B1 Available CO2 capture technology types

The three general approaches to capturing CO2 generated from fossil fuels (coal, natural gas, oil or biomass) are post-combustion, pre-combustion, and oxyfuel combustion. Refer to Figure B-1.

Figure B-1 CO2 capture approaches Source: WorleyParsons, 2009

Post-combustion CO2 capture involves separation of CO2 from flue gases produced by conventional fossil fuel combustion in air. Refer to Figure B-2. The flue gas is at atmospheric pressure and the CO2 concentration is typically 3 to 15 percent by volume, with the main constituent being nitrogen from the combustion of air. This results in a low CO2 partial pressure and a large volume of gas to be treated.

Figure B-2 Typical post-combustion CO2 capture Source: WorleyParsons, 2009

For existing power plants, current PCC systems would employ chemical absorption processes for separating CO2 from the flue gas streams such as amine-based scrubbing. One such chemical solvent is monoethanolamine (MEA), which is capable of a high level of CO2 capture (90 percent or more) due to fast kinetics and strong chemical reaction. Refer to Figure B-5.

Pre combustion CO2 capture involves reacting a fuel with oxygen and/or air in a gasifier to yield a synthesis gas (syngas), mainly consisting of carbon monoxide and hydrogen. Additional hydrogen and CO2 are produced by reacting the carbon monoxide with steam in a shift reactor. The CO2 is then separated, typically utilising a physical or chemical absorption process. The high concentration of CO2, typically 15 to 60 percent by volume, and the high pressures used, typically 4.0 Megapascals (MPa), are more favourable for CO2 separation. The resulting stream of hydrogen (H2) can then be used as a fuel; the CO2 is removed before combustion occurs. Refer to Figure B-3.

Figure B-3 Typical pre-combustion CO2 capture Source: WorleyParsons, 2009

Oxyfuel combustion, also known as oxyfiring, uses nearly pure oxygen instead of air for combustion of the fuel, resulting in a flue gas that is mainly water vapour and CO2 (more than 60 percent by volume). The water vapour is then removed by compression, cooling and condensation. In oxyfuel combustion, cooled flue gas is recycled back to the combustor to moderate the high flame temperature that results from combustion in pure oxygen. This process also requires the upstream separation of oxygen from air, with a purity of 95 to 99 percent in most cases. Refer to Figure B-4.

Figure B-4 Typical oxyfuel combustion CO2 capture Source: WorleyParsons, 2009

B1.1 Post-combustion CO2 capture

The following technologies can potentially be applied to PCC.

  • Absorption by regenerable solvent.
  • Adsorption on a solid bed.
  • Cryogenic separation.
  • Membrane separation.

The most common technique to remove CO2 from sour gas streams is absorption by regenerable solvent, using counter-current contact with the solvent. From the absorber, the CO2-rich solution is transferred to and stripped of the CO2 in a regenerator, usually by the application of heat and reduction of pressure. The regenerated lean solution is cooled and recirculated to the top of the absorber, completing the cycle.

The stripped acid gas, which is concentrated in CO2, is in a form ready for dehydration and compression to a pipeline liquid. The pipeline pressure is typically 13.8 MPa (2,000 psig), above the pressure for CO2 liquefaction.

B1.1.1 Chemical solvent absorption

Chemical solvent absorption is a chemical reaction that forms a loosely bonded intermediate compound. Chemical reagents are used to remove the acid gases by a reversible chemical reaction of the acid gases with an aqueous solution of various alkanolamines or alkaline salts in water. For CO2 capture application, a chemical solvent is exposed to a flue gas where it reacts chemically with CO2, separating it from the other gases. The intermediate compound is then isolated and heated causing it to break down into separate streams of CO2 and solvent.

Chemical solvents are more suitable than physical or hybrid chemical/physical solvents for applications at lower operating pressures. The chemical nature of acid gas absorption makes solution loading and circulation less dependent on the acid gas partial pressure. In a conventional amine unit, the chemical solvent reacts exothermally with the acid gas constituents.

Chemical solvents achieve a specified acid gas content in the treated gas, with fewer contacting stages than needed for physical solvents. Their relatively high alkalinity helps to achieve nearly complete CO2 removal. Two commonly used chemical solvents for acid gas removal are amines and hot carbonates.

Figure B-5 Typical amine-based chemical scrubbing Source: WorleyParsons 2009

Amines are classified as primary, secondary and tertiary, based on the number of amine groups, with advantages and disadvantages to each type.

Monoethanolamine (MEA) The MEA-based scrubbing process is a commercially available technology (Reddy et al. 2003). The solvent MEA, a primary amine, reacts with CO2 at around 40°C and near-to-ambient pressure, which is suitable for post combustion flue gas. Currently, MEA scrubbing technology is a state-of-art option for PCC. Nevertheless, the high heat of reaction with CO2 and the corrosivity of MEA are drawbacks that have restricted its use. Commercial applications have used formulations of proprietary corrosion inhibitors with MEA, such as the Fluor Econamine FGSM process.

