The promise of cellulosic biofuels

Switchgrass biofuel crop
Switchgrass biofuel crop

Will next-generation biofuels have a less destructive effect on agriculture? A study just published by US government scientists suggests that so-called ‘cellulosic’ ethanol has much better energy balance than today’s biofuels.[1] By energy balance, we mean the energy used to make the fuel compared to its energy value when burnt in a car’s engine. News summaries of the paper’s contents focused on one estimate that suggested that to make cellulosic biofuels might only need 6% of the energy value contained in the fuel. Depending on which crop is used, where it is grown, and how it is refined, most of today’s biofuels have only a weakly positive energy balance. So the paper gives hope that we might expect considerable progress towards carbon-neutral transport fuels when we can start refining all vegetable matter, not just foodstuffs, into fuels.

Cellulosic biofuels may well become important sources of motor fuels. There is certainly huge amounts of money flowing into the field. Unfortunately none of the news articles covering the US research pointed out the technology for turning cellulose into fuel is still a long way from commercial viability. Yes, we can turn grass into ethanol, but at prices which will double the price of petrol. And the greenhouse gas savings will almost certainly not be as attractive as the paper suggests, not least because the authors did not include the serious impact of nitrous oxide emissions from fertilised fields.


Today’s biofuels are made from food. Whether it be wheat, sugar beet or rapeseed, biofuel refineries use materials that could be otherwise used for human or animal feed. As is now well understood, the competition between food and fuel is being won by the motor car, not human beings. It is not even a fair contest: humans need about 2,000 calories a day in food, but a UK car driver needs 40,000 calories a day to keep his car on the road in terms of the energy in his petrol.

This is the root of the biofuels problem. Even getting a small fraction of total motor fuel demand from agricultural crops requires us to turn large percentages of arable land over the petrol feedstock. To power the UK’s cars would require us to turn over all the UK’s cropland to fuel plantations. And, if we are not careful, this might happen. Your 2,000 calories of food a day is worth about 3p to a wheat farmer, but about 10p to a petrol retailer. In the US and the EU we actually rig the market even more in favour of fuels, giving subsidies to biofuel producers and telling retailers that they must include a percentage of plant-sourced ethanol in the mix at the petrol pump.

Biofuels are tending to crowd out food production – a point that is increasingly understood. But they also save little in the way of greenhouse gases. Growing a crop of sugar beet in an East Anglian field requires tractor fuel and fertiliser. The fertiliser creates the greenhouse gas nitrous oxide. The beet needs substantial processing, including a substantial amount of heat, before it becomes usable sugar. Then the sugar needs to be fermented into ethanol. All in all, the greenhouse gas savings are pretty minor, at least in high latitudes. Brazilian sugar cane ethanol is much better because it grows without fertiliser and the waste green matter is used for fuel in the refinery.

As oil runs out, what is going to stop us devoting all our arable land to service the needs of the world’s one billion or so automobiles rather than its six and a half billion people? There are three possible answers:

  • Extending the area given over to arable crops. Most countries only allocate a small fraction of their total land area to the growing of any form of crop. Poor or inaccessible land is given over to grazing animals, forestry, or left as ‘set-aside’. We could increase the total amount of land under cultivation. England, for example, uses less than a quarter of its non-urban land for crops. Much of the rest is given over to animals. A large-scale switch to grains and seeds and away from animals would enable more food calories to be produced on the same area of land. But the Financial Times of 12 January reported that the sown area of US farmland will rise a surprisingly small 3.8% in 2008 even after the huge increase in the prices of agricultural commodities. It seems that the supply of crop-growing land is not very elastic.
  • We could extend the area under cultivation around the world. Farmers’ leaders consistently say that prolonged low prices for agricultural commodities have reduced the incentive for farmers in poor countries to use land for crops. This is a much stronger point than conventionally acknowledged: high prices will drag large acreages into production in less accessible and prosperous parts of the world. Will it be enough? It may be that commodities like jatropha do enable us to ‘grow’ motor fuels on land that is too dry or infertile for conventional agriculture, but the evidence as yet is far from compelling.
  • We can hope for advances in cellulosic ethanol production. That is to say, instead of using food products for fuels, we could use the woody and straw waste materials from agriculture. This is the focus of many research groups and the piece of US government research covered in this article suggests considerable grounds for optimism.

It sounds easy to make a petrol replacement from woody waste. You dump it in a tank in which enzymes eat away at the lignin and cellulose and turn them into simpler carbohydrates. These carbohydrates are then fermented into ethanol (better know as 'alcohol'). The process seems simple. In fact Henry Ford thought that all car fuel would eventually be made from waste vegetation.

But so far the advances have been quite slow. There’s no doubt that we will find a way of breaking down cellulose on an industrial scale. Sheep and cows do it, after all. However the enzymes needed to turn the fairly complex cellulose molecule into simpler carbohydrates are expensive to make, and, like making ethanol, the process still needs considerable amounts of heat. Very roughly, we might be talking about a manufacturing cost of £1.50 per litre at the moment, though it will come down.

The US government research suggests that eventually the net energy input from turning cellulose into ethanol will be very small. So, to put it another way, we will need very little fossil fuel to make a litre of ethanol, saving money, greenhouse gases, and reducing the oil import bill.

The researchers used a conventional American prairie grass as the source material. This grass (‘switchgrass’) grows on dry plains with low soil fertility. Unless the price of motor fuel rises to very high levles, it will not drive out the growing of agricultural crops. The project measured the typical yields from fields, calculated the fossil-fuel inputs necessary to grow and harvest the grass, and then estimated how much energy will be used to convert the cellulose to ethanol.

Their estimates of very low energy use to make cellulosic ethanol must be tempered by four important caveats:

  • The scientists do not appear to have calculated the impact of fertiliser use on the emissions of nitrous oxide. Although average fertiliser inputs were quite low (about 75kg of nitrogen fertiliser compared to an average of about 200kg to grow English wheat) the nitrous oxide emissions from this use may well mean that the net greenhouse gas savings from cellulosic ethanol are relatively small. Remember that the researchers were primarily looking at the impact on net energy need not on the relative greenhouse gas emissions of petrol versus ethanol.
  • Their estimates of the manufacturing energy going into the production process are, at best, guesses. The paper should have been a little bit more transparent about this. And the scientists do not appear to have included the impact of actually making the enzymes to crack the cellulose.
  • A large fraction of their projected savings come from the increase in soil carbon derived from turning the land over to permanent grassland. In other words, they are crediting the cellulosic ethanol with the carbon sequestration created by planting the grass. But this sequestration would have happened anyway if the grass had been allowed to grow uncropped. This is unfairly favouring the use of land for making fuels.
  • Lastly, the lands that the researchers expect to use for switchgrass ethanol production are in the parts of the US most likely to suffer from desertification as a consequence of climate change. Large areas of the Great Plains were desert and may well become desert again within the next few decades. (Readers are referred to pages 5-7 of the UK edition of Mark Lynas’s wonderful book, Six Degrees, due out in the US in the next few weeks.)

Cellulosic ethanol is one of the great hopes for technology fixes that may help reduce our use of fossil fuels. But the reality is that the greenhouse gas savings may not be anywhere near as much as this research suggests. Similarly, it may not decrease the pressure on food production as much as its proponents hope.

Footnote [1] M. R. Schmer and others, ‘Net Energy of Cellulosic Ethanol from Switchgrass’, Proceedings of the National Academy of Sciences, 105.2 (15 January 2008), 464-9; [accessed 14 January 2008].