Life Cycle Emissions and Health Cost Impacts of Producing Ethanol and Electricity from Willow and Switchgrass in the Riparian Buffers of Mid-Atlantic United States

Abstract

The United States (US Mid-Atlantic Region (MAR) has the potential to grow a variety of perennial feedstocks such as switchgrass and shrub willow to increase domestic energy production. These cellulosic feedstocks have also shown improved ecosystem services, such as soil carbon sequestration, nitrate leaching reduction, and flood mitigation along rivers and streams as partially harvested riparian buffers. To examine the effects on greenhouse gases (GHGs) and criteria air pollutants (CAPs) from using these feedstocks to produce ethanol or electricity, we conducted a comprehensive life cycle assessment (LCA) and estimated the impact on human health costs when land use is changed from corn production for ethanol. Results indicate up to 54% reduced GHG per hectare from using willow and switchgrass feedstock sources to produce ethanol instead of corn. However, there was a trade-off in terms of CAP emission, as grass-based energy emitted more NO_x and SO _x compared to the corn ethanol pathway, except for SO_X emissions from willow-based electricity. Electricity from cellulosic biomass had higher particulate matter (PM) emission compared to that from corn ethanol. Estimates for health cost to society ranged from $2498 ha^–1 for electricity from switchgrass to a net benefit of $448 ha^–1 for ethanol production from willow, depending on varying biomass yield under different market scenarios. Although using cellulosic feedstocks to produce bioenergy has great potential to reduce GHG emissions, CAP control measures are needed to manage CAP-induced health costs.

1. Introduction

The US is working to increase its domestic energy production and diversify its energy sources. In 2023, the US energy sector consumed approximately 99 billion GJ of energy, of which only 8.54 billion GJ (9%) came from renewable sources, such as hydroelectric, biomass, geothermal, solar, and wind. Among the renewable sources, biomass contributed 60% to total US energy production and 5% to renewable electricity production. Cellulosic biomass can potentially increase the renewable share of the total energy production in the US, in both electric and liquid fuel sectors. Investments in domestic renewable energy sources such as biofuels can create rural jobs in addition to mitigating greenhouse gas (GHG) emissions and combating energy security concerns amidst fluctuating oil prices in an uncertain future. − The share of US biomass-based electricity can be increased significantly if dedicated bioenergy feedstocks, such as perennial grasses and short rotation woody crops, are utilized. According to the Billion Ton study by the US Department of Energy, about 361 million dry Mg of energy crops will be available under the mature-market medium scenario, which can reach up to 579 million dry Mg under the mature-market high scenario. This can be utilized to meet the increasing US energy demands.

According to the US Department of Energy’s Annual Energy Outlook, residential and industrial electricity consumption will increase by up to 22% and 38%, respectively, by 2050, compared to 2022 level, driven by rising demand for space cooling and economic growth. On the liquid energy side, the US is a global leader in biofuel production, contributing over 38.1% (57.5 billion liters) of the world’s annual production, followed by Brazil at 35.6 billion liters. , The Environmental Protection Agency (EPA) Renewable Fuel Standard (RFS) is a policy designed to enhance U.S. energy security by promoting the development of advanced and cellulosic biofuels. Advanced biofuels are produced from renewable sources such as cellulose, sugar, or waste material, rather than corn starch. These biofuels are considered valuable domestic resources due to their ability to increase energy diversity, reduce carbon intensity (CI) compared to corn grain-based values, and their potential to produce biopower and other bioproducts. However, scaling up advanced biofuels remains challenging due to high production costs and insufficient investment in research and development (R&D). , These barriers make it difficult to meet RFS targets, , underscoring the need for further research into advanced bioenergy sources.

Following the US Department of Energy’s billion-ton report classification of biomass resources from agricultural lands, most cellulosic biofuels production commercialized so far has come from agricultural residues (e.g., corn stover). For instance, in 2022, about 1.4 million gallons of cellulosic ethanol were produced mainly from about 13,000 dry tons of agricultural residues, while energy crops remain largely in research pilot stages without significant commercial-scale biofuel production. However, dedicated energy crops have the potential to reach 65% of economically available agricultural biomass and 32% of overall national biomass potentially available under mature market conditions. Dedicated energy crops are divided into two main subcategoriesshort-rotation woody crops (like willow) and herbaceous energy crops (like switchgrass). Both willow and switchgrass have been identified as promising bioenergy options for the eastern US. Due to their high productivity, these crops can be grown on land unsuitable for food production such as on marginal land and abandoned mine land, which helps avoid the competition with lands growing food crops. Depending on the land types, perennial bioenergy crops can sequester enough carbon to have low carbon or potentially negative carbon emissions. − Perennial biomass feedstocks have also been shown to generate environmental benefits such as decreased nitrate leaching to nearby streams and improved water quality. − However, the calculation of environmental benefits requires a comprehensive life cycle assessment (LCA).

