Regional economic aspects of carbon markets and anaerobic digesters in the USA: the case of swine production
Abstract
The USA has significant potential to produce energy from anaerobic digesters (AD) due to the size of its agricultural sector. The use of ADs reduces greenhouse gas (GHG) emissions from manure management. The financial benefits to farmers come from the on-farm use, or off-farm sales, of biogas and its end products, namely renewable natural gas (RNG) or electricity. Current energy prices and policies in the USA are insufficient to trigger large-scale construction of ADs; however, payments to avoid GHG emissions and sequester carbon could become sufficiently high to prompt investment. This analysis quantifies the economic incentives necessary for the construction of ADs for swine producers and can easily be expanded to include other feedstocks. Various end-use pathways to produce RNG and electricity are considered to account for location, herd size, and other parameters to deliver a comprehensive analysis for the USA. The analysis and results are composed of a generic part to illustrate the effects of carbon payment on profitability in general as well as a specific analysis for states representing 83.6% of the US hog inventory. Our results indicate that carbon payments would be a stronger determinant than energy prices in farm-level decisions to install ADs, but that energy prices would be influential in determining the optimal biogas end use. The potential need for long-term contracts – both for energy and carbon payments – to reduce investment uncertainty and increase investment in ADs is also discussed.
Introduction
In the USA and globally, there is an increasing interest in involving farmers in greenhouse gas (GHG) mitigation efforts. For example, livestock producers can change manure management practices, for example by installing anaerobic digesters (AD) to reduce GHG emissions, control pathogens, and manage odor while producing organic fertilizer and energy.1 As climate change mitigation is a public good and GHG mitigating efforts usually involving higher production costs and/or capital investments, additional compensation may be needed for producers to remain profitable when adopting GHG-mitigating practices. Examples of such compensation are voluntary carbon programs, which pay farmers to reduce emissions and sell the generated credits to entities interested in offsetting their own emissions. (Here and elsewhere in the paper, the carbon price is in USD per metric ton of CO2 equivalent.) Payments could also include government subsidies for the adoption of GHG mitigation practices without a corresponding offset market. This approach has been adopted by the US Department of Agriculture (USDA) through programs administered by the Natural Resources Conservation Service (NRCS). In this paper, we quantify the financial incentives necessary for US swine producers to implement ADs on their farms and monetize the biogas-based production of electricity or renewable natural gas (RNG) both in the absence and presence of carbon markets.
The main GHG emissions from manure management are in the form of methane (CH4) and nitrous oxide (N2O). In 2020, 81.8% of emissions from swine operations were attributed to CH4 from manure management.2 Livestock producers can install ADs that capture CH4 to generate biogas to mitigate CH4 emissions. Biogas is produced during the anaerobic process and is composed of approximately equal parts of methane and carbon dioxide (CO2). It can be used to generate on-farm heat and/or electricity to lower on-farm energy expenses.3 Additional revenue is possible if the excess electricity can be sold to utilities. Biogas can also be upgraded to RNG and fed either into the natural gas grid or compressed as a transportation fuel in the form of compressed natural gas (CNG).
Anaerobic digesters are not used widely in the USA, mainly due to a mismatch of costs and benefits. Without carbon policies, the profitability of an AD depends primarily on energy prices, which, in the USA, have not been sufficiently high to trigger widespread adoption. Uncertainty of return arising from volatile energy prices also discourages investment in anaerobic digesters.4 Livestock producers would require either certainty of revenue (e.g., from energy production and/or carbon offset generation) or higher expected returns given uncertainty over the project period. Smaller farms also do not benefit from economies of scale.5 As of July 2023, there were 343 livestock ADs (operational projects and under construction), which avoid 10.49 million metric tons of CO2 equivalent (MMT CO2-e) on an annual basis (Fig. 1). Total emissions from US agriculture were about 629 MMT CO2-e in 2019.2 California has seen an increase in the use of ADs in recent years as illustrated by the amount of CO2 avoided (Fig. 1). The end uses associated with the majority of the emission reductions in California are CNG and electricity whereas for Missouri and Oregon, the majority of the emissions reductions are from RNG. Anaerobic digesters using dairy manure are responsible for 80.8% of the emissions reductions followed by swine manure with 15.7%.

Recent proposed and enacted laws in the USA have stimulated interest in biogas. For example, the Inflation Reduction Act (IRA) focuses on reducing CH4 emissions and includes tax credits and subsidies that could increase AD construction. (See for example ‘The IRA revolutionizes AD tax credits’ and ‘Why the AD industry is poised for expansion’ in BioCycle.Net on 23 August, 2022 and 6 December, 2022, respectively.) The IRA includes a production tax credit of up to 3.3 cents per kilowatt-hour (kWh) for qualifying projects in addition to investment tax credits of up to 50%. In addition, the US Environmental Protection Agency (EPA) has proposed a framework for renewable energy credits if electricity for electric vehicles is produced from biogas. (See Document 87 FR 80582 in the Federal Register from 30 December 2022: ‘Renewable Fuel Standard (RFS) Program: Standards for 2023–2025 and Other Changes.’) The Consolidated Appropriations Act of 2023, signed into US law at the end of 2022, could also potentially play a large role in the development of biogas markets by introducing more certainty in supplemental carbon markets. (See H.R.2617 – Consolidated Appropriations Act, 2023.) The new law authorizes the USDA to establish a GHG Technical Assistance Provider and Third-Party Verifier Program to provide technical assistance for producers and, perhaps most important, allows USDA to verify certifiers in much the same way that the USDA uses with other agricultural certification programs, for example organic labeling. If approved, USDA will not directly create a carbon marketplace but will help set the rules, thus alleviating much of the adverse selection and moral hazard concerns that currently exist in carbon markets.
