From Nature...puts in perspective how important the technology...

From Nature...puts in perspective how important the technology could be.

Interesting user comment at the bottom too.


Published online 20 February 2008 | Nature 451, 880-883 (2008) | doi:10.1038/451880a

News Feature

Energy: Not your father's biofuels
If biofuels are to help the fight against climate change, they have to be made from more appropriate materials and in better ways. Jeff Tollefson asks what innovation can do to improve the outlook.

Jeff Tollefson


D. SIMONDSBiotechnology has changed the way that drugs are discovered, designed and, often, made. It has spread new capabilities across the farms of much of the world, sometimes amid much controversy. Now some of its advocates are suggesting that it is poised to overhaul the energy sector as well, changing both the crops that are grown and the fuels that are made from them.

Entrepreneurs have attracted hundreds of millions of dollars for bio-energy companies working on 'second-generation' fuels produced from crop residues, grasses or woody materials that avoid the shortcomings of ethanol distilled from corn starch or biodiesel produced from oil crops. Some are offering hope for higher yields from less land, from more marginal land, and with less investment in terms of energy and fertilizer; others are promising fuels better suited to the needs of drivers and the existing fuel infrastructure. Even the major oil companies are getting involved.

“The market is slated to be so big that there will be opportunities for multiple approaches, and it will probably take many years before we settle on one absolute best approach,” says Doug Cameron, chief science officer for Khosla Ventures, a venture-capital firm in Menlo Park, California, that is backing a large number of start-up bioenergy companies.

Liquid biofuels will never make up a significant portion of the global transportation fuel supply unless biologists and engineers make the most of these opportunities. During the past four years alone, global ethanol production has more than doubled to nearly 50 billion litres (about 13.2 billion gallons) in 2007. Biodiesel, although starting out much lower, nearly quintupled to 9 billion litres (2.4 billion gallons) during the same period (see A growing concern). This rate of growth is not sustainable — and with current production methods far from desirable, because the agricultural techniques used often damage the environment on a scale that far outweighs any good achieved through the biofuels' use.

“We're not screwing around — the basic science is done.”
Pat Gruber
In the United States, which has ramped up production and is now home to the world's largest biofuel industry, roughly 23% of the corn crop goes to ethanol, which in turn provides 3% of the nation's transportation fuels, according to Alex Farrell, an energy and resource scientist at the University of California, Berkeley. Worldwide, biofuels make up less than 1% of transportation fuel, he says. Current technologies could push that as high as 2–3%, “but anything much larger than that will have to be based on significantly different technologies”.

Fuels for the future
Applying biotechnology to biofuels is not a new idea. In 1991 Lonnie Ingram, a microbiologist at the University of Florida, Gainesville, was awarded a patent for an engineered Escherichia coli bacterium that converts sugars into ethanol. But only now is the technology really coming into its own — not least because today's 'metabolic engineering' and 'synthetic biology' put much more ambitious fuels than ethanol on the table, or indeed into the pipeline.

“I've always been of the opinion that ethanol is for drinking, not driving,” says Jay Keasling, a chemical engineering professor at the University of California, Berkeley, who has pioneered the synthetic biology needed to get microbes to produce various new classes of molecule. To take one example: drinking is made easy by the fact that ethanol and water mix easily. That's good for scotch and soda, but bad for pipelines, which give the fuel a chance to get watered down and contaminated. In the United States, home to the largest biofuels industry in the world, trucks or train cars carry the fuel from the agricultural heartland where corn is grown to the coastal areas where fuel is needed most. In Brazil, the other biofuel giant, where sugar-cane ethanol is produced much more efficiently, the agricultural source and the urban users are relatively close together in the country's southeast — one of the many advantages the Brazilian industry enjoys.