Digycolamine (DGA® ): DGA, another primary amine, is similar to MEA in stability and reactivity, but can be used in much higher concentrations, up to 60 weight percent, requiring less energy and circulation, and providing a substantial savings in equipment cost. DGA has a lower vapour pressure and lower inherent corrosivity than MEA. The disadvantages are higher solvent cost and high heat of reaction with CO2.

KS® series solvents: Kansai Electric Power Company (KEPCO) and Mitsubishi Heavy Industries (MHI) have been developing stearically hindered amines, KS-1, KS-2 and KS-3. Among them, the most commonly known is KS-1. These amines are claimed to have an advantage (as compared to MEA) of a lower circulation rate due to a combination of higher CO2 loading differential, lower regeneration temperature and lower heat of reaction. KEPCO and MHI reported that the regeneration energy for the KS® series solvents is lower than that of MEA (Yagi et al. 2004). They are also non-corrosive to carbon steel at 130°C in the presence of oxygen. KS solvent based absorption systems have been utilised on chemical plants for CO2 separation. The first commercial plant using KS-1 has been in operation since 1999 at Petronas Fertiliser Kedah Sbn Bhd’s fertiliser plant in Malaysia (Kishimoto et al. 2009). Similar commercial systems are also being used by chemical plants in India. The KS solvent based system for coal-fired power plant application is still in the pilot stage. Hokuriku Electric Power Company has operated a test plant with KS® series solvents treating 50 Nm3/h of flue gas from a coal-fired unit at the Toyama-Shinko power station.

Cansolv: The Cansolv CO2 Capture System absorbs CO2 from a feed gas using Cansolv Absorbent DC-101, a patented amine-based regenerable solvent. The recovered CO2 can be dried, compressed and sequestered without further treatment. The amount of heat added on the regeneration step determines the extent to which the Cansolv absorbent is stripped of CO2 in the regeneration tower. The regeneration is typically sized to enable bulk removal (90 percent) of the CO2 in the absorber. However, the process is capable of CO2 purity to 99.99 percent (dry basis) if required (Cansolv 2009). The Cansolv technology can integrate CO2 and sulphur dioxide (SO2) capture in a single absorber, if desired. Cansolv has operated SO2 capture commercial plants since 2002. It has operated CO2 pilot plants at several locations, logging over 6,000 hours of operation. The two technologies will come together in an integrated system, in a plant designed to generate 50 tons per day of CO2, which will start up in 2009 (Shaw 2009).

HTC Purenergy: The HTC Purenergy CO2 capture process uses a proprietary amine-based HTC solvent to capture CO2 from industrial flue gases, in particular from fossil fuel power stations. The CO2 capture efficiency for flue gases from a gas-turbine exhaust is claimed to be typically 85 percent. The solvent is tailored to the specific requirements of the customer to reduce the cost of capture. HTC claims that steam consumption is reduced by approximately 50 percent and total solvent losses are up to 10 times less (HTC Purenergy 2006).

AMP, 2-amino-2-methyl-1-propanol: Abu-Zahraa et al. (2009) believe that using AMP as a solvent results in a substantial reduction in regeneration energy and the overall cost of CO2 avoided. S 37 percent reduction in the avoided cost with a flue gas recycle ratio of 45 percent is achieved using AMP as a solvent comparing to 10 percent using MEA solvent.

Chilled ammonia: The chilled ammonia process for CO2 capture is being developed by Alstom. It entails scrubbing cooled flue gas with slurry containing a dissolved and suspended mix of ammonium carbonate and ammonium bicarbonate in a counter-current absorber, similar to ammonia-based SO2 absorbers. Prior to entering the CO2 absorber, the flue gas is cooled to approximately 2°C in a direct contact cooler and mechanical chiller, condensing large quantities of water. The chilled flue gas then enters the absorber, where up to 90 percent of the CO2 is removed. CO2-rich slurry from the absorber, containing mainly ammonium bicarbonate, is pumped to a high pressure regenerator, where CO2 is released and separated from other gases. In laboratory tests co-sponsored by Alstom, EPRI and others, the process has demonstrated a potential for capturing more than 90 percent CO2 at a lower efficiency penalty than other CO2 capture technologies. The challenges are ammonia volatility and poor kinetics in the absorber. In February 2008, a pilot plant that uses chilled ammonia to capture CO2 from a 1.7 MW equivalent slip stream of flue gas from a coal-fired boiler was launched by Alstom and EPRI at the We Energies’ Pleasant Prairie Power Plant in Wisconsin. Also there is a 20 MW validation plant slated for completion mid-2009 at American Electric Power’s (AEP) Mountaineer Plant in West Virginia. The chilled ammonia system currently is not offered commercially (AEP 2008).

