Techno‐economic and life cycle analysis of renewable natural gas derived from anaerobic digestion of grassy biomass: A US Corn Belt watershed case study

Restoring native grassland vegetation can substantially improve ecosystem service outcomes from agricultural watersheds, but profitable pathways are needed to incentivize conversion from conventional crops. Given growing demand for renewable energy, using grassy biomass to produce biofuels provides a potential solution. We assessed the techno‐economic feasibility and life cycle outcomes of a “grass‐to‐gas” pathway that includes harvesting grassy (lignocellulosic) biomass for renewable natural gas (RNG) production through anaerobic digestion (AD), expanding on previous research that quantified ecosystem service and landowner financial outcomes of simulated grassland restoration in the Grand River Basin of Iowa and Missouri, United States. We found that the amount of RNG produced through AD of grassy biomass ranged 0.12–45.04 million gigajoules (GJ), and the net present value (NPV) of the RNG ranged −$97 to $422 million, depending on the combination of land use, productivity, and environmental credit scenarios. Positive NPVs are achieved with environmental credits for replacement of synthetic agricultural inputs with digestate and clean fuel production (e.g., USEPA D3 Renewable Identification Number, California Low Carbon Fuel Standard). Producing RNG from grassy biomass emits 15.1 g CO2‐eq/MJ, which compares favorably to the fossil natural gas value of 61.1 g CO2‐eq/MJ and exceeds the US Environmental Protection Agency's requirement for cellulosic biofuel. Overall, this study demonstrates opportunities and limitations to using grassy biomass from restored grasslands for sustainable RNG production.


| INTRODUCTION
Predominant land uses in the world's major agricultural regions are highly productive, but with negative outcomes for climate, soil, water, and biodiversity (Alexander et al., 2008;Lark et al., 2020Lark et al., , 2022;;Thaler et al., 2021).Financially viable alternatives are needed to meet the multiple demands of 21st-century society, which include ecosystem services and biodiversity preservation in addition to food and energy production (Mishra et al., 2019;Schulte et al., 2022).Over the past two decades, lignocellulosic biofuels, especially biofuels from perennial grasses, have been pursued as a pathway to address this need (Gelfand et al., 2013;Hallam et al., 2001;Jarchow et al., 2015;Meehan et al., 2013;Schmer et al., 2008;Tilman et al., 2006).Ecosystem services associated with grassland systems provide widespread public benefits, such as climate regulation, water purification, and recreational services, some of which can be monetized (Johnson et al., 2012;Meehan et al., 2013;Mishra et al., 2019).Strategically converting small areas of cropland, for example, where annual crops are underperforming or to protect key environmental resources such as streams and wetlands, to bioenergy grassland could substantially improve environmental outcomes while keeping landscapes in farmland (Brandes et al., 2016;Gelfand et al., 2013;Schulte et al., 2017;Ssegane & Negri, 2016).
Advantages of grassy feedstocks for renewable fuel production are that they can be grown in areas where annual row crop production is not profitable and/or areas of high conservation value (Brandes et al., 2016;Meehan et al., 2013;Mishra et al., 2019).Converting low-yielding cropland to grassland cover has the potential to improve the overall profitability of farm fields.The cost of grassland establishment and management can be favorable compared to the highly volatile production costs of row crops (Audia et al., 2022), and depending on local or regional market development, perennial grassland systems may outcompete annual grain systems in terms of profitability (Brandes et al., 2016;Gelfand et al., 2013;Manatt et al., 2013;Tilman et al., 2006).
Our goals for this paper are to inform renewable fuels policy and development by evaluating the economic feasibility of biofuel pathway we term "grass-to-gas," which includes renewable natural gas (RNG) production through the anaerobic digestion (AD) of harvested biomass from bioenergy grasslands.AD is a biochemical process in which microorganisms break down organic materials in the absence of oxygen to biogas, which mainly contains methane (CH 4 ).We evaluated the grass-to-gas pathway using a techno-economic analysis (TEA) and life cycle analysis (life cycle assessment [LCA]) to assess the net present value (NPV) and net carbon and greenhouse gas (GHG) emissions.We concentrated on RNG as the focal biofuel because, as of 2022, RNG comprises 90% of the advanced biofuel produced under the US Renewable Fuel Standard (US EPA, 2022a), while other nascent cellulosic biofuel projects have collapsed (Brown & Brown, 2013).Furthermore, RNG is incentivized by current and potential future state-level clean fuels policies in the United States, for example, the California Low Carbon Fuel Standard (Jaffe et al., 2016).
Previous studies have shown the possibility of producing RNG from grassy (lignocellulosic) biomass through AD (Braun et al., 2010).In this case, we simulated the AD of grassy biomass to produce biogas that is upgraded to pure methane, or RNG, by separating the RNG from the other gases that make up the biogas, primarily carbon dioxide (CO 2 ) but also small amounts of hydrogen sulfide, siloxanes, and water (Braun et al., 2010).
Through this analysis, we extend the work of Audia et al. (2022), who simulated a suite of financial and ecosystem service outcomes associated with strategically restoring and/or reconstructing grassland composed of native species as a biomass crop (hereafter referred to as grassy biomass) for the Grand River Basin (GRB), located in the US Corn Belt states of Iowa and Missouri.The GRB was chosen for these studies because it represents an agriculturally dominated watershed contributing to water quality impairments in the Mississippi River Basin.It also hosts projects that seek to simultaneously meet biofuel production, rural economic development, and conservation goals through grassland reconstruction and through a manureto-energy scheme (Betsy Freese, 2018).