This is why Amyris Biotechnologies, a start-up company in Emeryville, California, of which Keasling was a co-founder, and a number of other small companies in California and Massachusetts are designing microbes to make better fuels — fuels with higher energy contents that are better suited to pipelines and other infrastucture. University researchers are also in the race. Last month Jim Liao, a researcher at the University of California, Los Angeles, described an E. coli in which more than a dozen modifications to the metabolic pathways normally used to produce amino acids caused the bacteria to produce isobutanol, an alcohol with four carbon atoms to ethanol's two, and similar molecules. Even without optimizing every step of the process, Liao's lab was able to achieve a yield of 86% of the theoretical maximum (S. Atsumi, T. Hanai & J. C. Liao Nature 451, 86–89; 2008).

His process has been licensed by Gevo in Pasadena, California, a start-up company funded by Khosla Ventures, among others, which says that it hopes to begin commercial-scale production within a few years. Gevo chief executive Pat Gruber says that the technology could be retrofitted on an existing bioethanol plant for as little as US$20 million: “We're not screwing around — the basic science is done. We are going to try and get this stuff developed and into the marketplace.”

Another form of butanol has attracted oil-giant BP and DuPont, one of the world's largest chemical companies, which are working together on biofuels. Butanol has a higher energy density than ethanol, offering roughly 85% of the energy content of a standard petrol mix, compared with about 66% for ethanol; this offsets the fact that less butanol than ethanol can be made from a given amount of biomass (there are only so many carbon atoms to go round). Although other molecules could pack even more energy, DuPont officials say that they settled on butanol as a molecule that meets their needs and can be developed quickly.

DuPont has a track record in the sort of metabolic re-engineering required for such things: the E. coli that churn out propanediol, a chemical used in various materials and industrial processes, in its facility in Loudon, Tennessee, have had 30 changes made to their metabolic pathways. It took about 11 years for that project to get from the proof of concept to production, but officials say biobutanol could be sorted out much faster. “In the old days, it would take four to six months to clone a new gene,” says John Pierce, DuPont's vice-president for applied biosciences. “Now it takes two weeks and you usually do it by mail.”



View an enlarged version of the graph.
Butanol's advocates say that their fuel could be run through a refinery to produce longer chain molecules as desired. Amyris and LS9 of Cambridge, Massachusetts, are looking to skip this step and move straight to molecules more like those that engines already burn. Both companies specialize in synthetic biology — which aims to create new biological entities in a much more thoroughly designed way than traditional genetic engineering — and both have received backing from Khosla Ventures, among others.

LS9 was co-founded by George Church, a geneticist at Harvard Medical School in Boston, Massachusetts, and Chris Somerville, a plant biologist at Stanford University in California, who previously headed the Carnegie Institution's department of Plant Biology. Somerville is currently heading the new Energy Biosciences Institute (EBI), a partnership between the University of California, Berkeley, Lawrence Berkeley National Laboratory and the University of Illinois at Urbana-Champaign. The EBI was kick-started in 2006 with a $500 million pledge from BP.

Different pathways
LS9 wants to transform the fatty acids naturally produced by E. coli into specific hydrocarbon fuels; it has announced plans for a pilot plant and says that commercial production could begin within two to three years. Amyris decided to take yeast, which is known for its ability to make ethanol, and train it to produce longer-chain molecules such as gasoline, diesel and jet fuel.

The company has engineered roughly 1,000 strains of yeast so far and expects to build as many as 2,000 more in the months ahead. When a strain looks promising, the scale-up begins, as molecular biologists hand their product to chemical engineers and fermentation engineers. In addition to producing exactly what is needed, Jack Newman, Amyris's senior vice-president for research, says that the technology could prove to be extremely versatile. “You could have your plant making diesel fuel one day and almost literally turn around and make jet fuel the next.”

Wood pulp and fiction
All these companies are looking at traditional raw materials for their wonderbugs — sugars. And this means that although they may make more attractive and easily distributed fuels, they will not of themselves change the industry's productivity. To do that requires cheaper and more sustainable feedstocks than sugars and starches from cane and corn.

“I'm of the opinion that the crucial thing that needs to be done is not actually to make better fuels out of sugars but to make sugars more efficiently from cellulose,” says Lee Lynd, a biology and engineering professor at Dartmouth College in Hanover, New Hampshire. “The thing that limits corn ethanol, frankly, is not ethanol — it's corn. Those other molecules will become important in the broader scheme of things if and when we solve the problems with cellulosic biomass.”

