Insider Conversation: Can Biofuels Be Carbon Friendly?

moderates this blog, as well as Science's news Twitter and Facebook. Send him a nice email and he will solve all of your problems.

Tim SearchingerJohn Sheehan

Less than 2 years ago, Princeton agriculture expert Tim Searchinger published a paper in Science that sought to quantify how growing biofuels on cropland in the United States could lead to deforestation abroad. He estimated in some cases that indirect emissions could lead to a doubling of emissions associated with corn ethanol. Previously, researchers thought using the fuel could cut emissions by 30% since it would replace gasoline. Rarely do scientists have as immediate an impact on government policy. Since Searchinger’s paper was published, the U.S. Environmental Protection Agency, spurred by Congress, has been mulling whether to take so-called indirect land use into account when calculating the carbon footprint of biofuels for new regulations it is crafting, expected by December. Critics say Searchinger’s calculations were faulty and that uncertainties made it impossible to gauge their effects.

Right now, indirect land use related to biofuels isn’t included in proposed climate change legislation in the U.S. Senate, as well as proposed agreements that will be on the table in Copenhagen. In a recent policy piece published in Science, Searchinger and colleagues wrote that such a policy "erroneously treats all bioenergy as carbon neutral,” calling it a major "accounting error."

To discuss these issues, Insider conducted an e-mail conversation with Searchinger and John Sheehan, an expert on biofuels at the University of Minnesota, Twin Cities.


ScienceInsider: First for Tim: Can you explain the concept of indirect carbon impact? And for John: What do you make of Tim’s idea of accounting errors in his latest paper?
Tim Searchinger

All the potential benefits, and most of the potential costs, from bioenergy turn on land management. When cars burn biofuels instead of gasoline or diesel, what comes out of the tailpipe does not change. Similarly, when power plants burn wood instead of coal, what comes out of the smokestack is roughly the same amount of real carbon dioxide. The potential of bioenergy to reduce greenhouse gases results from the fact that growing plants absorbs carbon from the atmosphere. When plants are turned into fuel and then burned, the carbon released is just what the plants absorbed, potentially offsetting the emissions.

But that is just potential. It occurs, for example, if barren land is planted to crops because all that carbon taken up by the crops would otherwise be in the atmosphere. But when existing trees are burned for bioenergy, bioenergy takes carbon that would otherwise be stored and puts it into the air, just like burning coal or oil. And when forests are cleared to grow bioenergy crops, sure there is plant growth that takes up carbon, but that comes at the expense of not having ongoing forest growth that does the same and after losing much of the stored carbon in the forests. So bioenergy only reduces greenhouse gases if it results from additional plant growth or in some other way uses carbon that would not otherwise be stored (for example, by using the waste material left after timber harvest that would decompose rapidly anyway).

The problem with the treaties and laws limiting carbon dioxide from energy use is that they treat all bioenergy as carbon neutral. They implicitly assume that it all results from additional plant growth when in fact it often occurs by displacing stored carbon, such as forests.

The food problem is a subset of this issue. When existing corn is used for ethanol, what comes out of the tailpipe doesn't change and what is taken out of the atmosphere doesn't change either because the corn would be grown anyway. So the first question you have to ask is whether you get any additional carbon at all from this process and therefore reduce greenhouse gases. The reason that carbon in corn is not stored is that livestock and people consume it and burn it and put it back in the atmosphere. So it is wrong to assume that using crops gives you any direct benefit. Any benefit or cost depends on how it is replaced. If it is replaced by people eating less, you get a benefit. If it is replaced by other farmers boosting their yields more than they otherwise would, spurred by higher prices, you may get a benefit because higher yields absorb more carbon (but it comes at some greenhouse gas costs from fertilizer use and the like). But if corn is replaced by clearing other land to grow more food, then you have emissions. That does not mean farmers are penalized for land use elsewhere. It means they are not rewarded simply because their crops are used for fuel instead of food.

John Sheehan

I can state without qualification that I agree with the premise of Tim's recent article that there is a major flaw in current and proposed accounting systems for carbon. This is a flaw not just for biofuels. It is a flaw that creates a gaping hole in the global accounting of carbon. Broadly speaking, land-use related emissions of greenhouse gases are a large contributor to our annual release of carbon to the atmosphere.

