Mechanical gas separation for post-combustion CO2 capture

[Wilmot McCutchen]

At thermal equilibrium, all molecules in a gas mixture have the same kinetic energy, so the lighter molecules (low molar mass) must be going faster. The kinetic energy may be the same, but low molar mass molecules are on average faster, and have lower momentum.

 

Consider a mixture of CO2 and nitrogen at thermal equilibrium at room temperature (300 K). The average speed (vrms) of the N2 molecules is 517 m/s and that of CO2 molecules is 412 m/s. The CO2 molecules have higher momentum than the N2 molecules. Consider now an eddy in that gas mixture and continuous axial flow out of that eddy. It is a coherent vortex like the Red Spot of Jupiter. The axial flow should be rich in low momentum molecules, i.e. nitrogen. Self separation of a gas mixture can occur spontaneously given the proper flow geometry. But without a way of collecting the separation effects to prevent remixing, nothing on a macroscopic scale will be noticed.

 

Could the intrinsic momentum differences of the fractions in the gas mixture be exploited for mechanical gas separation in a way that does not depend on the conventional gas centrifuge of narrow cylinders rotating at high speed? For example, in a forced flow geometry comprising eddies having a 1mm radius, connecting to larger vortex structures in a tree network? In that 1mm eddy, the radial acceleration of the N2 is v^2/r or 27.2 million g. CO2 is only 17.3 million g.

 

Area-preserving fractal networks, like the root system of trees, might provide a way to strip out the N2 through connected vortex cores so as to leave behind the CO2. See George Johnson's very interesting essay on quarter-power scaling and fractal networks. http://hep.ucsb.edu/courses/ph6b_99/0111299sci-scaling.html See also http://www.freepatentsonline.com/7901485.pdf

[Enthalpy]

What you describe with hidden words is the presently standard method for isotope separation, especially for uranium enrichment, and the second-standard method.

And, sorry, I won't help you in that way.

[CaptainPanic]

Could the intrinsic momentum differences of the fractions in the gas mixture be exploited for mechanical gas separation in a way that does not depend on the conventional gas centrifuge of narrow cylinders rotating at high speed? For example, in a forced flow geometry comprising eddies having a 1mm radius, connecting to larger vortex structures in a tree network? In that 1mm eddy, the radial acceleration of the N2 is v^2/r or 27.2 million g. CO2 is only 17.3 million g.

I think you can do this, but to have such a small rotating system to separate it you will run into 2 problems:

 

1. It's bloody expensive. You will need thousands, if not millions of these little vortex structures, and all have to be connected to each other in a parallel setup, possibly with some of them in series too to achieve a good separation. It's true that nature comes up with fractal systems - but they are self-organizing (you don't have to build it). This one, you need to construct manually and that won't be cheap.

2. The smaller the vortex and the higher the bulk gas velocity, the higher the pressure drop. And pressure drop means loss of energy (kinetic energy becomes heat).

 

If you want to separate CO2 from nitrogen, I would just remove it by absorption. For example, CO2 will dissolve into water much better than nitrogen. Problem solved. And that solubility is temperature dependent, so you can also get it out again. Problem solved.

For other simple binary gas mixtures, cryogenic distillation can be a method.

 

What you describe with hidden words is the presently standard method for isotope separation, especially for uranium enrichment, and the second-standard method.

And, sorry, I won't help you in that way.

Not everybody on the internet is interested in uranium enrichment for terrorist purposes or to help evil regimes. Some people are just chemical engineers with new ideas. And not all systems with lots of centrifuges are for uranium enrichment.

 

Anyway, since wikipedia describes the process, one at least two websites, and even has a picture, I don't think this is a secret.

Edited September 16, 2011 by CaptainPanic

[Wilmot McCutchen]

The requisite millions of vortex structures would self-organize into an area-preserving fractal network, where vortices on many scales converge at the axis of rotation of counter-rotating centrifugal impellers. Pressure drop would be minimal because of the area-preserving character, where the cross-sectional areas of the eddies sum to the cross-sectional area of the axial exhaust conduit. The shear layer between the two boundary layers against the impellers is where the fractal network provides a sink flow path for nitrogen to axial extraction, assisted by a steam ejector. In addition to axial suction there is the backpressure radially in. The impellers preferentially advect the high molar mass constituents like CO2, SOx, mercury, radially outward as the nitrogen is advected radially inward through the vortex cores. So you see I was not talking about a cascade like the Dyson vacuum cleaner where there is a solid structure defining the vortex core, but a dynamic cascade. l am not sure if it would be appropriate to describe it as inverse Kolmogorov scaling -- the opposite of turbulent dissipation. Negative entropy through intelligent design.

[John Cuthber]

Or you could just sieve out the nitrogen.

http://en.wikipedia.org/wiki/Nitrogen_generator#Membrane_technology

 

It would be slightly better still if you separated the O2 from the N2 before the combustion process.

The flame would be hotter and so more of the energy could be recovered.

