Fighting carbon with carbon: new low-cost filter halves CO2-capture costs

Researchers in Wyoming report the development of a low-cost carbon filter that can remove 90 percent of carbon dioxide gas from the smokestacks of electric power plants that burn carbonaceous fuels. The cost of capturing CO2 with the new carbon filter was found to be around half that of the most cost-effective alternative currently available (amine absorption). When this technology is coupled to power plants that burn biomass, and the CO2 captured and sequestered, the energy from such a plant would be carbon-negative.

Ironically, the new filter is based on activated carbon, which can be made from charcoal - itself biomass which captured CO2 from the atmosphere. So in a sense, the researchers found a way to fight carbon with carbon...

Maciej Radosz and colleagues at Wyoming's Soft Materials Laboratory cite the pressing need for simple, inexpensive new technologies to remove carbon dioxide from smokestack gases. Coal-burning electric power plants are major sources of the greenhouse gas, and control measures may be required in the future.

The study describes a new carbon dioxide-capture process, called a Carbon Filter Process, designed to meet the need. It uses a simple, low-cost filter filled with porous carbonaceous sorbent that works at low pressures. Modeling data and laboratory tests suggest that the device works better than existing technologies at a fraction of their cost.

The scientists took current amine absorption techniques, which are known to separate CO2 well, as a reference for technical and economic comparisons with alternatives and with their own CO2 separating process. They calculated the efficiency and cost for the application of the different techniques to one type of flue gas. The reference cost for the amine technology to capture CO2 was found to be $47/ton, close to the often cited range of $40 to $50 per ton of compressed CO2:
energy :: sustainability :: biomass :: bioenergy :: biofuels :: carbon capture and storage :: geosequestration :: bioenergy with carbon capture :: carbon-negative :: activated carbon :: climate change ::

The alternatives were the following:

Ionic liquid absorption: too slow
Absorption in ionic liquids, which are known to be selective for CO2, was found to be problematic because CO2 sorption and desorption rates were observed to be very low. Therefore, the ionic liquid absorption option was abandoned. Instead, polymerized ionic liquids were found to be more attractive as solid materials for CO2 membranes and sorbents, as documented in other papers.

Pressure-induced transport: expensive for membranes
An example of a CO2-philic membrane-based alternative was investigated, where the CO2 driving force is due to a pressure difference between the permeate (ambient) and retentate. From among the many membrane materials that are known to be selective for CO2, relative to nitrogen, a recently synthesized membrane that was made of brominated poly(phenylene oxide) impregnated with 30% of silica particles was selected, because its permeability and its CO2/N2 selectivity were very high. This system was capable of recovering at least 90% of the CO2 with a purity of 90%; however, it was found to be very costly.

Because no firm basis for estimating the cost of the nanocomposite was available, the cost of recovered CO2 was difficult to estimate, but even for optimistic material cost assumptions, this cost is likely to be "much higher" than that of CO2 from the amine process.

Zeolite sorbents: problems with heat of sorption, moisture Sensitivity, and material and pressure costs
An example of a zeolite-13X PSA process was evaluated. In a first-pass economic evaluation, the sorption and desorption steps are assumed to be approximately isothermal, even though the CO2 heat of sorption on zeolite is substantial enough to cause the sorption temperature to increase. It has been reported that the heat of CO2 adsorption on zeolite is ~30 kJ/mol,56 which is ~10 times higher than that on activated carbon (~3 kJ/mol)35 at the same temperature 25 C and pressure 1 bar.

Another drawback of the zeolite sorbent is its moisture sensitivity, which requires much higher (say, over 300 C) drying temperatures than the minimum temperatures needed to remove CO2 alone, which means extra recovery costs.

However, ignoring these drawbacks in a first-pass economic evaluation lead to a cost of recovered CO2 that is approximately $70/ton, which is less than the membrane-recovered CO2, but is ~40% more than the amine benchmark cost.

The need to dry the zeolite will increase this cost. The main cost components are the steam cost, the compression cost, and the zeolite cost ($33/lb). A less-expensive sorbent, such as activated carbon (for example, $1-$2/lb), can reduce the material cost, but compression will still be required, if it is used in a PSA mode. Therefore, a PSA route was not evaluated further.

The researchers then looked at designing a capture process based on carbonaceous sorbents which overcome the above problems. The challenge was to make them at a low cost, with good CO2-selectivity, while at the same time ensuring that they are insensitive to moisture easy to regenerate. Ideally, such a sorbent should be selective to other flue-gas pollutants, such as NOx, SOx, mercury, and arsenic, which would allow for a multifunctional sorbent.

Some but not all carbon-rich materials, such as activated carbon, charcoal, other coal pyrolysis-derived materials, or even virgin coal, can satisfy these requirements and, hence, became the focus of the work. Four preliminary model carbon-rich materials were selected: activated carbon, charcoal, and virgin bituminous coal.

The bulk prices of these materials were conservatively estimated to be $1500/ton for activated carbon, $200/ton for charcoal, and $40/ton for coal.

The materials were then analysed for their sorption capacity, selectivity, rate, and thermal stability. The sorption capacity increases as the pressure increases and the temperature decreases. The activated carbon capacity was somewhat higher than that of charcoal, and much higher than that of coal, which correlates with the surface area and the degree and type of activation.

A more interesting trend emerged based on an ideal CO2/N2 sorption selectivity: increasing pressure substantially decreases the selectivity (it increases the nitrogen capacity to a far greater extent than it does the CO2 capacity), which points to a low-pressure sorption advantage.

