The Nexus of Biofuels, Climate Change, and Human Health: Workshop Summary (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

4 Biofuels Impact on Air, Atmosphere, and Health One of the major reasons for encouraging the use of biofuels has been the positive effects their use is expected to play in reducing green- house gases and also air pollutants, with concomitant improvements in health. The speakers in the workshop’s fourth session offered details on how the production and use of biofuels should affect greenhouse gas levels, air quality, and health. BIODISTALLATE FUELS AND EMISSIONS In the first presentation, S. Kent Hoekman, research professor in the Division of Atmospheric Sciences at the Desert Research Institute, discussed biodistillate fuels and emissions. The term biodistillate is a more general term than biodiesel, he explained, and it includes not only biodiesel but also related biofuels. Background Hoekman began by offering some basic background on biodiesel and other biodistillate fuels, beginning with the drivers. “Why are we interested in biodiesel? I think today the simplest and most direct answer is because it’s the law.” In particular, the U.S. Environmental Protection Agency (EPA), in implementing several congressional mandates, has requirements for the use of renewable fuels divided into conventional biofuels, cellulosic biofuels, biomass-based diesel, and other advanced biofuels. The ultimate drivers for the use of biofuels—that is, the reasons behind the political decision to impose requirements for the use of biofuels—include concerns about greenhouse gas emissions, the desire to 47 

48 THE NEXUS OF BIOFUELS, CLIMATE CHANGE, AND HUMAN HEALTH develop renewable or sustainable energy sources, the desire to develop secure domestic fuel supplies, and an interest in rural development. Interestingly, Hoekman said, neither air quality concerns nor health concerns have been major factors in the push to increase the use of biofuels. “They are somewhat important, but they have not been the main drivers.” Terminology Next, Hoekman went over some basic terminology related to biofuels in order to clarify exactly what is meant by various terms. According to ASTM International, biodiesel fuel refers to “mono-alkyl esters of long-chain fatty acids derived from vegetable oils and animal fats.” The term can also refer to trans-esterified triglycerides or to fatty acid methyl esters (FAMEs), which are both closely related to ASTM’s definition. Biodiesel fuel is sometimes referred to as B100. Renewable diesel is produced from the same feedstocks as biodiesel, Hoekman said, but it is produced through hydroprocessing technologies so that the product is a hydrocarbon (HC), not an ester. It is also referred to as “green diesel.” Co-processed renewable diesel is a form of renewable diesel that is produced by adding vegetable oils or animal fats to feedstocks that are being hydrotreated to produce diesel fuel, creating a single product that is a mixture of bio and fossil HCs. Cellulosic biodiesel fuel, or synthetic biodiesel, is produced by pyrolysis or gasification of lignocellulosic feedstocks, such as grasses and woods. The resulting liquid generally requires rather considerable additional processing or upgrading before it can be blended into petroleum fuel stock. Biodistillate Production Technologies A variety of different production methods are used to produce the different types of biodistillates. Hoekman illustrated them with a single figure that showed the feedstocks, processing methods, and resulting fuels (see Figure 4-1). Hoekman pointed out that, as the figure indicates, there are many different types of fats and oils that can be used to produce biodistillates. “And this is an abbreviated list shown here,” he said. The line at the top of the figure represents the traditional biodiesel production pathway that uses methanol in a trans-esterification pathway to produce biodiesel and 

BIOFUEELS IMPACT ON AIR, ATMOSPHEERE, AND HEALLTH 49 glycerin. “We haven’t heard a lo ot about that,”” he said, “butt glycerin is tthe main byproduct of o biodiesel production. About one-ttenth as muuch glycerin is produceed as biodiessel.” None off the rest of the production pathwaays produces glycerin, he noted. Furthhermore, he nnoted that othher than thhe biodiesel produced p in that t first prodduction pathwway, the rest of the fueels produced are a HCs. It is only o biodiesel that is oxygennated. Att present, he said, biodiesel remains bby far the m most commonnly producced biodistillaate. In the Un nited States, ssoy oil is the main feedstoock used to produce biiodistillates, withw some w waste cookingg oil, sunflow wer oil, and other oils used u as well. In I Europe, thee main feedstoock is rapeseeed, while ini much of thhe rest of the world w palm o il is the domiinant feedstocck. In the United States S producction of bioddiesel is at ann all-time higgh, with more m than 1 billion b gallonns produced iin 2011 and 2012, up froom next too nothing a decade d earlierr. To put thatt in context, Hoekman saiid, the tottal U.S. petrolleum diesel fuel fu productioon is about 600 billion galloons per year, and gasoliine is about double that am mount. FIGUR RE 4-1 Biodistiillate productio on technologiess beginning witth feedstock useed, processsing technology y used, fuel produced, and chhemical type. NOTE:: ag = agricultu ural, FAME = fatty f acid methyyl ester, F-T = Fischer-Tropscch, HC = hydrocarbon, h H2 = molecular hydrogen, MS SW = municipaal solid waste. SOURC CE: Hoekman,, 2013. 