 

However, it would be very expensive.

You also need to remember that carbon capture doesn't stop us running out of coal, oil or gas so it's not a sustainable technology.

In any event, I don't think your system would work, it reminds me of this

" Big whorls have little whorls

That feed on their velocity,

And little whorls have lesser whorls

And so on to viscosity."

[Enthalpy]

Not everybody on the internet is interested in uranium enrichment for [bad purposes].

Not everyone, sure. But if I put ideas on the Internet for someone, other people read them and make other uses of them.

Suggesting expensive methods, that are used for uranium enrichment, but are far worse than the existing methods already used for CO₂ scrubbing, makes me wary.

 

Anyway, [], I don't think this is a secret.

Existing published technology shouldn't be secret, sure. But I have no intention to improve that particular technology.

 

It would be slightly better still if you separated the O2 from the N2 before the combustion process. [] However, it would be very expensive.

It is done presently and is affordable. Rather cheaper than separating the dioxide after combustion. Sequestration makes electricity produced with coal more expensive, but still far cheaper than nuclear or wind energy - it could stay cheaper than gas without sequestration.

 

...Carbon capture doesn't stop us running out of coal.

Proven and accessible coal reserves can replace oil and gas and still supply the whole Mankind for 1-2 centuries, so it can be a transition to the next source, with sequestration. Not the solution I dream of, but an affordable and available one.

[Wilmot McCutchen]

 

Proven and accessible coal reserves can replace oil and gas and still supply the whole Mankind for 1-2 centuries, so it can be a transition to the next source, with sequestration. Not the solution I dream of, but an affordable and available one.

 

Sequestration in depleted oil reservoirs for enhanced oil recovery (EOR) can't be extrapolated to deep saline aquifers, which are full of brine. You'd have to get the brine out of there, and then where would it go? Not to the surface, surely, because we don't even have a solution for reverse osmosis reject brine yet, and this new ocean of 100,000 ppm brine will be a treatment nightmare. Where else, into the groundwater? Geological sequestration at utility scale has been condemned by petroleum engineering experts as "a profoundly non-feasible option for the management of CO2 emissions." http://twodoctors.org/manual/economides.pdf The available pore space has been overestimated. I agree with you that we seem to be stuck with coal for the time being.

[Wilmot McCutchen]

 

Suggesting expensive methods, that are used for uranium enrichment, but are far worse than the existing methods already used for CO₂ scrubbing, makes me wary.

 

 

I did not mean to suggest that any methods that are presently used for uranium enrichment could be used for CO2 scrubbing. I don't understand how you came to the conclusion that my suggestion is "far worse than the existing methods used for CO2 scrubbing." Apparently you are under the impression that I was talking about narrow spinning cylinders, which is not the case.

 

The existing methods for post-combustion CO2 capture you refer to are chemical capture by amine scrubbing, cryogenic separation, and membranes. Membranes are not feasible for flue gas, which comprises mercury vapor, water vapor, fly ash, and other aerosols that clog the pores. The work pushing utility-scale flow through the membrane is prohibitive. Cryogenic separation of hot dirty flue gas is not as good as air separation before combustion. Amine scrubbing will not work at utility scale because it will double water consumption of power plants. Also, heat stable salts are an unsolved problem.

 

Air separation for oxyfuel combustion has the same 30% energy penalty as amine scrubbing, but it avoids the water consumption problem. Wouldn't a solution that uses the intrinsic momentum differences of flue gas fractions to self-separate them be superior to these conventional methods? If so, then why not think about it? Why must we only consider amine scrubbing for carbon capture?

[CaptainPanic]

I did not mean to suggest that any methods that are presently used for uranium enrichment could be used for CO2 scrubbing. I don't understand how you came to the conclusion that my suggestion is "far worse than the existing methods used for CO2 scrubbing." Apparently you are under the impression that I was talking about narrow spinning cylinders, which is not the case.

 

The existing methods for post-combustion CO2 capture you refer to are chemical capture by amine scrubbing, cryogenic separation, and membranes. Membranes are not feasible for flue gas, which comprises mercury vapor, water vapor, fly ash, and other aerosols that clog the pores. The work pushing utility-scale flow through the membrane is prohibitive. Cryogenic separation of hot dirty flue gas is not as good as air separation before combustion. Amine scrubbing will not work at utility scale because it will double water consumption of power plants. Also, heat stable salts are an unsolved problem.

 

Air separation for oxyfuel combustion has the same 30% energy penalty as amine scrubbing, but it avoids the water consumption problem. Wouldn't a solution that uses the intrinsic momentum differences of flue gas fractions to self-separate them be superior to these conventional methods? If so, then why not think about it? Why must we only consider amine scrubbing for carbon capture?

Let's start off with a welcome to the forum! I just noticed the post count of you, and this must be one of your first discussions here.

 

It's always a good idea to discuss new methods to improve separations. Chemical separations are a huge energy consumption in the world, and improvements can give enormous benefits.