LCA is designed to evaluate all stages of a product’s life beginning from raw material collection to the end use of the product. LCA results vary based on different assumptions considered (e.g., biomass yield, process inputs) as well as allocation methods such as mass-, energy-, and market-based allocation. To increase transparency, it is extremely important to provide the key premises involved such as system boundaries, allocation methods, and general inputs. Existing LCA studies comparing biofuels from sugar and cellulosic sources reported that ethanol produced from corn grain is less carbon intensive than the fossil fuel life cycle but more carbon intensive than advanced biofuel sources. For example, corn-based ethanol has an estimated average range of emitting 59–68.9 gCO₂e/MJ. − Biofuel from cellulosic sources, on the other hand, has lower values ranging from 12 to 32.46 gCO₂e/MJ for willow, from 12 to 32.37 gCO₂e/MJ for switchgrass, from −7 to 35.3 for miscanthus, and from −3.7 to 32.1 gCO₂e/MJ for corn stover. ,− Most of these LCA studies were calculated on a per-energy basis (e.g., 1 MJ). Although this is valuable for making appropriate standardized emission comparisons across fuels and for regulatory purposes, normalizing by land area (e.g., one hectare, ha) is more appropriate for landscape level LCA because it allows comparing between different crop production in a specific land base. ,

Compared to corn, perennial biomass has been found to be a reliable and sustainable bioenergy feedstock on marginal lands. Both willow and switchgrass are known for their high productivity, with yields varying from 3.6 to 14.6 Mg ha^–1 for willow and from 1.26 to 10.7 metric ton (Mg) ha^–1 for switchgrass, depending on soil quality and fertilization inputs. , Data sets from the Policy Analysis (POLYSYS) model generated for the US Department of Energy’s recent national analysis of biomass resources known as the Billion-Ton 2023 Report indicate that the US Mid-Atlantic Region (MAR) produced 64.4 million metric tons (dt) of corn for ethanol in 2023 (Figure ). Under a scenario of mature bioenergy market demand with energy crop yield improvements of 1% per year, 9.9 million dry ton (dt) of switchgrass and 4.3 million dt of willow for bioenergy could be produced by this region with a price offering of $70 dt^–1 (while still producing 64.4 million dt of corn). We have previously shown that perennial feedstocks could be grown within 90 meters (m) of stream channels throughout the MAR to provide ecosystem services in addition to providing biomass for bioenergy. Riparian buffers are defined as areas close to a water body featuring trees, shrubs, and/or perennial plants that are managed for conservation purposes but do not exclude agriculture.

1.

US mid-Atlantic region (MAR), comprised of US Geological Survey Hydrologic Unit Codes (HUCs) 02, 05, 06, 0301, and 0410–0415. The map shows corn production in the year 2023. Note that many of the mountainous counties in the center of the region do not currently produce corn.

In the MAR, annual crop production is common in riparian areas. Replacing annual crops (such as corn) with woody and herbaceous crops (such as switchgrass and/or willow) could provide ecosystem service benefits along with biomass for biofuel and/or bioenergy. Producing biofuel from perennial crops would use fewer chemical inputs in the buffer strips and would help prevent or reduce sediment, nutrients, pesticides, and other materials into the streams, lakes, or wetlands. , Direct land use change from annual to perennial crops has been attributed to increase biomass yield by 19% and soil organic carbon (SOC) by 4%. Herbaceous bioenergy crops in buffer strips are also effective in reducing nitrogen leaching to nearby waterbodies. ,−

In this study, we conducted a comprehensive LCA of two perennial cellulosic biomass options for ethanol and electricity production: shrub willow (Salix species such as Salix amygdaloides or Salix purpurea) and switchgrass (Panicum virgatum) grown in place of corn. Both switchgrass and willow have high energy crop potential in the Northeast region and have been previously modeled for an examination of potential ecosystem services. ,, Along with GHG and criteria air pollutants (CAP) estimation, we translated the direct and indirect particular matter (PM_2.5) emissions and emissions of sulfur dioxide and nitrogen oxide emissions into potential human health impacts and associated costs using data sets from the US Environmental Protection Agency. Literature suggests that health monetized benefit may range from approximately $1300 from reducing a ton of NO_X from Ocean-Going vessels to about $450,000 from reducing a ton of PM_2.5 from Iron Steel industries (both $2010).