50 THE NEXUS OF BIOFUELS, CLIMATE CHANGE, AND HUMAN HEALTH Biodistillate Properties and Composition Biodiesel and renewable biodiesel differ from petroleum-based diesel in a number of ways. One of the most important is the presence of oxygen. Neither petroleum diesel nor renewable biodiesel contains oxygen, while biodiesel is roughly 11 percent oxygen by weight. Another important difference can be found in the energy content of the different fuels. Petroleum diesel has a high energy content of 130,000 BTU per gallon. Biodiesel is 6 to 7 percent less—121,000 BTU per gallon for biodiesel and 122,000 BTU per gallon for renewable biodiesel. With respect to the chemical composition of the various types of diesel fuel, two critical factors influence the physical properties and performance attributes of the fuels, including their emissions. The first is the length of carbon chains in the molecules of the fuels. In conventional diesel the chains are typically 12 to 24 carbons long, although some molecules are somewhat shorter or somewhat longer. Biodiesel, being made from fatty acids, tends to have molecules with carbon chains that are 16 or 18 carbons in length. The second important factor is the degree of unsaturation, which is, roughly speaking, a measure of how many fewer hydrogen atoms a molecule with a certain number of carbon atoms has than it could have if the carbon atoms were arranged to maximize the number of hydrogen atoms in the molecule. The degree of unsaturation is important, Hoekman explained. Having too much unsaturation makes for an oxidatively unstable product, while having too little unsaturation results in a product with poor low-temperature performance—that is, it tends to “wax up” when the temperature drops. Compared to biodiesel, conventional diesel has lower unsaturation overall, and it has more branching of the HCs, he said. “Those are important for physical and chemical properties.” Different oil feedstocks lead to biodiesels with different chemical compositions. For example, soybean oil is dominated by linoleic acid entities, consisting of an 18-carbon chain with two double bonds. “That’s rather highly unsaturated in terms of fuel stability,” Hoekman said. By contrast, rapeseed is mainly oleic acid, a molecule with an 18-carbon chain but only one double bond, so it is not so highly unsaturated. Emissions Standards and Controls Diesel engine and vehicle emissions are regulated by the EPA as well as by some states, most notably California, Hoekman said. Different 

BIOFUELS IMPACT ON AIR, ATMOSPHERE, AND HEALTH 51 sets of standards are defined for different applications and purposes. For example, there are different emissions standards for different engine sizes—light-duty, medium-duty, and heavy-duty application—and, in fact, each of these categories has subsets with their own sets of standards. There are also different standards for on- and off-road applications. Off- road applications make up a significant part of diesel fuel usage, he said, and they include railroads, mining, and farming. Historically, four different types of emissions have been regulated for diesel uses: HC, carbon monoxide (CO), nitrogen oxide (NOx), and particulate matter (PM). Of those four, Hoekman said, the latter two have been of the greatest concern and have been under regulatory scrutiny for the longest time, principally for reasons related to air quality. The emissions standards are not static sets of numbers, he said. They have been steadily evolving. In particular, emissions standards have become much more stringent during the past 25 years, and the maximum allowable emissions of NOx and PM have been reduced by almost two orders of magnitude during that time. For example, in the late 1980s, NOx was regulated at 10.5 grams per brake horsepower-hour; today, the standard is 0.2 grams per brake horsepower-hour. Similarly, the standard for PM went from 0.6 to 0.01 grams per brake horsepower-hour. Those emission standards apply only to new engines and vehicles, he noted. They are not automatically applied to fleet vehicles already in use, and fleet turnover is very slow, particularly for heavy-duty diesel vehicles. The large reductions in emissions required by the standards have been achieved by a combination of engine improvements and improve- ments in emission control systems. Engine improvements have included the adoption of high-pressure, common-rail fuel injection; variable injection timing; and electronic monitoring and control systems. A recently developed emission control system used to reduce particulate emissions is the particulate trap. Particulate traps require regeneration, Hoekman noted, and there have been some issues regarding the regeneration of those traps. To control NOx, engines are now being built with selective catalytic reduction systems. These use urea injection to reduce NOx to molecular nitrogen. Those are significant changes that have taken place in just the past couple of years, he said. Another change that has made it possible to dramatically reduce emissions has been the introduction of ultra-low-sulfur diesel (ULSD). “The primary reason for having that is to enable satisfactory long-term operation of those sophisticated emission control systems,” Hoekman 