Aqueous ammonia: This joint NETL-Powerspan development entails reacting ammonia with CO2 in the flue gas to form ammonium carbonate, and subsequently heating the ammonium carbonate to release a pure CO2 stream. Advantages include: (1) low theoretical heat of regeneration; and (2) multi-pollutant control with saleable by-products (ammonium sulphate and ammonium nitrate fertilisers) using Powerspan’s commercial ECO™ system. One technical challenge is degradation of carbonate in the CO2 absorber leading potentially to ammonia slip in the flue gas. Powerspan’s ECO2™ technology is still at pilot stage. An ECO2™ pilot unit has been installed at the First Energy Burger Plant and started operation in October 2008. It processes a 1 MWe equivalent slip stream to capture 20 tonnes per day (tpd) of CO2.

Two ECO2™ demonstration projects have been announced by Powerspan. In November 2007, NRG Energy, Inc. and Powerspan announced their intention to commercially demonstrate the ECO2™ process at NRG’s W.A. Parish plant in Texas. The ECO2 ™ demonstration facility will be designed to capture 90 percent of CO2 from a 125 MWe slip stream, and the captured CO2 (about 1 Mt of CO2 annually) is expected to be used for EOR in the Houston area. The Parish plant is expected to be online in 2012.

In June of 2008, Powerspan and Basin Electric Cooperative announced a partnership to commercially demonstrate CO2 capture technology for conventional coal-based power plants. The demonstration project would capture about one million tons per year of CO2 from a slip stream of the exhaust from Unit 1 at the Basin Electric Antelope Valley Station. The Powerspan technology would remove CO2 from the equivalent of a 120 MWe slipstream. The captured CO2 would then be fed into an existing CO2 compression and pipeline system owned by Basin Electric’s Dakota Gasification Company. Start of construction of the CO2 capture system is scheduled for 2009, with operation commencing in 2012.

B1.1.2 Cryogenic separation

Carbon dioxide can be separated from other gases by cryogenic distillation. For CO2 capture applications a cryogenic separation process requires pressures above 2.1 MPa (300 psia) at temperatures of approximately minus 55°C. Cryogenic separation is used commercially for purification of CO2 from streams that already have high CO2 concentrations (typically greater than 80 percent). It is not normally used for more dilute CO2 streams, although it has recently been claimed that CO2 can be captured (by freezing it as a solid) from atmospheric pressure flue gases with energy losses similar to those of other techniques. A major disadvantage of cryogenic separation of CO2 is the amount of energy required to provide the refrigeration necessary for the process, particularly for dilute gas streams. Another disadvantage is that some components, such as water, have to be removed before the gas stream is cooled. Cryogenic separation has an advantage of enabling direct production of liquid CO2, which is therefore ready for transport with no further processing. The most promising applications for cryogenics are expected to be for separation of CO2 from high pressure gases, such as in pre combustion capture processes, or oxyfuel combustion in which the input gas contains a high concentration of CO2.

B1.1.3 Membrane processes

Gas separation membranes

Gas separation membranes rely on differences in physical or chemical interactions between gases and a membrane material, causing one component to pass through the membrane faster than the other. Various types of membranes are currently available, including porous inorganic membranes, palladium membranes, polymeric membranes and zeolites. Membranes cannot usually achieve high degrees of separation, so multiple stages and/or recycle of one of the streams is necessary. This leads to increased complexity, energy consumption and costs. Several membranes with different characteristics may be required to separate high-purity CO2. Membranes could be used to separate CO2 at various locations in power generation processes, for example from the fuel gas in an IGCC.

Gas absorption membranes

Gas absorption membranes are micro porous solids that are used as contacting devices between a gas and a liquid. The CO2 diffuses through the membrane and is removed by an absorption liquid such as amine, which selectively removes certain components. In contrast to gas separation membranes, it is the absorption liquid, not the membrane, which gives the process its selectivity.

Other membranes

Efforts to develop membranes used for PCC are ongoing. For example, Membrane Technology and Research (MTR) is in the process of testing membrane technology for post combustion flue gas applications. The demonstration of a small-scale pilot is scheduled for 2009 at Arizona Public Services (APS) natural gas combined cycle Red Hawk plant, and for 2010 at APS Cholla coal-fired power plant (Merkel et al. 2009). Overall, membrane technologies for PCC are in an early stage of development.

B1.2 Pre-combustion CO2 capture

In general pre-combustion CO2 capture technologies are favouring high pressure and low temperature applications and are not suitable for working in oxidising atmospheres. Hence, pre-combustion CO2 capture technologies are not suitable for a retrofit application of a conventional pulverised coal fired unit. Retrofit to existing gasification units is possible. However, these account for a relatively small portion of the existing emission sources around the globe. This is further explained in the following sections.

B1.2.1 Chemical solvent absorption

In a chemical absorption process the acid gases react to an intermediate liquid solvent species and are removed from the bottom of the absorber column with the rich solvent.

Diethanolamine (DEA) is a secondary amine. Like MEA, it can absorb CO2. However, it is less reactive than MEA and is highly susceptible to oxygen degradation that precludes DEA utilisation for PCC application (Dupart et al. 1993).

