The Tipping Point: Economic Viability and Resilience of Dairy Manure Bioenergy Under Market and Policy Shocks

Abstract

This study evaluated the economic viability and resilience of anaerobic digestion (AD) systems on United States (U.S.) dairy, revealing substantial vulnerabilities to policy and market shocks. While optimal Renewable Natural Gas (RNG) systems demonstrated a 54.0% success probability and positive mean Net Present Value (NPV) ($392,000) under baseline volatility, their viability is catastrophically degraded by federal policy shocks, causing the success probability to plummet to 1.4%. Conversely, Combined Heat and Power (CHP) systems showed a lower baseline success rate (32.6%) and negative mean NPV ($−156,000) but exhibit more gradual vulnerability. These findings were derived from an integrated analytical framework combining deterministic optimization, Monte Carlo simulation, and a novel multidimensional resilience assessment. Deterministic analysis confirmed that revenue diversification is essential for viability, with optimal RNG and CHP configurations achieving breakeven at 655 and 1165 cows, respectively. Our novel Composite Resilience Index (CRI) revealed a counterintuitive finding: despite RNG’s superior baseline profitability, CHP systems achieve a higher overall resilience score (52.3 vs. 47.7) due to better stability and shock resistance. These results highlight the critical importance of incorporating uncertainty and resilience considerations beyond traditional NPV analysis for renewable energy investment decisions.

1. Introduction

The intensification of dairy farming in the United States (U.S.) has led to increasingly concentrated livestock operations, exacerbating environmental challenges and creating new categories of financial risk [1]. Accumulated manure can contribute to greenhouse gas emissions and pose risks to water quality through nutrient leaching if not managed appropriately. The United States Environmental Protection Agency reported that methane emissions from manure management reached 66.0 MMT CO₂ equivalent in 2021, a 69% increase from 1990 levels, with dairy cow manure being a primary contributor to this surge [2]. Anaerobic digestion (AD) presents a promising technological solution that addresses both environmental compliance and economic opportunity. By converting organic matter in manure into biogas, AD systems can reduce greenhouse gas emissions, minimize odor, and produce valuable outputs including bioenergy and processed co-products [3,4,5].

The economic viability of farm-based AD systems, however, has historically been a significant hurdle to widespread adoption. Numerous studies indicate that baseline AD systems, relying solely on energy sales, often struggle with economic viability due to high initial capital costs, substantial operational expenses and uncertain revenue streams [6,7]. Recent Techno Economic Analysis (TEA) continue to underscore these challenges, often indicating that AD systems are not self-sustaining, particularly on a small to medium-scale [5].

To overcome these barriers, the literature emphasizes two critical strategies: valorization of co-products and integration of environmental credit (EC) revenues. Co-products, such as separated solids for bedding or soil amendments, and concentrated nutrients for fertilizer applications, can provide significant additional revenue streams [8,9,10]. Similarly, ECs include Renewable Identification Numbers (RINs) under the Renewable Fuel Standard, Renewable Energy Certificates (RECs), and carbon credits [4]. These mechanisms can theoretically transform marginal projects into economically at-tractive investments.

While the inclusion of co-product revenues and ECs can substantially improve the theoretical economic viability of AD projects, the stability and longevity of these revenue streams are subject to considerable uncertainty and volatility. This volatility is a key aspect of the broader risks inherent in renewable energy investments, which include not only market price fluctuations but also uncertainties related to the stability and longevity of support mechanisms [11]. The perceived reliability of such support mechanisms is a critical determinant of investment in the renewable energy sector [11].

EC mechanisms are highly dependent on evolving political priorities and regulatory frameworks [12]. Studies reveal that environmental and energy policy uncertainties reached elevated levels during recent political transitions, with measurable negative impacts on renewable energy investment [13]. Historical analysis of renewable energy policy transitions demonstrates both gradual phaseouts and abrupt discontinuations that create distinct risk profiles for investors. Gradual policy erosion allows market adaptation but erodes long-term project economics through sustained uncertainty effects [14]. Conversely, sudden policy reversals, such as those documented during the 2017–2021 administrative transition where federal environmental priorities shifted dramatically, can create immediate viability crises for projects dependent on federal incentives [12,15]. These contrasting uncertainty patterns necessitate differentiated analytical approaches in TEA. Additionally, nascent markets for novel outputs like processed digestate fiber can also impact revenue projections [4]. The potential for EC devaluation due to additionality concerns or market saturation further compounds investment risk, as policies may be revised if credits fail to deliver measurable environmental benefits beyond baseline scenarios [14]. This multifaceted uncertainty landscape requires analytical frameworks capable of evaluating system performance under both gradual policy evolution and discrete shock events [16].