This paper contributes to the understanding of how the profitability of ADs changes in the presence of carbon payments, which is of increasing importance for US farmers. The analysis is not limited to a particular pathway in terms of biogas end use but covers electricity and RNG generation given energy prices and initial manure management system. There will be regional differences across the USA given variations in manure management systems, temperature, energy prices, and feedstock availability. Although presented for the case of swine manure, the analysis can easily be expanded to other feedstocks. Assessing the necessary incentives shows how farmers can be incorporated into a domestic carbon offset market. We account for heterogeneity in farm size, waste characteristics, geographic location, and end use, which are important determinants of AD adoption.6 The farm's initial manure management system will also affect significantly how much GHG mitigation a farm can receive credit for by installing an AD.3
Background
Absent carbon payments, the financial decision to construct an AD depends on the initial capital and operating cost as well as the net income from the various biogas uses. It is also important to determine the GHG emissions prior to the construction of an AD to obtain a baseline to which the emission savings are compared. Assessment of the baseline emissions is needed if participation in carbon markets is possible. This section provides an overview of (1) manure management systems and their associated emissions, (2) AD characteristics, and (3) biogas end use. Additional material can be found in the Supplemental Information (SI) available at https://github.com/SwineCarbonMarkets/GitHubAnalysis.
Manure Management Systems and Emissions
In 2019, the shares (weighted by hog inventory) of liquid/slurry, anaerobic lagoon, and pit storage for US swine manure were 13.8%, 15.6%, and 69%, respectively.2 These shares vary by region, however. For example, 49% of hog manure in North Carolina is stored in anaerobic lagoons as opposed to 4% in Iowa. Air temperature influenced the choice of current manure management systems across the states.7 Regions with overall higher yearly average temperature allow for the anaerobic decomposition to take place over a longer period throughout the year making anaerobic lagoons more prevalent.2, 7
The focus of this analysis is on the mitigation of CH4, which is produced from the anaerobic (without oxygen) decomposition of liquid manure in lagoons or pits. Emissions from manure management (prior to application to cropland) are influenced by a variety of factors such as feed composition, system used, storage time, and air temperature. Depending on the ambient temperature and the system used, there are regional differences in CH4 emissions per head of swine (Fig. 2). These differences play a key role in the presence of carbon markets because livestock producers with higher baseline emissions will benefit more from the construction of an AD because they are able to claim larger emission reductions and, thus, receive higher payments. This could be perceived as problematic because prior manure management system choices that led to higher GHG emissions are rewarded with a higher financial incentive.

Anaerobic digesters characteristics
In the USA, three AD designs are commonly used: (1) covered lagoon, (2) complete mix, and (3) plug flow.2 A covered lagoon is a passive system in which the AD temperature cycles with outside temperature and methane production ceases if the temperature is below 20 °C. Thus, for colder climates such as prevalent in Iowa and other Midwestern states, this type of design will not be optimal for methane production. The two other designs are actively heated to a range of 30–40 °C for mesophilic ADs and 50–60 °C for thermophilic ADs. The percentage of solids is lowest in a covered lagoon (0.5%–3%) and highest in a plug flow digester (11%–13%) with a complete mix digester covering the range in between. The hydraulic retention time – which is the time interval in which the feedstock remain in the AD – is at least 15 days for complete mix and plug flow and 40–60 days for covered lagoons. The feedstock characteristics used in an AD determines the biogas yield which can vary significantly. For example, the biogas yield from manure and corn stover is 333 and 585 cubic meters per kilogram of volatile solids (m3 kg−1), respectively.8 Co-digestion, i.e., the use of more than one feedstock can increase the biogas yield. Besides biogas, ADs also produce digestate, which is the nutrient-rich residue after anaerobic digestion. Digestate is either liquid or solid following a separation process. Liquid outputs can be used, for example, as organic fertilizer whereas solid outputs can be used as soil amendments or animal bedding, among other uses.
There are also economic and policy implications involved with the handling of manure due to its multiple uses as a fertilizer substitute, anaerobic digester feedstock, and/or source for carbon offsets.9-12 Besides producing a feedstock for heat and electricity generation as well as reducing GHG emissions, methane digesters can also reduce odor and surface water contamination;7 likewise, the digestate (a co-product of anaerobic digestion) itself can be used as an organic fertilizer. Application of manure to cropland also has water quality implications due to its nitrogen and phosphate content.