Methyldiethanolamine (MDEA) is a tertiary amine. In recent years, MDEA has acquired a much larger share of the gas-treating market (SFA Pacific 2002). Compared with primary and secondary amines, MDEA has superior capabilities for selectively removing hydrogen sulphide (H2S) in the presence of CO2. MDEA is resistant to degradation by organic sulphur compounds and has a low tendency for corrosion. Compared to MEA, it requires a relatively low circulation rate and consumes less energy. Several MDEA-based solvents that are formulated for high H2S selectivity are commercially available. However, MDEA is not suitable for post-combustion applications due to its oxygen-caused degradation. MDEA has been used for H2S removal in chemical plants and IGCCs.

The majority of chemical solvents are organic amine based. However, there are some alternative inorganic solvent systems such as Na/K carbonates.

Benfield Process. The Benfield process (UOP LLC 2000), also known as the hot carbonate process, uses an inorganic chemical solvent potassium carbonate (K2CO3) and catalysts. The process typically works at 70 to 120°C and 2.2 to 6.9 MPa. The Benfield process is widely used for purification of H2 streams, and is not considered a good option for PCC due to the low pressure of the flue gas. This process is commercially available. Hot carbonates are well suited for CO2 removal at moderate to high partial pressures of the feed gas. While hot carbonate plants have been used for bulk CO2 removal, their relatively high solvent circulation and heat requirements make them more expensive than other processes.

B1.2.2 Physical solvent absorption

Physical solvent scrubbing of CO2 is a well established technology, which is widely utilised to treat both natural and synthesis gas streams. In a physical absorption process the acid gases are physically absorbed into the liquid solvent and are removed from the bottom of the absorber column with the rich solvent. The solubility of individual gas compounds in a physical solvent follows Henry’s Law, and favours high pressure and low temperature operation. Physical solvents combine less strongly with CO2 than chemical solvents. The advantage of such solvents is that CO2 can be separated from them in the stripper mainly by reducing the pressure, resulting in much lower energy consumption. These solvents are better suited for applications at a higher pressure such as syngas streams in the coal-based IGCC process (typically 2.0 MPa or higher) and the concentrations of CO2 are about 35 to 40 percent. Hence, the CO2 partial pressure is much higher than that in normal combustion flue gas.

The physical solvents are regenerated by multistage flashing to low pressures. Because the solubility of acid gases increases as the temperature decreases, absorption is generally carried out at lower temperatures, and refrigeration is often required.

Several physical solvents that use anhydrous organic solvents have been commercialised. The following are some commercially available physical solvents that could be used for CO2 capture in applications, such as IGCC.

  • Rectisol. The Rectisol process uses chilled methanol as a scrubbing solvent. Typically, the process works at temperatures of -10 to -70°C and higher than 2.0 MPa. The process is licensed by Linde AG and Lurgi AG.
  • Selexol. The Selexol process uses dimethylether of polyethylene glycol as the solvent. Typical working conditions are -20 to 40°C and 2.06 to 13.80 MPa.
  • Fluor Process. The Fluor process uses propylene carbonate as the solvent. The solvent generally works below ambient temperature and at high pressure (3.1 to 6.9 MPa).
  • Purisol. The Purisol process uses n-methyl-2-pyrolidone as the solvent. The process condition generally works at temperature of -20 to 40°C and at high pressures (≥2.0 MPa).

Commercially available physical solvent scrubbing technologies generally require high pressure and low temperature, and hence are not considered to be preferable options for PCC processes. However, some efforts are being made to develop new solvents that are expected to be suitable for PCC such as ionic liquids.

B1.2.3 Hybrid physical/chemical solvent absorption

Hybrid solvents combine the high treated gas purity offered by chemical solvents with the flash regeneration and lower energy requirements of physical solvents. Some commercially available scrubbing technologies that use a mixture of physical and chemical solvents are as follows.

  • Sulfinol. The Sulfinol process is developed by Shell. The solvent is a mixture of diisopropanolamine (DIPA) and Sulfolane (tetrahydrothiophene dioxide). The former provides a chemical solvent and the latter a physical solvent. Meanwhile a modified solvent, known as Sulfinol-M, has been developed that uses MDEA as the chemical solvent. The Sulfinol process typically works at a pressure higher than 0.5 MPa (73 psia) and can be used for applications such as IGCC.
  • Flexsorb™ PS. The Flexsorb PS process is a mixed hindered amine/physical organic solvent version of the Flexsorb process developed by ExxonMobil. It is very stable and resistant to chemical degradation. It was developed to compete with the Sulfinol process. In one Canadian natural gas plant, Sulfinol-D was replaced with Flexsorb PS solvent to reduce the solvent circulation rate and reboiler duty (Korens et al. 2002).
  • Ucarsol™ LE. Ucarsol LE solvents offered by the Dow Chemical Company (2004) are used for high-efficiency acid gas removal. Ucarsol LE-701 is for selective H2S and controlled CO2 removal, while Ucarsol LE-702 is for complete acid gas removal, meeting low H2S and CO2 specifications. These are MDEA-based, physical/chemical hybrid solvents, which offer lower regeneration energy demand, lower hydrocarbon solubility and less degradation than other commonly used hybrid solvents, resulting in the potential for significant operating cost savings.
  • Amisol. The Amisol process was developed by Lurgi Germany. The process uses a mixture of MEA or DEA with methanol. The process works at ambient temperature and a pressure higher than 1 MPa (145 psia). The process has been applied down stream of a number of oil gasification units, but has not established a wide market.