This research addresses these limitations by developing and applying a comprehensive framework that explicitly incorporates deep uncertainties into AD system economic assessment. The study makes three primary contributions to the renewable energy economics literature. First, it develops a Composite Resilience Index (CRI) that moves beyond traditional expected value metrics to capture multiple dimensions of system robustness under uncertainty. While other models like Real Options Analysis (ROA) or simple volatility metrics exist, the CRI’s novelty lies in its integration of multidimensional performance metrics with system performance under discrete, plausible shock scenarios, moving beyond simple price volatility. Second, it quantifies the economic impact of plausible policy and market shocks on optimized AD configurations, providing empirical evidence of vulnerability patterns that are invisible in conventional analysis. Third, it demonstrates how resilience-based assessment can fundamentally alter technology selection decisions compared to traditional net present value analysis.

2. Materials and Methods

This section outlines the TEA framework used to evaluate the economic viability and resilience of AD systems on U.S. dairy farms under varying degrees of market and policy uncertainties. The framework combined deterministic optimization, scenario-based stochastic analysis, and assessment of a novel CRI.

2.1. System Boundary and Configurations

The AD systems analyzed were designed for manure management on dairy farms. The analysis encompassed four primary components: (1) the AD for converting dairy manure into digestate and raw biogas, (2) energy conversion options including either Combined Heat and Power (CHP) systems for on-site electricity generation or Renewable Natural Gas (RNG) systems for biogas upgrading to pipe-line-quality biomethane, (3) digestate processing units for mechanical solid separation and liquid nutrient recovery that generate co-product revenue, and (4) EC mechanisms providing revenue streams from RINs, RECs, carbon credits, and federal tax credit. The RNG system’s upgrading process includes the removal of CO₂ and impurities like H₂S, the costs of which are captured within the system’s Operating & Maintenance (O&M) cost functions. These components combine to yield the six configurations in

Table 1. System Configuration Matrix. This table outlines the six distinct Anaerobic Digestion (AD) scenarios analyzed, categorized by energy output type (Combined Heat and Power [CHP] vs. Renewable Natural Gas [RNG]) and the level of revenue integration. The “Base” level includes only energy sales; “Enhanced” adds revenue from fiber and nutrient co-products; “Premium” integrates all available environmental credits (Renewable Identification Numbers [RINs], Renewable Energy Certificates [RECs], Carbon Credits, and Tax Credits).

Configuration	Energy Product	Co-Products	Environmental Credits (ECs)	Revenue Streams
CHP-Base	Electricity	–	–	Electricity
RNG-Base	Biomethane	–	–	Biomethane
CHP-Enhanced	Electricity	Fiber, nutrients	–	Electricity, co-products
RNG-Enhanced	Biomethane	Fiber, nutrients	–	Biomethane, co-products
CHP-Premium	Electricity	Fiber, nutrients	RECs, Tax Credit, Carbon credit	Electricity, co-products, ECs
RNG-Premium	Biomethane	Fiber, nutrients	RINs	Biomethane, co-products, ECs

, representing a progression from basic energy recovery to fully integrated “premium” systems.

2.2. Economic Model

The economic model followed a sequential, three-phase analytical framework, as illustrated in Figure 1. Phase 1 consisted of a deterministic TEA to establish baseline performance, identify breakeven points, and select the optimal system configurations for further analysis. Phase 2 built on this by introducing uncertainty, using Monte Carlo simulations to assess the performance of these optimal systems under four distinct shock scenarios. Finally, Phase 3 used the full set of NPV distributions from the stochastic analysis to conduct the comprehensive resilience assessment, culminating in the final CRI score.

2.2.1. Deterministic Modeling Framework

An initial deterministic TEA was conducted to establish baseline economic performance and identify optimal system configurations for subsequent analysis.

Cost Functions

Capital costs $\left(C_i\right)$ and annual O&M costs (${\Omega }_i$) for each technological component $i$ were modeled as piecewise linear functions of farm size ($x$), defined as number of cows. This approach accounts for economies of scale and specific threshold effects. The capital cost function, and the annual O&M cost function, are defined in Equations (1) and (2), respectively:

$$C_i(x)=\left\{\begin{array}{c}{v}_{1,i}x+f_{1,i} if x\le {\alpha }_i\\ v_{2,i}x+f_{2,i} if x>{\alpha }_i\end{array}\right.$$

(1)

$${\Omega }_i(x)=\left\{\begin{array}{c}{w}_{1,i}x+g_{1,i} if x\le {\beta }_i\\ w_{2,i}x+g_{2,i} if x>{\beta }_i\end{array}\right.$$

(2)

where $v_{.,i}$ represents variable costs per cow and $f_{.,i}$ represents fixed capital costs; $w_{.,i}$ represents variable O&M cost per cow per year, $g_{.,i}$ represents fixed annual O&M costs and ${\alpha }_i$ and ${\beta }_i$ represents scale threshold parameters (number of cows) at which the cost structure changes for component $i$. Total system costs aggregated across all constituent technological components for each configuration.