Biogas use
In the absence of any carbon payments, biogas as well as the liquid and solid outputs from the AD are the only financially valuable products. Although an important reason for the construction of ADs has been odor control, which yields important nonmarket benefits (e.g., neighbor relations, work place amenities). The use of biogas resulting from ADs can be divided into three broad categories: (1) flaring, (2) electricity, and (3) RNG. Flaring is simply burning off the CH4, which is converted to mostly CO2 emissions. Due to the difference in global warming potential, the emissions are reduced 28-fold, flaring of 1 unit of CH4 saves about 28 units of CO2. Aui et al.8 assess the feasibility of a combined heat and power production using cattle feedlot manure, biomass, and glycerin as inputs. Although this is a possibility, we keep the analysis separate between electricity and RNG.
Although there is some electricity production occurring from biogas in the USA the majority of ADs convert biogas to RNG for injection either into the natural gas grid or for sale as CNG. Especially in California, CNG benefits from low carbon fuel standard (LCFS) regulation, which has contributed to greater construction of ADs in recent years.13, 14
Methods and model
This section outlines the various components of the model (Fig. 3) as well as the internal rate of return (IRR) calculations across various dimensions. In comparison with the return on investment (which requires the choice of a discount rate to calculate the net present value), the IRR has the advantage of avoiding the choice of a discount rate. The parameters and prices are assumed to be constant over the AD lifetime of 30 years. The subsections below include the presentation of (1) energy prices, (2) anaerobic digestion, (3) biogas use, and (4) IRR calculations. The unit of analysis is a swine farm characterized by inventory, farm type, and current manure management system. The temperature range is limited to 6–18 °C. This temperature range corresponds to the average temperature in 386 counties, which contain 95% of US swine inventory. Note that all data and supplemental information (SI) are available at www.github.com/SwineCarbonMarkets.

Four biogas uses are assessed based on cost data provided by the EPA:15 (1) Microturbine, (2) reciprocating engine, (3) CNG, and (4) RNG. The first two uses generate electricity whereas the last two produce natural gas. The EPA15 presents three additional uses, i.e., fuel cell, flaring, and gas turbine, which we do not include in this analysis for three reasons. First, electricity production through a gas turbine is used in landfills, which are characterized by a high volume of biogas, exceeding that of even the largest swine farms. Second, flaring is unlikely given the modeled carbon payments and the current incentives set forth in the IRA on payments for biogas electricity production. Finally, biogas use in fuel cells to produce electricity has the highest cost and, thus, seems unlikely as an alternative to microturbines or reciprocating engines.
Energy prices
As mentioned above, the analysis is separated into a general part – based on national energy prices – and a state-specific part. See Energy Information Administration (EIA) for Natural Gas Spot and Futures Prices (NYMEX) (EIA Series ID: NG_PRI_FUT_S1_D) and Average Price (https://www.eia.gov/electricity/sales_revenue_price/). Long-term natural gas prices in million metric British thermal units (MMBTU) are projected to be in the range of $3.41–$3.90 depending on economic growth and oil price trajectories.16 Natural gas spot prices have been outside this range in the past, i.e., $2.03 (2020) to $8.86 (2008). Thus, we conduct our analysis using natural gas prices ranging from $3.00–$5.00 in increments of $1.00 because annual natural gas prices have been above $5 only during the financial crisis and at the beginning of the war in Ukraine. Electricity prices are projected to be close to constant at $0.10 kWh−1 in the 2022 Annual Energy Outlook (AEO).16 There are large regional differences in the contiguous US with 2021 commercial prices ranging from $0.08 (Nevada) to $0.20 (California), so we use electricity prices of $0.10–$0.18 in increments of $0.04. Renewable natural gas derived from biogas potentially qualifies for renewable identification number (RIN) payments, specifically D3 RINs generated from cellulosic biomass, including biogas. More details are found in the description of the Renewable Fuel Standard Program (https://www.epa.gov/renewable-fuel-standard-program). The D3 RIN prices have been fluctuating between $0.60 and $3.30 (per unit of gasoline gallon equivalent) and, for this analysis, the RIN price is set to $3.00.
To illustrate the regional effects, IRR calculations are presented for the ten largest states in terms of hogs – representing 83.6% of inventory in 2021 – based on manure management system and energy prices. There are relatively large differences in energy prices in MMBTU across states over the years 2017–2022 (Table 1). The mean electricity prices over the period vary from $23.87 in Oklahoma to $31.04 in Minnesota. For natural gas prices, Oklahoma is lowest ($5.04), and Missouri is highest ($8.56). In terms of variability over 2017–2022, Missouri and Illinois are highest for electricity (range of $3.36) and natural gas ($2.21), respectively.