This study fills the gap in the literature in several important ways. Previous studies ,,, focused on differences in GHG emissions but lacked proper comparisons on CAP (such as SO_X, NO_X, particulate matter or PM) emissions and associated health costs. Additionally, most studies were conducted using an attributional LCA approach where the energy and emission of individual processes were allocated among different coproducts. In this study, we used consequential LCA that takes into account both the direct and indirect effects of crop substitution such as land-use change and associated demand mitigation, expanding the boundaries of the LCA. , Consequential LCA is defined as an attempt to provide information on direct and indirect environmental burdens as a consequence of a decisiongenerally associated with a change in demand for a product. For example, besides bioethanol production, corn also produces Distiller’s Dried Grains with Solubles (DDGS), a high-protein ingredient for animal feed (e.g., for cattle, swine). If land planted in corn were replaced by perennial crops in MAR’s riparian areas, additional land would be required to produce DDGS to meet the existing demand. There is a gap in the literature in comparing LCA results using such a harmonized scenario, and therein lies the primary novelty of this study. A second novel feature is the combined use of geospatial tools to assess riparian buffers (BioVEST), transportation costs, GHG emissions, and valuation. BioVEST is a software to conduct valuation and visualization of ecosystem services for water consumers. Finally, this study makes a significant contribution by providing an economic platform to evaluate multiple bioenergy pathways based on their production practices, associated air emissions, and potential human health impacts. Results from this study can provide an economic tool to policy makers regarding crop production in riparian areas.

2. Methods

2.1. Site Description

The US Mid-Atlantic region includes 10 million acres of formerly mined and other marginal agricultural lands that can be reclaimed and used to produce dedicated energy crops. This region has experienced significant water quality degradationin part, due to agricultureand this is expected to worsen in the future. As result, incentives and credits are available to support growing advanced perennial biomass crops; e.g., incentives to reduce nutrient runoff are available in many watersheds. MAR also includes ∼13 million ha of annual cropland in the riparian buffer strips with an average buffer width of 90 m (Table ). The analysis here builds on one published previously using the BioVEST model. Although the region currently produces lower amounts of bioenergy feedstock than the Midwest, it has substantial potential for future production of perennial feedstocks (e.g., switchgrass and willow) to support production of advanced biofuels.

1. Potential Mid-Atlantic Region (MAR) Land Conversion for Perennial Riparian Buffers Based on Calculations with the 2022 USDA Cropland Data Layer.

Buffer Width (m)	Annuals (million ha)	Small Grains (million ha)	Grassland/Pasture (million ha)	Total Buffer Area (million ha)	Corn Area (million ha)	Corn Proportion (%)
30	3.4	1	3.2	7.6	1.5	20
60	7.7	2.3	6.7	16.7	3.4	21
90	13.1	4	10.8	27.9	5.8	21
120	17.9	5.4	14.1	37.5	7.9	21

2.2. Life Cycle Assessment

2.2.1. Goal and Scope Definition

The goal of this study is to assess potential environmental and cost trade-offs of producing advanced biofuels and biopower from willow and switchgrass feedstock relative to ethanol production from corn grain. For this analysis, we conduct a consequential life cycle assessment study of producing ethanol and electricity from these biomass feedstocks by using harmonized scenarios. The five pathways considered in this study are (i) willow to ethanol, (ii) willow to electricity, (iii) switchgrass to ethanol, (iv) switchgrass to electricity, and (v) corn to ethanol. Corn grain to ethanol was considered the baseline pathway, as corn production is assumed to be the default land use in the riparian buffers. The harmonized demand assessment due to direct land use change is discussed in the life-cycle inventory section. LCA was conducted using Argonne National Laboratory’s Greenhouse Gases, Regulated Emissions, and Energy Use in Technologies (GREET) model. The metrics assessed include greenhouse gases (CO₂, CH₄, N₂O, VOC, and CO), criteria air pollutants (particulate matter or PM, oxides of nitrogen or NO_X, oxides of sulfur or SO_X). The functional unit was 1 ha of land, e.g., Mg CO₂e ha^–1, Mg SO_X ha^–1, or Mg PM_2.5 ha^–1. Results on an energy basisper 1 MJare available in the Supporting Information.

2.2.2. Life Cycle Inventory of Bioethanol and Biopower Pathways

Input required for biomass production is presented in the Supporting Information (Table S1). Average biomass yield across the MAR was calculated from the POLYSYS modeling output files generated for the DOE Billion-Ton 2023 Report (Table ). Three scenarios were considered, as described in the Billion-Ton report, and are as follows: (i) MMH = Mature Market High (3% annual energy crop yield improvements with increased harvest efficiency); (ii) MMM = Mature Market Medium (1% annual energy crop yield improvements with increased harvest efficiency); (iii) MML = Mature Market Low (no energy crop yield or technology improvements).

2. Main Assumptions for the Conversion of MAR Biomass To Bioethanol and Biopower.

Input	Scenario	Units	Willow	Switchgrass	Corn
Biomass yield	MMH	Dry t ha–1	18.0	19.2	11.0
	MMM	Dry t ha–1	16.0	15.1	9.78
	MML	Dry t ha–1	15.2	13.0	9.78
Ethanol yield from biomass		L t–1	93.7	88.2	112.73
Electricity yield from biomass		kWh gal–1	1.79	1.79
Soil carbon sequestration (from corn) ,−		t CO2e ha–1	0.72	1.17

^aMMH = Mature Market High (3% annual energy crop yield improvements with increased harvest efficiency).

^bMMM = Mature Market Medium (1% annual energy crop yield improvements with increased harvest efficiency).

