A seemingly innocuous cellulosic biofuel pathway
Cellulosic ethanol was once touted as the future source of ultra-low carbon biofuels. But it has struggled to reach commercialization, with production failing to match the U.S. Environmental Protection Agency’s (EPA) original volume mandates—often by substantial margins. While there are only a handful of facilities dedicated to cellulosic ethanol operating worldwide, a bright spot has been the emergence of corn kernel fiber-based cellulosic ethanol.
Though nominally a “waste” product, that doesn’t mean corn kernel fiber should get a free pass. Indeed, we’ve found that several waste and residue pathways that qualify for the advanced biofuel sub-target in the Renewable Energy Directive have indirect emissions as high as those from indirect land use change for food-based biofuels. That’s because diverting material that has productive use to biofuel necessitates replacing it. In the case of diverting corn kernel fiber to make cellulosic ethanol specifically, how it is removed and exactly what changes occur in the makeup of livestock feed as a result of diverting corn kernel fiber significantly influence the final climate impact.
Corn kernel fiber is the shell or bran around each corn grain, and it is mainly composed of cellulose. It’s considered waste because it’s not typically used in the production of corn ethanol, which is what a lot of U.S. corn is grown for. The EPA determined that it’s a crop residue and thus qualifies as a cellulosic biofuel feedstock in the Renewable Fuel Standard (RFS). When it isn’t converted into ethanol, though, corn kernel fiber becomes part of the distillers’ grains from corn ethanol production that’s typically fed to livestock. Another reason it’s considered waste is because, as fiber, it’s expected to have low nutritional value. Well, it turns out the nutritional content depends a lot on the kind of corn kernel fiber process used.
Cellulosic ethanol can be produced from corn kernel fiber in one of three ways: (1) Separating the bran from the dry corn before fermentation and processing it in a distinct process (pre-processing separation);(2) Adding enzymes to the corn ethanol fermentation vat to hydrolyze the cellulose into sugars, and fermenting them into ethanol alongside the corn starch (co-processing); and (3) Separating the bran from the distillers’ grains after fermenting the corn grain and then processing it (post-processing separation).
Pre-processing corn bran isn’t a true waste product—it’s a market product used in livestock feed, and it contains a lot more than cellulose. While a highly variable product, data for maize bran from starch production from the INRA-CIRAD-AFZ feed tables (Table 1) shows it typically has almost as much starch as cellulose. There’s also a significant amount of protein and even some fat in it. So, when this material is removed from the corn going into corn ethanol fermentation—and thus also from the distillers’ grains—there is a substantive loss in the amount of nutritious feed going to livestock. In contrast, in co-processing, the only material that is removed from the distillers’ grains is the cellulose that is converted to ethanol. Further, post-processing corn kernel fiber more resembles pure cellulose than pre-processing corn bran because virtually all the starch has been removed. Assuming the same degree of starch loss for corn bran as for corn grain in the fermentation process and no loss of other components, post-processing corn bran likely has around 2% starch content. (See Table 1. This is calculated by comparing maize bran from starch production, maize, and distillers’ grains from the feed tables.) Though this is consistent with claims from Edeniq that most corn kernel fiber has less than 5% starch, the company did not differentiate between type(s) of processing.
Understanding the loss of nutrients from corn kernel fiber removal doesn’t tell us the precise impact of low-fiber feed on livestock. For that, we’d ideally compare the metabolizable energy—the amount of calories that can be absorbed and used—of regular distillers’ grains with low-fiber distillers’ grains from each corn kernel fiber ethanol process for each type of livestock. While not all of that data is available, Table 1 illustrates that the corn kernel fiber removed during pre-processing contains a high quantity of metabolizable energy that could be utilized by each kind of livestock. For fiber removed co- and post-processing, we use the metabolizable energy of wheat straw as a proxy in Table 1, as wheat straw most closely resembles pure cellulose and the post-processing corn bran available in the feed tables. Wheat straw has nutritional value for cattle and even for pigs, but less than pre-processing separated corn bran. (While pigs only have one stomach, bacteria in their intestines can break down cellulose into digestible energy.) Broilers, a common type of chicken, can’t digest wheat straw.