52 THE NEXUS OF BIOFUELS, CLIMATE CHANGE, AND HUMAN HEALTH said. “It’s analogous to getting the lead out of gasoline so that catalytic control systems can function properly.” Effect of Biodiesel on Engine Emissions The traditional understanding of how using biodiesel in an engine affects tailpipe emissions comes from a 2002 EPA draft report (2002) that was “rather famous but never officially published in a final version,” Hoekman said. That report showed that as the blending level of biodiesel is increased from B0 (0 percent) all the way to B100 (100 percent), there are significant reductions in HC, CO, and PM emissions, while there is a slight increase in NOx emissions. “That increase in NOx has been a source of tremendous controversy in a lot of studies over the years,” Hoekman said. In most real-world applications, he noted, biodiesel is used at low concentrations, usually B5 to B20. So, it is in that range that it is most important to understand what happens to emissions. Recently, the Desert Research Institute, working on behalf of the Coordinating Research Council, conducted an updated literature review in order to examine more recent and comprehensive information concerning the effects of using biodiesel on engine emissions. The review, which Hoekman was a part of, examined more than 1,000 literature sources and analyzed the data with various sorting and statistical analysis methods. Focusing just on the data for NOx emissions, Hoekman noted that there was “tremendous scatter” in the data. That is, there was no smooth curve that could describe what happened to the emissions as the percentage of biodiesel increased from 0 to 100 percent; instead, the data points were scattered all over the graph. This was not particularly surprising, he observed, because the data were from a wide range of literature sources. The various studies were done under a very wide range of conditions, with many different forms of biodiesel blended into different base fuels, and many other differences as well. And the data scatter was just as large for HC, CO, and PM emissions, he noted. Still, it was possible to discern a trend in the NOx emissions. The data showed an upward trend in NOx as the percentage of biodiesel increased in the category of heavy- and medium-duty engine dynamometer emissions. However, the trend was the opposite for the category of heavy- and medium-duty chassis dynamometer emissions. The main difference between the tests lies in where the dynamometer is mounted during the test—coupled directly to an engine that is independent of a vehicle, 

BIOFUELS IMPACT ON AIR, ATMOSPHERE, AND HEALTH 53 versus coupled to the power train of a vehicle through the drive wheel or wheels without removing the engine from the frame of the vehicle. This difference in observed NOx effects, depending on the testing methodology used, is one illustration of why it is so difficult to determine the “true impacts” of fuel changes on engine emissions when applied across the entire vehicle fleet, Hoekman said. The study found significant decreases in HC, CO, and PM emissions with increasing percentage of biodiesel, which were in “reasonably good agreement” with the earlier EPA findings. “So, I think this is probably the best idea you can get as to the impact of biodiesel use on emissions across the whole fleet,” Hoekman said. Emissions from B20 Fuel Hoekman then focused specifically on the issue of emissions from B20 fuel because “that’s the upper end of the most common range of biodiesel usage levels.” He discussed a study by Robbins and colleagues (2011), in which the results were broken down by engine class: medium- and heavy-duty engine, medium- and heavy-duty engine on a chassis, and light-duty engine. Overall, there were large reductions in HC, CO, and PM emissions, with a slight increase in NOx emissions, although there was a large scatter in the results. The results in the study by Robbins and colleagues (2011) (HC, –17.4; CO, –14.1; PM, –17.2; NOx, +1.8) generally agreed quite well with the results from the earlier EPA study (EPA, 2002) (HC, –21.1; CO, –11.0; PM, –10.1; NOx, +2.0). Given that the review involved more than 1,000 individual studies, Hoekman said, it was possible to sort the studies according to various criteria, including the feedstock used for the biodiesel (soy oil, rapeseed oil, yellow grease, palm oil), the base fuel into which the biodiesel was mixed (No. 2 diesel, ULSD, California Air Resources Board [CARB] certified diesel), the engine year (as a proxy for the certification levels for the emissions and, particularly, the NOx certification level), and the test cycle load (light, medium, heavy). When Hoekman and his colleagues examined how these different criteria affected the emissions levels from the various biofuels, they found that there was so much data scatter in the results that it was difficult to detect significant effects across the whole fleet. To illustrate, he showed graphs of how B20’s effects on emissions varied by the type of feedstock used, the type of base fuel, the engine year, and the test 