B1.2.4 Pressure/temperature/electric/vacuum swing adsorption

Some solid materials with high surface areas, such as zeolites and activated carbon, can adsorb CO2 and be used to separate CO2 from gas mixtures by adsorption. The process operates on a repeated cycle with the basic steps being adsorption and regeneration. In the adsorption step, gas is fed to a bed of solids that adsorbs CO2 and allows the other gases to pass through. When a bed becomes fully loaded with CO2, the feed gas is switched to another clean adsorption bed and the fully loaded bed is regenerated to remove the CO2. In pressure swing adsorption (PSA), the adsorbent is regenerated by reducing pressure. In temperature swing adsorption (TSA), the adsorbent is regenerated by raising its temperature, and in electric swing adsorption (ESA) regeneration takes place by passing a low-voltage electric current through the adsorbent.

Recent investigations into adsorption technology have shown that CO2 recovery is also feasible under vacuum conditions (vacuum swing adsorption, VSA). Though VSA has not yet been commercially tested for CO2 recovery, it is a promising emerging technology with application in CO2 separation from blast furnace top gases, while the residual gases are recycled back to the furnace. This concept was recently tested with promising results at the Metallurgical Research Institute (MEFOS) experimental facility in Luleå, Sweden within a framework of ULCOS project (Air Liquide 2008, Danloy et al. 2009).

Pressure swing adsorption and TSA have been employed commercially for CO2 removal from syngas for hydrogen production. Electric swing adsorption is not yet commercially available, but it is said to offer the prospect of lower energy consumption than the other processes. Adsorption is not yet considered attractive for large scale CO2 removal from combustion flue gas because the capacity and CO2 selectivity of available adsorbents is low. However, it may be successful in combination with another capture technology. Some development efforts for new sorbents are being taken to develop adsorbents that can operate at higher temperatures in the presence of steam with increased capacity and improved selectivity, for example dry regenerable carbonate sorbent.

B1.2.5 Other technologies

Some other technologies are being developed, which do not fit in the categories mentioned above. Enzymatic CO2 capture process, developed by Carbonzyme Inc., uses an enzyme catalysed carbonic anhydrase-based liquid membrane biomimetic reactor. It is claimed that the technology is applicable to treating a large number of different flue gas streams (eg, flue gases generated by combusting fuels such as natural gas, oil or various ranks of coal). The process operates at moderate temperature and pressure. It has the ability to separate CO2 from other gases while using modest energy and employing no hazardous chemicals. However, the technology is still at an early development stage.

B1.3 Oxyfuel combustion CO2 capture

The oxyfuel combustion CO2 removal process for coal-fired boilers is a developing technology (Vattenfall 2008). This technology remains unproven at commercial scale in power generation applications. The first oxygen-fired pulverised coal (PC) pilot unit, a 30 MWth Alstom unit at the Schwarze Pumpe site in Germany, started operation in the summer of 2008. The beginning of engineering for the 300 MWe oxyfuel demonstration plant is planned for 2010. The start of demonstration plant operation is projected for 2015.

Oxygen combustion technology facilitates carbon capture in two major steps. Step one is accomplished within the oxygen combustion boiler system, in which flue gas with a high CO2 concentration is produced. Oxyfiring alone seldom produces a CO2 stream of sufficient purity to be reused or stored.

Step two processes include additional CO2 purification (as dictated by product CO2 specification), dehumidification and compression. In any coal or biomass fired application, particulate separation must be accomplished prior to CO2 purification. The purification step typically employs low temperature distillation (see cryogenic separation) to separate CO2 from the inert gases such as nitrogen and oxygen.

A cryogenic distillation process is utilised by commercially available air separation units (ASU) to produce the oxygen for the oxyfiring combustion process. A major disadvantage of a cryogenic ASU is the amount of energy required to provide the refrigeration necessary for the separation process. Several advanced concepts are being developed to reduce ASU parasitic load. Those include ion transfer membranes being developed by Air Products and oxygen transfer membranes being developed by Praxair. As part of the ongoing project, Air Products designed and commissioned in 2006 a 5.1 tpd prototype facility to test multiple membrane modules under commercially relevant operating conditions. The larger 152 tpd facility is expected to begin commissioning in late 2010. Overall, membrane technologies for oxygen separation are still in an early stage of development.