Revenue Functions

Annual revenues were derived from three distinct categories with configuration specific streams. The total annual revenue, $R_t\left(x\right)$, is the summation of energy, co-product, and environmental credit revenues, as expressed in Equation (3):

$$R_t\left(x\right)=R_{energy,t}\left(x\right)+R_{coproducts,t}\left(x\right)+R_{ECs,t}\left(x\right)$$

(3)

Energy revenues reflected either electricity sales (CHP systems) or biomethane sales (RNG systems), calculated based on herd size, manure production characteristics, biogas conversion efficiencies, and market prices. Co-product revenues included fiber sales and recovered nutrient sales where applicable. EC revenues encompassed RINs, RECs, carbon credits and tax credits as appropriate for each configuration.

Net Present Value (NPV)

The NPV for a given AD system configuration and farm size was calculated over the capital lifetime $T$. The $NPV$ is derived by discounting annual net cash flows at the real discount rate $r$, as formulated in Equation (4):

$$NPV\left(x\right)={\sum }_{t=0}^T{\displaystyle \frac{R_t\left(x\right)-{\Omega }_t\left(x\right)}{{(1+r)}^t}}-C_0(x)$$

(4)

Projects yielding an NPV > 0 are considered economically viable.

2.2.2. Stochastic Simulation Setup

Building directly on the deterministic results, the stochastic analysis involved four scenarios capturing various levels of market and policy shocks. Each scenario employed Monte Carlo simulation (10,000 iterations) with triangular probability distributions for all uncertain parameters, including prices, output yields, and credit values.

Uncertainty Specification and Distribution Selection

Triangular distributions were selected for their transparency and tractability in technology assessment contexts where historical data may be limited or non-stationary. For each uncertain parameter, the triangular distribution is defined by three values: a minimum (lower bound), a mode (most likely value), and a maximum (upper bound). The baseline deterministic values presented in Appendix A serve as the mode (most likely value) for all distributions, while the minimum and maximum bounds reflect plausible ranges based on historical price volatility, engineering uncertainty, and expert judgment (Appendix B). This approach assumes symmetric or near-symmetric uncertainty around baseline values, which is appropriate for technology and market parameters without strong directional bias.

While alternative distribution families such as lognormal distributions for energy prices might better capture empirical price dynamics in mature commodity markets, triangular distributions offer several advantages for this analysis: (1) they require fewer parameterization assumptions in contexts with limited historical data for novel revenue streams (e.g., digestate fiber markets, D3 RINs), (2) they allow transparent communication of uncertainty bounds to stakeholders, and (3) they avoid the right-skew bias of lognormal distributions that could overstate upside potential in nascent markets. Future research could employ empirically fitted distributions as longer time series become available for emerging AD revenue streams.

Scenario Definitions

The scenarios are grounded in historical precedents and plausible market dynamics to test the systems against realistic threats:

(A)

Baseline Volatility: Normal price variance based on triangular distributions (Appendix B), with no discrete shocks. This scenario represents routine market fluctuations without structural breaks.

(B)

Phased Policy Rollback: This scenario modeled a linear taper of federal incentives between project years 6 and 10, flattening at 10 percent of baseline thereafter. This five-year phase-down was modeled directly on the precedent set by the U.S. Congress in 2015 for the wind energy Production Tax Credit [17]. The legislation mandated a gradual reduction in the credit’s value over a multi-year period (e.g., to 80% in 2017, 60% in 2018, and 40% in 2019) before its expiration, providing a real-world example of a structured, long-term policy sunset.

(C)

Sudden Policy Shock: This scenario modeled an abrupt elimination of federal incentives in project year 5. The timing was chosen to emulate a potential, sharp reversal in federal energy and environmental priorities following a change in a U.S. presidential administration (which operates on a four-year cycle). Such sudden shifts have historical precedent; for example, executive order by Trump administration in 2016 to cancel the Clean Power Plan, introduction of 30% tariffs on imported solar panel in 2018 had an immediate impact on renewable energy development and installations [18].

(D)

Co-product market collapse: This scenario modeled a failure of the nascent market for AD co-products, with the shock occurring in project year 3. This early-stage timing reflects the high failure rate of new markets and the critical period for a project when debt service requirements were high, but cash flows are not yet stable [19,20]. The market for agricultural co-products like digestate fiber is characterized as underdeveloped in the U.S. and is subject to the inherent price volatility of all agricultural commodities, which face unpredictable shocks from weather, trade disruptions, and financial market spillovers [21,22].

Scenario Weighting and Aggregation

The four scenarios were designed as discrete stress tests rather than probabilistic forecasts, recognizing that assigning precise probabilities to structural policy shifts or nascent market failures would introduce false precision. However, for aggregation purposes in the CRI calculation, scenarios must be weighted. We employed equal weighting (25% each) as the baseline approach, which reflects a risk-neutral stance that treats all plausible shocks as equally important for resilience assessment. This conservative assumption prevents over-optimization for any single future state and aligns with decision-making under deep uncertainty, where probability distributions over scenarios are unknowable.

