The technical, economic, and environmental challenges of integrating sustainable aviation fuels (SAFs) have become a top priority for the renewable fuels industry. Greenhouse gas (GHG) emissions contributed by the aviation industry have grown steadily at a rate of 2.0% per year from 2000 to 2019 to a 2.8% global share of fossil fuel sourced emissions in 2019 ( Aviation, 2022). Most concerns, however, stem from the industry consensus that rapid growth will continue over the next 20 years. As of 2019, the International Civil Aviation Organization (ICAO) and Airbus estimate an annual growth rate of 4.3%, while Boeing’s estimate for commercial air transportation demand stood at 4.6% (Future of Aviation, 2019; Global Market Forecast, 2019-2038, 2019; Commercial Market Outlook, 2019-2038, 2019). However, the 66% reduction in global air travel observed in 2020 caused by the COVID-19 pandemic and resulting market changes (teleworking, increasing cargo demand, etc.) exemplified the air transportation industry’s exposure to market risks (Commercial Aerospace Insight Report, 2021). Nonetheless, the transience of events like COVID-19, persistence of long-term drivers for aviation industry growth, and global decarbonization goals combine to make a case for proactively addressing the sector’s climate change potential (Gössling and Humpe, 2020).

The prospect of the aviation industry negating nationally determined contributions (NDCs) submitted by countries under the Paris Climate Agreement (Paris Agreement, 2015) has already prompted action from several international organizations. Perhaps the most widely recognized countermeasures are being taken by the International Civil Aviation Organization (ICAO) with their Carbon Offsetting and Reduction Scheme for International Aviation (CORSIA). CORSIA initiatives aim to achieve carbon neutral growth after 2020 through a combination of operational improvements, decarbonizing propulsion, and purchasing carbon offsets (Climate Change Mitigation: CORSIA, 2019).

ICAO and others acknowledge radical technology shifts like electrification, fuel cells, or hybrid power that are promising for other transportation sectors as potential decarbonization strategies for aviation (Climate Change Mitigation: CORSIA, 2019; Driessen and Hak, 2021). However, the uniquely high energy density requirements of aircraft will likely relegate these strategies to long-term deployment timelines incompatible with CORSIA (Schwab et al., 2021). ICAO currently anticipates over 50% of their decarbonization goals will need to come from SAFs (Climate Change Mitigation: CORSIA, 2019). Moreover, national commitments such as the United State’s SAF Grand Challenge (SAF Grand Challenge Roadmap, 2022) and European Union’s ReFuelEU Aviation Initiative (Soone, 2022), which both set increasingly large SAF blending targets over time, underscore the role of SAF production as the leading strategy to decarbonize aviation.

However, with current global SAF production capacity estimated to be around 800 million gallons per year, amounting to less than 1% of total jet fuel demand, the SAF industry is clearly in its nascent stage (Newsom et al., 2023). Several barriers to adoption have been identified that have previously, and will continue, to challenge SAF market adoption. Primarily, limited feedstock availability, high collection costs, low yields, and immature process technologies have led to uncompetitive SAF prices (Chiaramonti, 2019). Additionally, when biomass resources are available, biorefineries are typically incentivized to produce sustainable diesel with additional tax credits over SAF (Rep Yarmuth, 2022). These challenges, as well as others, make attracting the capital needed to achieve economies of scale, reducing costs, and competing within the petroleum jet market difficult (Shahriar and Khanal, 2022).

Conversely, petroleum refineries, currently producing essentially all jet fuel globally, will be affected by the SAF market’s growth if some of its challenges are successfully addressed. The uncertain landscape in which refineries are currently planning and operating is further complicated by the decarbonization trajectories of the light/heavy duty and marine transportation sectors. The Energy Information Administration’s (EIA) Annual Energy Outlook (AEO) factors long-term market forces such as population growth, consumer behavioral shifts, and technology changes to project transportation fuel demands in the United States to 2050 as displayed in Figure 1 (Annual Energy Outlook, 2022, 2022). An alternative gasoline demand projection from the National Renewable Energy Laboratory’s (NREL) Automotive Deployment Options Projection Tool (ADOPT) that predicts a higher degree of vehicle electrification is also shown (Brooker et al., 2015). Combining these projections elucidates the broader industry’s expectation of gasoline demand either stabilizing or plummeting, with high degrees of uncertainty, alongside steady decreases in diesel and increases in jet fuel demands. These market shifts could set the stage for coordination between petroleum refiners and SAF producers to simultaneously meet increasing volumetric and decarbonization demands for aviation fuels (Global downstream outlook to 2035, 2019).

