I have been asked about this several times, so I figured I might as well post a reply. Recent news reports have Vinod Khosla making a new investment:
First, the obligatory hype we have all come to expect:
This technology could produce 50 billion gallons of ethanol from the world’s steel mills alone, turning the liability of carbon emissions into valuable fuels worth over $50 billion per year at very low costs and adding substantial value to the steel industry. The technology will also be a key contributor to the cellulosic biofuels business as it can convert syngas produced through gasification into ethanol.
But let’s strip out the hype and get right down to the fundamental problem with such approaches.
Background
Natural gas forms carbon dioxide and water vapor when burned with a sufficient oxygen supply. The heat released is a bit less than 1,000 BTUs per standard cubic foot of natural gas. If the natural gas is burned in an oxygen-restricted atmosphere, syngas – CO and H2 – can be produced. This is a partial combustion, or oxidation (partial oxidation = POX). Syngas can be used to make numerous compounds, including many that can be used as fuels (like diesel via the Fischer-Tropsch reaction).
CO can be combusted to form CO2, which would have been the end product if the initial reactant hadn’t been oxygen-deficient. The amount of heat released from this is about 340 BTU/scf, or about a third of the heat release of burning natural gas. The bacterium in question converts CO to ethanol, which seems like a pretty nifty trick. And from a purely scientific point of view, it is pretty interesting.
How About the Efficiency?
But this is an incredibly inefficient means of energy production/waste disposal. CO as a gas can be combusted for energy, or it can undergo various chemical reactions to produce useful chemicals. But if you ferment it into ethanol, you have now essentially destroyed the vast majority of the heating value initially present. How? These fermentations take place in water, and bacteria are not able to tolerate very high ethanol concentrations. Therefore, you are going to have a very dilute ethanol product that will have to now undergo a very energy-intensive distillation step. How energy intensive? If the ethanol concentrations attained are typical of bacterial fermentations, then it will take as much energy to distill off the ethanol as is contained in the ethanol. (This is exactly why I have never liked the BRI process).
So, where does that leave us? We have essentially run around in a circle in which no net useful energy was extracted. Let’s say 340 BTUs of CO get fermented to 340 BTUs of ethanol, and then it takes 340 BTUs of natural gas to purify the ethanol. In effect, what we have is an input of 680 BTUs of CO plus natural gas to produce 340 BTUs of ethanol. How about instead of this kind of system, we use the CO in a gas-phase reaction – keeping the product away from water – and then use the natural gas we saved in the distillation step to run a compressed natural gas vehicle?
From an economic perspective, this scheme is a non-starter. It is an interesting bit of scientific research, and certainly worth funding, but you aren’t going to see any ethanol fermentation plants commercializing gaseous waste streams. After all, ethanol can be produced in a gas-phase reaction without the need for an inefficient fermentation. But then that would not be “bio-ethanol”, and therefore wouldn’t qualify for subsidies. These are the sorts of schemes that our subsidy system encourages. I am just waiting for someone to figure out how to ferment 1 BTU of gasoline into 0.9 BTUs of ethanol – which would then be eligible for a “bio-ethanol” subsidy. Then again, that’s not too terribly far off from what we are doing now with grain ethanol.
Technical Details (My Calcs)
From Perry’s Chemical Engineers’ Handbook, (McGraw Hill, New York, 1997).
Enthalpies of Combustion
Natural gas = 802.6 KJ/mole
CO = 283.0 KJ/mole
1 KJ = 0.948 BTU
Natural gas = 760.72 BTU/mole
1 mole natural gas at STP = 22.4 liters
1 cubic ft = 28.3 liters
So, 1 mole = 0.79 cubic feet
Then natural gas = 760.72 BTUs/0.79 cubic feet = 961 BTUs/scf
And CO = 340 BTUs/scf
Natural gas/CO = 2.8; therefore the full natural gas combustion delivers 2.8 times the energy of the CO combustion.