Alternative weighting schemes could reflect different risk perspectives: a policy-stability optimist might downweight Scenarios B and C (gradual/sudden policy shocks) relative to Scenario D (market shocks), while a policy-pessimist would reverse these weights. Future extensions could employ scenario weights derived from expert elicitation or historical base rates of comparable policy transitions, though such weights would necessarily remain subjective.

The CRI aggregation proceeded in two stages: (1) within each scenario, metrics were normalized and weighted by dimension to produce a scenario-specific CRI score; (2) scenario-specific CRI scores were then averaged using equal scenario weights to produce the final aggregate CRI.

2.2.3. CRI Construction and Weighting

Traditional TEA treats expected NPV as a sufficient statistic, implicitly assuming risk-neutral investors and time-invariant distributions. Renewable-energy projects, however, are repeatedly shown to face deep uncertainties and structural breaks [23]. Resilience, defined as the ability of a system to anticipate, withstand, adapt to, and recover from shocks, offers a more robust and realistic lens through which to evaluate project viability [24].

The CRI developed for this study operationalizes this concept by translating principles from resilience engineering and financial risk management into a quantitative framework. The CRI aggregates 20 distinct metrics, each selected for its relevance in techno-economic and financial analysis, into seven conceptual dimensions that provide a holistic view of a project’s financial robustness.

Scenario Integration in CRI Calculation

The CRI integrated scenario-specific information through a two-tier approach. Six of the seven resilience dimensions (Resistance, Stability, Downside Protection, Recovery & Adaptation, Diversification, and Financial Strength) were calculated using performance metrics from Scenario A (Baseline Volatility) only, ensuring these dimensions reflect system performance under normal market conditions.

The seventh dimension, Shock Resistance, explicitly incorporated the discrete policy and market shocks (Scenarios B, C, D) by quantifying each system’s sensitivity to departures from baseline conditions. For each system, the percentage change in key metrics (probability of profitability, mean NPV, and tail risk) was calculated for each shock scenario relative to baseline. These three shock-specific impacts were then equally averaged to produce composite shock sensitivity metrics. Equal averaging of the three shock scenarios reflects the judgment that these structurally distinct risks warrant equivalent consideration in resilience assessment. The averaged shock sensitivities are incorporated into the Shock Resistance dimension after transformation to ensure higher values indicate greater resilience.

This approach allowed the aggregate CRI score to simultaneously reflect baseline performance and stress-test robustness, with the relative importance of these aspects determined by the dimensional weights.

Dimensional Definitions and Weights

The seven dimensions, their respective weights and their constituent metrics were as follows:

• Resistance (30%): Quantifies baseline profitability and likelihood of success. Metrics: Probability of positive NPV, mean NPV, conditional mean NPV if positive.
• Stability (20%): Measures predictability and volatility of financial returns. Metrics: Standard deviation, coefficient of variation, interquartile coefficient.
• Downside Protection (20%): Focuses on tail risk and catastrophic loss potential. Metrics: Value-at-Risk at 5th percentiles, Conditional Value at Risk, worst-case NPV.
• Recovery & Adaptation (10%): Captures ability to capitalize on favorable conditions. Metrics: Mean upside NPV above median, Ratio of 95th/5th percentiles, Normalized skewness.
• Diversification (8%): Assesses revenue stream concentration and policy dependency. Metrics: Herfindahl index of revenue sources, Inverse of policy dependency (defined as 1 minus the fraction of baseline revenue from policy-dependent credits), market stability (defined as the inverse of the historical coefficient of variation for the primary energy commodity).
• Shock Resistance (15%): Quantifies financial damage during discrete shocks. Metrics: Percentage NPV loss under scenarios B, C, D relative to baseline A.
• Financial Strength (12%): Uses the baseline mean NPV as proxy for fundamental viability. Metric: Baseline mean NPV.

Each metric was min-max normalized, then weighted by dimension. The dimensional weights reflect theoretical and empirical importance observed in energy finance literature, with higher weights assigned to dimensions capturing fundamental viability (Resistance, Downside Protection) and structural robustness (Shock Resistance, Stability). The final CRI is computed by aggregating the weighted normalized scores of all dimensions. The aggregation logic is mathematically defined in Equation (5):

$$CRI= \sum _{d=1}^7{w}_d\left(\sum _{m=1}^{M_d}{\displaystyle \frac{{metric}_{d,m}}{M_d}}\right)$$

(5)

where $w_d$ represents the weight for dimension $d$ and $M_d$ represents the number of metrics in dimension $d$. The CRI ranges from 0 to 1, with interpretation thresholds established through empirical analysis.

