Vorsana

Post-Combustion Carbon Dioxide Capture

Flue gas from coal-fired electric power plants is the main villain in global warming. Cars are not nearly as bad, so vehicle emission standards would have only a small effect on global climate change. The real problem is coal, and our demand for electricity.

The U.S. emitted 6,049 million metric tons of CO₂ in 2004, but by now we probably have yielded the gold medal to China. See the worldwide list here.

According to the IPCC Report on Carbon Capture (September 2005):

... the power and industry sectors combined dominate current global CO₂ emissions, accounting for about 60% of total CO₂ emissions. Future projections indicate that the share of these sectoral emissions will decline to around 50% of global CO₂ emissions by 2050 (IEA, 2002). The CO₂ emissions in these sectors are generated by boilers and furnaces burning fossil fuels and are typically emitted from large exhaust stacks. ... The largest amount of CO₂ emitted from large stationary sources originates from fossil fuel combustion for power generation, with an average annual emission of 3.9 MtCO₂ per source. Substantial amounts of CO₂ arise in the oil and gas processing industries while cement production is the largest emitter from the industrial sector. ... The ranges of the technical capture potential relative to total CO₂ emissions are 9–12% (or 2.6–4.9 GtCO₂) by 2020 and 21– 45% (or 4.7–37.5 GtCO₂) by 2050.

Coal is indispensable for our electricity. In the United States, for example, here are the figures in megawatt hours for the various forms of utility electricity production in 2006.

Nuclear is unpopular, and hydro is at its limit, with few damsites still available. Non-hydro renewables (wind, solar, and all others combined) are too small to matter. That leaves natural gas and coal, both of which are heavy emitters of carbon dioxide.

So, given our dependence on electricity and the huge streams of hot and dirty flue gas pouring into the atmosphere from coal combustion, how do we treat the flue gas to extract the carbon dioxide? That is called “post-combustion carbon capture.”

Present Methods of Post-Combustion Carbon Capture

Three carbon capture methods are presently known: membranes, compression, and amine scrubbing. All are unsatisfactory for dealing with large, hot, and dirty gaseous emission streams like flue gas from coal-fired power plants.

The National Energy Technology Laboratory run by the U.S. Department of Energy summarizes the problem of carbon capture from flue gas:

“The low pressure and dilute concentration dictate a high actual volume of gas to be treated. Trace impurities in the flue gas tend to reduce the effectiveness of the CO₂ adsorbing processes. Compressing captured CO₂ from atmospheric pressure to pipeline pressure (1,200 - 2,000 pounds per square inch (psi)) represents a large parasitic load.”

Translation: The presence of nitrogen ballast (N₂) in the flue gas (about three-quarters of its volume) means that carbon dioxide is protected from compression or chemical contact by a cushion of nitrogen molecules. NOx and SOx (nitrogen oxides and sulfur oxides) become heat-stable salts and corrosive acids during amine recovery, and along with the fine glassy dust (fly ash) of the flue gas these clog up and damage the equipment. The energy required for compression is prohibitively wasteful.

Carbon Capture by Membrane Separation.

Membranes may work in laboratory scale experiments, but could not possibly work on flue gas. The immense volumes to be filtered, and the pore-clogging fly ash and mercury in the flue gas, weigh heavily against membrane carbon capture.

Carbon Capture by Cryogenic Distillation.

Likewise carbon capture by compression. Flue gas is hot and very dirty, and the carbon dioxide in it is shielded from liquefaction by the presence of a large nitrogen ballast. Cryogenic distillation captures carbon by liquefaction of flue gas and separates out NOx and SOx by fractional distillation. Nitrogen, however, is very hard to liquefy and compressing it wastes energy which should go into compressing NOx, SOx, and CO₂. The small partial pressure of NOx and SOx in flue gas (which are very much less than 1% of the volume) and the small partial pressure of CO₂ (10 -15%) are both due to the high nitrogen ballast (75%).

Carbon Capture by Amine Scrubbing.

That leaves the third alternative, amine scrubbing, which is the only method presently being considered seriously for carbon capture from flue gas. Aqueous amine sorbents have been successfully used to clean carbon dioxide and hydrogen sulfide from natural gas and industrial waste streams, but the hope that this proven technology can be extended to flue gas runs into serious difficulties.

1. The science is not mature, and there is no time for study.

“The exact mechanism describing the chemical reaction of CO₂ with amine regents [sic] under the conditions typical of a CO₂ capture plant is the subject of much debate.” International Test Centre for CO₂ Capture (ITC) (Report of May 2005 §4.1.1).

The science is not mature and there are major unsolved problems. The situation is an emergency. Academic research, at its leisurely pace, and with its usual small and obscure results, cannot be expected to solve the problem of carbon capture from flue gas.

2. Amine scrubbing is very expensive.

The cost of amine scrubbing to capture carbon dioxide, then compressing it to pipeline pressure, is about $2000 per kilowatt, which makes it prohibitively expensive even if it were feasible for high volumes of flue gas, which it is not.