Making Gasoline from Carbon Dioxide: A solar-powered reaction turns a greenhouse gas into a valuable raw material, By Kevin Bullis Wednesday, on April 25, 2007 in Technology Review:
Researchers at the University of California, San Diego (UCSD), recently demonstrated that light absorbed and converted into electricity by a silicon electrode can help drive a reaction that converts carbon dioxide into carbon monoxide and oxygen. …
The UCSD system shows that it is possible to use solar energy to make carbon monoxide that then, together with hydrogen, can be converted into gasoline. Currently, carbon monoxide is made from natural gas and coal. But carbon dioxide is a more attractive raw material in part because it’s very cheap–indeed, it’s something industrial companies will pay to get off their hands, Jessop says. “There are very few chemicals which are cheaper than free, and carbon dioxide is one of them,” he says.
In the prototype device, sunlight passes through carbon dioxide dissolved in a solution before being absorbed by a semiconductor cathode, which converts photons into electrons. Aided by a catalyst, the electrons react with carbon dioxide to form carbon monoxide at the electrode. At the anode–a catalyst made of platinum–water is converted into oxygen.
To make a fuel, the carbon monoxide can be combined with hydrogen to create syngas in a well-known technology called the Fischer-Tropsch process, which has been widely used to make gasoline from coal. With the new process for creating syngas, however, fossil fuels could be unnecessary.
The system … is still a work in progress. In the first prototype, only about half of the energy needed for the reactions was supplied by the sun, with the rest coming from outside electricity. That’s because the researchers decided to prove the concept using silicon as the semiconductor. They are now working with a gallium-phosphide semiconductor, which has exactly the right electronic properties to drive the necessary reactions using sunlight alone.
At this early stage–Kubiak says that commercial systems could be 10 years away–the efficiency and economics of making fuels this way aren’t known. …
Researchers at the University of California, San Diego (UCSD), recently demonstrated that light absorbed and converted into electricity by a silicon electrode can help drive a reaction that converts carbon dioxide into carbon monoxide and oxygen.
A system like that should be quite workable. You end up with a gas as opposed to very dilute ethanol in water.
In fact, plants take sunlight and use that to convert carbon dioxide into long-chain carbon molecules and oxygen. In theory, you could do a lot of smart things with solar, including using it to make fuels. All of our biomass is captured solar energy.
But any system that takes a gas with decent BTU value and turns it into a liquid product at less than 4% concentration (and 96% water) is not going to be workable. It just takes too much energy to purify such a dilute stream.
Cheers, Rober
The process as specified does not list distillation as an energy input, and a blast furnace or other industrial source of CO probably has more than enough waste heat to perform the distillation.
The bigger concern is efficiency. The only reason to make ethanol is for vehicle fuel. Given the losses in fermentation plus the internal combustion engine, it is probably much more efficient to burn the CO in a gas turbine and use the electricity for various purposes, including electric vehicles.
The process as specified does not list distillation as an energy input, and a blast furnace or other industrial source of CO probably has more than enough waste heat to perform the distillation.
If they have heat of a high enough quality to run an ethanol distillation, they are almost certainly capturing and using that heat right now. Their engineers should be clever enough to have tapped those waste streams a long time ago.
Given the losses in fermentation plus the internal combustion engine, it is probably much more efficient to burn the CO in a gas turbine and use the electricity for various purposes, including electric vehicles.
Exactly my point. I can’t think of a much worse home for CO than being put through an aqueous fermentation. First, you have the gas/liquid mass transfer problem. But worse, you have the problem that bacteria die as soon as the ethanol concentrations builds to even mild concentrations.
Cheers, Robert
Veering off topic, have you looked at XL Dairy’s integrated dairy/ethanol/biodiesel facility?
http://www.xldairygroup.com/
I’ve read of integrated pilot plants, but this is full commercial scale. They claim EROEI=10 for corn ethanol (shipped from midwest to Arizona, no less).
Nice article, Robert!
It is amazing to see some of the hare-brained schemes people come up with and (worse) get other people excited about. Sometimes the lack of scientific training in this country is just too depressingly obvious.
I still don’t get why the idea that ethanol is a silver bullet is so prevalent. What is going on, is it the DNA of many generations of moonshine?
Keep up the good work, Robert! Keep explaining that ethanol is a bad fuel, but great for internal consumption: As you like to say: Cheers!
