A Proposal for Hydrogen, Synthetic Fuels and the Halving of US CO2 Emissions
14 March 2006
A flow-sheet for 100,000 gallons per day of CO2-free synfuel. Click to enlarge. |
A team from General Atomics is proposing the use of hydrogen provided from non-fossil sources (solar, wind or nuclear) and CO2 captured from coal-fired power plants or from the air to produce enough Fischer-Tropsch synthetics to meet the fuel needs of the transportation sector.
With such an approach, proposed in a poster session at the NHA hydrogen conference, the total net US release of CO2 could be halved, even factoring in the release of CO2 from the ongoing combustion of hydrocarbon—although not fossil—fuels, according to the researchers’ analysis.
The production rate of CO2 from coal power plants in the US is 1,875 million metric tons/year. If this CO2 were captured using proven absorption processes and used with hydrogen produced by solar, wind or nuclear energy to make synfuel, it would provide all the hydrocarbon fuel needed for our transportation economy.
Since that transportation economy produces 1,850 million metric tons of CO2 per year, this synfuel process would cut our CO2 production in half. We could shift from a petroleum-based transportation economy to a synfuel transportation economy.
This would reduce our petroleum use by 75%, and reduce our CO2 production by 50% with no increase in coal use. It would require significant quantities of hydrogen (255 million metric tons/year, or 25 times our current production) that would be produced from water using solar, wind or nuclear energy.
This hydrogen synfuel concept would allow us to significantly reduce our use of petroleum, and cut our CO2 emissions in half, while still using our existing hydrocarbon-based transportation infrastructure. It could provide a bridge to a pure hydrogen economy.
The Fischer-Tropsch process takes a synthesis gas (syngas) rich in hydrogen and carbon monoxide and converts it catalytically to liquid fuels and chemicals. The synthesis gas is produced by the gasification of carbon-bearing feedstocks (coal, biomass) or by the reforming of natural gas.
The gasification and reforming processes are energy, emissions and cost-intensive. The basic gasification reaction (for coal, for example) is:
2C + ½O2 + H2O → 2CO + H2
The Water-gas Shift reaction is then used to produce additional hydrogen:
CO + H2O → H2 + CO2
The reaction for producing Fischer-Tropsch products (generically [CH2]n) from synthesis gas (CO and H2) is:
CO + 2H2 → [CH2]n + H2O
The simultaneous Fischer-Tropsch and Water-Gas Shift reactions in the reactor leads directly to the complete reaction:
2C + H2O + ½O2 → [CH2]n + CO2
In summary, the process uses two carbons and half an O2 for every CH2 produced. Substituting such synfuels for petroleum-based fuels in transportations would triple US coal use and double current CO2 emissions.
Adding hydrogen from an external source into the process, however, cuts the carbon need in half compared to synfuel from standard coal gasification, and eliminated the production of CO2 from the process reactions.
Adding Water-split H2 into the F-T Process | |
---|---|
Gasification | C + ¼O2 + ½H2O → CO + ½H2 |
Water-splitting | 3/2H2O + Energy →3/2H2 + ¾O2 |
F-T Reaction | CO + 2H2 → CH2 + H2O |
Net Reaction | C + H2O + Energy → CH2 + ½O2 |
The US currently produces 11 million tons of hydrogen annually primarily through steam reformation of CH4 (methane). The process is fossil-dependent and produces 100 million tons/yr of CO2.
Using low-temperature electrolysis, high-temperature electrolysis or thermochemical conversion of water (assuming the electricity is provided by wind, solar or nuclear and the heat is provided by solar or nuclear), eliminates the generation of CO2 from hydrogen production.
If CO2 is used as the source for carbon in the FT process, the gasification step is replaced by a reverse water-gas shift reaction and the outcome becomes even more attractive.
Synfuel by CO2 Capture + H2 from Water-splitting | |
---|---|
Reverse water shift | CO2 + H2 → CO + H2O |
F-T reaction | CO + 2H2 → CH2 + H2O |
Water-splitting | 3H2O + Energy → 3H2 + 3/2O2 |
Net Reaction | CO2 + H2O + energy → CH2 + 3/2O2 |
No coal or methane is needed, and one CO2 is consumed for each CH2 produced. When the CH2 is burned, the process is net carbon-neutral.
