further removal as liquid acid. In this study we show that the reaction SO₂ + NO₂ → SO₃ + NO does not take place significantly in the absence of liquid water at a range of conditions relevant to the sour compression process. When liquid water is present, SO₂ is oxidised by NO₂ regenerating NO with the rate of conversion of SO₂ being dependent on the acid concentration in the liquid. The formation of small liquid droplets where very low levels of pH (?0) can be reached is shown to be of great importance to the SO₂ + NO₂ conversion process.     

Carbon dioxide capture with concentrated,aqueous piperazine

Concentrated, aqueous piperazine (PZ) has been investigated as a novel amine solvent for carbon dioxide (CO₂) absorption. The CO₂ absorption rate of aqueous PZ is more than double that of 7 m MEA and the amine volatility at 40 °C ranges from 11 to 21 ppm. Thermal degradation is negligible in concentrated, aqueous PZ up to a temperature of 150 °C, a significant advantage over MEA systems. Oxidation of concentrated, aqueous PZ is appreciable in the presence of copper (4 mM), but negligible in the presence of chromium (0.6 mM), nickel (0.25 mM), iron (0.25 mM), and vanadium (0.1 mM). Initial system modeling suggests that 8 m PZ will use 10–20% less energy than 7 m MEA. The fast mass transfer and low degradation rates suggest that concentrated, aqueous PZ has the potential to be a preferred solvent for CO₂ capture.     

Pathways towards large-scale implementation of CO2 capture and storage: A case study for the Netherlands

We sketch four possible pathways how carbon dioxide capture and storage (CCS) (r)evolution may occur in the Netherlands, after which the implications in terms of CO₂ stored and avoided, costs and infrastructural requirements are quantified. CCS may play a significant role in decarbonising the Dutch energy and industrial sector, which currently emits nearly 100 Mt CO₂/year. We found that 15 Mt CO₂ could be avoided annually by 2020, provided some of the larger gas fields that become available the coming decade could be used for CO₂ storage. Halfway this century, the mitigation potential of CCS in the power sector, industry and transport fuel production is estimated at maximally 80–110 Mt CO₂/year, of which 60–80 Mt CO₂/year may be avoided at costs between 15 and 40 €/t CO₂, including transport and storage. Avoiding 30–60 Mt CO₂/year by means of CCS is considered realistic given the storage potential represented by Dutch gas fields, although it requires planning to assure that domestic storage capacity could be used for CO₂ storage. In an aggressive climate policy, avoiding another 50 Mt CO₂/year may be possible provided that nearly all capture opportunities that occur are taken. Storing such large amounts of CO₂ would only be possible if the Groningen gas field or large reservoirs in the British or Norwegian part of the North Sea will become available.     

Heat of absorption of CO2 in aqueous ammonia and ammonium carbonate/carbamate solutions

A reaction calorimeter was used to determine the enthalpies of absorption of CO₂ in aqueous ammonia and in aqueous solutions of ammonium carbonate at temperatures of 35–80 °C. The heat of absorption of CO₂ with 2.5 wt% aqueous ammonia solution was found to be about 70 kJ/mol CO₂, which is lower than that with MEA (around 85 kJ/mol) at 35 and 40 °C. The value decreases with increased loading, but not to as low a value as expected by the carbonate–bicarbonate reaction (26.88 kJ/mol). The enthalpy of absorption of CO₂ in aqueous ammonia at 60 and 80 °C decreases with loadings at first, then increases between 0.2 mol CO₂/mol NH₃ and 0.6 mol CO₂/mol NH₃, and then decreases again. The behavior of the heat of absorption of CO₂ in 10 wt% ammonium carbonate solution was found to be the same as that of aqueous ammonia at loadings above 0.6 mol CO₂/mol NH₃. The heat of absorption increases with increasing temperature. The heats of absorption are directly related to the extent of the various reactions with CO₂ and can be assessed from the species variation in the liquid phase.     