2.2.4. Sensitivity Analysis of CRI Weights

The subjectivity of weight assignment is a known challenge in multi-criteria decision analysis. To address this and test the robustness of our findings, a sensitivity analysis was performed on the CRI’s dimensional weights. Three alternative weighting schemes were developed to simulate different investor priorities, based on the original weights:

• Equal Weights: All seven dimensions were weighed equally (14.3%), representing a neutral or agnostic perspective.
• Profit-Focused: We defined Profit-Oriented Dimensions (Resistance, Recovery & Adaptation, Financial Strength) and Risk-Mitigation Dimensions (Stability, Downside Protection, Diversification, Shock Resistance). We then applied a 3× multiplier to the original weights of Profit-Oriented Dimensions and a 0.5× multiplier to Risk-Mitigation Dimensions. The resulting normalized weights were: Resistance (46.2%), Recovery & Adaptation (27.7%), Financial Strength (9.2%), Stability (6.2%), Downside Protection (6.2%), Diversification (3.1%), and Shock Resistance (1.5%).
• Risk-Averse: This scheme modeled a conservative decision-maker by reversing the logic. We applied a 3× multiplier to the Risk-Mitigation Dimensions and a 0.5× multiplier to the Profit-Oriented Dimensions. The resulting normalized weights were: Stability (32.0%), Downside Protection (32.0%), Diversification (16.0%), Shock Resistance (8.0%), Resistance (6.7%), Recovery & Adaptation (4.0%), and Financial Strength (1.3%).

2.3. Data and Key Assumptions

The technical and economic parameters for this study were primarily derived from the techno-economic model detailed by a 2016 study [4], which builds upon the Anaerobic Digester System Enterprise Budget Calculator with inputs validated through industry partnerships and engineering assessments [25]. Key baseline assumptions included a project operational lifetime of 20 years and a real discount rate of 4%. Specific price assumptions for energy, co-products, ECs, and cost parameters are detailed in Appendix A and Appendix B. All financial values are presented in U.S. dollars.

3. Results

The results are presented in three sequential phases: deterministic optimization analysis establishing optimal configuration, stochastic analysis quantifying performance under various scenarios, and CRI assessment revealing multidimensional system robustness patterns.

3.1. Baseline Economic Performance and Breakeven Analysis

Under baseline economic conditions with stable market prices and policy support, both CHP and RNG options faced significant economic challenges across the farm size range analyzed (50–15,000 cows). Figure 2 presents the deterministic NPV results for all six configurations, revealing fundamental viability constraints that establish the foundation for subsequent uncertainty analysis.

Base CHP and RNG configurations demonstrated uniform negative NPV across all farm sizes, confirming that energy-only revenue streams cannot support economic viability under current cost structures. Co-product integration substantially improved economic performance for both energy options. The CHP-Enhanced configuration incorporating fiber and nutrient recovery generated positive NPV for farms larger than 2250 cows, while RNG-Enhanced systems showed breakeven at approximately 8500 cows.

EC inclusion enabled positive NPV achievement at practical farm scales. The RNG-Premium configuration achieved breakeven at 655 cows, while the CHP-Premium configuration reaches breakeven at 1165 cows. The superior breakeven performance of the RNG system was driven almost entirely by the high value of federal RINs, which provided a scalable revenue stream unavailable to the CHP system. Based on their financial superiority under deterministic conditions, RNG-Premium and CHP-Premium were identified as the optimal configurations for the subsequent stochastic analysis.

3.2. Vulnerability Assessment Under Shock Scenarios

The Monte Carlo simulation results revealed the risky financial status of both optimal CHP and RNG configurations. The results, summarized in

Table 2. Financial Performance of Premium AD Systems across Stochastic Scenarios. The table presents the statistical distribution of Net Present Value (NPV) outcomes derived from 10,000 Monte Carlo simulations for the optimal Combined Heat and Power (CHP) and Renewable Natural Gas (RNG) configurations. Metrics include: Mean NPV (average financial outcome); 5th and 95th Percentile NPV (tail risk boundaries); P(NPV > 0) (probability of the project breaking even); 5% Value-at-Risk (VaR) (maximum expected loss at 95% confidence); and Coefficient of Variation (CV) (measure of relative volatility). All monetary values are in millions of U.S. dollars ($M).