State | Temperature (in °C) | Electricity price | Natural gas price | ||||
---|---|---|---|---|---|---|---|
Mean | Min. | Max. | Mean | Min. | Max. | ||
Illinois | 11.56 | 28.69 | 28.35 | 29.26 | 7.37 | 6.38 | 8.59 |
Indiana | 11.50 | 29.28 | 28.90 | 30.03 | 6.55 | 5.99 | 6.92 |
Iowa | 9.18 | 26.32 | 25.37 | 26.74 | 5.80 | 4.92 | 6.97 |
Minnesota | 5.37 | 31.04 | 30.56 | 31.62 | 6.36 | 5.81 | 6.75 |
Missouri | 13.03 | 28.86 | 27.36 | 30.72 | 8.56 | 8.01 | 9.65 |
Nebraska | 9.70 | 26.43 | 24.55 | 27.78 | 5.46 | 4.77 | 5.86 |
North Carolina | 15.75 | 27.28 | 25.82 | 27.81 | 8.39 | 7.93 | 8.86 |
Ohio | 11.13 | 28.64 | 27.24 | 30.25 | 7.24 | 6.74 | 7.80 |
Oklahoma | 15.65 | 23.87 | 22.40 | 25.34 | 5.04 | 4.20 | 5.84 |
South Dakota | 7.66 | 29.62 | 28.97 | 30.77 | 5.40 | 4.66 | 6.11 |
Anaerobic digestion
The amount of biogas produced in an AD depends on the amount of manure – which is based on swine inventory and farm type – as well as the co-digestate. The bio-physical properties of the feedstock are taken from USDA NRCS17 and Aui et al.8 (Table 2). In this analysis, we focus on ADs tailored for two types of swine operations. The first type are breeding swine farms that are usually larger (in terms of inventory) and produce more manure due to the presence of lactating and gestating sows, which are generally larger animals (Table 2). The second type are market swine operations that grow swine to market size (farrow-to-finish operations). Farrow-to-finish operations can be further subdivided depending on the pig's life-cycle, but, for the purpose of this analysis, those stages are assumed to occur in the same facility. For GHG inventory purposes as well as for our analysis, the breeding swine are assumed to be 80% gestating, 15% farrowing, and 5% boars.2 Although, artificial insemination of sows is increasing, which leads to boars being obsolete on sow farms, EPA2 is still using boars for emissions calculation. For the farrow-to-finish distribution (labeled ‘Market Swine’ in Fig. 3), we assume an equal share among the four weight classes used by EPA.2 Since co-digestion is used in most digesters, we assume the use of corn stover and wheat straw at rates of 7.14% and 2.65% (weight basis), respectively. The quantity in metric tons of codigestate is denoted and the price per ton () is assumed to be $20 per ton.8
Feedstock | Feedstock (MT year−1) | TS (%) | VS (%) | CH4 yield (m3 kg−1 VS) |
---|---|---|---|---|
Gestating sow | 1.83 | 10 | 9.2 | 275 |
Lactating sow | 4.13 | 10 | 9.2 | 356 |
Boar | 1.39 | 10 | 8.9 | 356 |
Market (under 23 kg) | 0.47 | 10 | 8.8 | 356 |
Market, (23–54 kg) | 0.93 | 10 | 8.3 | 356 |
Market, (54–81 kg) | 1.61 | 10 | 8.3 | 356 |
Market, (over 81 kg) | 2.16 | 10 | 8.3 | 356 |
Wheat straw | 90 | 89.3 | 208 | |
Corn stover | 40 | 94.0 | 348 |
The current manure management system, in the absence of an AD, determines baseline GHG emissions, . Those emissions depend on farm inventory, farm type (i.e., breeding or market swine), manure management system, and temperature. In the presence of an AD, the emissions from the AD system depend on farm inventory, farm type (i.e., breeding or market swine), and AD type. The digester type is assumed to be a complete mix AD and the emissions are labeled . Given a carbon payment, the difference between and determines the yearly carbon revenue from constructing an AD (see SI for more information).
In this analysis, we set , , and to match closely Aui et al.8 We assume that the annual operating and maintenance cost of the AD amounts to 5% of investment cost.
Biogas use
We assume that the biogas produced contains 50% methane. This value is also consistent with the Anaerobic Digester Screening Tool Version 2.2 developed by the US Environmental Protection Agency for the Global Methane Initiative (GMI). The tool has also been a source for the feedstock characteristics besides the USDA NRCS.17 The energy content of biogas and methane is 0.0234 and 0.036 gigajoules (GJ) m−3, respectively. The EPA15 evaluated the cost and GHG emissions from various biogas uses in California. We use the information provided to estimate the relationship between biogas input flow (m3 year−1) and (1) capital cost, (2) operating and maintenance cost, (3) product output (GJ year−1), and GHG emissions (Table 3). The capital cost consists of a fixed part and a variable part determined by the intercept and slope of a regression line linking capital cost to biogas flow as determined by the amount of manure produced. The basis for the biogas flow and capital cost are presented in the SI.