Turning cellulose — the tough polymer from which the cell walls of plants are made — into biofuel is currently a major focus in the industry. It widens the possible range of feedstocks greatly, making it possible to use crops for biofuels that are not also food for humans. Legislators are keen to push the industry in that direction. The ethanol mandate enacted by the US Congress in 2007 requires that, of an annual ethanol production of 136 billion litres (36 billion gallons) required by 2022 — more than five times current US production — 44% must come from cellulosic foodstocks, with corn's contribution remaining static from the mid 2010s. Things might be able to go even further. A report by McKinsey & Company, a consultancy firm based in New York, suggests that, at oil prices above $70 per barrel, cellulosic feedstocks could supply half of the global transportation fuels by 2020 (although that figure explicitly ignores many real-world constraints).

The fly in this ointment is that the world cannot yet boast a single commercial-scale cellulosic-ethanol facility. Breaking cellulose down into sugars is not easy work, and can use up a lot of energy; what's more, not all the sugars produced are easily fermented. Despite the recent spike in oil prices on the international market, lenders and investors have hesitated to pump money into commercial-scale ventures, fearing technical risks and a potential drop in the price of oil. To help push the technology over the first economic hurdle, the US Department of Energy (DOE) is pumping $385 million into six demonstration projects. Work begun on the first of them, run by Colorado-based Range Fuels, in Georgia last autumn.


R. KALTSCHMIDT/LBNLE“I've always been of the opinion that ethanol is for drinking, not driving.”
Jay Keasling
Cellulose can be broken down with heat and catalysts. It can also be broken down with biology. This is where biotechnologies can come in. Dartmouth's Lynd is the co-founder and chief scientific officer for Mascoma Corporation in Cambridge, Massachusetts, which is pursuing work on a bacterium that can produce the enzymes to break down cellulose on demand as part of the process. Verenium Corporation, also in Cambridge, the first publicly held cellulosic-ethanol enterprise, is developing both new enzymes and organisms to break down the structural tissue. Other companies and researchers are looking to the microbes that break down cellulose in the guts of cows and termites, or to fungi — that is to say, the natural world's most accomplished consumers of the cellulose found in grasses and wood.

Finding a feedstock
If cellulosic technologies can be made to work, there's still the question of where the cellulose comes from. For the past ten thousand years, most human meddling with the proclivities of plants has been designed to make them better to eat — more sugar in the cane, more starch in the corn, more protein. Energy crops, though, require a different approach. No need for fruit or seed — just for fast growing cellulose, ideally in a form that requires little or nothing by way of extra inputs such as fertilizer. Cellulose that is particularly easy to break down gets bonus points.

Switchgrass, a hearty, fast-growing grass native to America's prairies, is one much-touted possibility, but even a brief survey indicates that there could be better options. Steve Long, a professor at the University of Illinois at Urbana-Champaign, has spent years studying Miscanthus giganteus, a relative of sugar cane that is native to East Asia. He has studied the grass in Denmark, England and now Illinois, and says that the average yield is double that of switchgrass and 50% higher than corn — without any fertilizers.

According to Long, meeting the ethanol goals laid out in 2006 by President Bush (which were a touch lower than the mandates actually passed by congress) with corn ethanol would require 25% of the nation's cropland. That figure drops to 15% if cellulosic technology allows you to make use of the corn 'stover' — the bits left over after harvest. It drops to 8% if you use M. giganteus — and such grasses could be grown on marginal land that is not being used for food production today.

And, according to Long, neither stover nor M. giganteus are the last words. “A lot of effort is being invested in modifying corn stover to make it easier for digestion, but I think we need to think much more broadly,” he says. “It seems very unlikely that we've come across the very best option.”

