Switchgrass as a biofuel

Report by Vincent P. Gutschick 11 March 2010

Global Change Consulting Consortium, Inc.

4904 Calabazilla Dr., Las Cruces, NM 88011 USA

(575) 5751-2269 vince@gcconsortium.com



This is a brief report generated for the Advisory and Workgroup Meeting of the Southwestern Biofuels Association, Monday, March 15th, 2010 and the subsequent 2010 Biofuels Policy Summit, April 13th & 14th. It is not a comprehensive overview of switchgrass as a biofuel. Rather, it is intended to provide important discussion points, most of which apply to the other biofuels as well.


Table of Contents (hyperlinks within this document; for using these in Word, see Appendix III at the very end of this document)

Brief Introduction

Biology and agronomic practices




Hazards to productivity

Practicalities of use

As direct combustion feedstock

For synthesis of liquid fuels


Direct investments and returns

Indirect economic effects

Water as a potential issue of cost and availability in our area

Transportation of the biomass

Need for new economic analyses and regulatory clarifications

Meeting demands for life-cycle energy return and GHG reduction

Energy per se

Energy per land area

GHG impacts

Other environmental impacts of biofuel production that play in the structure of regulations, current and future.


References Cited

Appendix I. Efficiency of using solar energy input in growing switchgrass

Appendix II. Ultimate energy efficiency of using biomass for vehicular transporations, as liquid fuel vs. for electricity generation

Appendix III. Using hyperlinks within this document



Brief introduction

Switchgrass (Fig. 1) is a perennial grass that can grow in most states of the US. Its attractiveness as an alternative energy source stems from its renewability, good net energy yield (see below), generally favorable impact on life-cycle greenhouse gas (GHG) emissions relative to fossil fuels, wide environmental tolerances, and harvestability with standard farming equipment. Its negative impacts include high land use per unit power provided (with consequent impacts on habitats and biodiversity), continuation of crop-derived water pollution, and some particulate air pollution, depending on mode of usage. One more impact, if the grass is grown in drier areas, is high consumption of water (relative to non-crop alternative of wind, solar, or nuclear power). The most promising avenue appears to be conversion of its cellulosic biomass to ethanol using advanced / synthetic enzymes. Some unique challenges exist in control of pests, diseases, weeds, and fire, while common agronomic practices might be adapted to deal with many of these. Another challenge is transportation cost, in energy and dollars, from point of growth to point of use. Biomass farms may be located at significant distances from conversion points as ethanol production facilities or powerplants.


Biology and agronomic practices

Botany. Switchgrass, Panicum virgatum, is a warm-season perennial grass native to North America, growing from southern Canada into Mexico. The mature plant commonly attains a height of 60 to 150 cm (2 to 5 ft.), with extremes to 220 cm (8 ft). Before European settlement, it was a major component of many grassland ecosystems in these regions. Switchgrass tolerates a broad range of soils, doing best on better-drained soils that are neutral to slightly acidic (pH 5.5 - 7). It is deep-rooted and spreads slowly by belowground rhizomes. Contributing to its success in diverse conditions is that switchgrass has evolved many ecotypes that are adapted to local conditions. The ecotypes differ genetically and physiologically from each other, a factor that both requires and enables breeding programs for optimizing its performance. Stand establishment from seed is relatively rapid, in about 3 years, and stands persist at least 10 years in trials with artificial plantings.

Switchgrass has the C4 photosynthetic pathway, in which a biochemical "pump" for CO2 precedes the universal C3 pathway of all photosynthetic vascular ("higher") plants. This pump enables C4 plants to have higher efficiencies than C3's in using sunlight, water, and nitrogen. It also makes them more cold-sensitive because the enzyme in the "pump" dissociates into inactive units at low but above-freezing temperatures. Consequently, the growing season varies from 3 months at the northern limits of its range to 8 months in the southern regions.