Numerous conversion pathways are being researched for drop-in SAF production such as alcohol-to-jet, oil-to-jet via hydroprocessing, and Fischer-Tropsch based gas-to-jet, yet no clear winner has emerged (Wang et al., 2016; Boyd, 2022). Supplementally, several refinery integration strategies including direct blending, co-location, co-processing bio-intermediates like pyrolysis or hydrothermal liquefaction oil, or equipment repurposing are also being considered (Holladay et al., 2020). Some studies have taken an integrated approach that analyzes SAF production alongside refinery integration with promising reductions in minimum feasible selling prices ranging between 3% and 34% reported (Tanzil et al., 2021; Su et al., 2022).

The goal of the modeling framework presented herein is to expand upon existing research into refinery integrated SAF production with a focus on drop-in SAF blending, refinery modeling, and risk analysis. A comprehensive refinery optimization model, consistent with industry practice, is utilized as a valuation tool that indicates the economic viability of SAF production under different market conditions. Furthermore, the refinery model is packaged into a Monte-Carlo framework that attempts to capture key economic drivers, their variability, and their impact on the risk profiles associated with different jet production strategies. More specifically, SAF integration is compared to a capital investment scenario which represents a more traditional approach refiners might consider to meet rising jet fuel demands. Mitigation costs needed to incentivize SAF integration are also deduced and converted to marginal abatement costs of CO₂ to supplement existing literature (Capaz et al., 2021). Finally, different incentive structures supplying mitigation costs are explored to understand what incentives could best de-risk SAF production. The market-sensitive incentive structures developed in this analysis could help CORSIA and other climate change initiatives succeed in decarbonizing liquid transportation fuels.

NREL’s Economic, Sustainability, and Market Analysis Group has developed a collection of refinery economic models within AspenTech’s Process Industry Modelling System (PIMS) software. These models utilize non-linear programming (NLP) to determine the most profitable unit capitalizations and operating parameters given pertinent constraints and feedstock/product pricing. Figure 2 is a graphical depiction of the refinery modelling framework taking inputs, optimizing refinery operations, and outputting optimal material/cash flows. Model inputs are formatted and packaged into “cases” that are individually optimized. The methodology of this analysis, along with others (Carlson et al., 2023; Jiang et al., 2023), can be most simply described as forming case stacks representing refinery bio-integration strategies and consequently observing how a refinery model adapts. Additionally, each strategy’s optimal profitability and environmental impact can be compared. A more in-depth explanation of the modeling framework can be found in (Carlson et al., 2023).

Several modifications were made to the refinery modelling framework presented in (Carlson et al., 2023) to better address the questions posed in this analysis.

- What is the expected value of a refiner investing capital to meet increasing jet demands?

Modifications enabled discounted cash flow (DCF) analysis so that capital investments and future cash flows could be simultaneously considered with net present value (NPV) as a consistent metric. The intention was not to compute NPVs based on the whole refinery’s cash flows for either capital investment or SAF integration strategies. Instead, NPVs attributed to changes in refinery cash flows, from either capital investments or SAF purchases, were obtained by first running baseline cases and rerunning with pertinent changes in place. By analyzing only changes in performance, the influence of certain nuisances and assumptions inherent in the modeling framework were minimized, thereby improving the generality of results.

Moreover, DCF analysis was embedded into a broader Monte-Carlo framework to supplement economic viability calculations with some uncertainty quantification. A primary objective of the analysis was to elucidate the financial risks of SAF production, thereby providing insight into potential mitigation strategies should they be deemed necessary. The framework was designed to allow many simulations with random variables simulating market risks to be sampled efficiently and formatted as input to the refinery model (Meins and Sager, 2015). Resulting NPV distributions would then offer insights into the expected economic performance (weighted average), risks (standard deviation), and risk-adjusted performance (weighted average/standard deviation or Sharpe ratio) of each strategy.

The iterative procedure used to generate one Monte-Carlo simulation is described below alongside novel components of the modelling framework.

1) A jet debottlenecking strategy was first devised using refinery model constraints.