Whichever crops look most promising, engineers will be eager to tinker with them. Keasling says that a decade of intensive biotechnology could now do for an energy crop as much as centuries of selective breeding have done for many food crops. New models of cultivation may help too — for example the replacement of monocultures, which are useful if you want a particular form of food, with polycultures that simply maximize biomass. David Tilman and his colleagues, at the University of Minnesota in St Paul, reported last year that plots with a diverse mixture of prairie grasses yielded on average 238% more energy than plots with a single crop (D. Tilman, J. Hill & C. Lehman Science 314, 1598–1600; 2006). All these approaches might be tailored to marginal lands where the soil wouldn't support food crops.

Or you could just do away with soil altogether: that's the appeal of algae. GreenFuel Technologies, based in Cambridge, Massachusetts, is developing algal bioreactors that tap into carbon-dioxide streams from coal plants to produce rich algal crops that can be harvested and turned into biofuels. A more ambitious approach would be to have the algae act as both feedstock and processor, secreting ready-made fuels. Tasios Melis at the University of California, Berkeley, is looking at getting algae to secrete hydrocarbons in a form that can be continuously collected.


As it is today: this VeraSun Energy plant near Aurora, South Dakota, produces 450 million litres of corn ethanol a year.E. LANDWEHR/APBotryococcus braunii, the green alga that Melis is working with, naturally secretes a 30-carbon terpenoid that can be processed into fuel, perhaps in order to reduce the effects of ultraviolet light. Melis says that he has devised a method for collecting the product — a sticky film to which the algae adhere — and is now working to increase production. More ambitiously, various groups, including Melis, are looking at having algae produce hydrogen.

Another approach, taken by Coskata in Warrenville, Illinois, (yet another Khosla Ventures start-up) and its partner General Motors, is to avoid putting effort into specialist feedstocks and instead develop a process that works universally. Coskata's process converts carbon-based feedstocks, which could be crops, agricultural waste or municipal waste, into carbon monoxide and hydrogen. This is the first step of processes by which coal or natural gas are turned into liquid fuels, a procedure that is normally taken as an alternative to biological means for producing fuels. Coskata says it has a way of combining the two approaches, with microbes that turn the 'synthesis gas' straight into ethanol. The company is currently developing a pilot project that it says will be able to produce ethanol for less than 26 cents per litre ($1 per gallon).

Scaling up
One thing all these innovations share is that they have yet to be attempted on a large scale. Biotech may work well for drug companies that sell small volumes at premium prices, but the economics of biofuels are more in line with the oil and gas industries, which sell in bulk for low prices and are dominated by the cost of raw materials and manufacturing.

ADVERTISEMENT

One of the problems is that microorganisms evolved to take care of themselves, not people. Re-engineering their metabolisms in such a way that they excrete fuels — which by definition are energy-rich compounds — means convincing them to forgo energy that they might otherwise use to their own ends. Moreover, in many cases the fuels can be toxic to the organisms themselves. All this provides ways for the engineering to come unstuck.

“If [the microbes] are unhappy with what they are doing, they are going to evolve away from what you want them to do,” says genome entrepreneur Craig Venter, whose company Synthetic Genomics in La Jolla, California, has an interest in biofuel production. “A key part of the future is going to be designing a system where they are not grossly unhappy with what they are doing.” Many researchers see hope in producing longer-chain biofuels precisely because it decreases the stress on the microbes doing the work. Ethanol is poisonous to its producers (which is why fermentation can't on its own produce hard liquor — the yeast dies). Longer-chain molecules will separate out from the medium in which their producers grow; this avoids the costly step of distillation.

Even if the microbial producers can be kept happy, there's a lot more work to do in producing systems that will work on an industrial scale in commercial refineries. “Among scientists it's considered very doable,” says Michael Himmel, who heads a team of researchers working on cellulosic ethanol at the DOE's National Renewable Energy Laboratory in Golden, Colorado. “It's difficult work and it's going to take some funding. But it's not cold fusion.” But what's doable in principle to a scientist is not always practical to an engineer.

Big oil companies, which would have the know-how for such engineering, are keeping an eye on the field. As well as BP's work with DuPont, Shell is partnering with Iogen in Ottawa, Canada, which ran one of the DOE's cellulosic pilot plants, and with an algal biofuel producer in Hawaii. But all this is small compared to the investments such companies make in their oil-based business.