B1.3.1 Chemical looping system

Alstom is in early stages of developing the limestone-based chemical looping system for existing and new pulverised coal-fired power plants. In a sense, Alstom’s chemical looping process is oxyfuel combustion without the ASU. The system operates as follows: solid limestone based oxygen carrier circulates between oxidiser and reducer and carries oxygen, heat and fuel energy. The carrier picks up oxygen in the oxidiser and leaves nitrogen behind. The carrier delivers oxygen to the fuel in the reducer. Heat generated by fuel oxidation in the reducer produces steam for power. Alstom is in the process of designing a 1,000 lb/hr of coal prototype plant, which is scheduled to start testing in 2011. Commercial operation of this technology is projected for 2019 (Alstom et al. 2009).

B1.4 Emerging technologies

Emerging technologies are mirrored in the US DOE (2007) Carbon Sequestration Technology Roadmap, shown here for CO2 capture in Figure B-6.

Figure B-6 US DOE CO2 capture pathways Source: US DOE Office of Fossil Energy, 2007

Analysis has shown that CO2 capture accounts for the majority of CCS system costs. Therefore, efforts are focused on improving efficiency and reducing costs for capturing CO2 from coal-fired power plants, since they are the largest stationary sources of CO2, although the technologies developed will be applicable to other sources as well. Figure B-6 highlights the critical challenges and R&D pathways related to CO2 capture. It is those promising technologies emerging from bench to pilot-scale that are of interest here.

B1.4.1 Ionic liquids

Ionic liquids (ILs) are organic salts with low melting points, many below room temperature. Even though they are liquids, they have negligible vapour pressure. Thus, they have an advantage over conventional solvents for absorption of CO2 from flue gas because they would not contaminate the purified gas stream. Typical ILs are composed of imidazolium, pyridinium, ammonium or phosphonium cations with any of a wide variety of anions. Their properties can be varied tremendously by the choice of anion, cation and substituents. Ionic liquids are typically combined with supported membranes in the CO2 capture application. This technology is still in laboratory stage and is being developed by the University of Notre Dame, Sachem Inc., and Merck (NETL 2007).

B1.4.2 Dry regenerable carbonate sorbent CO2 adsorption

In this process, the sorbent material (based on sodium carbonate) captures CO2 at a boiler typical flue gas exhaust pressure and temperature of approximately 60°C (carbonation). The sorbent is then regenerated at a temperature of about 120°C to yield a concentrated stream of CO2 for sequestration or other use. The regenerated sorbent is recycled to the adsorption step for CO2 capture. The process is expected to be less expensive and energy intensive than MEA technologies. This process is compatible with current power plant operating conditions and hence applicable for CO2 capture from coal and natural gas-fired power plants. The challenges for the technology include continuous circulation of large quantities of solids and requirements for contaminants. The developer of the technology is the Research Triangle Institute. Currently, the technology is in the small-scale technology demonstration phase, in which a 1 tpd CO2 capture facility is being built. The demonstration of the small-scale pilot is scheduled for 2010, and a large-scale demonstration at UNC Chapel Hill coal-fired plant is planned for 2012 (102 tpd of CO2 captured). It is envisioned that the technology will be ready for commercial offering in 2015 (Nelson et al. 2009).

B1.4.3 Amino acid salt CO2 absorption

This process, in development by Siemens AG, is a proprietary second generation PCC process, based on amino acid salt formulations. Amino acid salts are the basis of their solvent. Siemens claims the salts essentially have no vapour pressure; the benefits of which are no thermodynamic solvent emissions, no flammability, no explosion risk, no odour, and no inhalation risk. Being that the amino acid salt is a negative ion, it is less sensitive to oxygen in the flue gas; the benefit of which is a low rate of degradation. The fact that amino acids are naturally present in the environment means that they are biodegradable, non-toxic, and relatively environmentally friendly.

In its current state of development, Siemens has compared the amino acid salt CCS process plant efficiency to that of a non-CCS reference plant. The net efficiency of the non-CCS reference plant (800 MWe) is 45.7 percent. In earlier versions of the amino acid salt CCS plant the efficiency decline was greater than 10 percent. With process improvements the current version is projected to have an overall efficiency approximately 9.2 percentage points lower than the reference plant, or about 36.5 percent. In this comparison, compression of 99 percent pure CO2 to a pressure of 20 MPa (200 bar) is included (Kremer 2009).

Siemens claims the laboratory-scale pilot plant has been in continuous operation for nearly two years, since 2008. A pilot plant at the E.ON power plant at Staudinger (Germany) is planned to start operation in August 2009. Siemens expects, that after 1 to 2 years operation, this will lead to a demonstration project being commissioned in 2014 or later.

For a full-scale 800 MWe steam power plant, Siemens projects the investment cost for the CCS process plant would be in the range of 300 to 400 million Euros in which:

  • makeup solvent would cost about 5 Euros per litre for small quantities;
  • rate of solvent degradation would be approximately 75 percent less than with MEA; and
  • footprint size would be in the range of 20,000 to 40,000 square m.