Productivity. The plant can achieve high annual productivity of raw aboveground biomass, ranging from 6.7 to 13.5 tonnes per hectare per year (6 to 12 tons per acre per 3 years) as dry mass in US trial plantings at sites with good rainfall of 855 to 1100 mm (33.7 to 43.7 in.) (Kaiser and Bruckerhoff, 2009. Higher yields, averaging 14.6 tonnes/ha, were reported in another study over a wider geographic area (McLaughlin and Kzos, 2005). In New Mexico, yields should be expected to be considerably lower unless irrigation is used. The species' adaptation to North American conditions makes many ecotypes tolerant of drought. Irrigation is not necessary for yield in most years, though highest yields are attained with irrigation.

Inputs. Initial trials have all been rainfed, without irrigation. Standard planting and hay-harvesting equipment is usable. Nitrogenous fertilizer is applied at rates that have varied widely between trials (Barnhart and Gibson, 2007) , with averages near 170 kg per hectare (150 lb per acre). This input much exceeds the N removed in plant material (0.5 to 1% N) in a typical harvest of 6.7 tonnes/hectare (6 tons/acre). The impacts on water quality are discussed later in this report. Fuel usage for planting, tillage, and harvesting are modest, as noted in the reports on energy yield discussed below (XXX). No-till agriculture is an option with switchgrass, which has also been used for soil conservation by virtue of the continuous cover it affords.

Hazards to productivity. As all plants, switchgrass is susceptible to pests, diseases, and weed competition. Fungal and viral disease are common but appear not to depress yield severely or spread as epidemics, at least in small-scale tests. There is as yet no experience with large-scale monocultures of limited genotypic diversity. Switchgrass is additionally susceptible to wildfire. Nonetheless, burning of residues after harvesting is a common practice (which is likely to be curtailed in large-scale deployment of the crop to met EPA regulations on airborne particulate matter). Wildfire potential in large-scale plantings under extreme weather conditions cannot yet be assessed.


Practicalities of use

As direct combustion feedstock. Dried harvested material can be pelletized and used for combustion, including as an adjunct to coal in large power plants. Adjustments are needed in operating conditions because the combustion temperature is lower than that of coal or other fossil fuels and the fuel-air ratio is higher. Some small-scale trials have been favorable (http://www.iowaswitchgrass.com/).

For synthesis of liquid fuels. The principal route tested to date has been fermentation of the cellulosic component to ethanol. One recent study indicates a very favorable life-cycle energy return for this route, discussed below. Note that highly efficient cellulose fermentation methods have not yet been shown beyond the demonstration stage - that is, not to the scale of pilot plants or production facilities.

While ethanol has been fitted into the gasoline distribution system, problems remain. Energy density is lower, giving fewer miles per gallon (higher consumption as liters per 100 km in the standard European measure). Ethanol has limited solubility in gasoline hydrocarbons. With absorbed water, it also corrodes standard pipelines, so that it requires on-site blending prior to tanker delivery. As an alternative to conversion to ethanol, plant biomass can be thermally converted to liquids that closely approximate hydrocarbon fuels (http://aiche.confex.com/aiche/2009/webprogrampreliminary/Paper171370.html). In this process of hydropyrolysis, significant inputs of either fossil or solar energy are needed. The methods are still in development and are not yet practical and cost-competitive.



Direct investments and returns. A number of studies are completed or in progress for switchgrass and other renewables, all with a number of uncertainties. A study by the Center for Agricultural and Rural Development at Iowa State University (Secchi et al., 2008) projected a price of about $35 per oven-dry ton of switchgrass, while a price of $100/ton would be needed to support substantial plantings (46,000 ha) in the Upper Mississippi River Basin. A University of Tennessee Study (http://www.25x25.org/index.php?option=com_content&task=view&id=156&Itemid=56) projected prices of $44 to $88 per ton. Multibillion-dollar subsidies would consequently be needed

In contrast to market prices, the physical inputs on-farm are rather well quantified, from much experience in farming and energy conversion. Principal inputs are diesel fuel for planting, tilling, harvesting; nitrogenous fertilizer; process energy and amortized capital energy in constructing the conversion plant. The monetary costs of per unit of each of these inputs fluctuate, rendering a significant uncertainty in the total cost.

Another major cost is land rent. The land need not be of prime farmland quality but poorer land gives lower yields. Later in this report, the issue of land use for competing renewable energy technologies is explored.