Biogas use | Microturbine | Reciprocating engine | Compressed natural gas | Renewable natural gas |
---|---|---|---|---|
MT CO2e GJ−1 () | 0.084 | 0.091 | 0.051 | 0.042 |
O&M ($) () | 0.04461 | 0.05625 | 0.08924 | 0.15089 |
Output (GJ m−3) () | 0.00617 | 0.00832 | 0.00924 | 0.01348 |
Capital cost | ||||
Fixed () | 104 028 | 191 313 | 1 102 423 | 2 135 851 |
Variable () | 0.73466 | 0.67947 | 0.45657 | 0.99344 |
Delivering electricity produced from biogas to the power grid is less costly than transporting RNG to the natural gas grid. Thus the sale of CNG remains an important option. CNG stations in major swine-producing states in the Midwest, such as Iowa and Illinois, are sparse (see SI). North Carolina does not have many natural gas pipelines but Iowa, as the leading state in terms of swine production, has a substantial amount of pipeline potential serving as connection points.
Internal Rate of Return Calculation
All dollar values are assumed to be in real terms. The biogas installation and investment cost are in 2021 $ and were deflated from the original values in EPA15 using the GDP deflator. The remaining dollar values are from recent publications and, hence, if the year of those dollar values was not explicitly stated, the differences to 2021 will be relatively small.
Model simulation
The scenarios are simulated over a grid of preselected states and farm characteristics due to an otherwise infinite number of scenarios. The states can be thought of as representative of the macroeconomic and policy environments the swine farm is operating under. Natural gas and electricity prices are set to $3, $4, and $5 per million BTU, and $0.10, $0.14, and $0.18 per kWh, respectively. The sequence of carbon prices ranges from $0 to $100 in steps of $25. Ambient temperature is set to 9 °C and 18 °C, which are representative of the Midwest and the East Coast, for example, North Carolina, respectively. The farm characteristics differentiate between initial manure management system, i.e., anaerobic lagoon, liquid/slurry, deep pit, and deep pit but less than 1 month. The market and breeding swine farm sizes range from 1000 to 8000 head in steps of 500 head. All the results are presented for a complete mix digester with the four biogas end uses enumerated before. The complete set of results across all dimensions is available in the Supporting Information.
Results
The results are generally presented for electricity and natural gas prices of $0.14 kWh−1 and $4 MMBTU−1, respectively, which correspond approximately to the average retail prices in major swine-producing states (i.e., Illinois, Indiana, Iowa, and North Carolina). The codigestate is assumed to be corn stover. Given the large variability of distances to the electricity or natural gas grid, we do not include connection costs. We first present the results in the absence of a carbon payment to establish a baseline corresponding to the current situation in the USA. The effects of a carbon payment on the rate of return are presented in the subsequent section. We present the results for two manure management system/temperature combination, namely, deep pit at 9 °C and anaerobic lagoon at 18 °C, representing the situation in Iowa and North Carolina, respectively.
No carbon payment
Without a carbon payment, the current manure management system does not influence the profitability of an AD because the reduced emissions are not financially compensated. The use of a microturbine to generate electricity is not optimal for any farm size and the use of a reciprocating engine is preferred to generate electricity (Fig. 4). The results also suggest that production of renewable natural gas for pipeline injection or use in vehicles (CNG) leads to slightly lower returns compared to electricity production via a reciprocating engine for deep pits. For anaerobic lagoons, the returns are close to identical for CNG, RNG, and reciprocating engines. For market swine farms, only very large farms of 8000 head lead to a small positive IRR for deep pits. For breeding swine farms (deep pit), the profitability threshold is 5000 head with the production of RNG.

A US carbon tax as proposed in recent years could result in higher energy prices. For electricity and natural gas prices of $0.18 kWh−1 and $5 MMBTU−1, respectively, the threshold for a positive IRR is lowered for both types of swine farm. For breeding and market farms, the threshold is 4000 and 7000 head, respectively.
The findings, in the absence of a carbon payment, are twofold. First, the overall rate of return is small with the exception of large breeding swine farms. For market swine farms, a rate of return of 4% is never exceeded given our assumptions regarding energy prices, even in the case of high prices. Given current projections of energy prices, it seems unlikely that a further increase, sufficient to trigger AD investment, will occur. For breeding swine farms, an inventory of 8000 head is necessary to achieve returns of 4% or more. These results reflect the current situation (i.e., absence of carbon payments) in the USA, which results in low returns and economies of scale only realizable for large swine farms. Second, across the biogas uses analyzed, the maximum rate of return difference is 0.3 percentage points (excluding the microturbine use). This suggests that perceived energy price levels and fluctuations in the future are key for the decision regarding biogas end use.
Carbon payments
Given carbon payments, the manure management system from which the switch to an AD occurs matters due to emission differences. Anaerobic lagoons have higher CH4 emissions, so the construction of an AD is more profitable than a deep pit storage system (Fig. 4). We illustrate the profitability using carbon payments of $50 and $100. Besides the emission savings from switching to an AD, the GHG emissions from the biogas use manifests itself in the profitability because microturbines have high (on-farm) GHG emissions (Table 3). As in the situation where there is no carbon payment, microturbines are the least profitable at a carbon value of $50 compared to the reciprocating engine but become comparable at a carbon payment of $100 due to the higher emissions from reciprocating engines. At a $50 carbon payment, a positive rate of return is achieved for breeding and market swine farms bigger than 2000 and 3000 head, respectively, when switching from an anaerobic lagoon at 18 °C.