Scenario	System	Mean NPV ($M)	5th %ile NPV ($M)	95th %ile NPV ($M)	P(NPV > 0)	5% VaR ($M)	CV
A. Baseline Volatility	CHP	−0.52	−2.39	1.44	32.6%	−2.41	2.25
	RNG	0.39	−2.81	4.01	54.0%	−2.82	5.37
B. Phased Policy Rollback	CHP	−1.12	−2.50	1.32	18.7%	−3.16	1.89
	RNG	−0.89	−3.58	0.79	13.9%	−4.20	3.42
C. Sudden Policy Shock	CHP	−1.29	−2.97	0.57	13.3%	−3.45	1.76
	RNG	−2.35	−4.05	−0.58	1.4%	−5.89	1.98
D. Co-product Market Collapse	CHP	−3.55	−5.07	−2.0	0.0%	−5.23	0.98
	RNG	−1.31	−4.44	2.27	26.4%	−3.99	2.14

, revealed distinct risk profiles and highlight critical vulnerabilities that are highly dependent on the nature of the shock.

Under baseline market conditions (Scenario A), a fundamental risk-return trade-off emerged. The RNG system presented a positive expected outcome (mean NPV of $392,000) but was shadowed by extreme volatility (coefficient of variation of 5.37). This uncertainty was underscored by its wide range of potential outcomes, with a 5th percentile NPV of −$2.81 million and a 95th percentile NPV of $4.01 million. Conversely, the CHP system was unprofitable on average (mean NPV of −$516,000) but with significantly lower volatility. This suggested a more predictably negative result, with a much tighter, though still negative, distribution, ranging from a 5th percentile loss of $2.39 million to a 95th percentile gain of $1.44 million. Both systems faced severe downside risk, as shown by their 5% Value-at-Risk figures.

Financial performance degraded substantially under adverse policy scenarios (B and C). An abrupt federal policy shock was particularly catastrophic for the RNG system, whose probability of success plummeted from 54.0% to just 1.4%. The percentile data revealed the severity of this shock: the entire range of likely outcomes became negative, with the 95th percentile NPV collapsing to −$0.58 million. This extreme result revealed a critical vulnerability to federal incentives like RINs. While also damaged, the CHP system proved more resilient to policy shocks due to its diversified revenue from electricity sales and state-level credits.

In a reversal of fortunes, a co-product market collapse (Scenario D) exposed the CHP system’s primary vulnerability. Its financial viability was completely eliminated, with the probability of a positive NPV falling to zero. This was reflected in the NPV distribution, where even the 95th percentile outcome is a significant loss of −$2.0 million, confirming its heavy reliance on secondary revenue streams like fiber. The RNG system, while significantly impacted, was more resilient to this market-specific shock, retaining a 26.4% chance of profitability.

Ultimately, the analysis demonstrated that neither system was universally superior. Financial resilience was context-dependent, contingent on whether disruptions originate from policy shock or associated commodity markets.

3.3. Multidimensional Resilience Profiles

The multi-dimensional CRI analysis revealed a counterintuitive finding that is not apparent from NPV distributions alone: despite RNG systems’ superior baseline profitability (Table 2), CHP systems achieved higher overall resilience scores (CRI = 0.523 vs. 0.477) due to fundamental differences in their dimensional resilience profiles (Figure 3).

Figure 3 provides critical insight beyond the scenario-specific metrics in Table 2 by visualizing the complete resilience topology across all seven assessment dimensions simultaneously. The radar chart revealed a distinct dimensional pattern that explained the aggregate CRI inversion: RNG exhibited a “peaked” profile with exceptional performance in resistance (0.714) and recovery (0.601) dimensions but critical weaknesses in stability (0.512) and shock resistance (0.289). In contrast, CHP demonstrated a more “balanced” profile with moderate-to-strong performance across all dimensions, particularly excelling in stability (0.688) and diversification (0.613). This visual representation made immediately apparent that CHP’s resilience advantage stems not from superior performance in any single dimension, but from avoiding the catastrophic vulnerabilities that characterize RNG systems when facing their specific risk exposures.

This distinction highlights the difference between short-term viability (captured by baseline NPV metrics) and long-term resilience under uncertainty (captured by multidimensional assessment). The dimensional decomposition in Figure 3 demonstrates why equal-weighted or risk-averse investors might rationally prefer CHP despite its lower expected returns, the technology avoids the severe tail risks that make RNG vulnerable to specific shock types.

RNG systems dominated in resistance metrics (0.714 vs. 0.469), reflecting their superior baseline economics and higher probability of positive outcomes. They also exceled in recovery and adaptation (0.601 vs. 0.374), indicating better capacity to capitalize on favorable market conditions. Financial strength scores (0.539 vs. 0.448) confirm RNG systems’ fundamental economic advantages. CHP systems demonstrated superior stability (0.688 vs. 0.512), exhibiting lower volatility and more predictable outcomes. Their diversification advantage (0.613 vs. 0.535) stemmed from broader revenue streams spanning electricity sales, multiple environmental credits, and co-products. CHP systems also showed better shock resistance (0.456 vs. 0.289) when facing discrete negative events. An analysis of resilience scores across predefined scenarios revealed critical vulnerability patterns and the tradeoffs inherent in technology selection as shown in Figure 4.