| Land-use scenarios and grassy biomass production yields and costs
The hypothetical native grassland restoration scenarios evaluated by Audia et al. (2022) were designed to cost-effectively address water quality and biodiversity resource issues.Potential biomass revenue and the relative value of environmental enhancements associated with three land use scenarios were assessed: a "Baseline" scenario, based on 2016 land use/land cover data for the GRB; a "Buffered" scenario, which simulated the outcomes of replacing annual cropland located in riparian areas (i.e., within 20 m of a perennial stream) with grassland composed of native species; and a "Productivity-based" scenario, which simulated reconstructing grassland on low-productivity croplands (i.e., a yield potential of less than 0.5 based on the National Commodity Crop Productivity Index; Dobos et al., 2012).The estimated value of producing and selling grassy biomass ranged −$455 to $1291/ha/year in 2022 US dollars, depending on expected biomass yield and farmgate selling price.All environmental outcomes improved under the alternative compared to the Baseline scenario, including water quality, soil carbon storage, and pollinator habitat F I G U R E 1 Land use/land cover distributions and anaerobic digester sites in the Grand River Basin, located in southwest Iowa and northwest Missouri, United States, based on the National Land Cover Dataset (Wickham et al., 2021).measures, and were estimated to be of substantial value (Audia et al., 2022).Compared to the baseline, the buffered scenario resulted in a 7743 ha, or 0.4%, increase in grassland area at the watershed scale, with total potential revenue associated with grassy biomass harvests ranging from −$7.3 to $21.1 million.The estimated annual values for environmental enhancements for the buffered scenario were $1.7 million for nitrogen reduction, $0.1 million for phosphorus reduction, $0.5 million for sediment reduction, and $1.3 million for soil carbon storage.The productivity-based scenario simulated establishment of grassland on 91,274 ha, or 4.9% of the watershed, with an estimated value for harvested and sold grassy biomass of −$44.2 to $128.8 million.The estimated annual values for environmental enhancements for the productivity-based scenario were $18 million for nitrogen reduction, $1.4 million for phosphorus reduction, $2.5 million for sediment reduction, and $14 million for soil carbon storage.