^cMML = Mature Market Low (no energy crop yield or technology improvements).

Biomass to ethanol and electricity production data came from GREET. Input required for ethanol production at the biorefinery and electricity production at a 10 MW power plant is available in the Supporting Information (Table S2).

There were two primary assumptions for the output harmonization. First, the total quantity of outputs is assumed to be the same across all pathways. Second, if a certain pathway produces less of a certain output or does not produce at all, then the deficit must be met with additional production. Since electricity pathways do not produce ethanol, the deficit in demand created by not producing ethanol from those feedstocks is assumed to be met by producing additional conventional gasoline from fossil sources. Additionally, since ethanol from willow and switchgrass does not produce DDGS, we assumed that additional DDGS was produced from corn. On the other hand, the corn-to-ethanol pathway does not produce electricity as the perennial grass to ethanol pathways do. Therefore, additional electricity demand was assumed to be met by the use of electricity from the grid. Also, since dedicated electricity pathways produce more electricity per unit land than electricity coproduced via perennial biomass to ethanol pathways, we assumed that additional electricity from the grid would be required in ethanol pathways to harmonize the demand scenario. Table shows the energy required to harmonize all five scenarios.

3. Energy Required (10³ MJ ha^–1 of Fuel or Power) for Harmonized Scenarios .

Harmonized scenarios criteria	Yield	Willow|Ethanol	Willow|Electricity	Switchgrass|Ethanol	Switchgrass|Electricity	Corn|Ethanol
Ethanol produced	MML	114.6	0.0	92.6	0.0	88.8
	MMM	120.9	0.0	107.4	0.0	88.8
	MMH	135.8	0.0	136.4	0.0	100.2
Power produced	MML	9.2	50.8	7.4	40.9	0.0
	MMM	9.7	53.6	8.6	47.5	0.0
	MMH	10.9	60.2	10.9	60.3	0.0
Fuel required	MML	0.0	114.6	22.0	114.6	25.8
	MMM	0.0	120.9	13.5	120.9	32.1
	MMH	0.6	136.4	0.0	136.4	36.2
Power required	MML	41.7	0.0	43.4	9.9	50.8
	MMM	50.6	0.1	51.7	0.0	60.3
	MMH	49.4	0.1	49.4	0.0	60.3

^aMML = Mature Market Low (no energy crop yield or technology improvements). MMM = Mature Market Medium (1% annual energy crop yield improvements with increased harvest efficiency). MMH = Mature Market High (3% annual energy crop yield improvements with increased harvest efficiency).

2.2.3. Life Cycle Emissions

We used the GREET excel version 2022 database to carry out the life cycle assessment. The GREET database is developed and annually updated by the System Assessment Center of the Argonne National Laboratory. We followed hybrid allocation approach similar to that by Masum et al. where all processes are submitted to an energy allocation, except electricity displacement credit, biomass combustion, and transportation and distribution (T&D) processes. In our case, harmonized criteria inputs have been included, as shown in eq :

$$\mathrm{emissions}=({\mathrm{E}}_n+\mathrm{F\&P}+\mathrm{T\&}{\mathrm{D}}_{\mathrm{Biomass}}+\mathrm{FM}+\mathrm{Eco})\times \mathrm{energy}\mathrm{allocation}-\mathrm{DC}+\mathrm{T\&}{\mathrm{D}}_{\mathrm{Fuel}}+\mathrm{Fco}+\mathrm{Re}+\mathrm{Sc}$$ 1

E_n refers to emissions related to energy use in farming. F&P refers to the emissions from fertilizer and pesticide use. T&D_Biomass refers to the emissions related to the transportation of biomass from farms to the biorefinery and/or power plants, assuming an 80 km distance. FM and Eco refer to the emissions related to fermentation catalysts and energy use in conversion process, respectively. DC refers to the displacement credit for electricity and/or DDGS production. T&D_Fuel refers to the emissions from transportation of fuel from the biorefinery to gas pump. Fuel transportation assumes that 13.2% of ethanol transportation is conducted by barge, 78.9% by rail, and 7.9% by truck with travel distances of 520, 600, 800, and 80 miles, respectively, in addition to 30 miles trip by truck for ethanol distribution. Fco is the emissions during the combustion of biomass to produce electricity or combustion of ethanol. Re refers to the emissions related to mitigating the biofuel and/or electricity deficit from displacing corn by willow or switchgrass, which was filled by gasoline fuel on an energy basis. For comparison, we also show the results of a sensitivity analysis of corn-based ethanol as a replacement. Tables S3–S9 in the Supporting Information present details of these emission factors.

Sc refers to the carbon changes due to land-use change from corn to cellulosic biomass. Table presents the average soil organic carbon (SOC) sequestration amount due to the direct land use change from corn to willow and switchgrass taken from the Jager et al. ecosystem services benefit study, which is based on several other studies. ,,,− Across the MAR, the average SOC change for land converted from corn to switchgrass is an increase of 970 kg ha^–1 year^–1, and the average SOC change for land converted from corn to willow is an increase of 520 kg ha^–1 year^–1. These increases correspond to negative emissions, i.e., sequestration.