54 THE NEXUS OF BIOFUELS, CLIMATE CHANGE, AND HUMAN HEALTH cycle load. In each case, the error bars were larger than the effect sizes, so it was impossible to conclude that any of these factors had an influence on how B20 affected emissions. Moving on, Hoekman spoke briefly about mobile source air toxic (MSAT) emissions from biodiesel. There are dozens of MSATs, but those of greatest interest with respect to biodiesel are polycyclic aromatic HCs, aldehydes (formaldehyde, acetaldehyde, proprionaldehyde, and acrolein), and the total PM discussed previously. Oxygenated organics, such as biodiesel, might be expected to produce higher levels of oxygenated MSATs, Hoekman said, but there is very little relevant experimental data that address this issue. The existing data suggest that the use of biodiesel does not consistently increase emissions of these MSATs, he said. Hoekman concluded his presentation with a number of general observations about biofuels emissions. First, he said, although biodiesel— the FAME version—is currently the dominant form of biodistillate being produced, he believes that in the future the nonoxygenated biodistillates are likely to grow in use and perhaps even become the dominant form of biodistillate. Recent reviews of the biodiesel literature have confirmed what the EPA and others have been saying for many years—that the use of biodiesel reduces emissions of HC, CO, and PM while increasing NOx emissions by a small amount. Although data on the emissions of renewable biodiesel—that is, the non-oxygenated, HC biodistillates—are sparse, it does appear that renewable diesel provides emission reduction benefits that are just as big as, if not bigger than, those from biodiesel. Exhaust emission standards for diesel engines and vehicles have become much more stringent during the past 25 years, which has resulted in the development of advanced emission control systems that reduce emissions dramatically, much more so than a change in the fuel composition to include biodistillate fuels. Determining the effects of fuel-type fleet-wide emissions is difficult because of the variability of engine and vehicle types, test cycles, emissions control systems, and other factors. The variability in the data prevents drawing firm conclusions about the effects of biodiesel feedstock, base fuel type, the engine model year, or the test cycle on diesel emissions when using B20. In the case of aldehyde emissions, although the data are sparse, the use of biodiesel does not appear to affect the emissions in a consistent or significant way. The effects of biodiesel 

BIOFUELS IMPACT ON AIR, ATMOSPHERE, AND HEALTH 55 on polycyclic aromatic HC emissions are hard to ascertain, but the few data that exist suggest little effect, if any. Finally, Hoekman offered a recommendation. The various advanced diesel emission control systems, such as the selective catalytic reduction device and the particulate trap, have been in use for only a couple of years, he reiterated. “I believe that additional research and study monitoring is needed to assess the long-term effects of biodiesel and its impurities on the performance of those systems. If those systems fail over a shortened lifetime, that would have significant effect.” Discussion Following Hoekman’s presentation, there was a discussion period devoted to just his talk. The first question, which came from an audience member, was whether any work had been done to study the effect that contaminants in biofuels might have on health. In particular, the question concerned biofuels produced from things such as frying oils used for french fries and other foods, which could have a variety of contaminants. Hoekman replied that he was not sure if any data exist concerning the health effects of such contaminants. However, he noted that there is a rather long and stringent list of specifications for biodiesel fuel—as there is for petroleum-based diesel—because the diesel engines are expected to run for half a million miles or even longer on these fuels without breaking down. “The feeling is if there are excessive levels of salts or metals, they may impede the performance of emission control systems,” he said. “They may use some of the capacity of the trap, thereby reducing its overall efficiency.” In a follow-up question, Hoekman was asked about the difference between more versus less unsaturated feedstocks for biofuels. Plant- derived biofuels can vary in how unsaturated they are, depending on which plants they are derived from. So the question was whether that difference could lead to a difference in emissions from the biofuels. Hoekman answered that there has been quite a lot of work looking at how the extent of unsaturation affects total emissions and, especially, NOx emissions. The results have been somewhat equivocal, he said, but there is some evidence that the higher the unsaturation, the more NOx emissions there may be. However, he added, “what’s much more important with respect to unsaturation is the physical property—the oxidative stability of the fuel. Can you keep it out in the marketplace? Is it an acceptable fuel regardless of what happens when you burn it?” 