| Techno-economic analysis
The TEA evaluated the economic feasibility of RNG production through AD of grassy biomass hypothetically harvested within a 40-km radius of two biogas production sites (sites A and B) in the GRB, based on baseline, buffered, and productivity-based land use/land cover scenarios (Audia et al., 2022).Continuously stirred tank reactor (CSTR) anaerobic digesters with the capability of handling and processing grassy biomass were hypothetically added to sites A and B (Figure 1); these sites already process lignocellulosic manure feedstocks through AD for RNG production (Betsy Freese, 2018).The material and energy balance of the systems were based on the model by Aui et al. (2019), Mosleh Uddin et al. (2022), and Washington State University's Anaerobic Digester System Enterprise Budget Calculator (ADBC, 2021).Materials and energy balances were used to estimate the capital and operating costs.The study employs a discounted cash flow rate of return (DCFROR) analysis developed by the national renewable energy lab to estimate the NPV of the system (Gifford et al., 2010).
This study assumes the following about grassy biomass harvesting, handling, and processing: The biomass is collected and chopped from a circular area surrounding the anaerobic digester and is transported by truck to the digester; the grassy biomass is mixed with water and pumped into the digester; each digester consists of a CSTR operating at mesophilic (20-42°C) temperatures (Braun et al., 2010); a base case 101 tonnes (t) of biomass are processed per day across the system, which is equivalent to the total biomass available for the baseline 13.5 t/ha scenario to produce 430 million MJ of RNG per year while producing liquid and solid digestates as by-products.A fraction of the water stream in the digester output would be recycled to maintain a 40 wt.% solid content inside the digester; raw biogas, a mixture of CH 4 and CO 2 , produced by digesters is compressed and transported by trucks in gas containers to a central upgrading facility, located approximately equidistant from each of the two digester sites.The raw biogas is upgraded to RNG by separating CO 2 and other undesirable gases to meet pipeline quality natural gas specifications; the remaining material, that is, digestate, is split into a nutrient-rich liquid digestate that can be applied to cropland as a fertilizer and carbon-rich solid digestate that can serve as a soil amendment on degraded soils.Figure 2a illustrates the process flow of the biomass to the RNG system.The AD process consists of four primary unit operations: feedstock preparation, digestion, by-product separation, and RNG upgrading.The biomass to CH 4 yield is 382 m 3 /t of volatile solids (382 m 3 /t) (Braun et al., 2010), with raw biogas containing 60% CH 4 content for this study.The biogas to RNG yield is 97%, with a 2% CH 4 loss.The solid digestate has 28% solids content.
The study employs the nth plant assumption, which implies the technology is well established and commercially deployed in the industry (Snowden-Swan et al., 2016).Table S1 shows the financial assumptions employed in this study.The TEA profitability analysis assumes that the operation, which includes the anaerobic digesters and upgrading facility, has a 20-year lifetime (Fusi et al., 2016).The facility is financed with 40% equity and a 7.5% interest rate.We assumed the facilities would require 2.5 years to construct, and during half of the final year, they would receive only 50% of the potential revenue generated while incurring 75% of the variable and 100% of the fixed costs.The facilities would depreciate over 7 years following a double-declining balance schedule with a final salvage value of $0.We assumed the working capital was 15% of the fixed capital cost, which is a common assumption in TEA studies.The income tax rate was 21%, which is similar to the tax rate in Midwestern US states.
The DCFROR analysis utilizes the capital cost, operating cost, and financial assumptions to estimate the NPV of the facility over its lifetime for various land use scenarios.The capital cost, a one-time expenditure, includes equipment costs, installation costs for piping, building, instrumentation, indirect costs, land purchase expenses, and working capital.The equipment cost consists of the anaerobic digesters, heat exchangers, power generators, and various pumps, while the indirect cost includes the cost of engineering design, legal fees, and others.Table S2 shows the equipment cost information.The base costs of equipment, piping, and installation were derived from (Humbird et al., 2011) and the ADBC, and scaled accordingly using the power law Equation (1), where C b is the cost of the base equipment at a baseline mass M b and C s is the estimated cost of the equipment at the required mass capacity, M s .n is the economy of scale sizing exponent.A scaling exponent n value of 0.72 (Axelsson et al., 2012) was used to scale the power generator, while a scaling exponent value of 0.6 (Peters et al., 2003) was used for every other piece of equipment.
Operating costs include annual expenditures for process materials and utilities, labor and management expenses, and miscellaneous costs such as insurance and legal fees (Table 1).We consider farmgate feedstock prices ranging from $0 (base case) to $100/t.The base case assumes that the burden of collecting the grassy biomass is borne by the system operator similar to the collection of waste materials.However, we also consider scenarios with feedstock collection and transportation costs.Based on an Idaho National Laboratory study (Jacobson et al., 2014), collection costs were assumed to be $24.5/t.Transportation costs include fixed costs of $4.84/t for loading and unloading and $0.21/t/mi for distance-related costs.Transportation distance calculations were based on the collection radius of the circular area required to collect the target biomass production at a 13.5 t/ha biomass yield if 60% of the biomass is available.A tortuosity factor of 1.3 was included to account for road topology factors.Details of this equation are discussed in Wright et al. (2008).The cost of handling solid and liquid digestate was assumed to be $5.00/t and $2.64/t, respectively.Biogas loading and distance-based transport costs were $4.24/t and $0.09/t/ mi, respectively.The study assumed the cost of piping the RNG produced to an injection point on the natural gas grid to be $0.39/GJ(Astill et al., 2018).Gas cleaning costs were set at 0.08 cents/kWh (Astill et al., 2018).The RNG generates environmental credits, as presented in Section 2.4, of −$15.84/GJ and −$1.5/RIN.Transaction fees were set at 20% of the credits received.