It is hoped that the operating cost of amine scrubbing might come down to about $30 per ton of CO₂ captured and compressed, at Canadian energy costs. For the average source emitting 3.9 million tons per year, the best case price tag for amine scrubbing is an additional $120 million. That’s assuming it will work, which it won’t.

The theoretical maximum loading capacity of amines is a mole of amine for one mole of carbon dioxide. Actual performance is even lower, so you need even more amines. In addition, there is the replacement cost of amines which cannot be regenerated, which is an especially serious problem for flue gas scrubbing (see below) due to the presence of strong acids from NOx and SOx. So capturing the carbon dioxide in a typically enormous stream of flue gas (a million cubic feet per minute) would require an equally enormous purchase of amines.

Such a prospective bonanza to chemical companies from the plight of the planet should be taken into account in evaluating the pressures on researchers and policymakers, and the confident claims of the amine scrubbing proponents.

3. Corrosion is a serious problem.

Corrosion of the common materials of construction, such as carbon steel, in an amine scrubbing facility is a serious problem known from years of experience in natural gas sweetening. Carbon steel corrodes in the amine solution.

Flue gas presents an even more challenging corrosion problem because it is hot and contains oxygen, unlike natural gas. The rate of corrosion of carbon steel is higher in hot amine solutions. Flue gas also contains a significant fraction (~ 4%) of oxygen. The corrosion rate has been found to increase by 40% when oxygen content is increased from 0% to 10%.

4. Heat stable salts and fly ash sludge block the process.

Natural gas does not contain combustion products like NOx and SOx and fly ash. NOx and SOx turn into nitric acid and sulfuric acid when combined with water. Alkanoamines react with strong acids (sulfuric acid and nitric acid are very strong acids) to form heat stable salts, which effectively take the amines out of use, so more need to be added. During the process of regenerating the amine for re-use in the CO₂ stripper, a crust of heat-stable salts covers the heat exchange surfaces, reducing efficiency. A dilute sludge of fly ash and precipitated salts gums up everything.

If the aqueous amine approach from natural gas production were to be used for flue gas from coal, the additional acids, heat-stable salts, and fly ash sludge would be operating difficulties in addition to the already known corrosion problems and would drive up the operating costs considerably.

5. You have an even bigger problem with scrubber wastewater.

Amine scrubbing converts an air pollution problem into a water pollution problem. Large amounts of solution must be sprayed into the flue gas in order to get around the nitrogen ballast. Carbon dioxide molecules are in low concentration, like needles in a haystack, so to contact all of them a lot of amine solution must be used. What comes out is a voluminous stream of toxic wastewater. Fine solids suspended in this wastewater require a long time to settle, and there is also the dilute sulfuric and nitric acid from the NOx and SOx in the flue gas. Storing the wastewater requires lagoons or tanks, which are a waste of valuable space.

Conclusion:

From the foregoing discussion of amine scrubbing, membranes, and compression, it should be clear that there is in fact no present technologically feasible solution to carbon dioxide emissions from coal-fired power plants.

But blaming coal is not the answer, because alternative energy sources cannot satisfy our hunger for electricity. Wind and solar combined can only meet about 1% of the need. Either we cut back severely on electricity, or come up with another solution.

So What’s the Solution?

None of the presently known carbon dioxide capture methods is feasible unless the flue gas is first scrubbed of its particulates, NOx and SOx, cooled, and stripped of its nitrogen ballast.

That is what we are working on at Vorsana. The solution is actually very simple and inexpensive: highly turbulent scrubbing and radial counterflow in a rotating device.

An Alternative Method of Carbon Capture: Centrifugal Gas Separation.

Stripping the nitrogen ballast amounts to carbon capture, because what is left once the nitrogen and water vapor comes out is a concentrated stream of carbon dioxide, which can then be economically scrubbed of its particulates, NOx and SOx and compressed or disposed of by other means. Stripping the nitrogen ballast can be done by centrifugal gas separation, although not by presently known technology.

A useful illustration of centrifugal separation is the cream separator, which spins milk so it separates into cream and whey (skim milk). The cream, having a lower density than the whey, concentrates at the axis of rotation. If the raw milk were left in a glass without doing anything, the cream would rise to the top. The cream separator just exaggerates the acceleration and makes the separation happen faster. Centripetal acceleration caused by rotation of the milk in a drum squeezes the low density fractions like cream into the center, just like the acceleration due to gravity makes cream rise to the top. Heavy fractions, like water, displace the light fractions in rotation.

Density differences exist for the fractions in flue gas as well. It turns out that the bad stuff – carbon dioxide, NOx and SOx, mercury, and fly ash – is the heavy fractions. The good stuff – nitrogen and water vapor – is the light fractions. So, theoretically at least, spinning the flue gas fast enough should develop sufficient centripetal acceleration to separate the good stuff from the bad stuff.

What’s in coal-fired power plant flue gas?