Coastal spreading of olivine to control atmospheric CO2 concentrations: A critical analysis of viability

Qualitative proposals to control atmospheric CO₂ concentrations by spreading crushed olivine rock along the Earth's coastlines, thereby accelerating weathering reactions, are presently attracting considerable attention. This paper provides a critical evaluation of the concept, demonstrating quantitatively whether or not it can contribute significantly to CO₂ sequestration. The feasibility of the concept depends on the rate of olivine dissolution, the sequestration capacity of the dominant reaction, and its CO₂ footprint. Kinetics calculations show that offsetting 30% of worldwide 1990 CO₂ emissions by beach weathering means distributing of 5.0 Gt of olivine per year. For mean seawater temperatures of 15–25 °C, olivine sand (300 μm grain size) takes 700–2100 years to reach the necessary steady state sequestration rate and is therefore of little practical value. To obtain useful, steady state CO₂ uptake rates within 15–20 years requires grain sizes <10 μm. However, the preparation and movement of the required material poses major economic, infrastructural and public health questions. We conclude that coastal spreading of olivine is not a viable method of CO₂ sequestration on the scale needed. The method certainly cannot replace CCS technologies as a means of controlling atmospheric CO₂ concentrations.     

CO2 removal using membrane gas absorption

The objective of this study is to investigate the potential process for the removal of carbon dioxide (CO₂) from flue gas using fundamental membrane contactor, which is a membrane gas absorption (MGA) system. The experiments consisted of microporous polyvinylidenefluoride (PVDF) flat sheet membrane with 0.1 μm (as module I) and 0.45 μm (as module II) pore size. 2-Amino-2-methyl-1-propanol (AMP) solution was employed as the liquid absorbent. The effect of AMP concentration was studied with variation in the range 1–5 M. In addition, the experiments were carried out with 10%, 20%, 30% and 40% gas ratio of CO₂ to N₂ and pure CO₂ as well. Through contact angle measurement, membranes for module I and module II were obtained with CA values of around 130.25° and 127.77°, respectively. The mass transfer coefficients for module II are lower than those of module I for 1–5 M of AMP. Furthermore, the increase in CO₂ concentration in the feed gas stream enhanced the CO₂ flux as the driving force of the system was increased in sequence from 1 M to 5 M of AMP. However, after the particular percentage (40%) of CO₂ inlet concentration, the CO₂ fluxes seem saturated. The combination of AMP as liquid absorbent and PVDF microporous membrane in MGA system has shown the potential to remove the CO₂ from flue gas. In addition, the higher AMP concentration gave higher mass transfer coefficient at low liquid flow rates.     

Rate modeling of CO2 stripping from potassium carbonate promoted by piperazine

This work presents results from a rate-based model of strippers at normal pressure (160 kPa) and vacuum (30 kPa) in Aspen Custom Modeler^® (ACM) for the desorption of CO₂ from 5 m K⁺/2.5 m piperazine (PZ). The model solves the material, equilibrium, summation and enthalpy (MESH) equations, the heat and mass transfer rate equations, and computes the reboiler duty and equivalent work for the stripping process. Simulations were performed with IMTP #40 random packing and a temperature approach on the hot side of the cross-exchanger of 5 °C and 10 °C. A “short and fat” stripper requires 7–15% less total equivalent work than a “tall and skinny” one because of the reduced pressure drop. The vacuum and normal pressure strippers require 230 s and 115 s of liquid retention time to get an equivalent work 4% greater than the minimum work. Stripping at 30 kPa was controlled by mass transfer with reaction in the boundary layer and diffusion of reactants and products (88% resistance at the rich end and 71% resistance at the lean end). Stripping at 160 kPa was controlled by mass transfer with equilibrium reactions (84% resistance at the rich end and 74% resistance at the lean end) at 80% flood. The typical predicted energy requirement for stripping and compression to 10 MPa to achieve 90% CO₂ removal was 37 kJ/gmol CO₂. This is about 25% of the net output of a 500 MW power plant with 90% CO₂ removal.     

Novel lithium-based sorbents from fly ashes for CO2 capture at high temperatures

In this work several Li₄SiO₄-based sorbents from fly ashes for CO₂ capture at high temperatures have been developed. Three fly ash samples were collected and subjected to calcination at 950 °C in the presence of Li₂CO₃. Both pure Li₄SiO₄ and fly ash-based sorbents were characterised and tested for CO₂ sorption at different temperatures between 400 and 650 °C and adding different amounts of K₂CO₃ (0–40 mol%). To examine the sorbents performance, multiple CO₂ sorption/desorption cycles were carried out. The temperature and the presence of K₂CO₃ strongly affect the CO₂ sorption capacity for the sorbents prepared from fly ashes. When the sorption temperature increases by up to 600 °C both the CO₂ sorption capacity and the sorption rate increase significantly. Moreover when the amount of K₂CO₃ increases, the CO₂ sorption capacity also increases. At optimal experimental conditions (600 °C and 40 mol% K₂CO₃), the maximum CO₂ sorption capacity for the sorbent derived from fly ash was 107 mg CO₂/g sorbent. The Li₄SiO₄-based sorbents can maintain its original capacity during 10 cycle processes and reach the plateau of maximum capture capacity in less than 15 min, while pure Li₄SiO₄ presents a continual upward tendency for the 15 min of the capture step and attains no equilibrium capacity.     

