The Ethanol Debate....runs on
This posting is for posterity. I pay for the WSJ print and the WSJ online, but I have to pay for the archived article if I ever want to refer back to. . . so . . .
THE NUMBERS GUY
By CARL BIALIK
Digging Into the Ethanol Debate
June 9, 2006
President Bush announced in his State of the Union address1 in January that he backed funding for research into producing ethanol from corn and other farm products, with the goal of making a viable fuel alternative to gasoline for automobiles. Since then, Congress has wrangled over how to implement the idea.
Critics, meanwhile, have blasted the viability of ethanol. A central argument is that corn-based ethanol, the most-common form today, is literally a waste of energy. Detractors say that it takes more fuel to make ethanol -- growing the corn, bringing it to a processing plant and converting it to fuel -- than would be saved by using it.
That criticism has received attention in articles in the Washington Post2, the Louisville Courier-Journal3 and Cox News Service4 (all of which also included the pro-ethanol side). In April, Larry Kudlow said on his CNBC show, "So many experts believe it costs more energy to turn corn into ethanol-related gasoline than [is] actually produced."
Two prominent researchers are chiefly responsible for the energy-efficiency claim: Cornell University's David Pimentel and Tad Patzek of the University of California, Berkeley. In a co-written paper5 published last year in Natural Resources Research, Profs. Pimentel and Patzek wrote, "Ethanol production using corn grain required 29% more fossil energy than the ethanol fuel produced." By comparison, production of gasoline or diesel uses about 20% more fossil energy than the fuels produce. (For automobiles, ethanol is generally blended with gasoline in either 90-10 or 85-15 proportions, but the studies focused on the energy content of the ethanol itself.)
But the analysis stacks the deck against ethanol in a number of ways. Perhaps most important: The researchers attributed a wide array of energy costs to ethanol production, including the energy required to produce tractors used in cornfields and even all forms of energy consumed by workers for things such as food, transportation and police protection. Equivalent factors generally aren't included in comparable analyses of rival fuels like gasoline. Also, researchers didn't take into consideration the value of ethanol by-products, which can be used in cattle feed.
What's more, the skeptical analysis was based on all technology in use at the time, including old plants. Ethanol has become a hot business and a target of venture capitalists. There is reason to believe that ethanol production is only going to become more efficient, possibly at a faster rate than the more-mature petroleum industry. The newest plants incorporate technology to streamline the process and save energy and money. Researchers are also looking at methods to get ethanol from sugar cane and switchgrass, which appear to be more energy-efficient than those for corn. "There are a lot of new technologies," said Hosein Shapouri, an agricultural economist for the U.S. Department of Agriculture. "It's going to continue to improve the yield, and also lower the energy."
The Bush administration says ethanol is more energy efficient than the critics claim. Department of Energy spokesman Craig Stevens told me in an email, "Based on the vast majority of research and analysis, the department believes that the energy delivered by ethanol is greater than the fossil energy put into its production."
Other researchers have disagreed with Profs. Pimentel and Patzek. Michael Wang, a vehicle fuel-system analyst at Argonne National Laboratories in Lemont, Ill., calculates numbers that are frequently cited for the efficiency of producing petroleum and diesel fuel. He said those numbers don't include the energy needed for labor and to produce the equipment -- in large part because there aren't reliable, up-to-date estimates for that energy -- and therefore, neither should the ethanol numbers.
By his reckoning, it takes 0.74 BTU of fossil fuel to create 1 BTU of ethanol fuel, compared with a ratio of 1.23 BTUs to 1 BTU for gasoline -- that's 66% more than ethanol. (Dr. Wang's calculations are contained in a rather dense set of appendices to this report6; the conclusions are presented in a more user-friendly format in this brochure7.)
Prof. Pimentel defended his work in an interview. "I don't see how you could or should eliminate the labor of the farmer," he said. "He eats, sleeps, uses the highways, depends on the police force, fireman, and so forth."