56 THE NEXUS OF BIOFUELS, CLIMATE CHANGE, AND HUMAN HEALTH In response to a question about whether automotive manufacturers are willing to warranty engines used with a biodiesel mix, Hoekman said that his understanding is that most manufacturers of heavy-duty engines now accept up to B20. In fact, he said, the international standards organization ASTM has developed standards for biodiesel in the range of B6 to B20. “So, in that range, I believe biodiesel is accepted by all the U.S. heavy-duty engine manufacturers,” he said. However, he did not know whether manufacturers of light-duty vehicles have yet reached the same point. REGIONAL IMPACTS OF BIOFUELS ON HEALTH AND CLIMATE CHANGE IN BRAZIL In the next presentation, Elliott Campbell, an assistant professor at the School of Engineering at the University of California Merced Energy Research Institute, moved from the micro-level issue of tailpipe emissions to the macro-level question of how the use of biofuels might affect health and climate change on a regional scale. To begin, Campbell noted that much of the current discussion related to the effects of biofuels on climate change concerns those changes on the global scale. It concerns issues such as carbon cycle in the use of biomass feedstock and land-use change. But there is also emerging interest in examining the effects of biofuels use at a regional scale—for example, in studying how the broad use of E85 (a fuel with 85 percent ethanol and 15 percent gasoline) might affect air quality in a region such as Los Angeles and surrounding areas (Jacobson, 2007) or looking at the regional climate impacts from the widespread use of second-generation cellulosic biofuels products (Georgescu et al., 2011). Noting that there has been quite a bit of this sort of work done on the regional scale in the United States, he said that his talk would be focused instead on some emerging analysis of the regional climate and health impacts of biofuels production in Brazil. Background He began by providing some basic background on biofuels in Brazil. As can be seen in the top section of Figure 4-2, the consumption of liquid fuels has steadily increased during the past decade, as has the production, and in recent years production and consumption have been approximately 

BIOFUELS IMPACT ON AIR, ATMOSPHERE, AND HEALTH 57 equal. As can be seen in the bottom section of the figure, hydroelectricity produces a large majority of the country’s electricity, with fossil fuels a distant second. Nuclear power accounts for a very small percentage of the total electric power and is a smaller proportion than renewable energy sources other than hydroelectricity and nuclear. Two energy- related concerns in Brazil are the export of biofuels and diversifying local electric power production. Brazil produces large amounts of sugarcane whose sugar is used in the production of ethanol, and one issue of importance to Brazilian policy makers is what to do with the parts of the sugarcane plant that are not converted into ethanol. At present, Campbell said, about half of the sugarcane crop is subjected to pre-harvest burning, which makes the harvesting process much easier and less expensive and also returns nutrients to the soil. However, it produces massive amounts of air pollution during the time of the pre-harvest burning, and it is also wasting a large amount of energy that could be captured and used elsewhere. Reducing the amount of pre-harvest burning would lead to much larger quantities of available sugarcane residue, which could be turned into energy in two ways: it could be burned in electricity-generating plants, or it could be turned into cellulosic ethanol. Converting the residue into electricity has greater greenhouse gas benefits than using the residue to pro- duce ethanol, he said, and, furthermore, converting the residue into electricity could have a massive impact on Brazilian energy security. The Level of Direct Emissions After providing that brief overview, Campbell took a more careful look at the emissions caused by the field burning of sugarcane (Tsao et al., 2012). One way to understand these direct emissions, he said, is to use a bottom-up approach that combines emissions factors from the GREET (Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation) model from Argonne National Laboratory, maps of sugarcane production, and basic conversion factors generated from other agronomic and life-cycle assessment data. When the calculations are done, they show that field burning does indeed release massive amounts of various gases: CO, volatile organic compounds, NOx, PM, and carbon dioxide. So, field burning is a very important component of the direct emissions associated with the production of biofuels, he concluded. 

58 THE NEXUS OF BIOFFUELS, CLIMATEE CHANGE, AND D HUMAN HEAL LTH a b FIGURRE 4-2 Energy in Brazil: (aa) liquid fuelss production aand consumptiion (2002– –2014); (b) elecctricity generattion, by fuel typ ype (2001–20111). SOURC CE: EIA, 2012 2. 