If used to replace oil, the use of these synfuels in transportation would cut US carbon dioxide emissions in half.
Carbon dioxide is readily available, the General Atomics team points out, from flue gas from fossil power plants. A 1,000MW coal-fired power plant produces 5.5 million tons of CO2/yr. (14,500 tons/day).
Coal-fired plants, which account for 53% (0.38TWh) of US electricity generation, generate 2 billion tons of CO2/yr—meeting the total annual CO2 requirement to make synfuel for US transportation needs.
Carbon dioxide could also be captured from the atmosphere. Membrane separation of CO2 from air followed by its absorption by either amine or inorganic solvents is an emerging technology, but has been demonstrated on a laboratory scale.
Large airflow would be required due to low concentration of CO2 in air, and the resulting CO2 would be costly: about $0.10/kg (about $1.0/gallon cost added to the synfuel).
All the component pieces have been demonstrated, the General Atomics team points out. What is required is an integrated demonstration.
To actually implement the proposal would require enormous amounts of both feedstocks.
The US consumes 260 million gallons of transportation fuel per day. (13 million barrels of crude)
2.5 million tonnes of CO2 (2.5 billion kg) and 0.35 million tonnes of H2 (350 million kg) per day would be required to make synfuel to replace crude.
General Atomics naturally points to the potential for using nuclear reactors for high volume high-temperature electrolysis or thermochemical conversion of water to hydrogen.
But assuming sufficient hydrogen from renewable or zero-carbon processes, and factoring in a modest carbon credit, the production of synthetics in this manner could make economic sense as well as environmental sense, according to the team.
Resources:
Oh yeah, great! Let's just cover Utah in solar panels (more like Utah and Colorado) and Nevada in nuclear (sorry, make that nucular) waste, and viola!
What about if we use a perpetual motion machine to produce the hydrogen?
Posted by: An Engineer | 14 March 2006 at 11:17 AM
I have a better way to convert all that CO2 coal power plants are emitting into fuel. Algal biodiesel. GreenFuel Technologies.
Posted by: Cervus | 14 March 2006 at 11:27 AM
A little tiny 10 x 10 mile square of ocean is all that is needed to meet all of California's energy needs with wave technology. Renewable energy in sufficient quantities could easily be made available.
Why not stop at the hydrogen production, however? You could produce less hydrogen and power fuel cell vehicles directly.
Posted by: Rick | 14 March 2006 at 11:34 AM
Nobody asked the question: whether
... is either less expensive than converting to e.g. electricity (starting with the same energy sources but using plug-in hybrids instead of chemistry), or will achieve the reductions in pollution and GHG's that we need.I suspect the answer is "no to both".
Posted by: Engineer-Poet | 14 March 2006 at 12:29 PM
Two posts to biodiesel comment, four posts to plug-in comment.
Par for the course.
Mike could do a post about tulips and people would end up ranting about biodiesel and plug-ins.
Posted by: Joseph Willemssen | 14 March 2006 at 01:05 PM
Note two things about the first paragraph:
(a) the company proposing this is called "General Atomics". Three guesses what they are into.
(b) "non-fossil sources (solar, wind or nuclear)" - excuse me, but since when does uranium grow on trees? It's a fossil fuel, with very limited primary supplies. More to the point, how can anyone justify naming solar and wind and nuclear fission in the same context? Both are expensive, but nuclear has safety and massive radioactive waste and decomissioning issues. Talk about apples and oranges!
IMHO, the whole thing is just more smoke and mirrors from the nuclear lobby. Consider this: H2 production from electrolysis is ~50% efficient, reducing CO2 to CO requires a lot of H2, and Fischer-Tropsch is only ~50% efficient. The end result is burnt in diesel engines (40% efficient BEST case). Average well-to-wheels efficiency would be well below 10%!
Fair enough, CO2 emissions would go down. However, the production cost of the synfuel would be extremely high and, you'd end up with enormous piles of additional radioactive waste. Selling this concept as a contribution to the environmental debate is truly a new definition of chutzpah.