Nitrogen (N₂) is a harmless gas which constitutes 75% of the volume of flue gas from coal-fired power plants. This is referred to as nitrogen ballast. Nitrogen might be safely discharged to the Earth’s atmosphere, which is already 78% nitrogen. Water vapor (5%) produced by combustion is another harmless light fraction. That’s the white plume you see coming out of the smokestack. It is also why your car tailpipe drips water. So 80% of the volume of flue gas is harmless and requires no treatment at all, other than separating it from the carbon dioxide and other heavy fractions.

Once the nitrogen ballast and water vapor have been stripped, the remaining one-fifth of flue gas can be scrubbed to remove NOx, SOx, mercury, and fly ash using known methods or new ones. The toxic lagoon problem of known wet scrubbing methods will be minimized because less scrubbing solution will be necessary to contact the NOx and SOx. And once the nitrogen ballast has been stripped, there is no need for amine scrubbing.

The molar mass of nitrogen gas (N₂) is only 28 g/mol (grams per mole of gas); carbon dioxide (CO₂) is 36% denser at 44 g/mol. Sulfur dioxide is even denser (64 g/mol). Centrifugal gas separators which might exploit this 36% density difference are of two classes: mechanically driven and pressure driven.

Mechanically driven centrifugal gas separators.

Mechanically driven gas separators can exploit gas density differences as low as 1.5% , far beyond the performance required for flue gas separation. The ultracentrifuge is a very delicately balanced cylinder rotating at very high speed and generating very radial acceleration and high G force which radially stratifies gases by density within the cylinder. Such rotating cylinder gas centrifuges are used for separation of uranium isotopes to collect fissionable U-235 for bombs or peaceful purposes. Because of their extremely high rpm, gas centrifuge imbalances can cause catastrophic accidents. The separation effects are small for each device, so the output of one becomes the input of a second device, and so on, in what is known as a cascade.

Adapting conventional gas centrifuges of this type to flue gas would be impracticable. The huge volume of flue gas is an insuperable obstacle, and the fly ash, water, mercury, etc. which would condense under pressure might easily cause imbalances which would cause the centrifuge to destroy the facility.

The Vorsana radial counterflow device is a mechanically driven gas separator which operates on a different principle to perform separation at the comfortable 36% density difference between nitrogen and carbon dioxide, making it suitable for processing large volumes of hot and dirty flue gas with minimal rotation speed.

Pressure driven centrifugal gas separators.

Pressure driven devices include inertial collectors (also known as cyclones), and vortex tubes. Cyclones and vortex tubes are axial counterflow devices, wherein flow goes in opposite directions and the device is static. These have no moving parts.

1. Cyclones.

Inertial collectors, commonly known as cyclones, are used extensively to process gaseous emission streams to remove large particles of dust. Tangential feed swirls downward along a tank wall, then swirls upward to exhaust, meanwhile expelling most of its dust when it makes the turn. Pressure drives the flow, from compressing the feed or from sucking the exhaust, or both. The dust collects in a hopper at the bottom of the tank. Because both the downward and upward swirl are on the same axis of rotation, this is axial counterflow. Cyclones, even cascaded, are ineffective even for the relatively easy job of separating out >2.5 micron fly ash, which is relatively big dust. Nitrogen / carbon dioxide separation by cyclones has not been reported and would appear to be impossible.

2. Vortex tubes.

Another pressure driven centrifugal gas separation device without moving parts is the Ranque-Hilsch vortex tube. The effect of a vortex tube is to separate a high pressure mixed gas feed stream into a low pressure cool light fraction stream, which comes out the feed end, and a low pressure hot heavy fraction stream which comes out the other end of the tube. Residence time in the vortex tube is on the order of milliseconds – very short for doing effective separation by density. There are a variety of opinions about why thermal separation happens, but no consensus.

What is especially interesting about the vortex tube, and what has provoked a spirited discussion, is that a thermal separation happens without any work being put in, as in the thought experiment which is known as “Maxwell’s Demon.” In other words, disorder decreases without effort – an apparent violation of the Second Law of Thermodynamics.

In operation, pressurized feed gas is tangentially injected into one end of a static tube having both ends open. The feed gas swirls in a first vortex to the other end, as in a cyclone. At the other end is a conical central flow impedance. Hot gas exits the tube around the flow impedance. A recirculation flow rebounds from the other end in a second vortex inside the first vortex and exits the feed end cooler than the feed gas. Like the cyclone, the vortex tube is an axial counterflow device.

Application of the vortex tube has also been made to the separation of liquefied air into nitrogen and oxygen, to removing condensible vapors from natural gas, and to improving sorbent mixing for carbon dioxide scrubbing. The presence of fly ash in flue gas, the high energy requirement for pressurizing the feed, and the poor separation efficiency of vortex tubes due to their low residence time would appear to foreclose application of the Ranque-Hilsch vortex tube to the job of extracting nitrogen ballast from flue gas.

Conclusion:

Both conventional mechanically driven gas centrifuges and conventional pressure driven cyclones and vortex tubes are inadequate for separating nitrogen from carbon dioxide. What is needed is a new approach.