There is a significant difference in returns between the deep pit 9 °C and anaerobic lagoon 18 °C cases. For a breeding swine farm of 8000 head and at a carbon payment of $100, the difference in return is ranging between 8.6 and 10.4 percentage points (Table 4). The range of returns across the various biogas end uses is relatively small with a maximum of 2.4 and 0.7 percentage points for the anaerobic lagoon 18 °C and deep pit 9 °C cases, respectively. Plotting the internal rates of returns as a function of carbon payments and inventory, there is a tradeoff between both components. That is, a higher carbon payment reduces the swine inventory required for a given rate of return (Figs 5 and 6).
Anaerobic lagoon 18 °C | Deep pit 9 °C | ||||||||
---|---|---|---|---|---|---|---|---|---|
CNG (%) | MTU (%) | REC (%) | RNG (%) | CNG (%) | MTU (%) | REC (%) | RNG (%) | ||
Carbon Price: USD 50 | |||||||||
Breeding | 2000 | 0.7 | 1.2 | 1.6 | |||||
3000 | 3.2 | 3.6 | 4.1 | 2.4 | |||||
4000 | 5.1 | 5.3 | 5.8 | 4.2 | 0.5 | 0.1 | 0.8 | 0.1 | |
5000 | 6.5 | 6.6 | 7.2 | 5.7 | 1.9 | 1.4 | 2.1 | 1.5 | |
6000 | 7.8 | 7.8 | 8.3 | 6.9 | 3.0 | 2.5 | 3.2 | 2.6 | |
7000 | 8.9 | 8.7 | 9.3 | 8.0 | 3.9 | 3.4 | 4.1 | 3.5 | |
8000 | 9.8 | 9.6 | 10.2 | 8.9 | 4.8 | 4.2 | 4.9 | 4.4 | |
Market | 3000 | 0.2 | |||||||
4000 | 1.2 | 1.5 | 2.0 | 0.5 | |||||
5000 | 2.6 | 2.8 | 3.3 | 1.8 | |||||
6000 | 3.7 | 3.8 | 4.4 | 3.0 | |||||
7000 | 4.7 | 4.7 | 5.3 | 3.9 | 0.5 | 0.7 | 0.2 | ||
8000 | 5.6 | 5.5 | 6.1 | 4.8 | 1.3 | 0.8 | 1.5 | 1.0 | |
Carbon Price: USD 100 | |||||||||
Breeding | 1000 | 0.5 | 2.1 | 2.2 | |||||
2000 | 5.0 | 6.3 | 6.5 | 3.5 | |||||
3000 | 7.8 | 8.9 | 9.1 | 6.3 | 0.4 | ||||
4000 | 9.9 | 10.8 | 11.1 | 8.4 | 1.8 | 1.7 | 2.1 | 1.2 | |
5000 | 11.7 | 12.4 | 12.7 | 10.1 | 3.2 | 3.0 | 3.4 | 2.6 | |
6000 | 13.2 | 13.8 | 14.1 | 11.6 | 4.3 | 4.0 | 4.5 | 3.7 | |
7000 | 14.6 | 15.0 | 15.4 | 12.9 | 5.3 | 4.9 | 5.4 | 4.7 | |
8000 | 15.8 | 16.1 | 16.5 | 14.1 | 6.2 | 5.7 | 6.2 | 5.5 | |
Market | 2000 | 0.8 | 2.1 | 2.3 | |||||
3000 | 3.4 | 4.5 | 4.7 | 2.1 | |||||
4000 | 5.3 | 6.3 | 6.5 | 4.0 | |||||
5000 | 6.8 | 7.7 | 7.9 | 5.5 | |||||
6000 | 8.1 | 8.8 | 9.1 | 6.7 | 0.7 | 0.4 | 1.0 | 0.2 | |
7000 | 9.2 | 9.9 | 10.1 | 7.8 | 1.6 | 1.4 | 1.9 | 1.2 | |
8000 | 10.2 | 10.8 | 11.1 | 8.8 | 2.5 | 2.1 | 2.7 | 2.0 |


Biogas production generally consists of two systems: (1) AD, and (2) biogas end use (upgrading). Without carbon payments, the profitability only depends on the biogas end use choice and energy prices, which are too low to trigger investment. Our results suggest that the profitability differences across end uses is small, even in the case of fluctuating energy prices. Adding carbon payments makes the AD itself (from a systems perspective) generate revenue. Figures 5 and 6 indicate that carbon payments reduce the threshold for AD construction and that the various end uses only have a small influence on the herd size necessary to trigger investment.