Posted by: Rafael Seidl | 14 March 2006 at 02:39 PM
Shrub needs to get on national TV and explain this, especially the chemistry.
Posted by: t | 14 March 2006 at 03:07 PM
I'm glad to see this concept getting some attention. It's similar to an approach that I've been advocating, but more agressive, and arguably less realistic. I'll say more about that, but first some comments on the comments.
Producing enough hydrogen to convert that much CO2 to (CH2)n would be a daunting challenge, but the image of "covering Utah and Colorado" with solar cells and the state of Nevada with nuclear wastes is completely bogus. And what would 'an engineer' have us do instead? Keep burning coal, and ignore all the CO2 we're putting into the atmosphere?
'Cervus' may well be right about having a better way of converting CO2 into fuel using solar energy. But I'd say the jury is still out. The very high productivity of algae strains reported in the lab has been been slow to translate into working operations in the field. And it's not for lack of trying. Still, I'm rooting for GreenFuel Technologies to succeed.
I think Rick overestimates the power available in waves. By two or three orders of magnitude. An accounting problem? Area of coverage is almost irrelevant to wave energy. What matters is linear distance. Once you've extracted the wave energy along a front, there's no energy left to extract in the area behind that front.
'Engineer-Poet's comments, as usual, are on-point. But they don't consider the issue of time. In the long run, we'd be better off redoing our infrastructure and shifting toward electrified rail and plug-in hybrids. But that won't happen overnight. There will be a *lot* of gasoline and diesel-powered vehicles still on the road two decades from now. It would be really nice if we had a cleaner, less carbon intensive, way to fuel them.
Raphael's numbers for the efficiency of electrolysis and of F-T synthesis are too pessimistic. If electrolyis on the scale needed to convert a fraction of coal plants' CO2 into fuel is implemented, you can expect to see efficiencies approaching 100%. That's because high temperature steam electrolysis will be used. The electrolysis cells will operate inside a well-insulated box, with electrical "losses" from cell resistance supplying the heat. Aside from the very small amount of heat that leaks through the insulation, *all* of the electrical energy input goes, one way or the other, into splitting water.
No one has pointed out what seems to me the most obvious criticism of the concept. Carbon, in the form of coal, has a certain potential energy that we would like to tap. If we burn it in a new-generation coal-fired power plant (supercritical steam, not IGCC) we capture about 40% of that potential energy as electricity. We can then turn around and use power from other sources to "unburn" the coal and convert it to a clean liquid fuel, but the energy required to do so will be substantially greater than we got from burning the coal in the first place. From an energy point of view, we'd have come out ahead by using CTL technology directly to produce the fuel. That approach can capture about 70% of the chemical potential energy of the coal in the fuel produced. Then the energy we would have used to convert the coal plant's CO2 output to fuel can go instead to replacing the coal plant's electrical output--with a good bit left over.
There are several side-issues that complicate that analysis, however. The coal-fired plants are already there, and aren't going away any time soon. And, as it turns out, it's cheaper to build a plant for converting hydrogen and CO2 into fuel than a comparably sized CTL plant. CO2 and hydrogen are nice pure starting points, and the water gas shift reaction is a relative piece of cake. Coal, on the other hand, is coal, and what comes off in coal gasification is a devils brew that eats valves and catalysts. (There's a reason that IGCC has been slow to catch on, despite its high efficiency.)
Coal combustion, of course, also creates a devil's brew, but the process of extracting CO2 from the flue gas is extremely effective in separating the CO2 from everything else. So burning the coal can perversely end up being the most cost-effective way to obtain a very pure carbon feedstock for fuel synthesis. Not cheap in energy costs, of course, but in terms of capital cost.
Burning coal in a power plant, followed by "unburning" a portion of it into fuel using renewable energy, can also serve as a way to address the problem of intermittancy of solar and wind resources. If you're using renewable energy mostly to make hydrogen, it doesn't matter all that much if it's intermittant.