BIOFUEELS IMPACT ON AIR, ATMOSPHEERE, AND HEALLTH 59 Duuring the pastt decade, he said, s the emisssions of variious gases haave grown n despite the move m to mecchanization, aand he illustraated this withh a figure that showed d emissions over time in tthe Brazilian state in whiich most of o the sugarcane producttion takes pllace (see Figgure 4-3). Thhis increasse in emissio ons is due to o the expanssion of the aareas in whiich sugarccane is grown,, so that even with growingg mechanizatiion, the amouunt of fieldd burning hass been increassing. Reesearchers haave used vaarious methoods to estim mate the direect emissiions from fielld burning, Campbell said.. One such m method has beeen remotee sensing from satellites. This is a connvenient metthod becausee it makes it possible to o get estimates from many different suggarcane-growing regionns with relativ dditional efforrt over what it takes to gget vely little ad estimaates from just one area. However, H thee top-down ddata from suuch satellittes have provven to significcantly underesstimate the em missions verssus the botttom-up estim mates. “O One importantt question is ‘Why‘ the diffference?,’ beccause we wouuld love too be able to use u the remotee sensing dataa,” Campbell said. There aare a numb ber of factorss at play. One is that the remmote sensingg comes up wiith signifiicantly differeent estimates for the size oof the burnedd areas than tthe bottomm-up approach. There are also major diifferences in tthe estimates of FIGUR RE 4-3 Estimaated life-cycle emissions e of etthanol in Braziil from crop yeear 2000 to o 2008 (crop yeear is from Aprril to January tthe following yyear). NOTE: (a) Emission ns of volatile organic comppound (VOC), nitrogen oxiide (NOx),, particulate matter m less than n 10 microns inn diameter (PM M10), particulaate matter less than 2.5 microns in diameter d (PM22.5), sulfur oxxide (SOx), aand carbon monoxide (C CO). (b) Emisssions of blackk carbon (BC),, organic carbbon (OC), and a carbonyl su ulphide (COS), which are estim mated only forr the field-burniing phase. MT = metric toons. SOURC CE: Tsao et al., 2012. Reprintted with permisssion. Copyrighht 2012 Americcan Chemiccal Society. 

60 THE NEXUS OF BIOFUELS, CLIMATE CHANGE, AND HUMAN HEALTH the fuel load (the amount of biomass per unit area) and the emission factors (total emissions per kilogram of biomass). Regional Health and Climate Impacts With a growing understanding of the emissions from pre-harvest burning, the next question is, What might the health and climate impacts of these emissions be? To answer that question, Campbell said, one begins by trying to understand what the change in air quality is—not just how the emissions change, but how atmospheric species change in concentration and how those concentrations vary in space and time. The next steps in quantifying the human health effects are to determine the exposed populations, estimate the health effects, and determine a health baseline incidence. “You can try to gather these data at a variety of spatial scales in Brazil,” he said, including city-level, province-level, or country-level data. There are a number of studies that estimate the health effects of various levels of atmospheric pollution, Campbell noted, “and the change in the atmospheric concentration comes from these regional atmospheric models that I talked about previously.” To illustrate, he showed a map of PM levels in Brazil in January, during the sugarcane growing season, and in May, when the pre-harvest burning takes place. Combining models of this sort with estimates of the health effects for various levels of PM in the atmosphere, it is possible to derive estimates of the health effects of the pre-harvest burning of sugarcane based on the following approach: health effect = (air quality change) × (exposed population) × (health effect estimate) × (health baseline incidence). In looking at annual mortality changes, “you get somewhere between 20 and 4,000 deaths per billion gallons of ethanol,” Campbell reported. To put these numbers into perspective, it is helpful to review a study that compared deaths associated with operating gasoline-fueled vehicles versus those fueled with a blend of 85 percent ethanol (E85) based on modeling for 2020. Jacobson (2007) found that E85-fueled vehicles increase ozone-related mortalities by about 185 deaths per year, which corresponds to a 4 percent increase over the U.S. projected death rate of operating gasoline vehicles. Campbell noted that there is obviously much uncertainty in the exact number of deaths from the air quality effects of the pre-harvest burning, but from this preliminary analysis there appears to be significant potential for the health impacts to be quite large. 