2.3. Health Cost Assessment

Data on health benefit of reducing a ton of direct and indirect PM_2.5 and ozone precursors were obtained from the US Environmental Protection Agency, which used methods described in Fann et al. (Table ). For instance, the estimated monetary benefit average of reducing one ton of PM_2.5 emissions from refineries was reported to be $358,000. Economic values from the refinery sector were used for the life cycle CAP from displacing corn (pathway 5) to willow and switchgrass for bioethanol production (pathways 1 and 3). Industrial boiler sector numbers were used for the CAP delta of pathways 2 and 4. Additionally, a 90-m switchgrass buffer can reduce 45% of the nitrogen (N) and 50% of the phosphorus (P) runoff compared to corn production, and a 90-m willow buffer can reduce 20% of N and 8% of P runoff compared to corn production. ,− The estimated cost benefit assumption for water benefits due to phosphorus and nitrogen reduction was $67.4 for willow and $159.2 for switchgrass relative to corn (average in USD, 2019).

4. Monetary Benefits ($) of Reducing a Ton of PM_2.5 and Ozone Precursors in the Refinery and Industrial Boiler Categories .

	PM2.5-Related Benefits				Ozone-Related Benefits
Sector	PM2.5	SO2	NOx	NH3	NOx	VOC
2019, 3% discount rate, in 2025
Refineries	358,000	49,600	22,500	109,000	61,100	12,200
Industrial Boilers	169,000	37,200	13,400	75,800	61,700	12,500

The monetary benefit of transitioning from corn-to-ethanol production to either willow or switchgrass to bioenergy (ethanol or power) production is calculated using eq :

$$({Em}_{VOC}^{Corn|Ethanol}-{Em}_{VOC}^{Biomass|Bioenergy})\times {Ben}_{VOC}^{\sec tor}+({Em}_{{NO}_X}^{Corn|Ethanol}-{Em}_{{NO}_X}^{Biomass|Bioenergy})\times {Ben}_{{NO}_X}^{\sec tor}+({Em}_{{SO}_X}^{Corn|Ethanol}-{Em}_{{SO}_X}^{Biomass|Bioenergy})\times {Ben}_{{SO}_X}^{\sec tor}+({Em}_{{PM}_{2.5}}^{Corn|Ethanol}-{Em}_{{PM}_{2.5}}^{Biomass|Bioenergy})\times {Ben}_{{PM}_{2.5}}^{\sec tor}+({Em}_{{CO}_2}^{Corn|Ethanol}-{Em}_{{CO}_2}^{Biomass|Bioenergy})\times {Social}_{CO2}$$ 2

Em refers to the respective pollutant emissions (e.g., VOC) in tons per hectare for a given fuel pathway (e.g., corn to ethanol). Ben represents the monetary benefit of reducing a ton of PM_2.5 and ozone precursors, following the respective sector and emitter as outlined in Table . Social refers to the social cost of CO₂, which varies based on the discount rate: $17, $56, and $83 per metric ton of CO₂ for discount rates of 5%, 3%, and 3.5%, respectively. Equation was used to assess the health cost associated with each life cycle stage.

3. Results & Discussion

3.1. GHG Emissions

Lifecycle GHG emissions ranged from 6.7 to 17.1 MgCO₂e ha^–1 of land in the riparian areas (Figure ). When gasoline was assumed to meet the additional fuel demand, GHG emissions from the corn to ethanol pathway ranged from 14.2 to 17.1 Mg CO₂e ha^–1, depending on the market scenario-based yields. Generation of grid-electricity to meet the electricity production deficit in comparison with willow to electricity pathwaywhich had the highest electricity productionwas the most GHG intensive stage for corn-based ethanol. Since ethanol production from willow was higher than from switchgrass (except the MMH scenario), replacing gasoline added about 2–3 Mg CO₂e ha^–1 for corn-ethanol. Under MMH scenario, ethanol production was the highest for switchgrass. Still, GHG emissions were lower for the willow-to-ethanol pathway under all three scenarios, due to the lower farming energy and fermentation inputs. Compared to corn-based ethanol pathways, cellulosic ethanol provided between 7% and 54% GHG reduction depending on feedstock and market scenario. When corn ethanol is considered as a fuel replacement, the estimates were between 38% and 49%. This is in line with literature findings, where life cycle from cellulosic sources was found to be less GHG intensive than corn-based bioethanol. − In addition, sequestration of soil carbon (i.e., the increase in soil organic carbon and reduction of nitrous oxide (N₂O)) played a considerable role in reducing GHG from the cellulosic pathways. They contributed roughly 5% to 14.5% of total positive emissions in electricity from willow and switchgrass, respectively, under MMH considering gasoline replacement and under MML considering corn-based ethanol replacement, respectively.