Nonetheless, if we're going to be synthesizing fuel from coal, either directly or indirectly, it makes more sense to use hydrogen (and oxygen!) from non-carbon energy sources to increase CTL yields, rather than converting CO2. That's what I've been advocating. It has the same advantages for buffering intermittant energy resources, while giving a higher return on those energy resources. More "bang for the buck".
Posted by: Roger Arnold | 15 March 2006 at 12:43 AM
The second-to-last chemical equation:
Water-splitting 3H2O + Energy → CH2 + 3/2O2
I think it should be:
Water-splitting 3H2O + Energy → 3H2 + 3/2O2
Alternatively produce methanol:
CO2 + 3H2 + Energy → CH3OH + H2O
and
CO + 2H2O + Energy → CH3OH
Posted by: madScientist | 15 March 2006 at 12:22 PM
Thanks Roger,
That is a pretty good post, if I may say so at my own expense. I would agree that using renewable energy to produce renewable hydrocarbons is a good strategy. There are several reasons for continuing to use liquid fuels that are similar to today's fuels:
1. The possibility to gradually replace fossil fuel with renewable, at any ratio that is feasible.
2. The continued use of existing transportation infrastructure (I don't see the ICE going anywhere soon, but I digress).
3. The advantages of liquid fuel over both solid (ease of transportation) and gas (ease of handling and cost of transportation).
Once we agree on the above, the question becomes: What is the best way to achieve it? Part one: Where will the energy come from? Call me ignorant, but I just cannot get excited about nuclear. To me, the risks and the nuclear waste present a danger that is on par with concerns about GHG emissions. Solar (direct solar) is too diffuse (~1 kW/m^2) and is also intermittent. Wind fluctuate too much and too unpredictably. I like biomass, a self-replicating solar collector. Specifically I like waste biomass. There won't be enough waste to supply everything, but a large untapped potential is currently going to waste.
Part two: How to best convert biomass to fuel? An engineer would say, not ethanol: too much energy required for purification, the product is too volatile, and it is hygroscopic as well. Again, it should be blendable with existing fuel supplies. Biodiesel has potential, but it does require a mostly pure feedstock (i.e. energy demanding processing). It also requires a mostly pure alcohol (methanol typically, classified as a hazardous air polutant by Congress) and a mostly pure catalyst. Too much like organic chemistry practical, and too little like a refinery.
There are two promising technologies, according to this engineer: TDP and gasification/F-T. Thermal Depolymerization (aka Thermal Conversion) is certainly not everything its inventors claim it too be. From what I can work out, it converts mostly fat/oil waste into oil via two steps: hydrolysis (to release the fatty acid) and decarboxylation (to convert fatty acid to alkane/alkene, i.e. diesel grade oil). TDP thus directly competes with biodiesel. Compared to biodiesel, TDP has these important advantages:
1. Ability to process unpure waste (find a picture of the turkey guts to see what I mean).
2. Product that is chemically very similar to existing energy supplies.
Disadvantages of TDP:
1. The odor appears to be overwhelming.
2. A catalyst is required: sulfuric acid. This is something the inventors are not explicitly pointing out to interested parties.
3. That catalyst is going to require neutralization, i.e. a second chemical, most likely lime.
4. Finding a market for the byproducts (carbon, mineral and wastewater) is not as simple as the inventors would have you believe.
Yes E-P, TDP is profitable, if it is used to treat a feedstock that people would pay money to be rid of, such as sewage sludge. Using it to process animal feed (hello mad cow!) is not profitable. I believe the process cost, excluding feedstock, is about $60/bbl.
Gasification/Fischer-Tropsch is a well proven technology. Its most important drawback, I suspect is the need for a fairly dry feedstock. Drying biomass can be energy intensive. Nonetheless, the process should generate a lot of excess heat that could be used for drying the feedstock.
A coal based G/F-T process is profitable at crude oil prices of ~$40/bbl (those were the days, eh). Again, if you find a feedstock that would earn you a tipping fee, you'd probably be in business.
That leaves the question: What do we do when we have maxed out on waste products (probably suppling 10-20% of our oil needs)? I don't see energy farming as feasible, unless you are talking algae. One good source of algae would be the dead-spot in the Gulf of Mexico. If someone can work out a good way to harvest the algal biomass out of the dead-spot you could clean up the environment, produce fuel and recover fertilizer as a byproduct! Renewable fertilizer, anyone?