Carbon dioxide storage capacity in uneconomic coal beds in Alberta,Canada: Methodology,potential and site identification

Methodology is presented for a first-order regional-scale estimation of CO₂ storage capacity in coals under sub-critical conditions, which is subsequently applied to Cretaceous-Tertiary coal beds in Alberta, Canada. Regions suitable for CO₂ storage have been defined on the basis of groundwater depth and CO₂ phase at in situ conditions. The theoretical CO₂ storage capacity was estimated on the basis of CO₂ adsorption isotherms measured on coal samples, and it varies between ∼20 kt CO₂/km² and 1260 kt CO₂/km², for a total of approximately 20 Gt CO₂. This represents the theoretical storage capacity limit that would be attained if there would be no other gases present in the coals or they would be 100% replaced by CO₂, and if all the coals will be accessed by CO₂. A recovery factor of less than 100% and a completion factor less than 50% reduce the theoretical storage capacity to an effective storage capacity of only 6.4 Gt CO₂. Not all the effective CO₂ storage capacity will be utilized because it is uneconomic to build the necessary infrastructure for areas with low storage capacity per unit surface. Assuming that the economic threshold to develop the necessary infrastructure is 200 kt CO₂/km², then the CO₂ storage capacity in coal beds in Alberta is greatly reduced further to a practical capacity of only ∼800 Mt CO₂.     

Kinetics of sulfur dioxide- and oxygen-induced degradation of aqueous monoethanolamine solution during CO2 absorption from power plant flue gas streams

Studies of the kinetics of sulfur dioxide (SO₂)- and oxygen (O₂)-induced degradation of aqueous monoethanolamine (MEA) during the absorption of carbon dioxide (CO₂) from flue gases derived from coal- or natural gas-fired power plants were conducted as a function of temperature and the liquid phase concentrations of MEA, O₂, SO₂ and CO₂. The kinetic data were based on the initial rate which shows the propensity for amine degradation and obtained under a range of conditions typical of the CO₂ absorption process (3–7 kmol/m³ MEA, 6% O₂, 0–196 ppm SO₂, 0–0.55 CO₂ loading, and 328–393 K temperature). The results showed that an increase in temperature and the concentrations of MEA, O₂ and SO₂ resulted in a higher MEA degradation rate. An increase in CO₂ concentration gave the opposite effect. A semi-empirical model based on the initial rate, ?r_MEA = {6.74 × 10⁹ e^?(29,403/RT)[MEA]^0.02([O]^2.91 + [SO₂]^3.52)}/{1 + 1.18[CO₂]^0.18} was developed to fit the experimental data. With the higher order of reaction, SO₂ has a higher propensity to cause MEA to degrade than O₂. Unlike previous models, this model shows an improvement in that any of the parameters (i.e. O₂, SO₂, and CO₂) can be removed without affecting the usability of the model.     

CO2 capture from power plants: Part II. A parametric study of the economical performance based on mono-ethanolamine

While the demand for reduction in CO₂ emission is increasing, the cost of the CO₂ capture processes remains a limiting factor for large-scale application. Reducing the cost of the capture system by improving the process and the solvent used must have a priority in order to apply this technology in the future. In this paper, a definition of the economic baseline for post-combustion CO₂ capture from 600 MW_e bituminous coal-fired power plant is described. The baseline capture process is based on 30% (by weight) aqueous solution of monoethanolamine (MEA). A process model has been developed previously using the Aspen Plus simulation programme where the baseline CO₂-removal has been chosen to be 90%. The results from the process modelling have provided the required input data to the economic modelling. Depending on the baseline technical and economical results, an economical parameter study for a CO₂ capture process based on absorption/desorption with MEA solutions was performed.Major capture cost reductions can be realized by optimizing the lean solvent loading, the amine solvent concentration, as well as the stripper operating pressure. A minimum CO₂ avoided cost of € 33 tonne⁻¹ CO₂ was found for a lean solvent loading of 0.3 mol CO₂/mol MEA, using a 40 wt.% MEA solution and a stripper operating pressure of 210 kPa. At these conditions 3.0 GJ/tonne CO₂ of thermal energy was used for the solvent regeneration. This translates to a € 22 MWh⁻¹ increase in the cost of electricity, compared to € 31.4 MWh⁻¹ for the power plant without capture.     