The TNO Science & Industry (TNO) patented process (Versteeg et al. 2006) is based on the research carried out at the University of Twente (Holst et al. 2006). The University of Twente research involved addition of various salts to amino acid to produce precipitating solutions. Each were characterised for kinetics and activity and proposed for use in hollow fibre polymeric membranes. The TNO CO2-capture process, called DECAB, is utilising precipitating solvents of amino-acid salts. The process makes use of the fact that an amino-acid salt solution reacts with CO2, and at some point a precipitate is formed consisting of the amino-acid or CO2 containing species.

Preliminary economic evaluation reported by the TNO (Brouwer et al. 2006) shows that the DECAB process has the potential to substantially decrease the investment costs and the energy consumption of the capture process as compared to an MEA process. In June of 2009 Siemens and TNO signed a strategic cooperation agreement aimed at advancement of amino-acid salt-based carbon capture technology. The TNO process is reported to be testing various solvents since April 2008 in their CATO pilot plant at the Rotterdam site of E.ON Benelux. The partnership targets implementation of a full-scale demonstration plant based on amino-acid salts CO2 capture technology by 2014 (TNO 2009).

B1.4.4 US DOE R&D technologies

There are many carbon capture emerging technologies under R&D contract with the US DOE (2009). Only a brief summary of a few selected R&D technologies is included here.

  • Membrane Process to Capture CO2 from Coal-fired Power Plant Flue Gas (Membrane Technology & Research, Inc.)
    • The objective of this two-year (2008-2010) R&D program is to develop and validate that a membrane process can effectively and efficiently capture at least 90 percent of the CO2 from coal-fired flue gas at 50 to 60°C.
  • Chemical Looping Combustion Prototype for CO2 Capture from Existing PC Fired Power Plants (Alstom Power, Inc.)
    • The objective of this project is to develop and verify the high-temperature chemical and thermal looping process concept at a small-scale pilot facility in order to enable Alstom to design, construct and demonstrate a pre-commercial, prototype version of this advanced system.
  • Coal Direct Chemical Looping Retrofit to PC Power Plants for In-Situ CO2 Capture (Ohio State University)
    • The objective of this project is to investigate a Coal Direct Chemical Looping (CDCL) system to effectively capture CO2 from existing PC power plants by further advancing the novel CDCL technology to sub-pilot scale (25 kW).
  • CO2 Capture Membrane Process for Power Plant Flue Gas (Research Triangle Institute)
    • The objective of this two-year (2008-2010) R&D program is to develop an advanced polymeric membrane-based process that can be cost-effectively, easily, and reliably retrofitted into current PC power plants to separate and capture at least 90 percent of the CO2 from the plant’s flue gas at 50 to 60°C, the typical temperature of the exhaust gas emitted from the wet FGD system used in PC plants, with no more than a 35 percent increase in the cost of electricity (COE).
  • Novel Liquid Sorbents for CO2 Capture from Coal-Fired Power Plants (Reaction Systems LLC)
    • The objective of this project is to develop a novel CO2 scrubbing solution that has a high capacity for CO2, but requires little or no thermal energy for desorption. This process could capture 90 percent of the CO2 in the effluent of a coal-fired power plant without increasing cost by more than 25 percent (Wickham 2007).

B1.4.5 CO2 sequestration into cement

Several companies claim to have developed a process that essentially mimics marine cement, which is produced by coral when making their shells and reefs. The process takes calcium and magnesium in sea water and uses it to form carbonates at normal temperatures and pressures. One such company, Calera Corporation, claims for every ton of cement they make, they are sequestering a half ton of CO2. (Biello 2008) They claim to turn CO2 into carbonic acid and then make carbonate. Once dried, the Calera cement can be used as a replacement for Portland cement that is typically blended with rock and other material to make concrete, which is used in everything from roads to building construction.

Other companies are also pursuing this idea. Carbon Sciences in Santa Barbara, California, plans to use flue gas and mine slime – the water remaining after mining operations – which is often rich in magnesium and calcium, to create similar cements. Carbon Sense Solutions in Halifax, Nova Scotia, plans to accelerate the natural process of cement absorbing CO2 by exposing a fresh batch to flue gas.

B1.5 CO2 compression and dehydration commercially available technology

Conventional approach to CO2 compression and dehydration

The CO2 captured in a CO2 removal process needs to be compressed to a pressure suitable for pipeline transport and sequestration. The CO2 is conveyed as a liquid or as a dense phase supercritical fluid through increasing the pressure above 7.4 MPa (1,073 psi), the critical point pressure. The typical pipeline operation pressures are in the range of 13.8 to 20.7 MPa (2,000 to 3,000 psi). This allows for the CO2 to be pumped through the pipeline without further compression resulting in an energy saving. As the CO2 travels through the pipeline, the pressure drops. This drop needs to be considered in the initial compression of the CO2 and recompression stations along the pipeline. Additionally, the impact of elevation changes on the pipeline pressure needs to be taken into account. The critical point of CO2 is 7.38 MPa at 31.1°C.