Indirect economic effects. A consistent and large concern with all biofuel cropping is that it displaces food crops. Consequently it can raise the price of food, and globally as well as locally, given the globalization of grain markets (Miller, 2008). Switchgrass may have lower impacts on food prices than corn when either is used for biofuel production, but its displacement of crop area is not negligible (Secchi et al., 2008).

Water as a potential issue of cost and availability in our area. Irrigation water has not been supplied in trials to date, which have been done in areas with notably higher rainfalls than lower-elevation areas of New Mexico. Arid areas offer longer growing seasons and perhaps multiple crops per year, thus, higher annual yields. However, irrigation water would be a necessity for practical yields. Even though irrigation water is priced very low, its availability in the arid SW is very dubious.

Transportation of the biomass to a plant for combustion or for conversion to liquid fuel is another significant cost, one that plagues most renewables (crops, wind, many solar technologies) that are generated primarily in rural areas far from points of final use by consumers. There is one biorefinery in New Mexico, under construction (Fig. 2; the NM site is listed erroneously as in operation).

Need for new economic analyses and regulatory clarifications. Before switchgrass is considered for New Mexico or other areas of the arid SW, it is imperative to generate new economic analyses suited to our sites. In all areas of the US, it is still foreseen that economic subsidies will be needed for all biofuels, including switchgrass (see earlier). One may argue that this is useful, given that fossil fuels have many indirect subsidies, such as the costly operation of the US Seventh Fleet to support oil supplies. Nonetheless, the subsidy provisions remain uncertain for the future and across states they present a patchwork. So also do mandates for the fraction of energy from biofuels. The lack of a national energy policy, which would stabilize these subsidies and mandates, is seen as a hindrance to investment by primary energy producers, energy processors, and electrical utilities. These uncertainties may be more challenging to deal with than uncertainties in the costs of inputs and production.

One element in energy policy that is favorable to energy crops such as switchgrass, both economically and politically, is a new regulatory framework that requires life-cycle accounting for total greenhouse gas impacts (http://www.ucsusa.org/news/press_release/new-renewable-fuel-standard-favorable-review-from-UCS-0345.html) . The elements of GHG impacts are described in some more detail later in this report. What is important for economic viability is that the regulations also count indirect impacts. Perhaps chief among these is land conversion that must be done elsewhere to replace food crops that were displaced to grow the energy crops. These impacts are markedly lower for cellulosic ethanol production or hydropyrolysis of biomass, compared with current ethanol production from corn with fermentation of only the starches. Another regulatory consideration, that of airborne particulates released in energy use, may affect the use of pelletized dry switchgrass biomass in powerplants (or homes). In large powerplants, particulates are fairly readily controlled, but in some envisioned smaller plants serving local demand, this is less clear (literally). Favorable to biomass vs. coal but not yet in regulations is the much lower release of mercury per unit energy produced.


Meeting demands for life-cycle energy return and GHG reduction

Energy per se. Simulations show favorable local energy returns from growing and processing switchgrass, under the assumption of high conversion of cellulosic biomass to ethanol (0.34 L per kg of dry biomass). Nonconvertible lignin in the biomass is used for process heat, such as in final distillation. For switchgrass, the on-farm use of fossil energy for farm machinery and the pre-farm energy use in making nitrogenous fertilizer is less than half that for corn raised for ethanol production from starch, and still superior to that for corn raised for cellulosic ethanol. A recent study (Schmer et al., 2008) indicates that each unit of energy used to raise and process switchgrass for ethanol yields 5.4 units of energy in the final ethanol product. Also favorably, this same study used a more conservative estimate of biomass yields and final energy yields.

Energy per land area. The intensiveness of land use for energy cropping has been noted earlier. A more quantitative analysis may be pursued. Because all energy sources compete to a large extent, it is worth comparing the land use of energy crops such as switchgrass with land use of other renewable energy technologies. A simplified comparison is readily derived for solar photovoltaics. These have a high capital cost, which is declining markedly. Their efficiency varies by type, mainly crystalline vs. amorphous silicon. We can conservatively take 10% as a "first-law" efficiency, that is, electrical energy (enthalpy) out relative to incident solar energy. This is at the collector. Collectors may cover 80% of an installation. There is a further need for land area dedicated to energy storage of some type, to even out energy delivery during day and night and cloudy and sunny times. If this takes an equal area, then the energy yield on a total-land basis is down to about 4%. In the Appendix, I estimate that the annual average energy efficiency of growing switchgrass in mid-continental areas to make ethanol is 0.24%.