Prof. Pimentel added that he's studied the issue for over 20 years, and has no bias against ethanol -- quite the contrary: "I'd really like to support ethanol being a viable solution for our liquid-fuel needs, because I am an agriculturalist and a biologist. But I'm a scientist first."
His co-author on the study, Prof. Patzek, didn't respond to my requests for an interview.
There remain major challenges for ethanol. Among them: The high price of natural gas may force some plants to switch to coal, harming their environmental profile; the fuel has yet to prove its market viability for cars without subsidies; and the costs to revamp fuel stations for ethanol blends is steep.
When prompted by their students to investigate biofuels, Berkeley energy and resources professors Dan Kammen and Alex Farrell discovered the sharp disagreements among researchers. "It became pretty clear to us, as we were getting up to speed on ethanol, that there are a large number of divergent studies in literature, and it's not clear why they are divergent," Prof. Farrell told me. They attempted to reconcile disputing studies by comparing them8 side by side, tracing the numbers back to their original sources and converting everything to standard units. Their conclusion, published9 in Science in January, was largely in line with Dr. Wang's. (So was an analysis10 of published studies that appeared in March in Environmental Science & Technology, and funded in part by the environmental organization Natural Resources Defense Council.)
It can be disorienting to discover that reputable researchers can so seriously disagree on a single number. In an article11 last month, the Toledo Blade counted studies, as if that might help settle things. The newspaper noted Prof. Pimentel's work, and added, "Five other researchers have done studies and agree. Thirteen other studies, including one paid for by the Department of Energy, show the opposite."
A drawback of all the commonly cited numbers is that they generally rely on data from USDA surveys12 of farmers and ethanol producers. Such surveys are a few years old. That's not an unusual lag time for federal government surveys, but they don't capture the impact of new plants in the fast-evolving ethanol industry.
Broin Cos., based in Sioux Falls, S.D., has pioneered a method to convert corn to ethanol at 90 degrees, rather than the previous 230 to 250 degrees, improving energy efficiency by 10% to 12%, according to co-founder and Chief Executive Jeff Broin. And E3 Biofuels LLC is finding ways to get more out of all parts of the corn, by building plants near dairy farms and feeding cows the byproducts of ethanol processing, then using energy from the animal waste to help power the plants. "Wastes are converted to valuable products, such as biogas and biofertilizers, which replace fossil fuels and their derivatives," David Hallberg, president and chief executive of Omaha-based E3, wrote me in an email.
Vinod Khosla, a partner in the Menlo Park, Calif., venture-capital firm Khosla Ventures, has invested in several ethanol technologies and is an advocate for their promise. He said arguments against ethanol focus unjustly on older plants. "It's like saying, a power plant built in the '50s is very polluting, so all power plants are very polluting," Mr. Khosla told me.
* * *
Several readers responded to last week's column13 about political polling. Here are some excerpts:
Just about every last survey and poll has some element of sample bias. Only the U.S. Census Bureau has the resources to try to reduce sample bias to a negligible level. The trick is understanding what sample bias exists, using that knowledge for better filtering and interpretation of the data, and being "methodology agnostic" in order to select the optimal methodology for the task at hand rather than simply sell what best fits the business model of the research firm or pollster.
--James Chung
As someone actively involved with politics and public policy, I'm often confronted with misleading poll questions or unsubstantiated claims based on misleading poll results or survey research. The amount of bad data is a virus of sorts because it alters decision-making by candidates and lawmakers, and often for the worse. Interest groups manipulate the political debate by releasing alarming numbers that serve their ends and not the public's. Unfortunately, the media rarely question the confidence of the data or the biases in the research instrument because the results often make for alluring headlines and enticing soundbites regardless of their reliability.