Integration and evaluation of a power plant with a CaO-based CO2 capture system

This paper explores the integration and evaluation of a power plant with a CaO-based CO₂ capture system. There is a great amount of recoverable heat in the CaO-based CO₂ capture process. Five cases for the possible integration of a 600 MW power plant with CaO-based CO₂ capture process are considered in this paper. When the system is configured so that recovered heat is used to replace part of the boiler heat load (Case 2), modelling not only shows that this is the system recovering the most heat of 1008.8 MW but also results in the system with the lowest net power output of 446 MW and the second lowest of efficiency of 34.1%. It is indicated that system performance depends both on the amount of heat recovery and the type of heat utilization. When the system is configured so that a 400 MW power plant is built using the recovered heat (Case 4), modelling shows that this is the system with the most net power output of 846 MW, the highest efficiency of 36.8%, the lowest cost of electricity of 54.3 €/MWh and the lowest cost of CO₂ avoided of 28.9 €/tCO₂. This new built steam cycle will not affect the operation of the reference plant which vents its CO₂ to the atmosphere, highly reducing the connection between the CO₂ capture process and the reference plant which vents its CO₂ to the atmosphere. The average cost of electricity and the cost of CO₂ avoided of the five cases are about 58.9 €/kWh and 35.9 €/tCO₂, respectively.     

Assessment of basin-scale hydrologic impacts of CO2 sequestration,Illinois basin

Idealized, basin-scale sharp-interface models of CO₂ injection were constructed for the Illinois basin. Porosity and permeability were decreased with depth within the Mount Simon Formation. Eau Claire confining unit porosity and permeability were kept fixed. We used 726 injection wells located near 42 power plants to deliver 80 million metric tons of CO₂/year. After 100 years of continuous injection, deviatoric fluid pressures varied between 5.6 and 18 MPa across central and southern part of the Illinois basin. Maximum deviatoric pressure reached about 50% of lithostatic levels to the south. The pressure disturbance (>0.03 MPa) propagated 10–25 km away from the injection wells resulting in significant well–well pressure interference. These findings are consistent with single-phase analytical solutions of injection. The radial footprint of the CO₂ plume at each well was only 0.5–2 km after 100 years of injection. Net lateral brine displacement was insignificant due to increasing radial distance from injection well and leakage across the Eau Claire confining unit. On geologic time scales CO₂ would migrate northward at a rate of about 6 m/1000 years. Because of paleo-seismic events in this region (M5.5–M7.5), care should be taken to avoid high pore pressures in the southern Illinois basin.     

Prospects for cost-effective post-combustion CO2 capture from industrial CHPs

Industrial Combined Heat and Power plants (CHPs) are often operated at partial load conditions. If CO₂ is captured from a CHP, additional energy requirements can be fully or partly met by increasing the load. Load increase improves plant efficiency and, consequently, part of the additional energy consumption would be offset. If this advantage is large enough, industrial CHPs may become an attractive option for CO₂ capture and storage CCS. We therefore investigated the techno-economic performance of post-combustion CO₂ capture from small-to-medium-scale (50–200 MWe maximum electrical capacity) industrial Natural Gas Combined Cycle- (NGCC-) CHPs in comparison with large-scale (400 MWe) NGCCs in the short term (2010) and the mid-term future (2020–2025). The analyzed system encompasses NGCC, CO₂ capture, compression, and branch CO₂ pipeline.The technical results showed that CO₂ capture energy requirement for industrial NGCC-CHPs is significantly lower than that for 400 MWe NGCCs: up to 16% in the short term and up to 12% in the mid-term future. The economic results showed that at low heat-to-power ratio operations, CO₂ capture from industrial NGCC-CHPs at 100 MWe in the short term (41–44 €/tCO₂ avoided) and 200 MWe in the mid-term future (33–36 €/tCO₂ avoided) may compete with 400 MWe NGCCs (46–50 €/tCO₂ avoided short term, 30–35 €/tCO₂ avoided mid-term).     

