Yes. I have read through and commented on the drafts as he was working on the essay. We also spent another hour on the phone this week discussing some of these arguments. I obviously still have areas of disagreement, but I will let others make their specific criticisms known.
Whether readers agree or disagree with Mr. Khosla's ideas, I think we can all agree that debating the issues is the best way to get past hype and myth and into what can be scientifically supported.
I've kind of passed on the scientific debate over cellulosic ethanol, because it seems to presume future improvement. In that sense it is a moving target. Name any problem, and future innovation can fix it.
Well, I think so yes. There are cellulosic plants in production today, but they don't operate at commercial scales. They're tests. The tests have worked and now several companies are building large scale plants that should hopefully be profitable, touch wood.
You can guarantee that some variation of this will become known as the new, worse version of Monsanto's "Frankenfoods" or "terminator seeds."
Just as the petrochemical driven Green Revolution morphed into the reviled slogans above, so too with FTF (food to fuels). Right now biofuels are the darling of the sustainable development crowd, but that will change in short order when masses of economically marginal "consumers" begin to starve and die.
As you point out, "Trajectory" is what counts. And the Trajectory is to gradually convert food staples and cropland to motor fuels.
Imagine the bumpersticker
".....turns OUR food into gas...."
The first part might contain the name of a company, country, or even individual (if they become too closely identified with FTF). The last part might contain a motive or result, "so SUVs can drive, while people starve." etc.
If you dive into this, you and your companies will risk becoming the new Monsantos.
So it's mostly scaling up and increasing efficiency that is needed, rather than new breakthroughs in enzymes and stuff? If so, that is good news.
One thing I wonder about is his belief that cellulosic ethanol will outcompete biodiesel in gallons per acre. Currently, that is so not the case. Putting aside the ecological problems with palm oil, it's 5000 kg per hectare, far better than any currently producing ethanol crop. And what about biodiesel from algae? That's the fair point of comparison with cellulosic, which isn't quite there yet either.
Unless the problems of biodiesel from algae are more than efficiency and scaling...
But I think his rejection of EROEI is weird. What ultimately matters is how much energy we can use in total.
The recent post on nuclear power assumed a number of future improvements, in areas like commercialization of breeder reactors, greater uranium mining and processing efficiency, secure waste transport and disposal options, lowered plant decommissioning costs, and electrification of the industrial and transportation sectors. If it's okay to assume speculative improvements for nuclear power then why not for cellulosic ethanol?
But I am playing devil's advocate. Speculating on future trends is what entrepreneurs are in business to do. Governments shouldn't bet on one horse when there's a whole field of promising contenders and new breakthroughs are announced every month. Better they should all compete without subsidies or preferential treatment.
I certainly didn't say it was OK to assume advances. On the other hand, I think research, and piloting these things, is important. That's the only way you find out if you have a chicken.
What I have read is that if all the crop acreage in America was devoted to corn for ethanol, then it still wouldn't be enough to fuel America's truck fleet.
Can you do better than "What I have read"? Otherwise it is meaningless.
On the one hand, it is fairly clear that we can not replace every drop of oil we use from corn. On the other hand it seems equally obvious that ethanol, particularly if you are willing to consider cellulosic (I tend to agree with Odogragh here however), could play a very important role in offsetting declining oil supply.
I read somewhere that if you wanted to replace all global electricity consumption with solar technology, you would ahve to cover the entire earth with panels - so solar is worthless, right?
OK, it's one source and who knows how close it is, but let's see what happens...
Assume 4 sun hours per day (4 hrs at 1000 watts/sq.m), a 12% efficient panel will produce 1000x0.12x4x365=175,200 Whr per year. Take 4000 TW (4E15 Watts) divided by 175,200 = 22.8E9 square meters, = 22831 sq. km = 8800 sq. mi. for a square area that equates to 93 miles x 93 miles. To cover the electrical use for the entire country.
So unless the 4000TW number is off by orders of magnitude, we are nowhere near having to cover the US with solar panels to provide our current consumption (which BTW could be easily reduced substantially, but that's another topic). Does the rest of the world consume orders of magnitude more electricity per surface area than the US? I'd bet dinner not, and I only bet dinner when I know I'm going to win.
Take 4000 TW (4E15 Watts) divided by 175,200 = 22.8E9 square meters, = 22831 sq. km = 8800 sq. mi. for a square area that equates to 93 miles x 93 miles. To cover the electrical use for the entire country.
I wonder how much demand you could satisfy by covering a substantial fraction of all the roofs with solar panels? I realize currently that is cost-prohibitive, but I hope we see this happen someday.