The energy figures are not simply comparable, because liquid fuels and electrical energy are used differently. Consider then use for transportation. The most energy-efficient use of biomass for transportation is actually combustion in a central power plant. In Appendix II, I estimate that the effective efficiency of converting biomass to drivetrain energy is about 24%, in a hybrid vehicle. In Appendix I, I estimate that the efficiency of capturing solar energy to biomass energy with switchgrass is 0.48%. Thus, from solar energy to drivetrain energy, the net efficiency is about 0.24*0.0048 = 0.12%. For solar photovoltaics, the 4% efficiency at the plant output must be reduced, just as for bioenergy in Appendix II, to account for transmission line losses (10%) and battery charge/discharge efficiency (75%). The net figure is then about 2.5%. This is a factor of over 20-fold higher (2.5%/0.12%) than that for switchgrass. Consequently, the same amount of energy for transportation requires more than 20 times as much land when switchgrass is used vs. solar photovoltaics. Using switchgrass for ethanol doubles this figure, to 40-fold greater use. Note that the land-use figures for wind farms are also highly favorable (not presented here). Fossil fuels are even better than solar and wind energy as land-sparing energy technologies, as energy yield per unit of land disturbed. We may conclude that there is a serious need to consider how much value we place on the convenience of liquid fuels vs. the investment that enables electric vehicle use. In any event, we are committed, by the realities of our energy future, to use much more land for energy production.

GHG impacts. Renewable fuel legislation is now being abetted by regulations that account for GHG production or amelioration. Biofuels have long been expected to reduce the emission of greenhouse gases, particularly CO2, methane, and N2O. The realities are complex, as several reports have noted (Melillo et al., 2009), including candid reports from university consortia funded by the largest biofuels-promoting fossil-fuel corporation, BP (Davis et al., 2009). Biofuels largely recycle CO2 emitted in use of any fuel, fossil or bio-, but some fossil fuel is used in cropping and sometimes in processing. Consequently, biofuel use has a direct impact of increasing atmospheric CO2, but not as much as direct use of fossil fuels. In the study cited earlier (Shmer et al., 2008), the reduction is to a factor 1/5.4, or about 18%.

There are two major qualifiers to the immediate favorability of biofuels for GHG emissions, both of these being discussed in the reports just cited. First is indirect land use. If other (food) crops are displaced in current agricultural areas by biofuel crops, then new land will be put into cultivation for food crops elsewhere. This land must be cleared of existing vegetation, causing the release of CO2 and perhaps N oxides when the vegetation is burned or simply decomposed. Forest clearance, as in some South American farms, presents the worst case, with emissions as large as 90+ years of fuel use. Second is production of N2O and, less so, of methane. Some fraction of applied nitrogenous fertilizers always undergoes denitrification in the field or in waterways. This produces N2O, a greenhouse gas with 300 times the greenhouse effect of CO2, per unit mass on a 100-year basis (its residence time in the atmosphere is about 120 years). Fortunately, the N:C ratio in switchgrass is low, about 1:40, and the fraction of applied fertilizer N released as N2O is low, such that N2O accounts for perhaps 10-20% as much impact as CO2 (http://opencarbonworld.com/wiki/greenhouse-gas.html).

Other environmental impacts of biofuel production that play in the structure of regulations, current and future. The largest environmental impacts derive from the expansion of land use for (biofuel) crop growth. This necessarily entails further loss of habitat for native plants and animals. In turn, items of value to humans are adversely affected. Wild species have been characterized extensively as providing recreation (hunting, birding, etc.), breeding stock for crops (native plants, particularly outside the US), ecosystem services (flood control by riparian plants, pollination by insects, control of insect pests by birds and bats, etc.), other values as biodiversity (e.g., sources of medicines), and simple amenity value. Some but not all of these values can be monetized; others, including intrinsic human appreciation of nature (biophilia) must be accounted non-economically but realistically. Economic analyses of these environmental impacts are in various stages of completion and with various premises. In any event, policies will ultimately derive from such studies (Groom et al., 2007) and must be watched closely during the development of switchgrass and other bioenergy technologies.