All carbonaceous materials studied exhibited a rapid sorption rate. A typical time needed to nearly saturate these materials with CO2 was around 3 min at 25 C. This time increases as the temperature increases to 5-10 min at 75 C and 10-12 min at 110 C, with charcoal being on the low side and activated carbon being on the high side. Generally, the results suggested short sorption cycles at low temperatures.

The CO2 sorption was found to be reproducibly reversible, which suggested a good stability and easy desorption. 20 temperature cycles for activated carbon and 5 temperature cycles for charcoal between 25 C and 130 C did not affect the sorption capacity much.

These good qualities of the carbonaceous sorbents encouraged the researchers to design a filter based on them.

The new carbon filter
The scientists used the following design assumptions. The nominal CO2 recovery target for the filter was set at 90% and its purity target at 90%. In a first-pass approximation of the reference flue gas, the feed was assumed to contain 12% CO2, with the balance being nitrogen.

Low O2 sorption capacity was confirmed for activated carbon was determined to be as low as that of nitrogen, which suggested the CO2/oxygen selectivity would be similar to the CO2/nitrogen selectivity. Unless removed upstream of the carbon filter, which may be the case for existing power plants, SOx, NOx, and mercury were reported to have a high affinity for the activated carbon and, hence, were expected to be sorbed with CO2.

The sorption temperature of ~25 C was assumed not to change much during the sorption cycle, because the CO2 heat of sorption is on the low side. The sorption time is set at 2 min, the sorbent regeneration is set at 100 C, using direct-steam desorption for 2 min, based on preliminary breakthrough data taken in the laboratory.

The cooling-air stage time was also set at 2 min, which made the total cycle time for the preliminary example 6 min.

Without any attempt to optimize the vessel size, a cylindrical module was selected (3.5 m in diameter and 2.0 m in length). For the sorption-desorption-cooling cycle one would thus need 189 alternating vessels, 63 of which are in a sorption mode, 63 are in a desorption mode, and 63 are in an air-cooling mode.

Because the carbonaceous sorbents selected are known to be stable (that is, their capacity does not change much over time), it was assumed that no sorbent replacement would be required within 10 years. However, relaxing this assumption, for example, by replacing the sorbent more often, did not impact the cost of recovered CO2 too much.

Economic assumptions
To analyse the costs of capturing CO2 with the new filter, the following assumptions were made. Interest rates were set at 15%, the electricity cost at $0.07/kWh, the steam cost at $7/ton (or $3.2/MMBTU), and the annual maintenance and repair cost at 7% of the fixed capital investment. Manpower cost is a relatively minor component of the operating costs, and, hence, it was assumed to be approximately $1.5 million per year.

The carbon filter process was then evaluated for a vacuum regeneration case and a steam regeneration case, both before heat integration with the power plant and CO2 compression.

The scientists found that steam regeneration led to a significant cost reduction as compared to the amine benchmark.

Vacuum regeneration: comparable to benchmark
Flue gas at ~85 C is cooled to ~25 C with water before it is fed with a blower to the sorption unit. After the sorbent is almost saturated with CO2 for ~2 min, this unit switches to a 2 min regeneration cycle under vacuum, and then it alternates between the sorption and vacuum stages at ambient temperature.

The major cost items are associated with the vacuum pump. The total cost of the recovered CO2 is approximately $37/ton, which is comparable to that of the amine benchmark case.


Thermal regeneration with steam or hot CO2: much better than benchmark
An isobaric process with direct-steam or hot-CO2 regeneration is shown in Figure 2. Both sorption and desorption occur at ambient pressure. The feeding section and the sorption cycle are the same as those in the previous case. Instead of vacuum regeneration, however, the saturated sorbent bed switches to a steam heating cycle and then to an air-cooling cycle to bring the bed temperature to near-ambient temperature.

The major cost items were steam and electricity, and the total cost of the recovered CO2 is approximately $20/ton for activated carbon, which is much less than that for the amine benchmark.

The $20/ton low-pressure CO2 cost must be corrected for compression to make CO2 ready for transport. The compression cost, from ambient to a pipeline pressure (e.g., 2000 psi) was estimated to add $7/ton. Therefore, the total cost of compressed, pipeline-ready CO2 for a power-plant integrated activated carbon filter would be approximately $27/ton CO2.

Effect on electricity prices
Adding a carbon capture unit to a power plant will affect the electricity cost and, hence, the profitability. For the activated carbon filter, an approximate electricity cost change was plotted relative to carbon capture credits.

A run of costing models showed that credits of $30/ton can effectively reduce the electricity production cost by 10%, credits of $20/ton can leave the electricity cost unchanged, and zero credits can increase the electricity cost by ~30%.

Conclusion
The low-pressure carbon filter process proposed to capture carbon dioxide (CO2) from flue gas showed great potential to reduce costs. The filter is filled with low-cost carbonaceous sorbents, such as activated carbon, which has a high capacity to retain CO2 but not nitrogen (N2), which means a high CO2/N2 selectivity.

The carbon filter process can recover at least 90% of the flue-gas CO2 of 90% purity at a fraction of the cost normally associated with the conventional amine absorption process.

The filter can produce low-cost CO2, because it requires neither expensive materials nor expensive flue gas compression, and it is easy to heat integrate with an existing power plant or a grassroots plant without affecting the cost of the produced electricity too much.

References:
Maciej Radosz, Xudong Hu, Kaspars Krutkramelis, and Youqing Shen, "Flue-Gas Carbon Capture on Carbonaceous Sorbents: Toward a Low-Cost Multifunctional Carbon Filter for 'Green' Energy Producers." Industrial & Engineering Chemistry Research, May 21, 2008.