Bingo! Every time I hear someone talk about how you need to cover a sizable portion of Nevada with solar panels, it drives me nuts because that's not where cheap, thin-film panels will be installed--they'll go on rooftops of homes, businesses, municipal buildings, schools, etc.
The NanoSolar plant being built in/near silicon valley will produce 430MW of panels per year, supposedly at much cheaper prices than anything currently available. That's enough panels for one heck of a lot of rooftops, and from just one plant. It's a safe bet that if they come close to their price points and production volumes after they ramp-up that we'll see them getting tons of additional funding and building more plants.
I think it would be downright funny if all this talk about ethanol, hydrogen, CTL, etc. was largely derailed by one of the "old school" renewables, solar, that finally had the major price breakthrough we've all been hoping for since the 1970's.
Sorry Lou. Just trying to make a point. These blanket statements - about ethanol, solar, etc. - are inaccurate and don't help us to look at and compare solutions.
Why would you put all the solar panels in one 93 square mile area? For one it would become a huge terrorist target. Second it would be vulnurable to natural disasters (tornados, earthquakes, etc..)
I think a distributed solution of using business and household rooftops makes way more sense. Supplement that with wind and tidal power, and it could go a long way towards getting households and small business into a more sustainable model.
I wasn't saying that all of the panels should go in one area, I was just computing the total area required, since it was asserted above that we would have to cover the planet to meet our current consumption. One of the beautiful features of grid-tied solar systems is that the power source is distributed, not centralized, which is much more robust and secure. Also, you are generating the power close to where it is consumed, so transmission losses are minimized.
Rooftops are the perfect place for panels, and there's way more than enough existing roof space to cover that 8800 square miles. The only down side today is the high up-front cost, and the low supply of panels due to heavy solarization in Europe and Germany in particular.
National Geograpic, 8/05, said we need 10K square miles of solar panels, bigger than the area of Vermont, or roughly the size of Mendocino County, Ca, to satisfy all our electrical needs. Says all those panels could fit on less than quarter of the roof and pavement space in cities and suburbs.
Also, in California, at least, they figure 5 hours of sun/day; some are now starting to use 6 hours. Right now, I'm expected to be putting out 5.6KW/day; I'm actually at 9 +/- 0.3. Remains to be see what happens during 35 or 40 staight days of rain.
Your figure for electricity use is pretty similar to the one found here https://www.cia.gov/cia/publications/factbook/geos/us.html
I think at some point it will become apparant that we can live pretty comfortably with alot less electricity than that though, so we really won't need even as much generation as you calculated.
There is alot of energy in wind and solar, you just can't put them in your gas tank. Oh well.
So, 18.01% of 3,718,711 is 669,740 sq mi arable land, times 640 acres per square mile is 428633504.7 acres of arable land, times an average yield of 148.4 bushels times 3 gallons ethanol per bushel is about 190 billion gallons of ethanol. The U.S. uses about 150 billion gallons of gasoline per year. I don't have the exact energy conversion between ethanol and gasoline, but the amounts of energy are in the same ballpark. So we could theoretically not only fuel our trucking fleet but all of our automobiles, and an be energy independent without sacrifice if we just grow corn of every arable acre in the country. Fortunately, Canada is obligated by NAFTA to supply us with natural gas, so we can still make fertilizer to keep the corn yields up, and we can still eat if we can find an optimal deer population that will feed us but without eating so much corn to have an significant impact on yields.
This is looking easy. On an unrelated topic, I think I can run a marathon in about an hour and twenty minutes based on my time in the 100m when I was in high school.
So we could theoretically not only fuel our trucking fleet...
Ah, but the trucking fleet runs mainly on diesel, which is not included in the gasoline number. Diesel has an even higher energy content than gasoline, and diesel engines are far more efficient. So, to replace the 50 billion or so gallons of diesel we use each year would take around 100 billion more gallons of ethanol.
Regarding the energy content of gasoline versus ethanol, ethanol has about 65% of the BTUs of gasoline.
I need to be careful with sarcasm, I know it doesn't always work in writing as well as in speech, or with strangers. I hope you got that my commentary was tongue in cheek.
Actually this is useful. Firstly, I am happy to acknowledge that ethanol can not possibly, even in the wildest dreams of ADM, replace 100% of U.S. fuel consumption.
