The Environmental Benefits of Cellulosic Energy Crops at a Landscape Scale

1 Biofuels Feedstock Development Program, Environmental Sciences Division, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6335.

2 Institute of Agriculture, University of Tennessee, PO Box 1071, Knoxville, TN 37901-1071.

Sponsored by the Biofuels Systems Division, U.S. Dept. of Energy, under contract DE-AC05-96OR22464 with Lockheed Martin Energy Research Corporation. (Corresponding author: RLG)

From Environmental Enhancement Through Agriculture: Proceedings of a Conference, Boston, Massachusetts, November 15-17, 1995, Center for Agriculture, Food and Environment, Tufts University, Medford, MA.

Introduction: The Role of Cellulosic Energy Crops

This paper presents a broad overview of the potential environmental impacts of biomass energy from energy crops, particularly the cellulosic energy crops currently under development. We use the term energy crop to mean a crop grown primarily to provide a feedstock for biofuels such as ethanol or to be burned for heat or electricity. Energy crops currently in production include corn (Zea mays), sugarcane (Saccharum officinarum), and short-rotation plantations of poplar (Populus spp.), sycamore (Platanus occidentalis) and eucalyptus (Eucalyptus spp.). We use cellulosic energy crop to differentiate those grown for their cellulose content from those grown for carbohydrates (starch or sugars), such as corn or sugarcane. Cellulose and carbohydrates both can be converted to ethanol, but more cellulose can be produced per unit land area than carbohydrates. Therefore, cellulose-based ethanol production is a more efficient use of land.

Cellulosic energy crops currently under development in the U.S. include switchgrass (Panicum virgatum) and short rotation woody crops, especially poplar and willow (Salix spp.). Switchgrass is a drought-tolerant prairie grass with an extensive natural range in North America. It is the model herbaceous energy crop selected by the U.S. Department of Energy and is well suited for cellulose-based ethanol conversion (Wright et al., 1993). Cellulosic energy crops also are appropriate for producing heat and electricity.

From the perspective of the farmer, energy crop production generally is compatible with food crop production. Farmers most likely will use energy crops as part of a mix of several crops. However, introducing them into the larger farming scheme requires careful planning, because most cellulosic energy crops are perennials, with stands lasting five to twenty years.

Assessing the environmental impacts of biomass energy from energy crops is complex because two different kinds of impacts are involved: using biomass for energy must be considered in the context of alternative energy options, while the environmental impact of producing biomass from energy crops must be considered in the context of alternative land uses. Using biomass-derived energy can either reduce or increase greenhouse gas emissions; growing biomass energy crops can enhance soil fertility or degrade it. Therefore, one must know the specific circumstances to be able to make a statement about the environmental impacts of biomass energy. We focus on the environmental impacts of growing cellulosic energy crops, especially its impacts on a landscape or regional scale. However, we first compare the environmental advantages and disadvantages of using biomass-derived energy to those of other energy sources, including coal, gasoline, natural gas, nuclear power, hydropower, and photovoltaics.

Comparing Biomass Energy with Other Energy Sources

Coal. Biomass energy is most advantageous when compared with coal. Using it to displace coal in power generation has the greatest benefit in reducing greenhouse gases and air pollution. A recent analysis calculated that using the wood from 1 ha of short rotation woody energy crops instead of coal would displace 5.2 t (metric tons) of fossil C in CO₂ (Graham et al., 1992). Biomass burns cleanly, with much lower SO_x and somewhat lower NO_x emissions. Two additional environmental benefits come from the reduction in coal mining and the avoidance of coal ash disposal. Wood combustion produces much less ash than does coal, and the ash can be returned to agricultural soils.

Gasoline. Using biomass-derived ethanol to displace gasoline also has a considerable greenhouse gas benefit, but it is much smaller than the benefit when biomass displaces coal because gasoline has a much higher energy value per unit of fossil carbon. The CO₂ benefit of using ethanol made from cellulosic crops to displace gasoline is estimated to be about half as great as when it displaces coal (Graham et al., 1992). When the ethanol is produced from a crop grown with large inputs of fossil fuel, such as corn, the greenhouse gas saving may be very low; the current process of growing corn and converting it to ethanol produces 0.6 to 0.8 units of CO₂ for every unit of gasoline-CO₂ that is displaced (Marland and Turhollow, 1991). The low ratio is due to the heavy use of nitrogen fertilizers and the use of fossil fuels (coal) in the conversion.