| Life cycle assessment
The study conducts a LCA following the principles outlined in ISO 14040 and ISO 14044 (Finkbeiner et al., 2006) to understand the emissions associated with the RNG produced from AD and upgrading.LCA is an established method for evaluating the environmental impact of systems and processes (Finkbeiner et al., 2006).The LCA aims to estimate the emissions associated with 1 MJ of RNG produced via AD and biogas upgrading.The study uses 1 MJ of RNG produced as the functional unit to facilitate easy comparison with other studies.The LCA employed is based on a well-to-pump analysis using Argonne National Laboratory's GHGs, Regulated Emissions, and Energy use in Transportation, GREET.net (GREET, 2020).The system boundary for the LCA model is shown in Figure 2b.
The LCA system boundary includes grassy biomass cultivation, biomass transportation, the digester, biogas transportation, and biogas upgrading mineral fertilizers displaced by the digestates from the by-products of the process.
In the LCA, we assumed that the carbon emissions generated during the processing and combustion of RNG produced from biomass cultivation are compensated by the carbon uptake that occurs during the cultivation of the biomass (Paolini et al., 2018).The primary emissions from feedstock production are associated with fertilizers used in farming and fuel used to power equipment/vehicles during cultivation and feedstock collection.The data for the emissions associated with cultivation and feedstock collection were gathered from GREET.net (GREET, 2020).Other significant sources of emissions include fugitive emissions from RNG losses in the upgrading process, electricity for powering digesters and upgrading facilities, and the petroleum fuel used in transporting the biogas produced at each digester to the central upgrading facility.This study does not include a detailed assessment of landuse change emissions from the conversion of cropland to grassland.GREET.net biomass emission factors include contributions from direct and indirect land-use change emissions stemming from inputs and activities required for land conversion, and consequential impacts of replacing cropland products.
The LCA inventory data were gathered from GREET (Table 2).Grassy biomass production incurs emissions of 0.08 kg CO 2 -eq/kg.These emissions include allocations for machinery emissions associated with harvesting and collecting grassy biomass.Changes in CO 2 -eq associated with changes in land use were beyond the scope of this analysis, although Ssegane and Negri (2016) consider soil carbon storage for the GRB.Transportation emissions to the AD facility are 0.03 kg CO 2 -eq/kg assuming an average travel distance of approximately 40 km.Grid electricity emissions from the digester operation and RNG upgrading processes are 0.45 kg CO 2 -eq/kWh based on the US power grid mixture.Regional emission factors could be used to improve the accuracy of the emission analysis, T A B L E 1 Operating cost assumptions for renewable natural gas production from grassy biomass in the Grand River Basin.we choose to employ the national average value to facilitate comparison across studies.RNG loss and digestate displacement are major emission factors for prairie grass digesters.RNG losses have an impact factor of 30 kg CO 2 -eq/kg of CH 4 because of the greater global warming potential of CH 4 compared to CO 2 .RNG losses may be less critical for food or dairy digesters because the business-asusual scenario would result in higher CH 4 emissions.This is a topic of active research and has been discussed in recent studies (Ankathi et al., 2018;Lee et al., 2021).Liquid and solid digestate are assumed to replace synthetic nitrogen fertilizers.Hence, avoided emissions from conventional nitrogen fertilizers are added as credits with 11.7 kg CO 2 -eq/kg impact factor.The effective replacement factor of conventional nitrogen fertilizers will depend on the nutrient content of the digestate, farm management practices, and other factors beyond the scope of this study.

Parameters
The AD system generates revenues from federal and state incentive programs based on the net emissions for RNG production.The renewable fuels standard (RFS2) provides US Environmental Protection Agency Renewable Identification Number (RIN) credits for biofuels that meet GHG reduction levels compared to fossil-based alternatives.RIN credits vary depending on the fuel type, supply and demand market dynamics, and policy drivers.For this study, we assumed that RNG based on AD of grassy biomass qualifies for D3 RINs, which are in high demand and valued at $1.5/RIN (US EPA, 2022b).In addition, the California Air Resource Board (CARB) provides credits through the Low Carbon Fuels Standard (LCFS, 2021) for fuels with lower environmental impacts than fossil-based alternatives (California Air Resources Board, 2019).CARB provides a calculator for estimating the LCFS credit based on the fuel's carbon intensity score, which is equivalent to the lifecycle GHG emission.For this study, we calculated the LCFS credit to be −$15.84/GJ.The study includes scenarios where revenue only comes from RNG sales, RNG sales and credits from solid and liquid digestate, RNG sales, digestates credits, and D3 RIN credits, and lastly, RNG sales, digestate credits, D3 RIN and LCFS credits.The analysis is conducted for all land use scenarios.The study also includes a scenario analysis to assess the NPV of the grassy biomass AD and upgrading considering all farmgate prices discussed in Ssegane and Negri (2016).