Posted by: An Engineer | 15 March 2006 at 01:16 PM
madscientist, thanks for catching the typo. corrected.
Posted by: Mike | 15 March 2006 at 02:46 PM
Posted by: Engineer-Poet | 15 March 2006 at 02:48 PM
My own view is that there are several sectors that will always need fuels very concentrated in terms of energy... airlines, semi trailers, heavy duty trucks and construction equipment, etc. Also, if plug-in-hybrids are utilized, there will have to be a fuel to power the vehicle after the initial 20-30 miles of battery range is used up. Finally, the article fails to mention the need for fuel oil for heating in winter. These liquid fuel needs could concievably be met by this process.
However, I doubt that we could generate enough hydrogen to feed into the process to replace ALL transportation needs... at least not without dramatically lowering car liquid fuel needs with plug-in-hybrid technology. I also doubt that we could replace all of our heating needs in this fashion (mainly due to scalability).
Even though I doubt that this process could replace petroleum by itself, I think that it could play a significant role in a multipronged approach (that includes adoption of plug-in-hybrid tech, biofuels, wind, solar, nuc, etc.
Even if it just mitigates some of the use of biofuels, this may be helpful, as the growing of biofuels (especially on a massive scale) is not necessarially enviro-friendly. Just look at the rainforests cut down to make room for sugarcane-ethanol production in Brazil.
Posted by: BBM | 15 March 2006 at 04:58 PM
E.P. - no, I'm not assuming that at all. I'm assuming that all of "waste" heat from the electrolysis would be generated inside an insulated box, and would raise the temperature there until the contribution of thermal energy to the splitting of water matched its rate of production from internal resistance within the cells. Electrolysis is endothermic, and absorbs heat from its ambient environment. The higher the temperature, the more endothermic the reaction becomes.
A.E. - thanks for the compliment. I, too, favor waste biomass, but I also feel there's great potential for solar PV.
What I forsee are sparse arrays of tracking panels installed in space frames above agricultural fields and pastures. At noon, the panels would shade only about 25% of the land below. For most plants, available sunlight is not the limiting factor for growth rate, so the presence of the panels intercepting 25% of their noontime sunlight would have little effect.
The space frame--a rigid "tensigrity" structure of cables and rods--would be supported on poles high enough above the fields to be out of the way of farm equipment. The shadows of individual panels would move over the course of the day, allowing even sunlight, on average, to all parts of the field. The panels would use freznel lenses to concentrate sunlight by a factor of 400 onto small multi-junction solar cells. Conversion efficiency would therfore be about 30%. Each acre of covered field would produce, on average, about 2 MWh per day.
With plenty of electrolytic hydrogen and oxygen available to drive the reactions, none of the waste biomass would have to be burned in order to provide heat for gasification of the rest. The yield of synthesis gas per ton of biomass would be doubled, and the fuel synthesis process would produce no CO2 whatsoever.
Now I just need to find a forward-looking billionaire to fund the concept. Any angels out there?
Posted by: Roger Arnold | 16 March 2006 at 01:51 AM
excuse me, but since when does uranium grow on trees? It's a fossil fuel, with very limited primary supplies.
Eh. Fossil fuels are called, you know, "fossil", since they consist of what once was the fossilized remains of ancient (jurassic era?) flora and fauna.
Or are you claiming that jurassic era trees, dinosaurs and algae were made of uranium, as opposed to hydrocarbons, like every other form of living matter known, including you and me? ;-)
Posted by: jb | 16 March 2006 at 09:14 AM
As to the question on how to power the post-oil transportation, Nobel Laureate George Olah has recently published a book "Beyond Oil and Gas: The Methanol Economy", which as you can guess from the title, advocates an energy economy based on methanol. The reasons being that methanol is a liquid fuel that can be efficiently produced from almost any source, and can be used in ICE's as well as in fuel cells that can use methanol directly.
There is an interview with the guy in MIT technology review.
Posted by: jb | 16 March 2006 at 09:36 AM