I wish somebody would be able to make a high level “renewable fuel breakthrough” bullshit detector.
It would include something like this:
1. High level analysis of how chemical energy of starting materials
2. Amount of external energy (electricity, heat, etc) required to transform the starting materials into a usable fuel
3. Useful “waste” energy captured in the transformation process (i.e. heat)
4. Useful chemical energy stored in the transformed fuel
5. Utilisation efficiency of best fuel use cycle for the transformed “new” fuel (if it is indeed new and is not a direct replacement for say, NatGas, petroleum, etc)
My laymans take, without going into chemical energy calculations is:
– If you have a relatively non-reactive material (usually a byproduct of a spontaneous chemical reaction or burning), it will always require more energy to turn that material into a usable fuel, than what you get out of the fuel itself.
This seems fairly intuitive to me (esp. in light of laws of thermodynamics), but I could of course be overlooking something.
Also, one has to of course factor in energy quality factors.
I.e. sometimes if you have a high energy feed from say solar energy flow and can transform that to say, liquid carbohydron fuel, it may make sense up to a point.
But all of these “let’s start with CO2, water or something fairly inert and turn it into a fuel by using an extra source of energy” schemes seem to me like modern alchemy.
Will they become useful in the next 10-20 years?
I’m not saying breaking hydrogen from water using highly efficient blue leds might not make some sense one day.
However it’s still a highly inefficient conversion and a requires a source of pure feed (water) and a constant energy flow (sun).
Or to put it in other words, how big are the fuel creation cycles we need to create (in addition to the fairly slow biomass growth/photosynthesis) in order to replace the non-renewables? Including conversion to the types of fuels that we can actually use now (or in the next 20 years), without overhauling the whole infrastructure.
I’ve read of integrated pilot plants, but this is full commercial scale. They claim EROEI=10 for corn ethanol (shipped from midwest to Arizona, no less).
That’s what they project they will get if everything works out according to their most optimistic projections. It won’t. Until people actually build the plant and start seeing that they have been a little too idealistic about one thing or another, you will see them making these sorts of incredible projections.
E3 Biofuels did the same thing. They made some pretty high EROI projections. I have had indications that those have not been realized. Running a real plant is a lot harder than running one on a paper.
Cheers, Robert
I wish somebody would be able to make a high level “renewable fuel breakthrough” bullshit detector.
My formula:
Anytime it says “commercial systems could be X years away”… usually pins the detector needle.
Everything is hiding a fossil fuel input. Probably the simplest and most feasible completely renewable liquid fuel system would be some sort of hydroelectric power to fuel idea which would be totally unfeasible with today’s economy. Anything that makes a claim that it is more feasible, is doing it by hiding fossil fuels, burning off top soil and the wonder of hand harvested sugar cane which is prettymuch slavery-for-fuel.
You can also class the perpetrator:
1. Academics pulling PR stunts for research grants so they can noodle around the lab for another decade and then start another round of “in 10 years..”
2. Snake-oil based on making a quick buck on subsidies by repackaging cheap fossil fuels.
3. Naive billionaires.
Woops… By my scale renewable fuels are all bullshit. I guess that simplifies it.
“If they have heat of a high enough quality to run an ethanol distillation,they are almost certainly capturing and using that heat right now. Their engineers should be clever enough to have tapped those waste streams a long time ago.“
Using it for what? A steel mill doesn’t need much in the way of low-pressure steam, does it? I could see using a HRSG to generate some power (something Primary Energy might cook up) but if the value of the ethanol is greater than the value of the heat then it would make sense to use the nearly-spent steam to drive a distillation column instead of taking it all the way down to condenser temperatures.
“you have the problem that bacteria die as soon as the ethanol concentrations builds to even mild concentrations.“
Maybe some clever person can find a way to splice the CO-metabolism pathway into yeast.
(BTW: I’m completely locked out of my old Blogger account. All new material will be on WordPress or TOD.)
Blog: http://ergosphere.blogspot.com
EP
Using it for what? A steel mill doesn’t need much in the way of low-pressure steam, does it?