An impact with more immediate policy implications is runoff of excess N and P fertilizers to rivers and onward to coastal waters. A fraction of all applied fertilizer gets into surface waters directly or indirectly after leaching into groundwaters that supply springs that feed rivers. The fertilizer runoff from agriculture in the Mississippi River Basin to the Gulf of Mexico is enormous. In the Gulf, the nutrient input supports algal blooms, which, upon dying, create a large "dead zone"temporarily devoid of most sea life, including important food fish and crustaceans. The size of the zone, commonly over 20,000 km2 each year, is correlated with the fertilizer runoff. The US EPA has a task force (http://www.nola.com/news/index.ssf/2009/09/post_31.html) working on implementing a range of corrective actions, some of which may constrain fertilizer use and crop choice. Switchgrass cropping moderately reduces the N and P flux to the Gulf (Secchi et al., 2008), but the indirect impact of needing more food cropland elsewhere may exacerbate the dead zone problem and drive additional policy initiatives.



Switchgrass grown as an energy crop offers generally favorable results in net energy return and in low GHG impact (which is now becoming important in policy). Implementing it as an energy technology will require economic subsidies for the indefinite future. Switchgrass energy is practical to pursue with near-current technologies of farming and processing. Lan quality and siting is not critical agronomically, though the economics of land rent are important.

There are two basic ways to use switchgrass biomass for energy - direct combustion in an electric power plant and conversion to liquid fuel, particularly ethanol. Direct combustion is twice as efficient energetically but does not alleviate the developing crisis in energy for transportation without a concurrent push to develop electric vehicles as a major transportation component.

The overall attractiveness of switchgrass is a function of considerations in economics, energy policy, and environmental policy. While energy policy is firming up, environmental policy has taken new turns, particularly on greenhouse gas emissions and fertilizer runoff. These bear close watching and do not entirely favor switchgrass and most other biofuels.

Switchgrass is a viable component of our biofuels future. Nonetheless, biofuels are destined to be a bridge technology between current energy technology and future electricity-intensive technologies (J. Yang, DOE, pers. commun.). In the longer term (> 20 years), solar, wind, and nuclear energy technologies are favored by their low land use and better GHG impacts.

This report maintained a focus on switchgrass, but many considerations brought out here apply to other bioenergy technologies, such as algae and tree plantations, with only quantitative adjustments.


References Cited

S. C. Davis, K. J. Anderson-Teixeira, and E. H. DeLucia. 2009. Life-cycle analysis and the ecology of biofuels. Trends Plant Sci. 14: 140-146.

L. Gibson and S. Barnhart. 2007. Switchgrass. Report AG 200, Iowa State Univ., Ames, IA, USA.

M. J. Groom, E. M. Gray, and P. A. Townsend. 2008. Biofuels and biodiversity: Principles for creating better policies for biofuel production. Conserv. Biol. 22: 602-609.

J. Kaiser and S. Bruckerhoff. 2009. Switchgrass for biomass production by variety selection and establishment methods for Missouri, Illinois, and Iowa. Technical Note 35, USDA Iowa State Office, Ames, IA, USA.

S. B. McLaughlin and L. A. Kzos. 2005. Development of switchgrass (Panicum virgatum) as a bioenergy feedstock in the United States. Biomass and Bioenergy. Vol. 28. pp. 515-535.

J. M. Melillo,J. M. Reilly, D. W. Kicklighter, A. C. Gurgel, T. W. Cronin, S. Paltsev, B. S. Felzer, X. Wang, A. P. Sokolov, and A. Schlosser. 2009. Indirect emissions from biofuels: How important? Science 326: 1397-1399.

D. Miller. 2008. A note on rising food prices. Policy Research Working Paper 4682, World Bank, Washington, DC, USA.

M. R. Schmer, K. P. Vogel, R. B. Mitchell, and R. K. Perrin. 2008. Net energy of cellulosic ethanol from switchgrass. Proc. Nat. Acad. Sci. USA 105: 464-469.