 

The method you describe is not free of penalties either though. The big question is of course which is worse, and it's not so easy to calculate that. I certainly don't have the time to calculate it in my coffee breaks. And without calculations, all that remains is to be skeptical of each others' proposals.

 

First of all, you might already know that regardless of the method you choose to do your separation, you need to overcome the difference in entropy anyway. Splitting up a mixture into its pure components simply costs energy, always. So, there's no free lunch here.

 

It is my feeling (but I haven't calculated it yet) that the cyclones you describe require a relatively high gas velocity to get the G-forces to high enough values and that they will have a relatively high pressure drop as a result (difference in pressure before and after the cyclone). This means that you need to install compressors to pressurize the gas... and compressors are notorious energy consumers.

It's my feeling that you need to find out the pressure drop before we can say anything decent about this method of separation.

[Wilmot McCutchen]

Let's start off with a welcome to the forum! I just noticed the post count of you, and this must be one of your first discussions here.

 

It is my feeling (but I haven't calculated it yet) that the cyclones you describe require a relatively high gas velocity to get the G-forces to high enough values and that they will have a relatively high pressure drop as a result (difference in pressure before and after the cyclone). This means that you need to install compressors to pressurize the gas... and compressors are notorious energy consumers.

It's my feeling that you need to find out the pressure drop before we can say anything decent about this method of separation.

 

True. Thanks for your welcoming words. Let's assume that the area-preserving character of the self-similar vortex network means no pressure drop. The high gas velocity is not due to the impellers in the von Karman geometry, but to the intrinsic velocities of the molecules at thermal equilibrium. The necessary work for organizing the highly turbulent and high enthalpy flow can be compared to traffic enforcement by highways and traffic laws. Not nearly as much work as forcing the molecules, because you are just organizing the flow so the different momenta sort themselves into a radial spectrum, with steam at the center, then nitrogen, then CO2, then SO2. This spectrum is preserved in the zero pressure drop fractal vortex flow. In the case of turbine exhaust steam condensation, the expanding steam counter-rotates the organizing impellers and even harvests power. See http://www.freepatentsonline.com/7987677.pdf

Radial Counterflow Steam Stripper US7987677B2.pdf

[CaptainPanic]

If you approach this from the perspective of a single molecule, then what's the incentive for that molecule to travel in a circular motion in that vortex? Molecules only change direction because they bump into other molecules.

 

Molecules travel in a straight line until they hit something... and I do not yet understand why they would be separated individually.

 

The whole point of using a vortex is to separate something by a difference in density. But you cannot talk about density when you talk of a single molecule. The bulk gas has a density, but a molecule does not. A molecule only has weight.

[Wilmot McCutchen]

If you approach this from the perspective of a single molecule, then what's the incentive for that molecule to travel in a circular motion in that vortex? Molecules only change direction because they bump into other molecules.

 

Molecules travel in a straight line until they hit something... and I do not yet understand why they would be separated individually.

 

The whole point of using a vortex is to separate something by a difference in density. But you cannot talk about density when you talk of a single molecule. The bulk gas has a density, but a molecule does not. A molecule only has weight.

 

The counter-rotating centrifugal impellers of the open von Karman flow geometry provide surfaces for the molecules to bounce off and change direction. What happens to an individual molecule in the gas mixture would be difficult if not impossible to model, but bulk behavior of the ensemble of mixed molecules would be more predictable. Where the impellers come to close separation and high relative velocity (i.e. high shear rate) there is swirl forced in the gas mixture, and the vortices can't have a diameter greater than the impeller separation, even at the largest scale. It is this forced swirl of the bulk that provides the incentive for a particular molecule to change direction. Even if it does not bounce of an impeller, it will bounce off other molecules, and the high momentum molecules in the vortex shells absorb momentum from the low momentum molecules in the vortex cores, acting effectively as solid surfaces. With low momentum molecules continuously stripped out through the self-similar area-preserving vortex network in the shear layer between the impellers and axially exhausted, the remaining molecules of the bulk will have high momentum. Momentum diffusion from the high momentum molecules into the impellers will drive counter-rotation to force source flow of the bulk, and the steam ejector communicating with the axial exhaust conduit will force sink flow of the nitrogen ballast to axial extraction. The bulk will churn in the scrubber until eventually each molecule finds its exit.

 

Looking from the point of view of an individual molecule, on entering the von Karman geometry it has two possible eventual exits: axial and peripheral. It may have a long residence time in the scrubber and a tortuous flow path, but eventually it will either go out the axial exhaust conduit at the axis of rotation of the impellers, or it will exit the space between the impellers at their periphery.

 

Vortex separation by density is actually separation by momentum. The Vortex Tube thermal separation of gases is an example: high momentum gas molecules go out the hot end and low momentum molecules go out the cold end, without any forcing by moving parts. I believe this is called the Ranque effect. Maxwell's Demon (molecular separation without work) may be smart geometry where self-organized bulk separation of gases is able to manifest if the Reynolds number is high enough.