--John Colyandro, Texas Conservative Coalition Research Institute14
In this age of touting facts supported by (supposedly reliable) percentages and figures, there couldn't be a better time to sift through misleading polls. Especially when they come from mainstream sources, what is to keep the numbers from becoming facts? Moreover, it is when the numbers are most outlying that we tend to spread the word, defending our claims with, "I read it in an article." No longer can we simply segregate printed material as reliable and televised material as questionable. Sadly, we have surpassed the age when most news in print could be taken at face value.
--Donald Cunningham
I also received this letter about two articles I mentioned revisiting old statistics claiming that unmarried women at age 40 are extremely unlikely to get married:
There is no new evidence that Newsweek was "wrong." The Census Department demographer quarreled with the study at the time, and that's great -- so it's either 2.8% or 17%. But the fact that Newsweek found that many of the people it actually spoke to had gotten married tells us nothing about the real numbers or actual statistics. These women specifically interviewed are a small and nonrandom sample. For all we know, Newsweek, or the Census Department demographer, were right and are right today.
--Bob Barnes
That's true: The follow-up interviews are anecdotal, and it's even possible -- though not mentioned in the articles -- that exposure in the Newsweek article helped the cover subjects meet new people. The follow-up articles do offer some numerical information, as well.
And I got this letter in response to letters last week about box-office figures:
Number of tickets sold would be a helpful metric but "popularity" is a more difficult measure. This could be the percentage of the moviegoing public which saw that movie -- rather like television's Nielsen ratings, which shows the total viewers and the size of the audience for that show relative to the television universe expressed as a percentage. ... The real message here is the one you conveyed indirectly: these numbers are contrived as part of a publicity machine to generate sales.
--Jeffrey Atwood
Write to Carl Bialik at numbersguy@wsj.com15
URL for this article:
http://online.wsj.com/article/SB114970102238673892.html
Hyperlinks in this Article:
(1) http://www.whitehouse.gov/news/releases/2006/01/20060131-10.html
(2) http://www.washingtonpost.com/wp-dyn/content/article/2006/02/17/AR2006021702109.html
(3) http://www.courier-journal.com/apps/pbcs.dll/article?AID=/20050421/NEWS01/504210377/1008
(4) http://www.coxwashington.com/reporters/content/reporters/stories/BC_ETHANOL_ADV16_COX.html
(5) http://www.springerlink.com/(kmeezp55npe5ne3o2flnbsei)/app/home/contribution.asp?referrer=parent&backto=issue,6,6;journal,4,48;linkingpublicationresults,1:105547,1
(6) http://www.transportation.anl.gov/pdfs/TA/339.pdf
(7) http://www.transportation.anl.gov/pdfs/TA/345.pdf
(8) http://rael.berkeley.edu/ebamm/
(9) http://rael.berkeley.edu/ebamm/FarrellEthanolScience012706.pdf
(10) http://www.ce.cmu.edu/~gdrg/readings/2006/03/21/Hammerschlag_EthanolsEnergyReturnonInvestmentSurveyofLiterature1990-Pres.pdf
(11) http://www.toledoblade.com/apps/pbcs.dll/article?AID=/20060507/BUSINESS01/605070343
(12) http://www.usda.gov/oce/reports/energy/index.htm
(13) http://online.wsj.com/public/article/SB114910775055167791.html
(14) http://www.txccri.org/
(15) mailto:numbersguy@wsj.com
(16) http://online.wsj.com/public/page/0,,2_1125,00.html
(17) http://online.wsj.com/public/page/0,,2_1125,00.html
(18) mailto:numbersguy@wsj.com
(19) http://online.wsj.com/email
(20) http://WSJ.com/NumbersGuy
(21) http://online.wsj.com/xml/rss/0,,3_7028,00.xml
(22) https://users2.wsj.com/WebIntegration/WebIntegrationServlet?call=A_EC
(23) https://users2.wsj.com/WebIntegration/WebIntegrationServlet?call=A_EC
(24) http://online.wsj.com/reg/page/0,,5_3017,00.html
(25) http://online.wsj.com/article/0,,the_daily_fix,00.html
(26) mailto:numbersguy@wsj.com
THE NUMBERS GUY
By CARL BIALIK
Digging Into the Ethanol Debate
June 9, 2006
President Bush announced in his State of the Union address1 in January that he backed funding for research into producing ethanol from corn and other farm products, with the goal of making a viable fuel alternative to gasoline for automobiles. Since then, Congress has wrangled over how to implement the idea.