Ethanol has air pollution benefits over gasoline as an automobile fuel. It emits less CO, SO₂, and hydrocarbons, but slightly more NO_x (Graboski, 1993). It emits more aldehydes, but mostly as acetaldehyde, which is less reactive and less toxic than the formaldehyde produced by the combustion of gasoline (Macedo, 1993). In general, ethanol emissions are less toxic than gasoline emissions (La Rovere and Audinet, 1993). Compared with reformulated gasoline to which oxygenates have been added, biomass derived ethanol has no air pollution advantages or disadvantages other than the reduction in CO₂ (Tyson et al., 1993).

Natural gas. The environmental advantages of displacing natural gas with biomass derived electricity are less than for displacing coal, oil or gasoline. Natural gas contains twice as much energy as coal per unit of carbon. Also, natural gas power plants are more energy efficient than coal-fired plants (Turhollow and Perlack, 1991).

Nuclear and renewable energy sources (hydro, wind, photovoltaics). Biomass energy has less of an environmental advantage when compared with greenhouse-neutral energy types such as nuclear, wind, hydropower, and photovoltaics. It releases some greenhouse gases, because fossil fuels are likely to be used to produce energy crops (to power tractors, manufacture fertilizers, etc.). Although biomass releases only 1/10th to 1/20th as much greenhouse gases per unit of power as does coal, the emissions still are not zero (Turhollow and Perlack, 1991). However, biomass energy avoids the safety and waste disposal problems of nuclear power. Also, it does not impose the loss of land and the damage to fisheries that occur with dams and hydropower.

Land requirements. Biomass energy's universal disadvantage is its high land requirement. An efficient biomass power plant (33 %) will require between 200 and 400 ha of land in energy crops per MW of baseload power, depending on the biomass yield. Cellulose-to-ethanol technology is expected to require 150 to 300 ha per million liters of annual capacity, assuming moderate cellulosic energy crop yields (10 to 20 dry t/ha); U.S. ethanol production from corn requires close to 300 ha per million liters (Hohmann and Rendleman, 1993). It is important to quantify the environmental impacts of major shifts in land use to grow energy crops. Shifting from current agriculture to energy crops could change soil erosion patterns, water quality of regional streams, wildlife populations, and regional air quality. Characterizing these impacts is challenging because they depend on many site- and crop-specific factors.

Factors Determining the Environmental Impacts of Energy Crop Production

Crop Factors

The type of crop grown is a decisive variable in predicting environmental impacts from energy crop production; crops have different effects on erosion, water availability and quality, wildlife habitat, and air quality. For example, growing corn is likely to cause more soil erosion and use more fertilizers than growing short rotation poplar. However, because trees use more water than herbaceous crops, they may reduce stream flow. Wildlife will differentiate among crop types; for example, tree crops can provide habitat for forest bird species (Wright et al., 1993; Hoffman et al., 1995; Tolbert and Schiller, 1996). Perennial grasses enhance soil carbon more than do annual crops. Tree crops release more hydrocarbons into the air than do herbaceous crops (Perlack et al., 1992).

How the crop is managed also is important (Cook et al., 1991). Interplanting a cover crop between trees early in short-rotation woody crop production is likely to reduce erosion compared with leaving the soil bare. The types and amounts of pesticides and fertilizers applied and the timing of applications will affect water quality. Harvesting trees during the winter reduces the loss of nutrients from the site because the leaves are not removed. Burning crop residues in the field, as is done with sugarcane, can harm air quality locally and regionally (La Rovere and Audinet, 1993).

Site factors

Physical characteristics of the land will strongly influence the productivity of energy crops and therefore the likelihood that they will be grown (Liu et al., 1992; 1993). Soil type, climate, and topography will affect erosion and runoff (Office of Technology Assessment, 1993; Hoffman et al., 1995; Tolbert and Schiller, 1996). Soil type also will influence the need for fertilizers and the rate at which pesticides and fertilizers leach to groundwater. A high organic matter content increases the soil's retention of pesticides and nutrients. In a warmer climate, pesticides break down and volatilize more rapidly.

The previous use of the land is an especially important consideration when developing policies to promote or discourage energy crops. The difference between the environmental effects of the former land use and of the energy crop determines the environmental value of the energy crop. For example, growing switchgrass rather than soybeans (Glycine max) has many environmental advantages. Compared with soybeans, switchgrass will increase soil carbon, reduce erosion, improve water quality, and provide better animal habitat. Thus, a policy that encouraged the production of switchgrass on former soybean land would have environmental benefits, but a policy that encouraged conversion of forests to switchgrass would be environmentally damaging. Valuable forest habitat would be lost and water quality would likely be degraded.