If it’s low-pressure steam, you aren’t going to be able to drive an ethanol distillation anyway. Realize that we are talking about a very dilute ethanol concentration, so the lower half of the column is going to be essentially boiling up water. You aren’t going to do that with low-pressure steam once you take the heat transfer aspects of the exchanger into account. They might get away with 60# steam and a very large heat exchanger, but I bet I could look around the plant and find a much better use of 60# steam. Bottom line is that it is an incredibly inefficient usage of BTUs, and certainly not something one would scale up to make large quantities of ethanol.
Maybe some clever person can find a way to splice the CO-metabolism pathway into yeast.
It never ceases to amaze me how alike we think. That was exactly what I was thinking when I said that the research should be funded.
Cheers, RR
A steel mill doesn’t need much in the way of low-pressure steam, does it?
Incidentally, here is a much better usage of low-pressure steam. If you have low-pressure steam and a CO waste-stream, burn that CO and turn the low-pressure steam into high pressure steam, and then generate electricity with it. That is what an engineer who is not distracted by ethanol subsidies would do if they were interested in efficient BTU capture.
Cheers, Robert
By my scale renewable fuels are all bullshit. I guess that simplifies it.
Come, come, no need to get cynical.
There are some real low hanging fruit as far as biofuels are concerned. Unfortunately the powerful agricultural lobby has hijacked the idea of biofuels and use it add more subsidies, below the radar, if possible. They have been spectacularly successful, to the degree that most people believe biofuel means growing a dedicated crop for biofuel production. It is this line of thinking that poses the greatest threat to biofuels and has poisoned many (including Rohar) to the idea.
The reality is that the US produces 250 million tons a year of waste, of which 65% is renewable and organic, in other words ideal feedstock for biofuel production. Add in the plastic, and it is more than 75% that can be used to produce fuels.
If you want to simplify things, just converting the waste paper into fuel leave you with a feedstock of 84 million tons a year.
And this is free feedstock. In fact, in some locations, people will pay you to take it off their hands.
Come, come, no need to get cynical.
My position needs clarification. I am only refering to large scale “bio”fuel production as an oil replacement based on subsidies not the best use for the energy source.
Renewable fuels as recycling is a good idea if there isn’t a better and simpler use for the product like structure heating or cogeneration. Also, if it actually has positive EROEI, doesn’t stink up the neighborhood, cause a lot of pollution and doesn’t cause soil degradation or a dozen other real-world factors.
Biofuels from food without fixing the agricultural energy inputs first? Biofuels from industrial CO? Biofuels from cutting down rain forest? Biofuels from hand harvested sugar cane?
Nope.
As oil replacements from coal, NG or industrial gasses, non-bio products make more sense, but there aren’t subsidies for these and that leads to these schemes.
As far as biomass “recycling” there is a lot of bad information on soil stewardship and if anything we should be adding organic matter to farmland.
Anyone who thinks it is worth hauling straw to a processing plant for cellulosic ethanol should come out to the farm. I’ll light a round bale of wheat straw on fire (~1/2 tonne) and you can light 40L of gasoline. The discussion of the real energy value in low-grade biomass will be solved right there.
Conservation is good. Renewable electricity in as many places as possible is good. Moving shipping from trucks to electric rail is good. Diesel in passenger vehicles is good. True recycling is good.
The schemes to jump on biofuel subsidies with bad ideas is what I have a problem with.
“If it’s low-pressure steam, you aren’t going to be able to drive an ethanol distillation anyway. Realize that we are talking about a very dilute ethanol concentration, so the lower half of the column is going to be essentially boiling up water. You aren’t going to do that with low-pressure steam once you take the heat transfer aspects of the exchanger into account. They might get away with 60# steam and a very large heat exchanger….“
I’m afraid I can’t take your word for that.
1. 60 PSIA steam condenses at almost 300 F (147°C @ 400 kPa). You’re talking about a temperature drop of almost 50°C between the condensing steam and the still’s stripper column. I’d believe 10°C (a reasonable figure for a condenser), but not 50.
2. A vacuum still would run even cooler.
3. Others have investigated distilling fuel ethanol with heat from a salt-stratified solar pond collector. These cannot reach such high temperatures, but all they require is getting close to the boiling point of water at the still’s pressure. That’s set by the vapor pressure of ethanol at the condenser temperature.