Critics, meanwhile, have blasted the viability of ethanol. A central argument is that corn-based ethanol, the most-common form today, is literally a waste of energy. Detractors say that it takes more fuel to make ethanol -- growing the corn, bringing it to a processing plant and converting it to fuel -- than would be saved by using it.
That criticism has received attention in articles in the Washington Post2, the Louisville Courier-Journal3 and Cox News Service4 (all of which also included the pro-ethanol side). In April, Larry Kudlow said on his CNBC show, "So many experts believe it costs more energy to turn corn into ethanol-related gasoline than [is] actually produced."
Two prominent researchers are chiefly responsible for the energy-efficiency claim: Cornell University's David Pimentel and Tad Patzek of the University of California, Berkeley. In a co-written paper5 published last year in Natural Resources Research, Profs. Pimentel and Patzek wrote, "Ethanol production using corn grain required 29% more fossil energy than the ethanol fuel produced." By comparison, production of gasoline or diesel uses about 20% more fossil energy than the fuels produce. (For automobiles, ethanol is generally blended with gasoline in either 90-10 or 85-15 proportions, but the studies focused on the energy content of the ethanol itself.)
But the analysis stacks the deck against ethanol in a number of ways. Perhaps most important: The researchers attributed a wide array of energy costs to ethanol production, including the energy required to produce tractors used in cornfields and even all forms of energy consumed by workers for things such as food, transportation and police protection. Equivalent factors generally aren't included in comparable analyses of rival fuels like gasoline. Also, researchers didn't take into consideration the value of ethanol by-products, which can be used in cattle feed.
What's more, the skeptical analysis was based on all technology in use at the time, including old plants. Ethanol has become a hot business and a target of venture capitalists. There is reason to believe that ethanol production is only going to become more efficient, possibly at a faster rate than the more-mature petroleum industry. The newest plants incorporate technology to streamline the process and save energy and money. Researchers are also looking at methods to get ethanol from sugar cane and switchgrass, which appear to be more energy-efficient than those for corn. "There are a lot of new technologies," said Hosein Shapouri, an agricultural economist for the U.S. Department of Agriculture. "It's going to continue to improve the yield, and also lower the energy."
The Bush administration says ethanol is more energy efficient than the critics claim. Department of Energy spokesman Craig Stevens told me in an email, "Based on the vast majority of research and analysis, the department believes that the energy delivered by ethanol is greater than the fossil energy put into its production."
Other researchers have disagreed with Profs. Pimentel and Patzek. Michael Wang, a vehicle fuel-system analyst at Argonne National Laboratories in Lemont, Ill., calculates numbers that are frequently cited for the efficiency of producing petroleum and diesel fuel. He said those numbers don't include the energy needed for labor and to produce the equipment -- in large part because there aren't reliable, up-to-date estimates for that energy -- and therefore, neither should the ethanol numbers.
By his reckoning, it takes 0.74 BTU of fossil fuel to create 1 BTU of ethanol fuel, compared with a ratio of 1.23 BTUs to 1 BTU for gasoline -- that's 66% more than ethanol. (Dr. Wang's calculations are contained in a rather dense set of appendices to this report6; the conclusions are presented in a more user-friendly format in this brochure7.)
Prof. Pimentel defended his work in an interview. "I don't see how you could or should eliminate the labor of the farmer," he said. "He eats, sleeps, uses the highways, depends on the police force, fireman, and so forth."