BIOFUELS IMPACT ON AIR, ATMOSPHERE, AND HEALTH 61 Estimating climate effects caused by the emissions from pre-harvest burning is even more difficult. “It requires advanced climate modeling,” Campbell said, “but if you use climate forcing factors based on emissions for black carbon from the fields and from the boilers at the ethanol refineries, the climate forcing per unit energy of ethanol for sugarcane can increase from what we think it is now to something that may exceed regulatory thresholds.” Indirect Emissions In addition to the direct emissions from pre-harvest burning, there are also indirect emissions caused by indirect land-use change, meaning, in essence, a change in the use of land from forest or some other non- cultivated land to cultivated cropland. Such land-use change generally involves the clearing of forest or other land, which in turn involves cutting down and burning trees or other vegetation, which releases carbon and other elements into the atmosphere. Calculations show that rangeland converted to biofuels production in Brazil led to a significant amount of PM released into the atmosphere, Campbell said, but that the conversion of forest to rangeland accounted for far more emissions. When the indirect emissions from indirect land-use change are included in the calculations for total emissions in the life-cycle of sugarcane ethanol, a very different picture emerges. For PM (in particular, PM2.5, which refers to particles less than 2.5 micrometers in diameter), including indirect land-use change in the calculations may nearly double the estimated emissions due to the production of biofuels. To put this in perspective, Campbell compared the indirect emissions of PM from biofuels production to the emissions of PM caused by Amazonian deforestation. “Adding a billion gallons of ethanol is potentially on the order of all the emissions from deforestation in roughly the last decade,” he said. In summing up his presentation, Campbell offered the following takeaway messages:  The emerging trade regime for biofuels, with Brazilian biofuels being exported to developed nations, presents important “leakage” challenges with respect to regional health and climate impacts.  Previous burning estimates may underestimate the burned area by a factor of four. 

62 THE NEXUS OF BIOFUELS, CLIMATE CHANGE, AND HUMAN HEALTH  Sugarcane regional health impacts are potentially much larger than those of other biofuels, although a great deal of work remains to produce better estimates.  The regional climate impacts from biofuels production may mean that sugarcane ethanol, instead of providing a significant reduction in climate impacts relative to the fossil fuels that it replaces, is actually causing climate warming, at least at a regional scale. Finally, Campbell offered several research recommendations:  Regional climate and health impacts research should focus on Brazil, given the potential for relatively high impacts there.  The critical research gaps that should be addressed include top- down studies of burned area (in order to resolve the difference from bottom-up estimates), observation-based emissions factors (because they vary widely), and three-dimensional atmospheric modeling.  Integrate research and policy to address leakage issues for regional impacts (e.g., air quality, aerosol forcing) in addition to the current focus on global impacts (e.g., carbon dioxide). Discussion Campbell’s presentation was followed by a discussion period. Christopher Portier, director of the National Center for Environmental Health, Centers for Disease Control and Prevention, began by asking whether Campbell’s climate forcing model with which he examined the effects of field burning included carbon black. It only included carbon, Campbell replied, which is one of the weaknesses of that analysis. There have been some very simple ballpark estimates of the climate forcing of carbon black that have been public, he said, and applying those estimates implies that field burning “can potentially exceed some thresholds for life-cycle greenhouse gas emissions.” However, Campbell said, what is really needed is to run a regional climate model “because the climate forcing from these species varies so much depending on the domain, the time of year, the timing of the emissions, all of these kinds of factors.” Thus, the jury is definitely still out on the effects of these carbon black emissions. Carlos Santos-Burgoa from the Pan American Health Organization asked whether there were any changes that could be made to the sugarcane ethanol production process that would improve the emissions. 

BIOFUELS IMPACT ON AIR, ATMOSPHERE, AND HEALTH 63 One approach would be to continue reducing the amount of pre- harvest burning, Campbell said. There has been a pretty dramatic shift over the past decade away from pre-harvest burning and toward mechanization. “Brazil has a voluntary program that’s trying to move toward those better cultivation approaches,” he said, “but it’s unclear what the future trends would be if sugarcane cultivation expanded rapidly to try and meet demand for export of biofuels.” Roundtable member Bernard Goldstein referred to Campbell’s statement that the residue-based ethanol made in Brazil and shipped to the United States has little impact on the energy security of the United States but would have massive impact on Brazil’s energy security if it were not exported. Goldstein then suggested that economists would likely respond to this situation by saying that energy security is not being priced appropriately. Al McGartland responded that the EPA does include an analysis of energy security in the ethanol regulations when it sets a mandate. There are a variety of security benefits to ethanol production, he said, including making the economy less vulnerable to the price of oil and potentially decreasing military expenditures in the Middle East if imported petroleum is less vital to the national interest. The value of these security benefits “is not a trivial number” for the United States, he said, although he was not familiar with the case of Brazil. Goldstein then asked if it is correct that a drop in the price of oil would mean there would be less of a push for biofuels. McGartland replied that this is the case. Right now, biodiesel would not be made without a mandate because it costs more than petroleum-based diesel, but if the price of regular diesel fuel went up enough, biodiesel would be made without a mandate. Goldstein next asked how to get air pollution and other health issues considered more in the discussion concerning biofuels. Stephen Reynolds suggested that life-cycle assessments (LCAs) might offer a way in. “The whole concept of life-cycle assessments has become pretty popular these days,” he noted. “In fact, it’s mandated by certain regulatory agencies that some form of LCA must be done to evaluate the broader impacts of fuels or other technology introductions.” These LCAs have tended to focus more on greenhouse gases and energy balances and not so much on “mobile-source air toxics,” Reynolds said. So, one place to start getting more attention to health issues in the discussion on biofuels would be to work to get more consideration of toxic emissions in LCAs. 