2.

Life cycle GHG emissions of bioethanol and biopower harmonized scenarios from willow, switchgrass, and corn. Biomass yield assumptions: Low = Mature Market Low, Medium = Mature Market Medium, High = Mature Market High.

Compared with ethanol pathways, electricity produced from cellulosic biomass generated higher GHG emissions, primarily due to the larger quantity of fuel replacement. Switchgrass-based electricity was 31%, 35%, and 41% more GHG-intensive compared to switchgrass-based ethanol under the MML, MMM, and MMH scenarios, respectively. When corn-ethanol was used to meet the demand gapinstead of gasolineGHG emissions decreased by 0.5% to 0.7% for switchgrass-electricity and about 4% increase for willow-electricity.

3.2. Criteria Air Pollutants–SO _x , NO_x, and PM_2.5 Emissions

Life-cycle SO_X emission was the lowest, 4.5 kg SO _x ha^–1, for the willow to electricity pathway under the MML scenario (Figure ). Compared to the corn-based ethanol pathway, there was a 30.5% reduction in SO _x emissions for this pathway, when the ethanol gap was fulfilled by gasoline. The same estimate was 15.3% when corn-ethanol was used instead of gasoline to make up the demand. All other pathways incurred more SO _x emissions from the corn-ethanol baseline. Compared to switchgrass pathways, willow pathways overall presented lower sulfur emissions per hectare. Since willow has a lower sulfur content (500 ppm) than switchgrass (1,100 ppm) on a wet basis, the result is reflected in the biomass combustion category, which is the most SO _x intensive stage in both electricity pathways.

3.

Life-cycle SO _x emissions of bioethanol and biopower harmonized scenarios from willow, switchgrass, and corn. Biomass yield assumptions: Low = Mature Market Low, Medium = Mature Market Medium, High = Mature Market High.

Although the ethanol pathways from cellulosic feedstocks produced the lowest GHG emissions among all scenarios, they were the most SO _x intensive. This is primarily due to noncombustion emissions during the fuel conversion process; ethanol production from cellulosic biomass generates 1.19 g of SO _x L^–1 of ethanol during biomass conversion stage. Meeting the power demand also played a significant role in increasing the SO _x emissions for these pathways as well as corn-ethanol pathway, for which it was the most SO _x intensive stage. Compared to corn-ethanol, perennial feedstock farming reduced SO _x emissions by around 50% for switchgrass and 85% for willow, primarily due to lower fertilizer inputs.

Corn to ethanol life-cycle NO _x emission was the lowest among all pathways, ranging from 18 to 22 kg of NO _x ha^–1, depending on market scenario and gasoline replacement option (Figure ). Compared to corn-ethanol, NO _x emissions increased by 39% (under MMH) to 41% (under MML) for willow to ethanol. For switchgrass-ethanol, the same estimates were 49% and 41%, respectively. Switching gasoline with corn-based ethanol as the fuel replacement did not have a significant impact on total NO _x emissions for cellulosic ethanol. Overall, the switchgrass and willow pathways were comparable, with switchgrass pathways producing slightly higher emissions than willow pathways. Noncombustion emissions were the most NO _x intensive stage, as ethanol production from cellulosic biomass generates 2.91 g NO _x L^–1 of ethanol.

4.

Life-cycle NO _x emissions of bioethanol and biopower harmonized scenarios from willow, switchgrass, and corn. Biomass yield assumptions: Low = Mature Market Low, Medium = Mature Market Medium, High = Mature Market High.

For the dedicated electricity pathways, most NO _x emissions occurred because gasoline was used as a replacement to fill the coproduct deficit. NO _x emissions were reduced when corn ethanol was considered as the replacement. This change was not observed in the willow-ethanol pathway because it produced the most ethanol, and therefore needed no fuel replacement.

Particulate matter (PM) emissions were the lowest for willow-ethanol, 1 to 1.2 kg PM_2.5 ha^–1, depending on market scenario (Figure ). As we noted before, this pathway had the highest ethanol production and did not need additional gasoline replacement. Because the primary source of PM was gasoline replacement in other pathways, willow to ethanol was the least PM intensive pathway in our study. Compared to the corn-ethanol pathway, PM emissions were 59.4% to 62.5% emissions lower. Unlike SO _x and NO _x where cellulosic ethanol pathways showed the highest emissions, PM_2.5 intensity was the lowest.

5.

Life cycle PM_2.5 emissions of bioethanol and biopower harmonized scenarios from willow, switchgrass, and corn. Biomass yield assumptions: Low = Mature Market Low, Medium = Mature Market Medium, High = Mature Market High.

By contrast, the switchgrass-electricity pathway had the highest PM emissions (7.1 to 8.4 kg PM_2.5 ha^–1 based on market scenario) because of gasoline replacement, which constituted 84% of all PM emissions in this pathway. Biomass combustion for electricity generation was responsible for 13% of the total PM emissions. Overall, the PM emissions were at least 171% higher compared to the corn-ethanol pathway. When corn-based ethanol is considered a fuel replacement, the same estimate was 46%.