Under baseline conditions, the RNG system (orange line) demonstrated a marginal resilience advantage with a mean CRI of approximately 0.501, while the CHP system (blue line) started at approximately 0.498. Their overlapping 95% confidence intervals suggest this difference may not be statistically significant, but the RNG system is the clear top performer under normal conditions.

This advantage was immediately lost under policy stress. In the “Phased Policy Rollback” scenario, the RNG system’s resilience dropped sharply, crossing below the CHP system, which had a more moderate decline. This gap widened even more in the “Sudden Policy Shock” scenario, which is clearly the worst-case outcome for RNG. Its CRI plummets to its lowest point (0.218), while the CHP system, though degraded, maintained a significantly higher resilience score (0.31).

The roles completely reversed in the final scenario. The “Co-product Market Collapse” was catastrophic for the CHP system, causing its CRI to crash to its lowest point (0.171%). In contrast, the RNG system, which is less reliant on these co-product markets, actually showed a slight recovery, ending with a mean CRI of 0.32.

This analysis demonstrated that each system has a distinct “Achilles’ heel”: RNG’s resilience is highly vulnerable to policy stability, while CHP’s resilience is critically dependent on co-product market stability.

3.4. Sensitivity of Resilience Rankings to Weighting Schemes

The sensitivity analysis confirmed that the central finding, the superior resilience of the CHP system, is robust and not an artifact of the initial weighting scheme. As shown in

Table 3. Composite Resilience Index (CRI) Sensitivity to Alternative Weighting Schemes. This analysis tests the robustness of the resilience ranking between CHP and RNG systems by applying four distinct weighting profiles to the seven CRI dimensions: “Original Weights” (baseline assumptions); “Equal Weights” (neutral perspective); “Profit-Focused” (prioritizing return metrics); and “Risk-Averse” (prioritizing stability and protection metrics). A higher CRI score (0–1 scale) indicates superior system resilience.

Weighting Scheme	CHP-Premium CRI	RNG-Premium CRI	Ranking
Original Weights	0.535	0.464	CHP > RNG
Equal Weights	0.518	0.482	CHP > RNG
Profit-Focused	0.367	0.633	RNG > CHP
Risk-Averse	0.632	0.368	CHP > RNG

, the CHP-Premium system maintained a higher CRI score under both the “Equal Weights” scenario (CRI: 0.518) and the “Risk-Averse” scenario (CRI: 0.632). The “Risk-Averse” findings were particularly strong, as this scheme assigns a combined 64.0% of total importance to the ‘Stability’ and ‘Downside Protection’ dimensions alone.

The ranking inverted only under the “Profit-Focused” scheme, which reallocates weights to heavily prioritize returns. This result allows for a more nuanced conclusion: while RNG may be preferred by investors focused purely on maximizing expected returns (CRI: 0.633), CHP represents a more resilient choice for a majority of decision-makers, particularly those with a lower appetite for risk.

4. Discussion

This study provided a comprehensive TEA of AD systems on U.S. dairy farms, moving beyond deterministic evaluations to explore critical pathways to economic viability under significant market and policy uncertainties. The introduction of the CRI revealed important tradeoffs between baseline profitability and system robustness that have fundamental implications for investment decisions, policy design, and the broader deployment of agricultural bioenergy systems.

4.1. Theoretical Implications for Renewable Energy Investment Analysis

The divergence between NPV-based and resilience-adjusted technology rankings demonstrates a critical gap in renewable energy investment literature [26,27]. Traditional approaches that assume risk neutrality and focus on expected value optimization may systematically undervalue technologies with superior risk management characteristics [28]. The CRI addresses this limitation by incorporating concepts from complex systems resilience theory [29,30], recognizing that energy infrastructure investments often prioritize stability and downside protection over pure return maximization [31].

This framework contributes to the growing literature on deep uncertainty in renewable energy systems [32] by providing an operational methodology for incorporating discrete shock scenarios into technology assessment. The demonstration that structural breaks can eliminate investment attractiveness regardless of baseline performance supports theoretical arguments about the limitations of probability-based approaches when underlying distributions are unstable or unknown.

The multidimensional resilience assessment also addresses recent calls in the renewable energy economics literature for more comprehensive risk evaluation frameworks that move beyond simple volatility measures to capture the complex risk-return tradeoffs characterizing clean energy investments under policy uncertainty [33].

4.2. Revenue Structure Analysis and Risk Concentration

The detailed revenue composition analysis revealed that apparent diversification benefits may be illusory when revenue streams share common risk factors. CHP systems’ revenue distribution across electricity sales (45.2%), co-products (31.4%), and environmental credits (23.4%) initially suggested superior risk management. However, the complete viability collapse under co-product market failure demonstrated that diversification provides limited protection when risks are correlated. An analysis of historical data indicates a positive correlation between agricultural commodity prices and energy prices, suggesting that a downturn in the agricultural economy could simultaneously depress co-product values and energy revenues [34,35]. This challenges the conventional wisdom that revenue diversification necessarily reduces risk, suggesting instead that effective risk management requires understanding the underlying correlation structure of revenue streams [36].