Prof. Pimentel added that he's studied the issue for over 20 years, and has no bias against ethanol -- quite the contrary: "I'd really like to support ethanol being a viable solution for our liquid-fuel needs, because I am an agriculturalist and a biologist. But I'm a scientist first."
His co-author on the study, Prof. Patzek, didn't respond to my requests for an interview.
There remain major challenges for ethanol. Among them: The high price of natural gas may force some plants to switch to coal, harming their environmental profile; the fuel has yet to prove its market viability for cars without subsidies; and the costs to revamp fuel stations for ethanol blends is steep.
When prompted by their students to investigate biofuels, Berkeley energy and resources professors Dan Kammen and Alex Farrell discovered the sharp disagreements among researchers. "It became pretty clear to us, as we were getting up to speed on ethanol, that there are a large number of divergent studies in literature, and it's not clear why they are divergent," Prof. Farrell told me. They attempted to reconcile disputing studies by comparing them8 side by side, tracing the numbers back to their original sources and converting everything to standard units. Their conclusion, published9 in Science in January, was largely in line with Dr. Wang's. (So was an analysis10 of published studies that appeared in March in Environmental Science & Technology, and funded in part by the environmental organization Natural Resources Defense Council.)
It can be disorienting to discover that reputable researchers can so seriously disagree on a single number. In an article11 last month, the Toledo Blade counted studies, as if that might help settle things. The newspaper noted Prof. Pimentel's work, and added, "Five other researchers have done studies and agree. Thirteen other studies, including one paid for by the Department of Energy, show the opposite."
A drawback of all the commonly cited numbers is that they generally rely on data from USDA surveys12 of farmers and ethanol producers. Such surveys are a few years old. That's not an unusual lag time for federal government surveys, but they don't capture the impact of new plants in the fast-evolving ethanol industry.
Broin Cos., based in Sioux Falls, S.D., has pioneered a method to convert corn to ethanol at 90 degrees, rather than the previous 230 to 250 degrees, improving energy efficiency by 10% to 12%, according to co-founder and Chief Executive Jeff Broin. And E3 Biofuels LLC is finding ways to get more out of all parts of the corn, by building plants near dairy farms and feeding cows the byproducts of ethanol processing, then using energy from the animal waste to help power the plants. "Wastes are converted to valuable products, such as biogas and biofertilizers, which replace fossil fuels and their derivatives," David Hallberg, president and chief executive of Omaha-based E3, wrote me in an email.
Vinod Khosla, a partner in the Menlo Park, Calif., venture-capital firm Khosla Ventures, has invested in several ethanol technologies and is an advocate for their promise. He said arguments against ethanol focus unjustly on older plants. "It's like saying, a power plant built in the '50s is very polluting, so all power plants are very polluting," Mr. Khosla told me.
* * *
Several readers responded to last week's column13 about political polling. Here are some excerpts:
Just about every last survey and poll has some element of sample bias. Only the U.S. Census Bureau has the resources to try to reduce sample bias to a negligible level. The trick is understanding what sample bias exists, using that knowledge for better filtering and interpretation of the data, and being "methodology agnostic" in order to select the optimal methodology for the task at hand rather than simply sell what best fits the business model of the research firm or pollster.
--James Chung
As someone actively involved with politics and public policy, I'm often confronted with misleading poll questions or unsubstantiated claims based on misleading poll results or survey research. The amount of bad data is a virus of sorts because it alters decision-making by candidates and lawmakers, and often for the worse. Interest groups manipulate the political debate by releasing alarming numbers that serve their ends and not the public's. Unfortunately, the media rarely question the confidence of the data or the biases in the research instrument because the results often make for alluring headlines and enticing soundbites regardless of their reliability.
--John Colyandro, Texas Conservative Coalition Research Institute14
In this age of touting facts supported by (supposedly reliable) percentages and figures, there couldn't be a better time to sift through misleading polls. Especially when they come from mainstream sources, what is to keep the numbers from becoming facts? Moreover, it is when the numbers are most outlying that we tend to spread the word, defending our claims with, "I read it in an article." No longer can we simply segregate printed material as reliable and televised material as questionable. Sadly, we have surpassed the age when most news in print could be taken at face value.