Although GHG emissions were reduced by switching land use from corn to perennial feedstocks, CAP emissions increased. Two major observations can be made from these results: (i) CAP emission reduction systems such as scrubbers can be installed at the refinery or power plant to minimize the damage and (ii) system boundary decisions regarding coproducts are important when comparing life-cycle emissions. Using county-level biomass production projections from the Billion Ton Study, US cellulosic feedstocks such as corn stover, wheat straw, forest residues, and switchgrass generated lower air pollutant emission (e.g., PM_2.5, CO, SO _x , NO _x , VOC) per unit volume of ethanol produced compared to the traditional corn-based ethanol. However, the system boundary in that study only included farming activities, while downstream activities, such as the transportation of feedstock to the biorefinery and the end use of biomass, such as combustion and/or conversion to fuel, were excluded. Our study also observed lower emissions from farming activity for cellulosic biomass production. Ultimately, the conversion/combustion process was the biggest CAP contributor, which is supported by other studies ,, along with demand harmonization. Besides system boundaries, different allocation methods and displacement criteria can also provide different LCA results; PM₁₀, VOC, and total energy and water consumption estimates are provided in the Supporting Information (Figures S2–S4).

Our LCA estimates normalized to land area are difficult to compare with existing literature due to the differences in system boundary and demand harmonization. In terms of energy-normalized results, literature shows GHG results ranging from 7.2 to 32.46 g CO₂e MJ^–1 of ethanol. , Our estimates indicate emissions from 3.3 to 21 g of CO₂e MJ^–1 for ethanol and electricity production from willow. For switchgrass, literature estimates vary from 12 to 32.37 g CO₂e MJ^–1 of ethanol, ,, whereas estimates from our study ranged from 6.5 to 29.4 g CO₂e MJ^–1 for biopower and bioethanol. Finally, for corn-based ethanol, literature estimates range from 59 to 68.9 g CO₂e MJ^–1, which is comparable to our estimate of 65.7 g CO₂e MJ^–1. − Emissions estimates based on unit energy production are provided in the Supporting Information (Figures S5–S12).

3.3. Health Cost Assessment

Switching land management from growing corn to perennial feedstocks was associated with higher health costs to society, except when the land is used for willow production and the biomass is used for ethanol production (Figure ). By switching the land to willow for ethanol production, at least $231 and up to $448 ha^–1 benefit can be achieved depending on biomass yields (Table ). Switching the land to switchgrass for ethanol production has the potential to bring a $304 ha^–1 benefit to society if a higher biomass yield is attained (19.2 dry metric ton ha^–1).

6.

Health benefit assessment value per hectare of displacing corn by willow and switchgrass to bioethanol or power production considering 2020 dollars 3% discount rate for the CO₂ social cost portion. Data labels result from considering gasoline and corn-based ethanol as replacement, respectively. Biomass yield assumptions: Low = Mature Market Low, Medium = Mature Market Medium, High = Mature Market High. Deviation bars accounts for other discount rate: 2.5% and 5% as adopted in regulatory analysis.

The main economic burden from switching the land from the corn to willow and switchgrass would come from gasoline replacement in electricity pathways and noncombustion emissions in ethanol pathways. The social cost of carbon due to the GHG emissions of additional gasoline combustion was the primary contributor, constituting about 67% of all costs in electricity pathways, bringing the net social health benefit to negative. For ethanol pathways, the noncombustion SO _x and NO _x emissions during fermentation of biomass constituted at least 82% of all costs. On the other hand, the main economic benefit would come from using less material or energy inputs required to produce bioenergy from cellulosic pathways compared to corn-ethanol. The positive economic health impact from the water benefits category due to less fertilizer useand therefore reduced phosphorus and nitrogen leakage to watershedswas negligible.

Replacing gasoline with ethanol showed the opposite impact for electricity and ethanol pathways, increasing the benefit in electricity pathways but decreasing in ethanol pathways. This is not surprising since combustion of gasoline was the primary reason for increased carbon costs in electricity pathways. Since life cycle GHG emissions of ethanol were lower than those of gasoline, meeting the additional fuel demand with ethanol decreases the social health cost. However, the net social benefit remains negative for all pathways and scenarios under this assumption.

A few studies have analyzed the monetary value of producing certain crops for biofuel production. ,, Unlike the present study, which provides the monetary impact considering the direct land use change from the annual crop, Thakrar et al. assessed air quality impact of switchgrass production and found a damage average cost to the human health to be U$33.01 Mg^–1 of biomass. If a similar system boundary is considered, i.e., energy in farming and farming input, but adding water benefit and soil carbon sequestration categories from switching the land from annual to perennial crops, the monetary benefit estimated by this study to human health would range from U$43.1 Mg^–1 to U$67.07 Mg^–1 of biomass (Table ). Hill et al. estimated the human health cost for the production and combustion of gasoline, corn ethanol, and cellulosic ethanol from switchgrass and found that cellulosic ethanol cost was lower than gasoline and corn. This estimate is comparable to our results for replacing corn with willow to produce ethanol for all biomass yield scenarios. Although these results are comparable to those of other studies, differences in assumptions and system boundary must be noted.