The structural differences between CHP and RNG revenue portfolios create fundamentally different risk profiles that cannot be adequately captured through traditional variance-based risk measures. This insight has implications for renewable energy project finance more broadly, suggesting that portfolio optimization approaches should explicitly consider scenario-based stress testing rather than relying solely on historical correlation patterns.

4.3. Policy Design Mechanisms

The quantified vulnerability patterns provide empirical foundation for specific policy design mechanisms that address identified risk concentrations.

• For RNG Systems: The catastrophic vulnerability to federal policy shocks suggests the primary policy goal should be enhancing stability. Mechanisms could include long-term, fixed-price RIN contracts, graduated phase-out schedules with binding commitments similar to the 2015 Production Tax Credit extension, and federal loan guarantees to provide stability during policy transitions.
• For CHP Systems: The vulnerability to co-product market collapse indicates policy should prioritize market development. Mechanisms could include standardized product certification programs to build consumer trust, public procurement initiatives to create anchor demand, and temporary minimum price guarantees during the market development phase.

The analysis also revealed the importance of timing in policy support mechanisms. The critical vulnerability periods identified (year 3 for co-product markets, year 5 for federal policies) correspond to periods when debt service requirements are high but operational cash flows remain uncertain [37]. Support mechanisms that provide enhanced stability during these vulnerable periods, such as revenue insurance or stepped guarantee programs, could substantially improve investment attractiveness while minimizing long-term public commitments.

Risk-sharing mechanisms emerge as potentially more effective than traditional grant or tax credit approaches, given the substantial tail risks identified. Public-private partnership structures that share both upside potential and downside risks could align public policy objectives with private investment incentives, while managing fiscal exposure to policy and market uncertainties [38].

4.4. Limitations and Future Research Directions

Several methodological limitations suggest important directions for future research. The CRI weighting scheme, while tested for robustness, still reflects subjective judgments. The scenario specifications, while grounded in historical precedent, cannot capture all possible disruptions. Most importantly, the static analysis framework assumes fixed technology choices. Integrating ROA with the resilience framework could provide insights into the value of operational flexibility and optimal investment timing.

4.5. Resilience Analysis in the Context of Real Options

It is important to distinguish the CRI framework from ROA, another prominent method for valuing investments under uncertainty. ROA applies financial options pricing theory to quantify the value of managerial flexibility, such as the option to defer, expand, or abandon a project as new information becomes available. In ROA, uncertainty increases the value of these flexibility options. The CRI, in contrast, measures the inherent robustness of a fixed system configuration to shocks; it does not value the option to change that configuration mid-project.

The two frameworks are best viewed as complementary. The CRI can serve as a screening tool to identify the most inherently robust technological configurations. A subsequent ROA could then be applied to the most resilient option to determine the value of adding specific flexibility pathways.

Table 4. A Comparative Analysis of Investment Evaluation Frameworks. This table distinguishes the proposed Composite Resilience Index (CRI) from traditional Net Present Value (NPV) analysis and Real Options Analysis (ROA). It contrasts the frameworks based on their primary optimization goals, their mathematical treatment of deep uncertainty, and their implicit assumptions regarding management flexibility and risk preferences.

Feature	Net Present Value (NPV)	Composite Resilience Index (CRI)	Real Options Analysis (ROA)
Primary Goal	Maximize expected wealth	Assess multidimensional robustness	Value managerial flexibility
Treatment of Uncertainty	Priced as risk via discount rate	Modeled via simulation & discrete shocks	Valued as a source of opportunity
Key Output	Single point estimate of value ($)	Normalized composite score (0–1)	Expanded NPV (Static NPV + Option Value)
Implicit Assumption	Passive management; static path	Downside protection is a key priority	Active, strategic management; dynamic path
Primary Insight	RNG is superior based on higher expected return.	CHP is superior based on greater stability and shock resistance.	The ability to defer investment until policy is clearer, has a quantifiable economic value.

provides a comparative summary.

5. Conclusions

This study demonstrated that even optimized AD systems face substantial economic challenges under uncertainty. The core finding was that a trade-off exists between baseline profitability and long-term resilience: RNG systems are more profitable under stable conditions but are catastrophically vulnerable to federal policy shocks, while CHP systems are less profitable but more resilient due to diversification and lower policy dependence. This implies that for RNG, policy stability is the critical factor for investment, whereas for CHP, co-product market development is paramount. This reversal from NPV-based rankings demonstrates that incorporating resilience considerations can fundamentally alter optimal technology choices, guiding policymakers to prioritize stability and market creation, and providing investors with a framework that explicitly incorporates risk preferences beyond simple expected value.