“I could look around the plant and find a much better use of 60# steam.“
Heck, so could I. Run it through a couple more turbine stages before sending it to the still!
“Bottom line is that it is an incredibly inefficient usage of BTUs“
Blast furnaces yield plenty.
“and certainly not something one would scale up to make large quantities of ethanol.“
Now that’s true. The supply is limited by the waste CO from blast furnaces and similar industrial processes. But if all you can manage is low-tech stuff, converting the CO and waste heat to ethanol might be the best-paying option. It beats throwing it all away!
“If you have low-pressure steam and a CO waste-stream, burn that CO and turn the low-pressure steam into high pressure steam, and then generate electricity with it.“
My CRC handbook says blast-furnace gas is typically around 90 BTU/scf (60% N2, 11.5% CO2, 27.5% CO, 1% H2). This is not going to reach a very high combustion temperature because of all the inert constituents. If I was going for maximum efficiency I would try to burn it in a gas turbine and generate steam off a bottoming cycle, but the low (and variable) energy content of the gas would probably give migraines to the combustor and control-system engineers. Compressing such a great volume of fuel (about 40% of the air volume) would require a new design, and unburned CO in the exhaust would be a particular headache.
CO-metabolizing bugs work cheap, and might do a better job of CO cleanup than anything but the best control system. Using the heat of the gases to run the still (and perhaps generate power to run the required fans) would turn the whole thing into a money-making affair.
You’re talking about a temperature drop of almost 50°C between the condensing steam and the still’s stripper column. I’d believe 10°C (a reasonable figure for a condenser), but not 50.
But there is a lot more to heat exchanger design than delta T. You have to take mass and surface area into account. The governing equation is Q = U * A * delta T = mass flow * Cp * delta T. The problem here is that your mass flow * Cp on the inside of the distillation tower is going to be huge. Therefore, you are either going to need a huge heat exchanger with an incredibly high steam flow (large capital) or a very high delta T. Remember, inside that column you have mostly water (and a high ratio of water to produced ethanol), which has a very high heat capacity.
My CRC handbook says blast-furnace gas is typically around 90 BTU/scf (60% N2, 11.5% CO2, 27.5% CO, 1% H2). This is not going to reach a very high combustion temperature because of all the inert constituents.
You might be surprised by that. I have done these sorts of dilution experiments before. The heat capacity of those gases is low, so combustion, especially that involving direct heat transfer, will heat them up nicely. Besides that, this same situation works against your bacterial fermentations. If you are dealing mostly with nitrogen, that is mostly what the bacteria will encounter (which adds to the gas/liquid mass transfer problem).
CO-metabolizing bugs work cheap, and might do a better job of CO cleanup than anything but the best control system.
If all you were going to do is clean up the CO stream and dispose of it, then you might have a winner. It is the subsequent distillation that’s going to be the money pit. You will produce ethanol in a concentration that is treated as a waste stream (for good reason).
Cheers, Robert
I think this is along the lines of cellulosic ethanol where the feedstock and the industry of the feedstock is not understood by the promoters.
The initial smelting of iron is Fe2O3 + CO
Even with my NoKwowledgeOfTheSteelIndustry and EvenLessOfChemistry an alarm goes off when someone says that an industry who’s first step of smelting is rust + carbon monoxide -> iron + carbon dioxide has vaste quantities of CO they don’t know what to do with. It seems much the same as Iogen thinking there is waste wheat and barley straw or corn stover.
Besides smelting, it looks like they need to boil water at a steel mill and they can do it with CO. Are they supposed to divert the CO from the boiler to make ethanol?
This is about the same as thinking there is all of this forestry biomass that isn’t being used when there is 16 trillion BTU’s being used for heat/electrical generation as of 2003.
The NG price has made it feasible for even a small millshop to replace NG heat with a biomass boiler and pay for it in a short period.
USDA is now saying they were over-estimating how much corn stover can be pulled from farmland. They are still completely off. The answer is none. In the article “In other words, when factoring in soil quality…”. D’oh! you mean you just don’t want your topsoil to blow away, you want to be able to actually grow something on it? Oh.. that changes the numbers.