| RESULTS
The total amount of area devoted to bioenergy grassland in 40 and 80 km radii of each potential biofuel production site varied by scenario (Table 3).For the baseline scenario, the amount ranged from 1054 to 5700 ha, with site B having the most bioenergy grassland area within both 40 and 80 km radii.In the buffered scenario, the amount ranged from 1340 to 9883 ha, with site A having the most bioenergy grassland area within 40 and 80 km radii.For the productivity-based land use scenario, the amount ranged from 8108 to 66,848 ha, with site D having the highest area of bioenergy grassland in a 40-km radius and site B having the most in an 80-km radius (Table 3).However, these land-use scenarios overlap each other.Hence, the TEA was conducted using a 40-km radius for only sites A and B, which account for the most land area without overlap.

| Techno-economic analysis
Simplified material and carbon flow data for producing RNG from grassy biomass are shown in Figure 3.The AD system generates 22.9 t/day of RNG from 101.2 t/day of biomass with 60% moisture content.It also generates about 15.1 t/day of solid digestate and 211.1 t/day of liquid digestate.RNG carbon accounts for 43.6% of the total carbon in the biomass.
The baseline scenario's fixed capital investment (FCI) is $19.5 million.The FCI varies with the capacity of the digester sites and the upgrading facility.The FCI ranged from $17.96 to $28.92 million, with the minimum capital cost reported for the baseline land use 6.7 t/ ha biomass yield scenario, and the maximum FCI for the productivity-based land use 13.5 t/ha biomass yield scenario.This is expected as the equipment cost scales with the digester sites' capacity and biogas upgrading facility.Operating cost for producing RNG from grassy biomass under a baseline land use and digestate credits only (no environmental credits) scenario resulted in an operating cost of $10.12 and $9.25/GJ RNG, respectively (Figure 4a).Scenarios with digestate and RINs credits reduced operating costs to −$5.52/GJ RNG, whereas the all credits scenario reduced the operating cost to −$18.9/GJRNG (Figure 4a).
The NPV of selling RNG at different grassy biomass yield rates ranges from −$110 to $413 million (Table 4).NPVs range from −$110 to −$45 million for scenarios with no environmental credits.In general, scenarios that included all available credits are profitable; scenarios with higher grassy biomass productivity tend to be more profitable (see Table 3) productivity-based land use scenarios are more profitable than the baseline or buffered scenarios due to the availability of more biomass per hectare.More specifically, the productivity-based land use scenario in combination with the no environmental credits scenario incurs the lowest NPVs because it represents the largest amount of land converted to bioenergy grassland and the largest capital cost.The opposite is true for the baseline scenario, where digestate credits improved the revenue, but the NPV of all land use and biomass yield scenarios resulted in negative NPVs.For scenarios with digestate and either RINs or LCFS credits, the baseline and buffered 6.7 t/ha biomass yield scenario resulted in negative NPVs, but other scenarios resulted in positive NPVs.The power law of economies of scale supports this result.With all environmental credits considered, all land use scenarios resulted in positive NPVs.Unlike scenarios without environmental credits, the productivity-based land use scenario resulted in the largest NPVs because it has the largest amount of grassy biomass available.The productivity-based land use and all environmental credits scenario generate the largest NPVs ranging from $6.0 to $413 million.Increasing the yield and collection of grassy biomass from 6.7 to 13.5 t/ha increases the NPV.Increasing yields within the productivity-based land use scenario would generate the most economical return.This conclusion is also supported by the power law of economies of scale that dictates higher returns on a per-unit-of-RNG basis at increasing scales.We also assessed the effect of variation in farmgate prices for grassy biomass, as explored in Audia et al. (2022), on NPV (Figure 4b).For the base case analysis, we assumed $0 farmgate prices and only accounted for the collection cost from the field.All scenarios described above that resulted in negative NPVs F I G U R E 4 Operating costs for renewable natural gas production from grassy biomass under a baseline (no credit), liquid and solid digestate credits only, liquid and solid digestate and USEPA renewable identification number (RIN) credits only, and all environmental credits scenario.LCFS, California Air Resources Board's Low Carbon Fuel Standard.(b) Variation of the net present value of the anaerobic digestion and biogas upgrading facility to grassy biomass farmgate prices for scenarios with no environmental credits and all environmental credits.Base = baseline scenario, Prod = productivity scenario, Buff = buffered scenario, 6.5 t/ha = 6.5 t/ ha grassy biomass yield scenario, 13.5 t/ ha = 13.5 t/ha grassy biomass yield scenario.remained as we increased farmgate prices from $22 to $100/t.The NPV of all scenarios with all environmental credits remained positive except for the Baseline land use 6.7 t/ha biomass yield and environmental credits scenario, which was only profitable at farmgate prices less than $66/t of grassy biomass.