--Donald Cunningham
I also received this letter about two articles I mentioned revisiting old statistics claiming that unmarried women at age 40 are extremely unlikely to get married:
There is no new evidence that Newsweek was "wrong." The Census Department demographer quarreled with the study at the time, and that's great -- so it's either 2.8% or 17%. But the fact that Newsweek found that many of the people it actually spoke to had gotten married tells us nothing about the real numbers or actual statistics. These women specifically interviewed are a small and nonrandom sample. For all we know, Newsweek, or the Census Department demographer, were right and are right today.
--Bob Barnes
That's true: The follow-up interviews are anecdotal, and it's even possible -- though not mentioned in the articles -- that exposure in the Newsweek article helped the cover subjects meet new people. The follow-up articles do offer some numerical information, as well.
And I got this letter in response to letters last week about box-office figures:
Number of tickets sold would be a helpful metric but "popularity" is a more difficult measure. This could be the percentage of the moviegoing public which saw that movie -- rather like television's Nielsen ratings, which shows the total viewers and the size of the audience for that show relative to the television universe expressed as a percentage. ... The real message here is the one you conveyed indirectly: these numbers are contrived as part of a publicity machine to generate sales.
--Jeffrey Atwood
Write to Carl Bialik at numbersguy@wsj.com15
URL for this article:
http://online.wsj.com/article/SB114970102238673892.html
Hyperlinks in this Article:
(1) http://www.whitehouse.gov/news/releases/2006/01/20060131-10.html
(2) http://www.washingtonpost.com/wp-dyn/content/article/2006/02/17/AR2006021702109.html
(3) http://www.courier-journal.com/apps/pbcs.dll/article?AID=/20050421/NEWS01/504210377/1008
(4) http://www.coxwashington.com/reporters/content/reporters/stories/BC_ETHANOL_ADV16_COX.html
(5) http://www.springerlink.com/(kmeezp55npe5ne3o2flnbsei)/app/home/contribution.asp?referrer=parent&backto=issue,6,6;journal,4,48;linkingpublicationresults,1:105547,1
(6) http://www.transportation.anl.gov/pdfs/TA/339.pdf
(7) http://www.transportation.anl.gov/pdfs/TA/345.pdf
(8) http://rael.berkeley.edu/ebamm/
(9) http://rael.berkeley.edu/ebamm/FarrellEthanolScience012706.pdf
(10) http://www.ce.cmu.edu/~gdrg/readings/2006/03/21/Hammerschlag_EthanolsEnergyReturnonInvestmentSurveyofLiterature1990-Pres.pdf
(11) http://www.toledoblade.com/apps/pbcs.dll/article?AID=/20060507/BUSINESS01/605070343
(12) http://www.usda.gov/oce/reports/energy/index.htm
(13) http://online.wsj.com/public/article/SB114910775055167791.html
(14) http://www.txccri.org/
(15) mailto:numbersguy@wsj.com
(16) http://online.wsj.com/public/page/0,,2_1125,00.html
(17) http://online.wsj.com/public/page/0,,2_1125,00.html
(18) mailto:numbersguy@wsj.com
(19) http://online.wsj.com/email
(20) http://WSJ.com/NumbersGuy
(21) http://online.wsj.com/xml/rss/0,,3_7028,00.xml
(22) https://users2.wsj.com/WebIntegration/WebIntegrationServlet?call=A_EC
(23) https://users2.wsj.com/WebIntegration/WebIntegrationServlet?call=A_EC
(24) http://online.wsj.com/reg/page/0,,5_3017,00.html
(25) http://online.wsj.com/article/0,,the_daily_fix,00.html
(26) mailto:numbersguy@wsj.com