Sensitivity analysis
As highlighted in the previous section, the profitability of the end uses is relatively close. Thus, a sensitivity analysis with regard to energy prices is warranted. To do this, we vary the electricity-to-natural-gas price ratio measured in GJ. For the results presented in the previous section, that price ratio was 10.26. (Given the electricity and natural gas prices of 0.14 $ kWh−1 and 4 $ MMBTU−1 as well as the energy content of 1.0550559 GJ MMBTU−1 and 0.0036 GJ kWh−1, the ratio is 10.26) Over the period 1997–2022, the price ratio fluctuated from 3.37 (2005) to 19.57 (2020) indicating that the value chosen to present the results is close to the average. The historical data were obtained from the Federal Reserve Economic Data (FRED), i.e., Henry Hub Natural Gas Spot Price (FRED Series ID: DHHNGSP) and Average Price: Electricity per Kilowatt-Hour in US City Average (FRED Series ID: APU000072610). For a sensitivity analysis, we have assessed scenarios with a higher and lower electricity price of $0.18 and $0.10, respectively. The upper and lower bounds are represented as the shaded areas in Fig. 4. The narrow band suggests that energy prices play a minor role – at least for anaerobic lagoons – in deciding to construct an AD once carbon payments are in place.
State-specific results
We assessed the optimal end use in each state given the average, minimum, and maximum energy prices experienced in ten states (Table 1). For all states except North Carolina and Oklahoma, the initial manure management systems and codigestate is assumed to be a deep pit and corn stover, respectively. The initial manure management system for North Carolina and Oklahoma is an anaerobic lagoon. In addition, Oklahoma uses wheat straw as a codigestate.
For average and maximum energy prices, CNG is the dominant the end use optimal in all states given carbon prices of $50 and $100. In the absence of carbon payments, RNG is more profitable for smaller farms, with the exception of Oklahoma. Without a carbon payment, electricity production is never optimal and the production of RNG is optimal for small swine farms. For North Carolina and Oklahoma, microturbines become the optimal end use at carbon prices of either $50 or $100.
The situation is different if minimum energy prices are assumed in each state. In that case, electricity production becomes optimal for all states across all carbon payments (including no carbon payments at all). Specifically, a reciprocating engine becomes the most profitable use in all states except North Carolina and Oklahoma. Even in those two states, electricity production from using a reciprocating engine still dominates most scenarios with some nuances. For North Carolina and low swine inventory, electricity generation using microturbines is more profitable than using a reciprocating engine for a carbon payment of $100. For Oklahoma and no carbon payments, the optimal use in terms of profitability is a reciprocating engine. This also holds for carbon payments of $50 and smaller farms (i.e., below 6000 and 8000 head for breeding and market swine farms, respectively). For a carbon payment of $100, microturbines become the optimal use for all farm types.
Note, though, that Oklahoma has a very high density of natural gas pipelines compared with other states. The connection cost with the natural gas grid is not explicitly modeled in this report and the returns across the various end uses are relatively close to each other, so each individual project likely needs a more detailed evaluation of the optimal end use. High connection cost in other states due to longer connection distances may make electricity or CNG production more profitable, even if this analysis suggests that RNG is optimal. For a carbon price of $50 and a 6000 head swine farm, the distribution of results by state, end use, and farm type are illustrated in Fig. 7. North Carolina and Oklahoma have much higher returns, which is related to their initial manure management system.

Independent of the energy cost, there is also variability in the average size of swine operations. For example, swine in North Carolina are much more concentrated on large farms above 5000 head than in Midwestern states such as Iowa and Minnesota, except for Missouri and Illinois (see Supporting Information for additional information). Hence, there will also be differences between states in the adoption rate, depending on the industry structure in each state.
Discussion
There are caveats regarding the assessment of AD construction and end-use presented in this paper. Those limitations do not change the results that carbon payments can spur the construction of ADs. First, the results presented in the previous section assume perfect foresight in terms of carbon and energy prices. From the financial economics literature, it is well known that uncertain revenues from a project hinders and delays investment.4 The analysis also suggests that carbon payments have a larger impact on the IRR, and thus on the decision to construct an AD, than energy prices. Hence, long-term certainty of carbon payments is more important for than energy prices. Price and revenue stability can be achieved through long-term contracts between energy end users and/or entities purchasing carbon credits. Second, our analysis does not take into account the connection cost to a local natural gas pipeline. Incorporating that cost can substantially limit the profitability of pipeline injection of RNG if swine farms are not close to injection points. This is likely because pipelines connect cities to the natural gas grid whereas livestock farms are constructed away from cities due to odor. Third, our model also does not include methane emissions from leakage of anaerobic digesters or long manure storage periods.19
There is also the possibility of unintended consequences if manure becomes more valuable due to carbon offsets. That is, there is potentially the incentive for livestock producers to increase the herd size to produce more biogas. That would, of course, be counter to the additionality requirement of carbon offsets. Installing biogas facilities on a large scale can also result in changing crop production in the surrounding areas. Manure is not the only feedstock for an anaerobic digester, which can also include food waste, agricultural residues, or other organic material such as maize silage. This can incentivize the growing of feedstock for the biogas digester as opposed to other uses.20 There have also been newspaper reports about the unintended consequences on land-use in Europe. See for example ‘How a false solution to climate change is damaging the natural world’ in The Guardian on 14 March 2014 or ‘Biogas boom in Germany leads to modern-day land grab’ in Der Spiegel on 30 August 2012. A carbon price or tax affecting the electricity sector could also have indirect effects on digester adoption. A capped sector such as electricity producers could pass the higher cost of electricity production to consumers (including farmers), which would make it more profitable for farmers to adopt methane digesters.21 A carbon tax would result in the same effect by making the biogas produced more valuable.3 As pointed out by Key and Sneeringer,3 this could also lead to unintended consequences if livestock producers are increasing the methane output to produce more biogas. In short, the adoption of more digesters creates the potential for complicated feedbacks in other input sectors that are difficult to discern.