However, using your numbers, it does seem that it could make a difference. I have argued before that 10% of consumption on a BTU basis would make ethanol a huge contributer to replacing declining oil reserves. Even 10% of gasoline use would mean a lot, both in terms of its contribution as a fuel and as a replacement for the lightest refinery products - freeing up much needed capacity.
So, 10% of 150 billion gallons times 1.3 (to convert from volume to BTU) equals about 20 billion gallons. This is about 10.5% of the 190 billion gallons figure you cite for all US agricultural land.
Now, I also acknowledge that taking a mid-range EROEI figure of of 1.25, the net contribution to energy is only about 2% of gasoline use with the remainder being converted from other fuel sources.
My point is not that it is a good thing to do. It may not be. However, this analysis reveals a lot more about the actual tradeoffs than dismissing it because it can't solve all of our energy problems single handedly.
For the record, my sarcasm wasn't directed at you. I was taking to opportunity to do the overly optimistic "anything is possible" tone of cornucopianism (if that's a real word).
I would like to note that Mr. Khosla must agree with your assessment of there not being enough land for corn ethanol, which only provides 150 bushels per acre, as he states the following in his essay.
"Ethanol, butanol and any liquid fuel that has a shot at replacing our gasoline needs has to scale up to 2,000-3,000 gallons per acre."
Using your own calculations if cellulose ethanol is scalable to 2,000 bushels per acre as Mr. Khosla seems to believe, then it would require less then 10% of the U.S. arable land to completely replace our gasoline consumption with cellulose based ethanol. 10% seems like a much more reasonable number and something that may be achievable.
However this changes the main question to, "What is the evidence that Cellulose Ethanol can be scaled to 2,000 bushels per acre, given current methods." I think this question deserves much consideration, because it is at the heart of the debate.
If Mr. Khosla is still reading, I would recommend that he consider hiring Jack and Sailorman as consultants to his project. They would help greatly with efforts to win over the skeptical. And their style and methods are superb. There are others, but they are the only ones whose work on ethanol I am familiar with.
This is my first post. I want to make a small correction to Jack's post.
No! You do not need to cover the entire earth with panels to replace the entire electricity demand with solar. For rough calculations... The earth is receiving an average of 1,000 watts/m2 with half the earth's surface lit at any given time, 5x10^14 m2. This yields 2.5x10^17 average watts at any moment. With 8760 hours available per year, this yields 2.2x10^21 watt/hr or 2.2x10^12 GW/hrs!! Solar cells vary in their efficiency from 6% to 30%. Most solar cells have around 14% effeciency. Thus using 14% effeciency that gives 3.1x10^11 GW/hrs per year.
In 2003, the world used 14,781 billion killowatt/hrs or 1.4x10^7 GW/hrs for easy comparison above. That is 4 orders of magnitude lower.
Maybe you meant only on available land mass? The oceans make up 71% of the earth's surface. Thus that reduces available energy to 9x10^10 GW/hrs per year. Obviously, it is still orders of magnitude more than needed.
So what percentage of land mass would be needed? .02%!!
Welcome to TOD and thanks for the post. I was trying to be sarcastic or rhetorical, which can be hard in writing. Your post and others have added nicely to this discussion.
my understanding is that it would only take a PV panel array the size of, say, arizona to power the usa. extrapolating and recognizing that the usa is the big gran electricity consumer du monde, it would probably only take an array the size of all of the world's deserts to power the world. and i understand that there is pretty good sunlight in the desert.... perhaps not enough facts, but that (backing up WAGs with fact) is not my role in this forum.
-PoP
Check out the report on Xethanol, one of the companies supposedly producing ethanol from biomass. Seems they are a total sham, with no scientists on staff, no R & D, and one plant supposed to be poised to make ethanol has no water or sewer hookup. http://sharesleuth.com/2006/08/moonshine_blindness.html
This doesn't undermine the potential of another company to do it right, but there's a lot of hype about innovation that isn't being matched with results (or even effort).
Yes. I have read through and commented on the drafts as he was working on the essay. We also spent another hour on the phone this week discussing some of these arguments. I obviously still have areas of disagreement, but I will let others make their specific criticisms known.
Whether readers agree or disagree with Mr. Khosla's ideas, I think we can all agree that debating the issues is the best way to get past hype and myth and into what can be scientifically supported.
Am I being unfair?
So this is not exactly Star Trek technology.
I believe in counting chickens when they are hatched. In this case, I think a commercial plant, esp. with open books, would be "hatched."
You can guarantee that some variation of this will become known as the new, worse version of Monsanto's "Frankenfoods" or "terminator seeds."