Finally, the location of the energy crops in relation to other land uses will strongly influence water quality and wildlife impacts. Perennial energy crops such as trees or grasses that receive low levels of fertilizers or pesticides can serve as filters if planted along streams. These crops can absorb nutrients coming from more heavily fertilized conventional crops upslope and can catch sediment as it moves downslope. Both actions can improve local water quality (Ranney and Mann, 1994). If streamside planting is extensive and nonpoint-source pollution from agriculture is a regional problem, perennial energy crops could improve regional water quality (Office of Technology Assessment, 1993). Planting energy crops on a small amount of land adjacent to streams could have a much larger influence on water quality than planting two or three times as much land upslope.

If planted in a landscape dominated by annual crops, energy crops may enhance regional and local wildlife. In particular, woody energy crops add structural diversity to agricultural landscapes, which should enhance biodiversity on a regional scale. Measurements of abundance, type, and number of bird species have shown that woody crops can serve some of the habitat functions of natural forests (Office of Technology Assessment, 1993; Wright et al., 1993; Hoffman et al., 1995). However, woody energy crops are not a substitute for natural forests (Cook et al., 1991; Tolbert and Schiller, 1996). Producing energy crops can harm wildlife if the crops displace a food source in the original land use. For example, birds migrating from Canada to Mexico and South America use the corn left in the fields of the Midwestern U.S. If switchgrass displaced large areas of corn in this region, this food source would disappear, with no obvious replacement.

Quantity of land dedicated to energy crops

Clearly, the more land converted to energy crops, the greater their regional impact. However, their impact cannot simply be assumed to be proportional to the amount of land planted to them, particularly for wildlife and water quality impacts. Even an impact such as erosion, which can be calculated on a per hectare basis and does not depend on land use changes elsewhere, will not simply be proportional to the amount of land planted to energy crops in a region, because soil type, topography, and former land use all will vary within the region (Graham and Downing, 1993).

Interactions among factors

These factors cannot be considered in isolation, because they interact strongly in affecting the environment. For example, soil, climate, topography, crop type, and crop management all will affect energy crop productivity and therefore the quantity of land needed to produce a specific supply. Even former land use by itself can affect energy crop productivity. For example, soil compaction as a result of pasture use may reduce expected energy crop yields. Assessing the potential environmental impacts of energy crop production requires an integrated approach that considers all these factors. To be successful, however, the approach must also include the economic forces and policies that will control where energy crops will be grown and what land uses they will displace.

Quantitatively Predicting Environmental Impacts

It is difficult to predict quantitatively the environmental impacts of cellulosic energy crop development on a landscape or regional scale because most cellulosic crops are not yet planted in regionally significant amounts. Thus, the necessary empirical information on environmental impacts is lacking and a modeling approach is needed to predict the effects these crops might have if planted on the scale of a major crop such as corn, or even a less important crop such as barley (Hordeum vulgare) or oats (Avena saliva). We have developed a six-stage modeling approach for assessing regional or landscape scale environmental impacts. It includes economic considerations, since economics will determine where energy crops are profitable, what conventional crops they will displace, and what management regimes will be used to produce them. After briefly outlining the approach, we present some results on the environmental consequences of growing switchgrass to supply bioenergy conversion facilities in North Dakota and Tennessee.