The devil is in the details, however. Our ability to measure, let alone project, what the indirect land-use emissions for biofuels would be is sketchy at best. We are in the early days of scoping out how to do such analyses. And this is more than just an analytical problem. It is a genuine political problem. That is what makes the whole debate over biofuels so difficult. We not only have to sort out genuine analytic, economic, and scientific uncertainties about the interactions between biofuels and land-use change, but we have to recognize and deal with the political and ethical influences globally that lead to land-use change.

It's a simple matter of what I call "bathtub dynamics." When biofuels are burned, they reduce the amount of fossil carbon that would flow into the bathtub. If biofuels are produced by clearing forests and just dumping the aboveground carbon back into the atmosphere, then the bathtub fills up with carbon emitted from the sequestered carbon in the forest. Likewise, if biofuels cause other land to be cleared, there will be additional flow into the bathtub.

But if biofuels genuinely capture carbon beyond what was being captured on the land before the land was put into biofuels production, then there is a net savings. That's common sense. My biggest concern about the debate on biofuels is that we are not sufficiently focused on policies that encourage truly low carbon biofuels, and we publish results that focus on the worst case scenarios.

ScienceInsider: So there's a lot of uncertainty in greenhouse gas calculations on indirect emissions from biofuels. But with negotiations on emissions reductions, credibility is particularly important. What should negotiators in D.C. and in Copenhagen do about biofuels?
Tim Searchinger

Uncertainty means be cautious. Roughly a dozen major scientific assessments have now noted that because any use of productive land to produce biofuels has a high risk of creating large emissions through land-use change, we shouldn't be pursuing that route. Among these studies are those by our National Academy of Sciences, SCOPE, a special U.K. government review of biofuels called the Gallagher Report, Dutch reviews, and studies by the Joint Research Centre of the European Union.

To give you some idea of the risk, if palm oil grown in peatlands in Southeast Asia replaces only perhaps 5% of the soybean or rapeseed oil diverted to biodiesel, then the emissions from that decomposing peat alone cancel out any benefit from not using fossil fuels. And in reality, that percentage will probably be much more than 5%, and there will be lots of other land-use change as well to replace the other 95% of the vegetable oil. The alternative is to pursue biofuels that present little risk of significant land use change emissions. That means using timber and crop residues or trying to grow high-yielding grasses or trees on truly marginal and degraded land, and that is the recommendation of these many reviews.

However, the first policy that negotiators need to fix is the accounting error. This error applies not just to indirect land-use change but direct land-use change. According now to two separate modeling analyses published in Science, this error would lead to the loss of most of the world's natural forest because clearing those forests for bioenergy becomes one of the cost-effective means of complying with laws to reduce greenhouse gas emissions. Obviously, the emissions “cuts” these forms of bioenergy would achieve are not true in scientific fact. And the way to fix the accounting error is to count the very real emissions from using bioenergy and then provide a credit to that bioenergy which results from a source that really is "additional" carbon.

Turning to John's response, I appreciate his agreement and that the error is common sense. Yet I do not think we are focusing on worst case scenarios. Modelers working on this are doing the best they can to predict as accurately as possible what the likely consequences of various flawed policies, and for the most part they are noting the very real uncertainties. The reason using existing cropland for biofuels tends not to show up as yielding large reductions in greenhouse gas emissions is simply because those croplands are already absorbing large quantities of carbon. An Iowa corn field today, whatever critics say of its environmental impacts, is an extraordinary technological achievement and produces more food on an acre than almost any acre known in history—certainly any nonirrigated acre. Using that crop for fuel is not only unlikely to yield much greenhouse gas benefit, but much of that "benefit" for greenhouse gases may come at the expense of reduced food consumption (which would be good if that came from my plate but not when it comes from the plates of billions of poor, who unfortunately have less money to buy it than I do).