You can grow an equivalent or better crop with no urea fertilizer by maintaining organic levels and summerfallowing versus continuous cropping and 100lbs/acre of nitrogen. This would mean pulling 1/2 the farmland out of production, finding something cheaper than diesel to cultivate the land and composting all the biomass possible. That is sustainable. If you think it’s a bad idea, how do you think the we farmed and ate for 10,000 years up until 1975 or so?
As a counter-point to fallow, a legume in the crop rotation requires no additional nitrogen and increases soil nitrogen levels, but it doesn’t help with weed or disease control or add substantial organic matter.
My agricultural experience is north of the corn belt, but I would think that in a corn/soybean rotation, the best place for corn stover is dumping it on the soybean stubble one field over.
From the Corn stover collection project paper:
“No-till fields have the least requirement for residue…. Too much residue also causes problems, reducing the yield of the next crop (Smith, 1986).“
Go read the paper.
Go read the paper.
That paper is prime feedstock for biomass.
No-till has been promoted by Cargill, Monsanto, etc. It guarantees high fertilizer and chemical inputs and the price of diesel makes tillage expensive. The previous farming methods dropped the organic content down to where wind erosion made fallow impossible.
Besides farming, I owned a high clearance spraying business and my brother does it now. The high clearance sprayer was designed for pre-harvest Roundup as part of low and zero till perennial weed control, but the sprayer is used through pre-seeding burn-off in low-till right through the crop cycle.
Zero and low-till is great for the fertilizer business and at $4.50/acre and 150 acres/hour it’s pretty good for the custom sprayer. In 1975 my dad would fallow 1/4-1/3 and the reason it’s not done now is diesel is 4 times the price. The only way to survive is to seed every acre because the cost of cultivating summerfallow plus getting no crop off of it for the year is too high compared to 100lbs of urea (or better) even at $550/tonne.
This is in a location with high organic matter and soil erosion isn’t a problem with fallow. In locations without that level of organic matter, the point is moot because wind erosion makes fallow impossible and rotating a legume is the only real option.
Our farmland is leased out since 1992. The operators have practiced low-till since then. It’s basically a disaster and after 15 years the soil is like roadbed. You need to keep organic matter high and cultivate or you quickly go down a spiral of perennial weeds (thistle,quackgrass) with massive root structures that you need to kill with Roundup.
The high clearance sprayer that our company owns retails at $325,000, plus another large investment in a tractor/trailer to haul it around and pumps/tanks, etc. There is good money in the business for the investment and zero and low till methods plus farmers cropping way more than they can manage on their own created the business.
It’s not sustainable without cheap fossil fuels.
You’ve thrown around a lot of terms of art (like “fallow” used as a verb) but you haven’t explained why you come to the conclusion you do. Even the demand for additional diesel for cultivation argues against what you say is your preferred option, so again: what are all the factors, and what weight does each of them have?
The experience you claim runs counter to the references I’ve read which address the issue of tillage and soil carbon. Tillage exposes soil to erosion and carbon to oxidation; tilled soils have less carbon. If this is only true for certain patterns of (mis)management, I want to know more of the specifics and some references so I can be sure I have this right in the future.
This thread is probably too old to get an answer, but I was recently at the Materials Research Society’s spring conference and one of the talks was by Chris Sommerville and he mentioned production of syngas and taking it on to fuels as having a very high initial investment cost on the order of 100,000$ per barrel per year. Which seems ludicrous to me, but I haven’t been able to find numbers on it anywhere. Does anybody know current numbers?
It would have been $100,000 per daily barrel. That’s probably about right. The latest published numbers I saw from the DOE, which are a couple of years old, were around $20,000 per daily barrel for an oil refinery or ethanol plant, about $40,000 for GTL, $60,000 for CTL, and I believe around $100,000 for BTL.
Cheers, Robert
Great, thanks for the response. I actually just found an article in this weeks Science Magazine that talks (albeit generally) about it, citing 100,000$ per barrel per day.
The link is:
http://www.sciencemag.org/cgi/reprint/320/5874/306.pdf
though I don’t know what access is like (I am at a university so get the articles for free) but I can send you a copy if you are interested. Thanks again