1. Characterizing the region. In the first stage, we characterize the region's climate, topography, soil quality, and the types, location, management practices and profitability of current land use. If possible, we determine the relationship between profit and soil type. Understanding this relationship greatly improves the projections of where energy crops will be grown.
2. Developing energy crop management scenarios and production costs. In the second stage we determine management practices for energy crops and estimate production costs. This determination is based in part on our characterization of the region. Some energy crops are more appropriate for certain soils and climates than others.
3. Modeling crop yields and on-site environmental impacts. The third stage predicts variations in crop yield associated with soils and climates. It also predicts crop-specific, on-site environmental parameters such as erosion, runoff, and losses of nutrients and pesticide. This stage uses EPIC (Erosion Productivity Impact Calculator), a crop simulation model developed by the U.S. Department of Agriculture (Williams et al., 1989). The model is sensitive to many factors that control regional environmental impacts: soil type, topography, climate, crop type, and crop management. The model can predict not only crop yield but also the rates of erosion, runoff, and nitrogen and phosphorus losses in runoff or groundwater, all as functions of climate, soil, and topography. Changes in the environmental values associated with switching land to a specific energy crop can easily be calculated with EPIC. The crop yield information provided by EPIC is combined with empirical crop yield information to predict variations in yield associated with different soil and climate conditions.
4. Calculating probable farmgate biomass price. Using information generated in Stage 1 on the profitability of conventional crops and information on energy crop yields and production costs, we next calculate the price of biomass at the farm that gives the farmer the same profit as conventional crops. It does not include the cost of transporting the biomass to where it will be used. The break-even farmgate price is used to identify the lands most likely to be converted to energy crop production.
5. Determining where land use change will occur. In the fifth stage we predict land use changes, assuming that with transportation costs to the conversion facility being equal, land with the lowest farmgate price for biomass crops will be converted first, since it will produce the least expensive biomass for a conversion facility.
6. Evaluating environmental impacts. Regional impacts on soil fertility, water quality, and air quality largely depend on how much and what type of land is converted. They can be calculated by linking the environmental impacts per hectare determined in Stage 3 to the land-use changes predicted in Stage 5. Regional wildlife impacts depend not only on how much and what type of land is converted, but also on the location of that land in relation to other land uses. To evaluate these impacts, one must create maps of the changes in landscape pattern created by the projected land use changes. These maps are then used as inputs to spatial models of animal behavior and habitat to examine wildlife impacts. Effects on regional water quality depend on how much and what type of land is converted, where it is in relation to other land, and the topographic position of the land in relation to streams and lakes. Besides mapping projected land use changes associated with biomass production, one must also link those maps with topographic maps of the same area. Water quality models such as AGNPS (Agricultural NonPoint Source pollution), which predicts stream water quantity and quality as a function of topography, climate and land use pattern (Engel et al., 1993), can then be used to predict regional impacts on water quality.

Some results from two environmental analyses using this six-stage approach are shown in Fig. 1. Figs. 1a and 1b show the predicted impacts on regional soil erosion and nitrate loss from growing switchgrass for a hypothetical biomass energy facility near Fargo, North Dakota (English et al., 1993). The figures were created by comparing the expected water erosion and nitrate loss in runoff under two scenarios: 1) only conventional crops (wheat, oats, barley) are grown on all cropland within 50 km of the facility; 2) this land also produced switchgrass to supply an energy facility needing varying amounts of biomass (1,000 to 10,000 t/day). The corresponding land requirement ranged from 47,000 to 320,000 ha. Switchgrass production is predicted to reduce soil nitrate loss and water erosion in this region but the absolute amounts are low because the land is flat and rainfall is low.

Figs. 1c and 1d show some results from an analysis of the regional implications of growing switchgrass in two multi-county areas in Tennessee, one centered in Memphis, the other in Nashville (Graham and Downing, 1993). Converting 20% of the cropland in each region to switchgrass is predicted to reduce regional erosion by almost 20% (Fig. 1c). The absolute reduction in erosion is predicted to be greater in the Memphis region because of its significantly higher initial erosion rates. Increased switchgrass production reduces nitrate losses in both regions. Because of high predicted nitrate losses from soybeans, the benefits of producing switchgrass are higher in the Memphis region because of the greater displacement of soybean in that area (Fig. 1d). Thus, conversion of 20% of the cropland to switchgrass production is predicted to reduce regional nitrate loss in runoff by almost 20% in the Memphis region, but by only 10% in the Nashville region.

The results illustrate the potential environmental benefits of switchgrass production, but also emphasize the regional differences that may be expected. Biomass energy crops have many environmental benefits if they are well managed and well suited to the site; however, they are not universally advantageous. The government faces a serious challenge in developing policies and regulations that take advantage of the potential environmental benefits of energy crops and minimize their potential negative impacts. The value of the environmental benefits needs to be quantified and incorporated in cost comparisons between biomass and other energy sources.