John Sheehan

In the face of uncertainty, Tim is invoking the precautionary principle with respect to the risks of aggravated land-related greenhouse gas emissions associated with biofuels. But the risk equation has two sides to it. There are risks to falling prey to what I call "paralysis by analysis." One of the studies Tim has referenced basically concludes that the only thing policymakers can do is work on the assumption that biofuels, at this moment in time, are worse than petroleum fuels because of the future risk of land clearing and other GHG emission increases. In that case, we do nothing and suffer the consequences of dealing with a burgeoning demand for liquid transportation fuels globally without offering any alternatives to petroleum.

Policymakers certainly have my sympathy. Never before have we faced such complex issues. From a technical point of view, we are managing and monitoring our land at best like someone who is trying to drive a car by looking in the rear-view mirror. And that is the most positive spin I can put on it. In reality, our ability to see where we are going with sustainable land management is more like trying to drive blindfolded.

To carry the car analogy even further, the lack of jurisdictional infrastructure for international land management means we are trying to drive with a broken steering wheel. Ultimately, the political problem that regulators face here in the U.S. as they struggle with the California low-carbon fuel standard and the revised EPA Renewable Fuel Standard is that they are trying to regulate biofuels producers with respect to indirect changes over which they (both the regulators and the biofuels producers) have little control.

I would encourage policymakers to adopt flexible policies that take different approaches to dealing with direct versus indirect land-use effects. We know and understand direct emissions far better than we do indirect effects. So, tough and clear hurdles for reduced direct emissions are appropriate. On the indirect land-use side, policymakers should encourage mitigating or offsetting strategies from biofuels producers that help to address external land-use problems while we try to fix the international policy problems.

Finally, as to the question of whether modelers are focusing too much on worst case scenarios, this is a point on which Tim and I probably just disagree. This is where we need more open and transparent discussion of assumptions and approaches. Frankly, the models we are using now are too complex and impossible to understand.

Tim Searchinger

I do not believe it is logically or practically legitimate to distinguish direct from indirect land use.

The problem is that life-cycle analyses for biofuels literally assign a carbon credit to the biofuel that cancels out all the emissions from burning the fuel in the car. But that credit is invalid because the carbon in the plants is not necessarily additional. Or put another way, as I have said, just switching the use of already existing carbon from food to fuel does not result in any direct net gain. You have to examine indirect effects to determine if there is a gain or a loss.

Administratively, regulating direct but not indirect land use would have largely meaningless real-world consequences. For example, nearly all the world's palm oil today is used for vegetable oil from plantations previously carved out of forest. If the law limits direct land-use change, a palm oil producer cannot clear a new acre to make palm oil and sell that for biodiesel. But the company can take all the existing palm oil from already cleared forests and sell that for biodiesel, and then clear new land to make more palm oil and sell that for food. This approach doesn't do you any good.

John Sheehan

Separating direct and indirect effects is far from an ideal solution to proper regulating of carbon emissions from biofuels. But it may be the only thing we can do in the interim. I think Tim and I have a different understanding of what I mean by "direct emissions." These direct estimates of carbon emissions can indeed be designed to account for diverting land currently in food production to biofuels production. Such analyses do not require making the erroneous assumption of carbon neutrality for biomass.

And, if I had limited my proposal [in my last response] to just direct land-use regulation, I would agree with Tim that this would not be meaningful with respect to preventing international "shell games" of the sort he describes for palm oil. Such approaches are at the heart of what is wrong with the current European focus on accepting biodiesel only made from previously cleared land. My point was that regulators should be flexible in reducing or eliminating the indirect carbon emission penalty based on offset strategies offered by the biofuels producers. In the absence of any sound policy controls on international land use, I think this at least offers a way forward.

We have enough areas of agreement that I believe we can begin to focus on real solutions to both our energy security and climate change problems and that these solutions will involve, at least to some extent, biofuels. There are some transportation needs that simply must have a low-carbon liquid fuel alternative. Tim has pointed out that there are probably "safer" options for biofuels that minimize the risk of indirect land use. We should pursue those. But I think it would be a mistake if we narrowed the field to just [making fuel out of agricultural] residues and wastes a priori. Finally, I think we need to broaden the discussion of biofuels beyond the question of carbon emissions. After all, carbon emissions are just one lens through which to look at our global land-use problems. Biofuels have done us all a big favor by turning a spotlight on what my colleague Jon Foley refers to as the "other inconvenient truth": land use.

Posted in Climate, Policy