Just as the petrochemical driven Green Revolution morphed into the reviled slogans above, so too with FTF (food to fuels). Right now biofuels are the darling of the sustainable development crowd, but that will change in short order when masses of economically marginal "consumers" begin to starve and die.
As you point out, "Trajectory" is what counts. And the Trajectory is to gradually convert food staples and cropland to motor fuels.
Imagine the bumpersticker
The first part might contain the name of a company, country, or even individual (if they become too closely identified with FTF). The last part might contain a motive or result, "so SUVs can drive, while people starve." etc.
If you dive into this, you and your companies will risk becoming the new Monsantos.
One thing I wonder about is his belief that cellulosic ethanol will outcompete biodiesel in gallons per acre. Currently, that is so not the case. Putting aside the ecological problems with palm oil, it's 5000 kg per hectare, far better than any currently producing ethanol crop. And what about biodiesel from algae? That's the fair point of comparison with cellulosic, which isn't quite there yet either.
Unless the problems of biodiesel from algae are more than efficiency and scaling...
But I think his rejection of EROEI is weird. What ultimately matters is how much energy we can use in total.
But I am playing devil's advocate. Speculating on future trends is what entrepreneurs are in business to do. Governments shouldn't bet on one horse when there's a whole field of promising contenders and new breakthroughs are announced every month. Better they should all compete without subsidies or preferential treatment.
On the one hand, it is fairly clear that we can not replace every drop of oil we use from corn. On the other hand it seems equally obvious that ethanol, particularly if you are willing to consider cellulosic (I tend to agree with Odogragh here however), could play a very important role in offsetting declining oil supply.
I read somewhere that if you wanted to replace all global electricity consumption with solar technology, you would ahve to cover the entire earth with panels - so solar is worthless, right?
I see no facts criticised with no facts. Save the atta-boys for something worth reading not put-downs.
I believe your solar assertion, while I realize that you were using it to make a point, is erroneous.
http://www.quaker.org/tqe/2006/TQE145-EN-Hybrid-Details.html
OK, it's one source and who knows how close it is, but let's see what happens...
Assume 4 sun hours per day (4 hrs at 1000 watts/sq.m), a 12% efficient panel will produce 1000x0.12x4x365=175,200 Whr per year. Take 4000 TW (4E15 Watts) divided by 175,200 = 22.8E9 square meters, = 22831 sq. km = 8800 sq. mi. for a square area that equates to 93 miles x 93 miles. To cover the electrical use for the entire country.
So unless the 4000TW number is off by orders of magnitude, we are nowhere near having to cover the US with solar panels to provide our current consumption (which BTW could be easily reduced substantially, but that's another topic). Does the rest of the world consume orders of magnitude more electricity per surface area than the US? I'd bet dinner not, and I only bet dinner when I know I'm going to win.
I wonder how much demand you could satisfy by covering a substantial fraction of all the roofs with solar panels? I realize currently that is cost-prohibitive, but I hope we see this happen someday.
The NanoSolar plant being built in/near silicon valley will produce 430MW of panels per year, supposedly at much cheaper prices than anything currently available. That's enough panels for one heck of a lot of rooftops, and from just one plant. It's a safe bet that if they come close to their price points and production volumes after they ramp-up that we'll see them getting tons of additional funding and building more plants.
I think it would be downright funny if all this talk about ethanol, hydrogen, CTL, etc. was largely derailed by one of the "old school" renewables, solar, that finally had the major price breakthrough we've all been hoping for since the 1970's.
Why would you put all the solar panels in one 93 square mile area? For one it would become a huge terrorist target. Second it would be vulnurable to natural disasters (tornados, earthquakes, etc..)
I think a distributed solution of using business and household rooftops makes way more sense. Supplement that with wind and tidal power, and it could go a long way towards getting households and small business into a more sustainable model.
It doesn't solve the liquid fuel problem though.
Rooftops are the perfect place for panels, and there's way more than enough existing roof space to cover that 8800 square miles. The only down side today is the high up-front cost, and the low supply of panels due to heavy solarization in Europe and Germany in particular.
Also, in California, at least, they figure 5 hours of sun/day; some are now starting to use 6 hours. Right now, I'm expected to be putting out 5.6KW/day; I'm actually at 9 +/- 0.3. Remains to be see what happens during 35 or 40 staight days of rain.
Rat
I think at some point it will become apparant that we can live pretty comfortably with alot less electricity than that though, so we really won't need even as much generation as you calculated.
There is alot of energy in wind and solar, you just can't put them in your gas tank. Oh well.