CO2 capture from hot stove gas in steel making process

The capture of CO₂ from a hot stove gas in steel making process containing 30 vol% CO₂ by chemical absorption in a rotating packed bed (RPB) was studied. The RPB had an inner diameter of 7.6 cm, an outer diameter of 16 cm, and a height of 2 cm. The aqueous solutions containing 30 wt% of single and mixed monoethanolamine (MEA), 2-(2-aminoethylamino)ethanol (AEEA), and piperazine (PZ) were used. The CO₂ capture efficiency was found to increase with increasing temperature in a range of 303–333 K. It was also found to be more dependent on gas and liquid flow rates but less dependent on rotating speed when the speed was higher than 700 rpm. The obtained results indicated that the mixed alkanolamine solutions containing PZ were more effective than the single alkanolamine solutions. This was attributed to the highest reaction rate of PZ with CO₂. A higher portion of PZ in the mixture was more favorable to CO₂ capture. The highest gas flow rates allowed to achieve a desired CO₂ capture efficiency and the correspondent height of transfer unit (HTU) were determined at different aqueous solution flow rates. Because all the 30 wt% single and mixed alkanolamine solutions could result in a HTU less than 5.0 cm at a liquid flow rate of 100 mL/min, chemical absorption in a RPB instead of a packed bed adsorber is therefore suggested to capture CO₂ from the flue gases in steel making processes.     

Geoelectrical methods for monitoring geological CO2 storage: First results from cross-hole and surface–downhole measurements from the CO2SINK test site at Ketzin (Germany)

The feasibility of monitoring CO₂ migration in a saline aquifer at a depth of about 650 m with cross-hole and surface–downhole electrical resistivity tomography (ERT) is investigated at the CO₂SINK test site close to Ketzin (Germany). The permanent vertical electrical resistivity array (VERA) consists of 45 electrodes (15 in the injection well Ktzi201 and 15 in each of the two observation wells Ktzi200 and Ktzi202), successfully placed on the electrically insulated casings, in the depth range of about 590–740 m with a spacing of about 10 m. The three Ketzin wells are arranged as perpendicular triangle with distances of 50 and 100 m.First synthetic modelling studies indicate an increase of the electrical resistivity of about 200% caused by CO₂ injection, corresponding to a bulk CO₂ saturation of 50%, which is in good agreement with laboratory studies. Finite difference inversion of field data delivers three-dimensional resistivity distributions between the wells which are consistent with the reservoir modelling studies.To increase the limited observation area provided by the cross-hole measurements, additional surface–downhole measurements were deployed. A main CO₂ migration in SE–NW direction is deduced from surface to downhole resistivity experiments.The first cross-hole time-lapse results show that the resolution and the coverage of the electrode array in the Ketzin setting are sufficient to resolve the expected resistivity changes on the characteristic length scale of the electrode array. Significant resistivity changes could be measured, however, detailed information on the CO₂ plume could not be resolved yet by VERA under the existing geological circumstances.     

Capture of carbon dioxide from flue or fuel gas mixtures by clathrate crystallization in a silica gel column

A column of silica gel was employed to contact water with flue gas (CO₂/N₂) mixture to assess if CO₂ can be separated by hydrate crystallization. Three different silica gels were used. One with a pore size of 30 nm (particle size 40–75 μm) and two with a pore size of 100 nm and particle sizes of 40–75 and 75–200 μm respectively. The observed trends indicate that larger pores and particle size increase the gas consumption, CO₂ recovery, separation factor and water conversion to hydrate. Thus, the gel (gel #3) with the larger particle size and larger pore size was chosen to carry out experiments with concentrated CO₂ mixtures and for experiments in the presence of tetrahydrofuran (THF), which itself is a hydrate forming substance. Addition of THF reduces the operating pressure in the crystallizer but it also reduces the gas uptake. Gel #3 was also used in experiments with a fuel gas (CO₂/H₂) mixture in order to recover CO₂ and H₂. It was found that the gel column performs as well as a stirred reactor in separating the gas components from both flue gas and fuel gas mixtures. However, the crystallization rate and hydrate yield are considerably enhanced in the former. Finally the need for stirring is eliminated with the gel column which is enormously beneficial economically.