5. Monetized Health Value of Switching Corn to Willow and Switchgrass for Bioethanol and Power Production .

Harmonized scenarios criteria	Yield	Farming boundary only (USD/metric ton of biomass)	Life-cycle boundary (USD/metric ton of biomass)
Corn to Willow|Ethanol	High	51.32	24.93
	Medium	50.77	23.73
	Low	54.14	15.22
Corn to Willow|Electricity	High	49.57	–124.48
	Medium	49.11	–125.66
	Low	52.29	–131.90
Corn to Switchgrass|Ethanol	High	43.10	15.83
	Medium	53.83	–0.66
	Low	67.07	–24.28
Corn to Switchgrass|Electricity	High	43.61	–130.11
	Medium	53.82	–141.67
	Low	66.32	–158.37

^aBiomass yield assumptions: Low = Mature Market Low, Medium = Mature Market Medium, High = Mature Market High.

The findings of this study provide some important conclusions and grounds to guide future bioenergy production strategies. First, while GHG emissions from advanced biofuels from switchgrass and willow are lower than those of corn ethanol, there is a trade-off in terms of higher CAP emissions that are harmful to human health. Thus, air pollutant reduction technologies like vent or stack scrubbers and baghouse filters should be considered for advanced biofuels production. Prior studies have also concluded the need for emission controls in biorefineries. − Second, dedicated electricity production from willow and switchgrass scenarios would produce high levels of CAPs (in particular, PM_2.5) which, in turn, would elevate health cost without much reduction in GHGs compared to corn-ethanol. This indicates that unless advanced technologies are installed to offset harmful combustion emissions, cellulosic crops might be better directed for ethanol production instead of electricity generation. Diverting biomass for pellet production may reduce up to 73% of NO _x and 99% of sulfur dioxide. Ash content from biomass burning can be used as fertilizers in agricultural field which may reduce emissions from traditional fertilizer production, , or used in cement or nanotechnology application in industrial catalysis. Other options are the coupling of air staging technology to reduce up to 50% NO _x emissions in combustion processes, or coupling combined technology to reduce both PM_2.5 and NO_x. This is not a comprehensive list but gives a few options that biorefineries and/or power plants can consider to reduce criteria air pollutant emissions.

Future research can study the social-health-cost impact of using these technologies. Future studies can also address technoeconomic implications of such technologies, i.e., economic return of technology installation cost toward social health benefit. Additionally, future studies should delve deep into the spatial characteristics of the region and analyze the impact of utilizing multiple riparian buffer lengths, e.g., 30, 60, and 90 m. Biomass yields, carbon sequestration potential for switching crops, and therefore, the health cost estimates can vary by their locations. Despite these limitations, this study provides a comparatively new and strong platform for studying bioenergy crop production. This study focuses on harmonizing the total energy products, which is an important factor in improving energy security. Above all, the health cost analysis provides a fresh look at the comparison of multiple bioenergy pathways.

Glossary

Abbreviations

$

United States Dollar

DDGS

Distiller’s Dried Grains with Solubles

EISA

Energy Independence and Security Act

GHG

Greenhouse Gas

GREET

Greenhouse Gases, Regulated Emissions, and Energy Use in Technologies

ha

Hectare

kg

kilogram

LCA

Life Cycle Assessment

Mg

Metric ton or tonne

MJ

Mega Joule

NO _x

Oxides of Nitrogen

PM_2.5

Particulate Matter

RFS

Renewable Fuel Standard

SOC

Soil Organic Carbon

SO _x

Sulfur Dioxide

VOC

Volatile Compounds

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.est.5c05846.

• Life cycle inventory of farming corn, switchgrass, and willow, life cycle inventory of energy production from corn, switchgrass, and willow, emission parameters for inputs, transportation, biomass and ethanol combustion, displacement and replacement credits, life cycle PM₁₀ and VOC emissions, life cycle energy and water consumption, life cycle GHG, SO _x , NO _x , PM_2.5, PM₁₀, and VOC emissions on an energy basis, and life cycle energy and water consumption on an energy basis (PDF)

This study was sponsored by the U.S. Department of Energy (DOE), Bioenergy Technologies Office (BETO) under the Geospatial Analysis and Ecosystem Services Portfolios project and was performed by Argonne National Laboratory operated by UChicago Argonne, LLC under Contract No. DE-AC02-06CH11357, and Oak Ridge National Laboratory operated by UT-Batelle LLC under Contract No. DE-AC05-00OR22725.

The authors declare no competing financial interest.