The options available to achieve the required CO2 pressure are:

  • compression only; and
  • compression and pumping.

B1.5.1 Compression

Compressor type selection is dependent on the inlet volumetric flow rate, starting and final pressures, and gas composition. For an amine chemical absorption/regeneration process the starting pressure is approximately 0.18 MPa (26.3 psia) and for Oxyfuel CO2 capture, the starting pressure is approximately 0.10 MPa (15 psia). For IGCC CO2 capture, depending on the technology chosen, the starting pressure can range from 0.14 to 0.345 MPa (20 to 50 psia).

Three compressor types that may be considered are:

  • reciprocating compressor;
  • multi-stage, integrally-geared centrifugal compressor; and
  • single-shaft, multi-stage centrifugal compressor.

Reciprocating compressors are applicable to CO2 compression. Typically these are more limited in capacity as compared to the centrifugal compressors, such that more trains in parallel may be needed. Due to the nature of their construction and operation, reciprocating compressors generally have lower reliability than centrifugal compressors.

In addition, due to their lubrication, reciprocating compressors have the characteristic of inherent and unavoidable lube oil carryover into the gas. Centrifugal compressors typically have a dry gas seal arrangement and do not have this characteristic. This is of particular concern to a triethylene glycol (TEG) dehydration system, which is commonly used in this service, due to lube oil causing severe glycol foaming problems and operating upsets in the contactor.

MAN Turbo AG is one of the global suppliers with experience in the area of CO2 compression to dense phase conditions, having supplied (as Borsig) the compressor at the Great Plains Synfuels Plant for Dakota Gasification Company in Beulah, North Dakota (USA), for the transport of CO2 to Canada for EOR via the “Weyburn” pipeline. For that application, MAN Turbo supplied three 8-stage integrally-geared compressors.

MAN Turbo claims that, while they are generally interchangeable, the integrally-geared centrifugal compressors have an advantage with relatively high molecular weight gases, such as CO2 (44), while the in-line centrifugal compressors have an advantage with relatively low molecular weight gases, such as natural gas (19).

Compression configurations

Based on the compressor stage gas inlet temperature compression configurations can be classified as:

  • isothermal compression with intercoolers after each stage; and
  • adiabatic compression, with reduced number of intercoolers, or cooling only after the final stage that enables compression heat recovery at a relatively high temperature.

In general, isothermal compression requires less compression power, as the gas temperature entering each compressor stage is maintained constant and relatively low by the means of interstage cooling. In adiabatic compression gas temperatures could reach 200°C, making adiabatic compression a possible choice for systems that would benefit from compression heat recovery for feedwater preheating, or steam generation.

B1.5.2 Compression and pumping

In a compression and pumping process the CO2 stream is compressed, dehydrated, chilled and then pumped to a required pressure. Compression and pumping systems requires energy for low pressure compression, chilling and pumping. Combined compression and pumping processes are reported to require less power compared to compression only. However, in other publications the compression and pumping option is shown to require more power then compression only, depending on system configuration. Presence of impurities in the product CO2 stream such as nitrogen will increase the energy requirement for chilling at a given pressure. Thus the compression plus pumping option is typically utilised in conjunction with CO2 low temperature purification systems. The most likely application is the oxycombustion process, which could produce a product stream with CO2 purity greater than 90 percent and containing 2 to 3 percent of oxygen. The product CO2 stream needs to be distilled at a low temperature to separate non-condensable gases. The CO2 purification is accomplished in a low temperature distiller, in which liquid CO2 is collected on the bottom of the distiller column and then pumped to a specified pressure.

B1.5.3 Dehydration options

During staged compression of the captured CO2 product stream, the moisture content is first reduced by cooling the gas below its dew point and knocking out the water and then finally by dehydration. The main processes which are typically utilised for gas dehydration use glycol or a solid absorbent (e.g. mole sieve).

The TEG is the most widely used fluid in dehydration absorption systems, since it offers the best combination of ease of operation and economics. It is the most common dehydration method used for natural gas.

Dehydration by TEG contacting is the standard method of achieving the CO2 transport and sequestration moisture specification. Although not usually required, TEG dehydration can achieve the near-bone-dry specification of 20 ppm by volume moisture content (1 lb/MMscf), equivalent to a water dew point of -55°C (-67°F). More typically, the moisture specification, as used by Kinder Morgan for EOR, is 633 parts per million by volume (ppmv) (30 lb/MMscf), equivalent to a water dew point of -22°C (-8°F), for reasons of corrosion (Havens 2007).

Mole sieve has a higher life-cycle-cost than glycol and is usually used when completely dried gas is required, such as low temperature liquefaction utilised for CO2 compression and pumping.