You can if we had electric vehicles. Has everyone seen this?
http://www.teslamotors.com
I can. Here are the raw numbers I will start with:
Total land are in the U.S. = 3,718,711 sq mi. http://en.wikipedia.org/wiki/United_states
Percent arable land in the U.S. = 18.01% https://www.cia.gov/cia/publications/factbook/geos/us.html
Average corn yield per acre (2005) = 148.4 bushels http://www.usda.gov/nass/aggraphs/cornyld.htm
So, 18.01% of 3,718,711 is 669,740 sq mi arable land, times 640 acres per square mile is 428633504.7 acres of arable land, times an average yield of 148.4 bushels times 3 gallons ethanol per bushel is about 190 billion gallons of ethanol. The U.S. uses about 150 billion gallons of gasoline per year. I don't have the exact energy conversion between ethanol and gasoline, but the amounts of energy are in the same ballpark. So we could theoretically not only fuel our trucking fleet but all of our automobiles, and an be energy independent without sacrifice if we just grow corn of every arable acre in the country. Fortunately, Canada is obligated by NAFTA to supply us with natural gas, so we can still make fertilizer to keep the corn yields up, and we can still eat if we can find an optimal deer population that will feed us but without eating so much corn to have an significant impact on yields.
This is looking easy. On an unrelated topic, I think I can run a marathon in about an hour and twenty minutes based on my time in the 100m when I was in high school.
Ah, but the trucking fleet runs mainly on diesel, which is not included in the gasoline number. Diesel has an even higher energy content than gasoline, and diesel engines are far more efficient. So, to replace the 50 billion or so gallons of diesel we use each year would take around 100 billion more gallons of ethanol.
Regarding the energy content of gasoline versus ethanol, ethanol has about 65% of the BTUs of gasoline.
However, using your numbers, it does seem that it could make a difference. I have argued before that 10% of consumption on a BTU basis would make ethanol a huge contributer to replacing declining oil reserves. Even 10% of gasoline use would mean a lot, both in terms of its contribution as a fuel and as a replacement for the lightest refinery products - freeing up much needed capacity.
So, 10% of 150 billion gallons times 1.3 (to convert from volume to BTU) equals about 20 billion gallons. This is about 10.5% of the 190 billion gallons figure you cite for all US agricultural land.
Now, I also acknowledge that taking a mid-range EROEI figure of of 1.25, the net contribution to energy is only about 2% of gasoline use with the remainder being converted from other fuel sources.
My point is not that it is a good thing to do. It may not be. However, this analysis reveals a lot more about the actual tradeoffs than dismissing it because it can't solve all of our energy problems single handedly.
The average yields cited per acre are in places that have 8-10 feet of rich soil and sufficient water.
"Ethanol, butanol and any liquid fuel that has a shot at replacing our gasoline needs has to scale up to 2,000-3,000 gallons per acre."
Using your own calculations if cellulose ethanol is scalable to 2,000 bushels per acre as Mr. Khosla seems to believe, then it would require less then 10% of the U.S. arable land to completely replace our gasoline consumption with cellulose based ethanol. 10% seems like a much more reasonable number and something that may be achievable.
However this changes the main question to, "What is the evidence that Cellulose Ethanol can be scaled to 2,000 bushels per acre, given current methods." I think this question deserves much consideration, because it is at the heart of the debate.
No! You do not need to cover the entire earth with panels to replace the entire electricity demand with solar. For rough calculations... The earth is receiving an average of 1,000 watts/m2 with half the earth's surface lit at any given time, 5x10^14 m2. This yields 2.5x10^17 average watts at any moment. With 8760 hours available per year, this yields 2.2x10^21 watt/hr or 2.2x10^12 GW/hrs!! Solar cells vary in their efficiency from 6% to 30%. Most solar cells have around 14% effeciency. Thus using 14% effeciency that gives 3.1x10^11 GW/hrs per year.
In 2003, the world used 14,781 billion killowatt/hrs or 1.4x10^7 GW/hrs for easy comparison above. That is 4 orders of magnitude lower.
Maybe you meant only on available land mass? The oceans make up 71% of the earth's surface. Thus that reduces available energy to 9x10^10 GW/hrs per year. Obviously, it is still orders of magnitude more than needed.
So what percentage of land mass would be needed? .02%!!
- Jack
-PoP
http://sharesleuth.com/2006/08/moonshine_blindness.html
This doesn't undermine the potential of another company to do it right, but there's a lot of hype about innovation that isn't being matched with results (or even effort).
:-)