Appendix I. Efficiency of using solar energy input in growing switchgrass

What is the solar energy input at places where switchgrass trials have been done?

See the map at http://rredc.nrel.gov/solar/old_data/nsrdb/redbook/atlas/serve.cgi

4.5 kWh m-2 d-1, or 16.2 MJ m-2 d-1. Per year, this is 5.91 GJ m-2

Assume the biomass yield averaged about 14 tonnes/ha, or 14 Mg/ha, which is 1400 g m-2. (This is higher than the values in trials analyzed for the link above, but consistent with some other trials.)

The embodied energy in biomass for combustion in air is closely 19 kJ g-1. The 1400 g per m2 then represents an energy value of 27 MJ m-2.

This energy is converted to final fuel energy with slightly more than 50% efficiency [1 kg of dry biomass with an energy content of ca. 19 MJ, yields about 0.34 L of ethanol, with an energy content of 10 MJ], giving us an annual energy yield per ground area of 14 MJ m-2.

The energy efficiency is the ratio between this figure and the total solar energy, or

14/5910 = 0.0024 = 0.24%.

Note that the highest one-crop energy conversion (to biomass only) ever observed was 6.6% in sugarcane in Hawaii. The average efficiency of earth's vegetation in capturing energy in sunlight is 0.30%.

The low efficiency for switchgrass as an ethanol source derives in good part from its short growing season and the need to convert it to ethanol.


Appendix II. Ultimate energy efficiency of using biomass for vehicular transporations, as liquid fuel vs. for electricity generation

Gasoline as a liquid fuel is converted to mechanical energy in an automobile engine at an average rate commonly quoted as approximately 18% at the beginning of the drivetrain (http://www.fueleconomy.gov/FEG/atv.shtml). The figure varies considerably with the degree of urban vs. highway driving. Regenerative braking in a hybrid vehicle recoups energy and raises the effective efficiency projected to the beginning of the drivetrain by about 1/3, to about 24%.

Assuming the same efficiency for ethanol use in an engine, and 50% energy conversion from plant biomass to ethanol, and a multiplier of 0.9 to account for energy use in distribution, the overall efficiency from biomass to drivetrain is about 11%. This assumes use of the fuel in a hybrid vehicle.

Electricity generation from biomass to electric power in a coal-fired power plant ranges from 35 to 45%. The lower figure would apply to a power plant using or co-fired with biomass extensively at lower combustion temperatures. Transmission line losses average about 7% nationally; we might inflate this to 10% for longer lines linking biomass power plants to sites of use.

Consider use of the electricity in an electric or plug-in hybrid vehicle. The charge/discharge cycle of the vehicle's batteries has an efficiency ranging from 60 to 90%, depending on the type of battery (various sources) so that the net energy efficiency from biomass to battery-delivered energy in an electric vehicle is about 0.35*0.9*0.75 = 24%. Electric motors operate over a wide range of energy efficiencies, from zero at dead start up to 90% at full operating speed. In a vehicle, we might assume 75% as an average (http://www.fueleconomy.gov/Feg/evtech.shtml). We then credit the regenerative braking by multiplying the effective efficiency at drivetrain entry by 4/3, giving us a final energy efficiency equivalent to 24%, more than twice that for using biomass to make ethanol for a comparable vehicle.

These calculations do not account for the significant capital energy in making the batteries for electric vehicles, either plug-in hybrid or pure electric. Nor do they account for the lower load-carrying capacity of electric vs. liquid-fueled vehicles that may affect the number of vehicles needed nationally.


Appendix III. Using hyperlinks within this document

This document has been produced in the format of Microsoft Word 2003. Highlighted items, as well as entries in the Table of Contents, constitute hyperlinks for easy navigation. The printed version will lack this convenience but is fully accessible nonetheless.

Clicking on a hyperlink (CTRL-click) leads you to the referenced location in the document. You can return to the original location, where you find the hyperlink: in Word, go to View Toolbars Web; a Web toolbar will be inserted in your window where Word is open; click on the back-arrow (lower left item in image below, showing left side of Word toolbar area):