64 THE NEXUS OF BIOFUELS, CLIMATE CHANGE, AND HUMAN HEALTH An audience member elaborated on the importance of persuading decision makers—and especially the economists who advise them—that air quality needs to be taken into account as a serious policy concern. “I think that air quality is one of a class of health-related issues that traditionally have been marginalized,” he said, “in part because although economists will pay lip service to issues of health and productivity, I’m not sure that the current economic theory really believes it.” The result is that economic models often minimize or leave out health considerations. If people are removed from a labor market that is already glutted with free labor, for example, that does not have much overall economic impact. Thus, more work needs to be done to health and productivity and economic development. “Some progress is being made in health and productivity studies of the type that are being done in industry to relate health promotion to productivity,” he said, “but that so far has been very narrow and very rich-country-oriented. I think we need a far more robust economic approach.” Another audience member immediately challenged those comments. “With all due respect to the last speaker, most of what he said is really quite wrong from an economics perspective. The environmental economics community . . . would never think about valuing health in the way that was described. We know that it is absolutely wrong to think about the benefits from loss of life or illness as something as simple as lost life years or lost productivity. It’s wrong from an economic analysis perspective. It feels in-the-gut wrong, and it is wrong. So, I don’t want to give you a big long lecture, but that notion is really not an accurate depiction of the state of the field.” In reality, the commenter said, the benefits of improved human health are very well understood as belonging in cost-benefit analysis, and there are many studies doing this. Goldstein offered a relevant anecdote. Forty years ago, as a young investigator at the U.S. Office of Management and Budget (OMB), he gave a talk about the effects of sulfur dioxide and mentioned the infants who had died in the London great smog event of 1952. “I was asked by the OMB economists whether they were male infants or female infants, and when I looked horrified, and . . . asked why would that be a question, it was pointed out that females did not contribute to gross domestic product, but males did. So, OMB has come a long way.” 

BIOFUELS IMPACT ON AIR, ATMOSPHERE, AND HEALTH 65 REFERENCES Campbell, E. 2013. Regional impacts of biofuels on health and climate change. Presentation at the Institute of Medicine Workshop on the Nexus of Biofuels Energy, Climate Change, and Health, Washington, DC. Georgescu, M., D. B. Lobell, and C. B. Field. 2011. Direct climate effects of perennial bioenergy crops in the United States. Proceedings of the National Academy of Sciences of the United States of America 10.1073/ pnas.1008779108. EIA (U.S. Energy Information Administration). 2012. Brazil. Available at http://www.eia.gov/countries/cab.cfm?fips=BR (accessed July 29, 2013). EPA (U.S. Environmental Protection Agency). 2002. A comprehensive analysis of biodiesel impacts on exhaust emissions: A draft technical report (EPA420-P-02-001). Available at http://www.epa.gov/otaq/models/ analysis/biodsl/p02001.pdf (accessed October 10, 2013). Hoekman, S. K. 2013. Biodistillate fuels and emissions in the United States. Presentation at Institute of Medicine Workshop on the Nexus of Biofuels Energy, Climate Change, and Health, Washington, DC. Jacobson, M. Z. 2007. Effects of ethanol (E85) versus gasoline vehicles on cancer and mortality in the United States. Environmental Science and Technology 41:4150–4157. Robbins, C., S. Hoekman, E. Ceniceros, and M. Natarajan. 2011. Effects of biodiesel fuels upon criteria emissions. SAE Technical Paper 2011-01- 1943, doi:10.4271/2011-01-1943. Tsao, C.-C., J. E. Campbell, M. Mena-Carrasco, S. N. Spak, G. R. Carmichael, and Y. Chen, 2012. Increased estimates of air-pollution emissions from Brazilian sugar-cane ethanol. Nature Climate Change 2:53–57.