Conclusion
There is a large potential for US livestock producers to reduce emissions from manure management by constructing anaerobic digesters to produce energy. The biogas end product can be either renewable natural gas or electricity – with various technology options available for the production – there is flexibility depending on the energy and carbon prices as well as potential policy instruments in place. That is, livestock producers integrated into a carbon market will have multiple technology pathways available between manure production and sale of energy.
The analysis shows that the construction of an AD is much more likely for a wider range of farm sizes given carbon payments because otherwise, of the two types of facilities analyzed, only large breeding swine farms can benefit from economies of scale. The production of electricity leads to a higher return given the current price ratio of electricity to natural gas. A higher natural gas price (with or without a subsidy) could shift the end use away from electricity. Although the internal rates of returns are relatively close to each other for electricity and CNG production. Given carbon payments, the construction of ADs is more profitable if the initial manure management system is GHG intensive such as anaerobic lagoons which are common in warmer climates such as North Carolina. A large hurdle is the initial capital cost of building a digester, which can be overcome by organizing the construction cooperatively, i.e., locating an AD at a central location that is accessible to multiple small producers.
There are also other regional aspects that may determine the construction of anaerobic digesters as well as the biogas end use. For example, there may be state incentives and policies that favor electricity over RNG production or vice versa, independent of energy prices. One such example would be the low carbon fuel standard (LCFS) in California, which has spurred not only AD construction but also compressed natural gas as the end use.
Integrating livestock producers into carbon markets, which would subsequently lead to local energy production, can also spur rural support of the energy transition. CNG production on livestock farms can support shifting the heavy freight transportation sector away from fossil fuels. The CNG fuel station network is sparse in Iowa and Illinois as well as the western parts of North Carolina where swine production dominates. Electricity production from biogas could support moving the light-duty vehicle sector towards electric vehicles. There is also the possibility that energy consumers have a high willingness to pay for low-carbon energy sources. There are many opportunities and challenges ahead for US agriculture, with biogas production having significant economic and environmental potential.
Acknowledgements
This research was supported in part by the US Department of Agriculture (USDA) Economic Research Service (ERS) Cooperative Agreement Number 58-3000-1-0089 ‘Swine production and carbon markets.’ The findings and conclusions in this publication are those of the authors and should not be construed to represent any official USDA or US government determination or policy. The views expressed herein are those of the authors and do not necessarily reflect the views of the USDA.
Biographies
Jerome Dumortier
Jerome Dumortier is associate professor at the O'Neill School of Public and Environmental Affairs at Indiana University Indianapolis. His research focuses on energy and agricultural economics. He is co-editor-in-chief of BioEnergy Research and Applied Economic Perspectives and Policy.
John Crespi
John Crespi is director of the Center for Agricultural and Rural Development and professor in the Department of Economics at Iowa State University. His areas of expertise include agricultural economics and industrial organization.
Dermot J. Hayes
Dermot J. Hayes is Charles F. Curtiss Distinguished Professor in Agriculture and Life Sciences in the Department of Economics and professor and Pioneer Hi-Bred International Chair in Agribusiness in the Ivy School of Business at Iowa State University.
Molly Burress
Molly Burress is deputy director for data management in the Market and Trade Economics Division (MTED) of the US Department of Agriculture's (USDA's) Economic Research Service. Previously, she served MTED as assistant director for communications and as managing editor of Web communications.
Adriana Valcu-Lisman
Adriana Valcu-Lisman is an economist with the Market and Trade Economics Division at the US Department of Agriculture's (USDA's) Economic Research Service. She was a postdoc with the Center for Agriculture and Rural and Development and a scientist in the Department of Natural Resources and Ecology Management at Iowa State University.
Jan Lewandrowski
Jan Lewandrowski currently contracts with the US Department of Agriculture's (USDA's) Economic Research Service as a senior climate advisor. He previously spent 32 years at USDA as a resource economist and has extensive experience leading research and policy support activities related to agriculture and forestry.
Open Research
Data Availability Statement
All data and code are available at www.github.com/SwineCarbonMarkets.