References

1. Cook, J.H., J. Beyea, and K.H. Keeler. 1991. Potential impacts of biomass production in the United States on biological diversity. Annual Review of Energy and Environment 16:401-431.
2. Engel, B.A., R. Srinivasan, J. Arnold, C.C. Rewerts, and S.J. Brown. 1993. Nonpoint source (NPS) pollution modeling using models integrated with geographic information systems (GIS). Water Science and Technology 28:685-690.
3. English, B.C., K.D. Dillivan, M.A. Ojo, R.R. Alexander, and R.L. Graham. 1993. An evaluation of the regional supply of biomass at three Midwestern sites. In Proceedings of the First Biomass Conference of the Americas, Burlington, Vermont, August 1993. NREL/CP-200-5768. National Renewable Energy Laboratory, Golden, Colorado. pp. 106-111.
4. Graboski, M.S. 1993. An overview of the effect of fuel properties on emissions from biomass fuels. In Proceedings of the First Biomass Conference of the Americas. Burlington, Vermont, August 1993. NREL/CP-200-5768. National Renewable Energy Laboratory, Golden, Colorado. pp. 1651-1662.
5. Graham, R.L., and M. Downing. 1993. Renewable biomass energy: Understanding regional scale environmental impacts. In Proceedings of the First Biomass Conference of the Americas, Burlington, Vermont, August 1993. NREL/CP-200-5768. National Renewable Energy Laboratory, Golden, Colorado. pp. 1566-1581.
6. Graham, R.L., L.L Wright, and A.F. Turhollow. 1992. The potential for short-rotation woody crops to reduce U.S. CO2 emissions. Climatic Change 22:223-238.
7. Hoffman, W., J. Beyea, and J.H. Cook. 1995. Ecology of agricultural monocultures: Some consequences for biodiversity in biomass energy farms. In Proceedings of the Second Biomass Conference of the Americas: Energy, Environment, Agriculture, and Industry, Portland, Oregon. NREL/CP-200-8098. National Renewable Energy Laboratory, Golden, Colorado. pp. 1618-1627.
8. Hohmann, N., and C.M. Rendleman. 1993. Emerging technologies in ethanol production. Agricultural Information Bulletin No. 663. U.S. Dept. of Agriculture, Economic Research Service, Washington, D.C.
9. La Rovere, E.L., and P. Audinet. 1993. Environmental benefits of the Brazilian ethanol program. In Proceedings of the First Biomass Conference of the Americas, Burlington, Vermont, August 1993. NREL/CP-200-5768. National Renewable Energy Laboratory, Golden, Colorado. pp. 1533-1546.
10. Liu, W., V.D. Phillips, and D. Singh. 1992. A spatial model for the economic evaluation of biomass production systems. Biomass and Bioenergy 3:345-356.
11. Liu, W., R.A. Merriam, V.D. Phillips, and D. Singh. 1993. Estimating short-rotation Eucalyptus saligna production in Hawaii: An integrated yield and economic model. Bioresource Technology 45:167-176.
12. Macedo, I.C. 1993. Fuel ethanol production in Brazil. In Proceeding of the First Biomass Conference of the Americas, Burlington, Vermont, August 1993. NREL/CP-200-5768. National Renewable Energy Laboratory, Golden, Colorado. pp. 1185-1193.
13. Marland, G., and A. F. Turhollow. 199 1. CO₂ emissions from the production and combustion of fuel ethanol from corn. Energy 16:1307-1316.
14. Office of Technology Assessment. 1993. Potential environmental impacts of biomass energy crop production-Background paper. OTA-BP-E-118. U.S. Congress, Washington, D.C,
15. Perlack, R.D., J.W. Ranney, and L.L. Wright. 1992. Environmental emissions and socioeconomic considerations in the production, storage, and transportation of biomass energy feedstocks. ORNL/TM-12030. Oak Ridge National Laboratory, Oak Ridge, Tennessee.
16. Ranney, J.W., and L.K. Mann. 1994. Environmental considerations in energy crop production. Biomass and Bioenergy 6:211-228.
17. Tolbert, V.R., and A. Schiller. 1996. Environmental enhancement using short-rotation woody crops and perennial grasses as alternatives to traditional agricultural crops. In W. Lockeretz (ed.), Environmental Enhancement through Agriculture. School of Nutrition Science and Policy, Tufts University, Medford, Massachusetts. pp. 209-216.
18. Turhollow, A.F., and R.D. Perlack. 1991. Emissions of CO₂ from energy crop production. Biomass and Bioenergy 1:129-135.
19. Tyson, K.S., C.J. Riley, and K.K. Humphreys. 1993. Fuel cycle evaluations of biomass-ethanol and reformulated gasoline. NREL/TP-463-4950. National Renewable Energy Laboratory, Golden, Colorado.
20. Williams, J.R., C.A. Jones, J.R. Kinery, and D.A. Spanel. 1989. The EPIC crop growth model. Transactions of the American Society of Agricultural Engineers 32:497-511.
21. Wright, L.L., J.H. Cushman, A.R. Ehrenshaft, S.B. McLaughlin, S.A. Martin, W.A. McNabb, J.W. Ranney, G.A. Tuskan, and A.F. Turhollow. 1993. Biofuels Feedstock Development Program Annual Progress Report for 1992. ORNL- 6781. Oak Ridge National Laboratory, Oak Ridge, Tennessee.