Table 1. Composition of corn kernel fiber in different processes, in wheat straw (a similar cellulosic material), and the metabolizable energy of each of these materials in various livestock.
|Corn kernel fiber separated pre-fermentation (corn bran)||Corn kernel fiber fermented with starch||Corn kernel fiber separated post-fermentation||Wheat straw|
|Cellulose and hemicellulose||50%||100%||72%||70%|
|Cattle ME (MJ/kg||11.7||6.3|
|Adult pig ME (MJ/kg)||9.5||3|
|Broiler ME (MJ/kg)||8||—|
Given these impacts on livestock nutrition, what are the carbon implications? If less metabolizable energy is delivered in distillers’ grains when corn kernel fiber is removed, that energy must be replaced. In a previous blog, I described the evidence that corn oil pressed from distillers’ grains is replaced by additional corn grain. This is consistent with the EPA’s argument that farmers are unlikely to notice the change in distillers’ grains quality and will simply supply additional feed to their animals. That additional feed is mostly corn. If the same is true for corn kernel fiber, then any nutrition removed from distillers’ grains for cellulosic biofuel production will be replaced by the equivalent amount of corn grain in terms of metabolizable energy for each type of livestock. Table 2 shows these substitution ratios, with calculations based on data in the feed tables.
When classifying corn kernel fiber as a crop residue, the EPA argued that there would be no displacement effect because low-fiber distillers’ grains would go to pigs and chickens with no loss of metabolizable energy. While this seems logical at first, further consideration suggests otherwise. For one, we see that pigs likely do get nutritional value from corn kernel fiber. That means the EPA’s assumption only fits if low-fiber grains are fed to poultry. Historically, though, cattle received the vast majority of domestically consumed distillers’ grains, with only around 5% fed to poultry. This is likely because most poultry farming is in the Southeast U.S. and California, far from the bulk of corn ethanol production in the Midwest. Cattle production, in contrast, is more spread out across the country, and a great deal of it is in the Midwest. Pig production is also concentrated in the Midwest. It thus seems unlikely that the majority of low-fiber distillers’ grains will be transported from corn ethanol facilities in the Midwest to poultry farms in the Southeast.
As shown in Table 2, if low-fiber distillers’ grains are supplied to the same mix of livestock that typically receives distillers’ grains, pre-processing corn kernel fiber ethanol has a whopping 36 gCO2e/MJ indirect emissions, and cellulosic ethanol from co- and post-processing fiber separation has 18 gCO2e/MJ. Add these to the direct processing emissions from even the most efficient corn kernel fiber ethanol plants supplying to California’s Low Carbon Fuel Standard, and corn kernel fiber wouldn’t qualify under the 60% GHG reduction threshold for cellulosic biofuel in the RFS.
Table 2. Substitution ratios of corn grain for corn kernel fiber for different processes and displacement emissions when corn kernel fiber is replaced by additional corn in feed for various livestock.
|Corn kernel fiber separated pre-fermentation||Corn kernel fiber fermented with starch or separated post-fermentation|
|Displacement emissions (gCO2e/MJ)||Cattle||37||20|
|Average livestock mix||36||18|
I’m not saying corn kernel fiber ethanol shouldn’t be supported in the RFS. However, corn kernel fiber’s diversion emissions drive its carbon intensity above that of other cellulosic ethanol feedstocks like agricultural residues and energy crops. Fortunately, the risks of significant food market and land use change impacts from fiber are much lower than for food-based biofuel feedstocks. But maybe we should reconsider support for pre-processing separated fiber, and see corn kernel fiber ethanol as a stepping-stone technology to get us closer to a cellulosic biofuel industry capable of processing large quantities of even lower-risk corn stover and cellulosic energy crops, rather than as a long-term solution to the food vs. fuel debate.