| Life cycle analysis
We estimate the life cycle carbon intensity of RNG production from grassy biomass (Figure 5).Results indicate a GHG emission of 29 g CO 2 -eq/MJ RNG without digestate credits.RNG loss, grass production, and biogas upgrading power use are the primary sources of GHGs within the system boundary.GHG emissions associated with grassy biomass production depend on agricultural management practices, including prairie establishment, cultivation, harvesting, and preprocessing.Grassy biomass production contributes 27% of the total GHG emissions.Reported RNG emissions are based on a 2% fugitive emission rate in the upgrading process, but they can range from 0% to 5%.Emissions from RNG loss contribute 45% to the total GHG emissions because of the greater global warming potential of CH 4 compared to CO 2 .Upgrading electricity emissions are based on the US grid mixture, with only about 20% coming from renewable sources.Biogas upgrading power contributes 23% to the total GHG emissions.Emissions associated with upgrading power can be reduced if a larger percentage of the electricity consumed is sourced from renewable sources.Digestate emissions have a significant impact on reducing AD system emissions by displacing fertilizers employed in agricultural applications.Digestate displacement factors depend primarily on the nitrogen content of the digestate (Aui et al., 2019).Digestate credits of 14 g CO 2 -eq/MJ RNG reduced the GHG emissions to 15 g CO 2 -eq/MJ RNG.The GHG emission result from this study falls within the range of reported RNG emissions (−300 to 120 g CO 2 -eq/MJ RNG) in the California Air Resources Board database.They are lower than conventional natural gas (90 g CO 2 -eq/MJ) and comparable to landfill gas (25 g CO 2 -eq/MJ) and other biowaste pathways (Ankathi et al., 2018;Lee et al., 2021).

| Sensitivity analysis
We conducted a sensitivity analysis on the combined baseline land use 13.5 t/ha biomass yield all environmental credits scenario to estimate the impact of several factors on the economic potential and environmental impacts of the system.Biogas yield, operating capacity, RIN credit price, and LCFS credit are the key drivers to process economics (Figure 6a).Biogas yield significantly impacts NPV because of its connection to RNG and environmental credit revenue.Altering the biogas yield from 421.6 to 328 m 3 and 421.6 to 515 m 3 CH 4 /t results in a respective 45% decrease and increase in the NPV.Operating capacity, RIN, and LCFS credits have similar large impacts on the NPV.The RIN credit price varies due to market dynamics, including supply and demands of RIN credits, RFS obligations, market speculations, and government policies.The LCFS credit also varies due to market dynamics, but it can also change based on the environmental performance of the technology.Thus, the LCFS credit has the greatest potential to vary based on the facility.Other factors like the total capital investment, farm gate feed price, and digestate credit prices have minor impacts on the NPV.Greenhouse gas emissions of the system are most sensitive to CH 4 losses and equivalent fertilizer replaced by digestates in the broader economy (Figure 6b).A reduction of CH 4 losses from 2% to 1% results in a 40% reduction in GHG emissions, and an increase in CH 4 losses can increase GHG emissions by 124%.CH 4 losses result in significant GHG emissions because of its higher global warming potential compared to CO 2 .A 20% increase and 20% decrease in the emission savings from the digestate can reduce and increase the GHG emissions by 17%, respectively.Changes in the emissions associated with F I G U R E 5 Life cycle carbon intensity of renewable natural gas (RNG) from grassy biomass compared to RNG from waste sludge, landfill gas, and fossilbased compressed natural gas (CNG).grassy biomass and biogas transportation have little on GHG emissions.The sensitivity analysis indicates that if the CH 4 losses can be kept at the barest minimum, a reasonable reduction in GHG emissions can be accomplished.