A graphical depiction of the Monte-Carlo framework is provided in Figure 3 and more detailed descriptions of each modification to the original modelling framework described in (Carlson et al., 2023) are provided below.

Results indicate that refineries deploying capital to produce more fossil jet fuel is a risky proposition given the uncertainties present in the market as evidenced by Figure 10. Moreover, the distribution of mitigation costs shown in Figure 8 indicate that directly blending SAFs with refinery jet blendstocks is probably economically viable given current incentives likes LCFS and RIN credits. More specifically, producing SAFs with sufficiently low production costs and high carbon intensity reductions, thereby generating more LCFS credits, can be profitable if the final price is approximately $2.60/gal, given the specifics presented in this analysis.

However, a comparison between the capital investment and Incentive Structure 1 strategy distributions in Figure 10 highlights the inherent risk that biofuel producers take on when their business is supported with fixed incentives within the context of greater market volatility. Even when mitigating 32% of the HEFA cost, or 49% of the FTJ cost, the variability in economic outcomes resulting from Incentive Structure 1 observed in Figure 10 would likely be considered too risky to invest in SAF integration projects. Alternatively, both quantitively (Table 1) and qualitatively (Figure 10), Incentive Structure 2 clearly produces a preferable distribution of economic outcomes when compared to Incentive Structure 1. Incentive Structure 2 removes much of the risk associated with Incentive Structure 1 and even overcompensates in terms of expected NPV for both HEFA and FTJ integration scenarios.

Yet, it is important to note where risk is borne in the case of Incentive Structure 2, as the underlying sources of uncertainty are still present. To reiterate, mitigation costs could be composed of feedstock cost reductions, technology improvements, and SAF purchasing incentives. Both feedstock cost reductions and technology improvements are longer-term trends that are not freely changed in step with oil market volatility. Therefore, any variability in mitigation costs would likely need to come from a sufficiently large fund that supplies mitigation credits when oil prices are low and collects debits when oil prices are high. The fund would effectively serve as an economic buffer to absorb risk for SAF producers, refiners, and airline companies. The key difference between current biofuel incentive structures, represented by Incentive Structure 1, and the proposed fund, represented by Incentive Structure 2, is the introduction of bidirectional payments.

The question of how efficient Incentive Structure 2 was compared to Incentive Structure 1 then arose, or in other words, how did resulting MACs compare? The expected MAC for HEFA observed in Figure 11A was $136/Ton-CO2e which was 13% lower than the fixed MAC ($156/Ton-CO2e) attributed to Incentive Structure 1. Similarly, Figure 11B implied an expected MAC of $182/Ton-CO2e for FTJ which was 7% lower than the fixed Inventive Structure 1 MAC of $195/Ton-CO2e. For both SAFs, the total, long-term cost of providing variable rather than fixed incentives was lower.

It is also worth noting the variable mitigation cost formulas (Equations (8 and (9) implemented in Incentive Structure 2 both exceeded the economic performance target, being the expected NPV for the capital projects strategy, as evidenced by Figures 10A,B. Refined mitigation cost formulas that more closely matched the capital investment strategy NPV would likely bring the expected MACs attributed to Incentive Structure 2 down even further while still more effectively de-risking SAF production for industry adoption.

In summary, preexisting refinery optimization models were further developed to consider capital investments and market risks via Monte-Carlo simulation. The resulting framework enabled economic value and risk comparisons between refinery investments to produce more petroleum jet and SAF integration to meet rising jet demands. This framework could be leveraged to address similar questions faced by refiners interested in balancing future investments with decarbonization goals.

Two distinct SAF purchasing incentive structures were also tested within the modelling framework. A traditional, fixed incentive per volume of SAF, was shown to produce untenable economic risk for SAF producers and refiners to sustainably integrate. Results indicated that variably incentivizing SAF purchases in step with changing crude oil prices could be a more efficient and effective way to support SAF deployment relative to fixed incentive structures.

The concept of market-sensitive, variable incentives could hypothetically be generalized to any biofuel being integrated into existing fossil-fuel markets which are more mature and accordingly able to absorb risk. Moreover, support from an effective incentive structure will allow SAF or other biofuel producers to more quickly gain the operating experience needed to realize nth plant prices thereby reducing broader decarbonization timelines. For these reasons, mitigating pricing risks upfront as suggested here, in a fair and sustainable way could be integral to the growth of the SAF and greater biofuels industries.