D. SIMONDSBiofuels will never take over the whole liquid-fuel market, let alone amount to a large proportion of total energy use. But they, and other technologies, have a part to play. A decade and a half after receiving his patent on an ethanol-producing bacterium, Ingram is still a biofuel enthusiast. But he's also a realist. In the same law that expanded the ethanol mandate, Congress also increased the fuel-efficiency requirements for vehicles by 40%, pushing the average efficiency required by 2020 up to 6.7 litres per hundred kilometres (35 miles per gallon) from the present average of 9.4 litres per hundred kilometres (25 miles per gallon). The technologies to do that are already available — Japan had an average fuel efficiency of 5.1 litres per hundred kilometres in 2002. And as Ingram points out, “If we increase gas mileage by 1 mile per gallon, that is about equal to all the ethanol we are making right now from corn.”

Jeff Tollefson covers climate, energy and the environment for Nature. See Editorial, page 865.

Comments
Reader comments are usually moderated after posting. If you find something offensive or inappropriate, you can speed this process by clicking 'Report this comment' (or, if that doesn't work for you, email [email protected]). For more controversial topics, we reserve the right to moderate before comments are published.

I was surprised by the lack of mention of Searchinger and Tilman's recent Science papers on the enormous breakeven times of biofuels. We unfortunately don't have decades to solve our greenhouse gas emissions problem. Biofuels will also have a tough time supplying anything but a small fraction of our energy because plants are so inefficient. Fundamentals of Renewable Energy Processes gives the solar energy to ethanol conversion efficiency of sugarcane (one of the best) as 0.13%. If you compare this to the 30% of a Stirling dish, you've got a factor of 231. Worse, ethanol, butanol, etc. are converted into mechanical energy far less efficiently (much less than half) than the output from a Stirling dish (heat engines are subject to the Carnot limit after all). All of this translates into enormous differences in land area to create the same amount of work. In my opinion, the land area requires should be a primary criteria for evaluating renewable fuels.
Report this comment 20 Feb, 2008 Posted by: Earl Killian There is talk in this article of growing sources of cellulose on “marginal land” or with no fertilizer inputs. However, unless you recycle the wastes from the energy production process back to the soil the cellulose was grown in it will eventually become depleted for e.g. phosphate – and growth of plants that even tolerate low phosphate levels will become restricted. This is not a small problem since a Hubbert linearization analysis of mineral phosphate extraction predicts that this resource is 75% depleted! ( See www.energybulletin.net/33164.html and www.energybulletin.net/40300.html ) Phosphate extraction from lower quality sources would require greater energy inputs. This means that, if any phosphate for fertilizer is available in 10 years time, it will be exceedingly expensive. The looming phosphate shortage has enormous implications for biofuels and, more importantly for the continuation of western industrial agriculture.
Report this comment 20 Feb, 2008 Posted by: Michael Lardelli The article failed to discuss that there is one cellulosic feedstock that is readily available and at present has negative environmental value, the 2 billion tons of straw available world wide. Straw eventually becomes carbon dioxide without providing value, and at the same time temporarily binds nutrients, causing farmers to overfertilize. Straw also harbors pathogens requiring fungicide use which is why it used to be burnt in the field. The use of this byproduct would not necessitate putting new land in cultivation. The major problem with all the crops being used as cellulosic feedstocks is that they have not been domesticated for use as biofuel feedstocks. The main problem is lignin, which prevents metabolism of cellulosics by the wonderful microorganisms being developed (as described by the author). All the cellulosics described must be pretreated with heat and acid before the microbes can be used. Reducing and modifying lignin by genetic engineering of the cellulosic crops, whether switchgrass, Miscanthus, or straw from crops, can substantially reduce the amount of pre-treatment, as has already been demonstrated in the pulp and paper industry in transgenically modified trees. The potential uses of biotechnology for developing biofuel feedstocks is described in a recent review: “Transgenics are imperative for biofuel crops. Plant Science 174: 246-263 (2008).”
Report this comment 23 Feb, 2008 Posted by: Jonathan Gressel