| DISCUSSION
Over 72% of the Corn Belt is devoted to cropland that produces corn (Zea mays L.) and soybean (Glycine max [L.] Merr.) for biofuels, feed, and export markets (Green et al., 2018).The RFS2 mandate for biofuel production in the United States stimulated the conversion of land to corn for ethanol due to its established infrastructure supporting high levels of productivity (Lark et al., 2020;Wright et al., 2017).As of 2019, about 40% of the corn produced in the region was used for ethanol production, 38% was used for feed, 14.5% was exported, and the remainder was used for food, seed, and industrial uses (Saavoss et al., 2021).Livestock production also represents a large portion of the economy and, in 2021, cattle and hog farming added almost $100 billion to the regional economy (USDA ERS, 2022).The values of total corn and soybean produced were almost $71.1 and $49.2 billion, respectively, that same year (USDA ERS, 2022).While compelling and resilient in terms of macroeconomics, the existing corn-soybean-biofuel-livestock system is often not lucrative for famers and is associated with major negative impacts to air, soil, water, and wildlife (Alexander et al., 2008;Lark et al., 2020Lark et al., , 2022;;Schulte et al., 2022).Strategically integrating perennial cover within the dominant land use matrix of corn and soybean is widely accepted strategy for shoring up many environmental shortcomings (Brandes et al., 2016;Gelfand et al., 2013;Jarchow et al., 2015).A major remaining challenge, however, is to meet environmental goals while not financially overburdening famers, who already face difficult economics (Audia et al., 2022;Brandes et al., 2016;Hallam et al., 2001;Schulte et al., 2022).The grass-to-gas renewable fuels pathway, which includes producing RNG from grassy biomass via AD and biogas upgrading, investigated was conceived to simultaneously address the dual of farm finances and the environment.Ssegane and Negri (2016) address changes to farm finances, water quality, and soil carbon storage in a companion study that simulated the strategic restoration of diverse, native, grassland cover within the GRB.The potential financial benefit to farmers for producing and selling grassy biomass to an energy market ranges from −$445 to $1291/ha/year in 2022 US $.Positive net revenue can be generated for farmers at biomass selling prices above $88/Mg in the 6.7/Mg/ha biomass yield scenario and above $49/Mg in the 13.5/Mg/ha biomass yield scenario.These net returns can on average compete with the production of annual crops in the productivitybased land use scenario where bioenergy grasslands are established on poorly performing cropland.These areas comprise 91,274 ha or 4.9% of the watershed.Ssegane and Negri (2016) further found disproportionate reductions in annual nutrient and sediment loss, and thereby expected improvements in water quality with increased grassland land cover.Soil carbon storage was estimated to increase by 1.13% over 10 years.The projected annual value of these enhancements, in terms of benefits to society, was $18 million for nitrogen reduction, $1.4 million for phosphorus reduction, $2.5 million for sediment reduction, and $14 million for soil carbon storage.We sought to build on this analysis by estimating RNG production, GHG emissions, and NPV of the prospective grass-to-gas pathway from the standpoint of an RNG producer.
Broadly, the process economics for the grass-to-gas pathway in the United States strongly depends on biomass yield and environmental credits.Credits for liquid and solid digestate from AD of agricultural feedstocks also improve the process economics of the system.More specifically to grass-to-gas within the GRB of Iowa and Missouri, our results indicate the pathway is unprofitable for all land use scenarios if any of the digestate or renewable fuels credits are absent (Figure 4).Only scenarios with all four credits-liquid digestate, solid digestate, LCFS, and RIN-resulted in positive NPVs.The NPV for scenarios with all credits ranged from $6 million for the baseline land use 6.7 t/ha biomass yield scenario to $413 million for the productivity-based land use 13.5 t/ha biomass yield scenario.Sensitivity analysis showed that significant drivers of process economics include biogas yields, facility operating capacity, LCFS credits, and RIN credits (Figure 6).The aforementioned are also correlated with the RNG output of the facility.Hence, land use scenarios with more biomass yield represent the most attractive economic scenarios, given that more biomass resulted in more feedstock and, subsequently, more RNG produced, which attracts credits.
Simulated GHG emissions associated with RNG produced via AD and biogas upgrading of grassy biomass in the GRB are 15 g CO 2 -eq/MJ RNG.This value compares favorably to the life cycle emissions of fossil natural gas, which are 90 g CO 2 -eq/MJ (Figure 5).With an 83% reduction, RNG from grassy feedstocks exceeds the US Environmental Protection Agency's 60% lifecycle GHG reduction requirement for cellulosic biofuel (US EPA, 2022c).CH 4 losses, biomass harvest and transportation, and electricity consumption contribute largely to the GHG emissions (Figure 5).Sensitivity analysis indicated that reducing CH 4 losses in the upgrading process is especially important in reducing GHG emissions (Figure 6).
In addition to comparing favorably to fossil natural gas, this pathway produces lower emissions than some other biofuel pathways.Emissions from corn stover or corn grains to ethanol range from 23 to 62.64 g CO 2 -eq/MJ ethanol, as reported in GREET (Wright et al., 2008).Emissions from soybean-based biodiesel are 19.6 g CO 2 -eq/MJ (Wright et al., 2008).Uddin et al. (Ankathi et al., 2018) reported GHG emissions from AD of manure and biochar amended AD of manure as −33 g CO 2 -eq/MJ RNG and −43 g CO 2 -eq/ MJ RNG, respectively.Emissions from the manure-based pathways are less than the emissions recorded in this study because they account for emission avoidance from conventional manure management and additional carbon sequestration from biochar application to soil.
A limitation of our current analysis is that we do not fully account for changes in CO 2 -eq due to changes in agricultural land use, which can be substantial (Gelfand et al., 2013;Jarchow et al., 2015;Meehan et al., 2013).The production of grassy biomass typically reduces tillage and use of fertilizers compared to annual crop production, and contributes to increase carbon storage in soil (Jarchow et al., 2015;Lee et al., 2021).Ssegane and Negri (2016) estimate a limited amount of carbon storage as soil organic carbon over a 10-year horizon, but noted that substantially greater accrual is likely over multiple decades.Regardless, reductions in CO 2 -eq associated with reduced nitrous oxide (N 2 O) emissions are likely to be more impactful, given N 2 O is nearly 300 times as potent a GHG as CO 2 (Shine, 2009) and strongly associated with nitrogen fertilizer application (Lu & Tian, 2013).Land use change was beyond the scope of this study and not estimated in Ssegane and Negri (2016) due to a dearth of data from the study region.This is an important area for future research: ongoing scientific studies building on Jarchow et al. (2015) and Bendorf et al. (2021) in the study region are better quantifying GHG emissions as well as soil organic carbon dynamics.
Our results combined with those of Ssegane and Negri (2016) underscore the need for further scientific, technical, and policy development to fully realize grass-to-gas value chain and address level economic and environmental shortcomings of the existing Corn Belt system.Scientists are already investigating means to improve AD efficiency, reduce emissions to air and water pollution, and improve the process economics through the further production of high-value renewable chemicals and products (Schulte et al., 2022).Engineers are deploying AD systems at multiple scales to reliably process biomass mixtures into pipeline quality RNG (Braun et al., 2010).Farmers are learning how to establish and manage native grass plantings on their farms (Grudens- Schuck et al., 2017).Markets for grassy biomass could further foster emerging ecosystem and/ or commodity markets such as carbon credits, nutrient trading, and hunting leases that could benefit farmers and/or landowners in the region and beyond (John & Mcisaac, 2017;Meehan et al., 2013;Mishra et al., 2019;MR/GMWNTF, 2008).Our integrated analysis when combined with that of Ssegane and Negri (2016) revealed potentially lucrative scenarios, where biomass sales, RNG sales, and environmental credits for digestate replacing purchased agricultural inputs and renewable fuels achieve positive NVPs for farmers and RNG producers in the GRB, as well as environmental benefits to be broadly enjoyed society.Robust policies in support of environmental credits are needed, however, to realize the economic and environmental value proposition associated with grass-to-gas.
of the grassland is enrolled in the USDA Conservation Reserve Program (FSA, 2022).Water quality impairments are pervasive throughout the watershed, and include bacterial, nitrogen and phosphorus nutrient, and heavy metal contamination and low levels of dissolved oxygen (Iowa Department of Natural Resources [IDNR], 2022; Missouri Department of Natural Resources [MoDNR], 2020).Habitat quality and loss are also of concern in the region (Missouri Department of Conservation [MDC], 2015; Simplified flow diagram for anaerobic digestion of grassy biomass to renewable natural gas, and solid and liquid digestate.(b) The life cycle assessment system boundary for producing renewable natural gas (RNG) from grassy biomass.

F
Material and carbon process flow diagram for grassy biomass to renewable natural gas (RNG) production.

T A B L E 4
Net present value in millions of US dollars from renewable natural gas sales based on different grassy biomass productivity, environmental credit, and land use scenarios.

F
Sensitivity analysis of (a) net present value and (b) greenhouse gas emissions (GHG) of renewable natural gas production from grassy biomass to several key parameters of the system.The numbers in parenthesis represent the range of values evaluated.Orange and blue bars, respectively, indicate lower-and uppercase assumptions.EF, emission factors.
Life cycle assessment emission factor inventory (from GREET.net).
T A B L E 2Abbreviation: RNG, renewable natural gas.

Bioenergy grassland area based on land use and harvest distance scenarios (thousand ha) Biomass yield scenario (t/ha) RNG produced (million GJ)
Estimated production of renewable natural gas (RNG) produced from different land use, harvest distance, grassy biomass yield, and harvest distance scenarios.
T A B L E 3

Land use scenario Baseline Buffered Productivity based Biomass yield (t/ha) Credits NPV (MM)
Note: Numbers in parentheses are negative values.Assumptions: 10% internal rate of return and $2.84/GJ price for renewable natural gas.Abbreviations: LCFS, low carbon fuel standard; RIN, USEPA renewable identification number.