Back
Jul 03, 2025
By Scott Gerlt • ASA Chief Economist
Biofuel markets are driven by policy, and these policies often challenge the markets. Favorable announcements can be few and far between at times. Yet, EPA’s proposed blending levels for 2026 and 2027 that were released on June 13 provided an unequivocal boost to soy-based biofuels. The proposal contains blending levels that would significantly boost soybean oil demand and reduced credit generation on imports that would help give preference to domestic feedstocks. If EPA adopts this proposal later this year, it will significantly boost domestic soybean oil demand.
Under the Renewable Fuel Standard, fuel refiners are required to either blend biofuels into their fuel or purchase credits from those that do. The renewable volume obligations set the minimum blending levels. Table 1 shows the required blending levels for 2025 and the proposed levels for 2026 and 2027. Biomass-based diesel includes biodiesel and renewable diesel and is represented by a D4 renewable identification number. One RIN is equal to the energy in one gallon of ethanol (equivalence value) and is the credit that can be traded to demonstrate blending compliance. EPA is proposing an increase of 2.26 billion gallons of BBD between 2025 and 2026, which is a 67% increase!
BBD D4 RINs can be used to meet the D4 blending requirements or for higher categories such as the D5 or D6. The Advanced D5 category is composed of biofuels other than corn ethanol that have at least a 50% reduction in GHG emissions. Cellulosic (not shown) and BBD requirements are part of the D5. After subtracting out those specific requirements, the Advanced gap is left. BBD RINs more than those needed to meet the D4 requirements can be used to fill the advanced gap. EPA essentially left this gap unchanged from 2025 in the proposal.
The last category shown in Table 1 is conventional (D6). Under the RFS, this is the only category that corn ethanol qualifies for. Any biofuel that can achieve at least a 20% reduction in GHG emissions can generate D6 RINs. EPA left the D6 unchanged at 15 billion RINs.
Although biomass-based diesel generally generates a D4 RIN, that RIN can also be used towards advanced and conventional blending obligations. In essence, the D4 RVO is the minimum blending amount for BBD. Figure 1 shows how EPA assumed the BBD RINs would be used. For instance, in 2026 about 7.1 billion RINs would be used for D4 compliance, 350 million for D5 compliance and 1.2 billion for D6 compliance. The D6 number is large, as EPA set the conventional gap at 15 billion gallons. Corn ethanol fills most of this, but due to the blend wall EPA assumes they will only generate 13.8 billion gallons/RINs. BBD must backfill the other 1.2 billion RINs to meet the obligation.
The EPA proposal includes several EV changes that will impact biofuel markets. First, EPA is considering reducing the EV for renewable diesel from 1.7 to 1.6. For comparison, biodiesel generates 1.5 RINs per gallon. EPA reports that they made an error when initially calculating the EV for renewable diesel by not accounting for the fact that a portion of the hydrogen used in the fuel making process is derived from fossil fuels. This change would correct that error. It would also put RIN generation, and thereby credit value, closer to parity between the two biofuels.
The other EV change that would alter the status quo is partial RIN generation for biofuel imports and biofuels produced from imported feedstocks. In either case, the biofuel would generate 50% of the RINs typical for the fuel. For instance, domestic renewable diesel produced from imported used cooking oil would generate 0.8 RINs per gallon.
This change has the potential to propel soy value. Historically, imported biofuels and biofuels produced from imported feedstocks received the same state credits, federal tax credits and RINs. In fact, certain types of feedstocks like UCO and tallow were generating much higher credit values in California’s Low Carbon Fuel Standard than soybean oil. Given the fixed federal blending levels since 2023, this resulted in feedstock imports displacing soybean oil in biofuel production. Domestic soybean oil became the residual feedstock.
If EPA adopts this proposal, the order of preference for feedstocks would change. Domestic soybean oil would lead to higher value for biofuel producers than imported feedstocks or fuels. Imports would still be allowed, but they would become residual suppliers instead of the preferred suppliers.
Figure 2 demonstrates the fact that EPA still expects imports as the RVO proposal does not ban them, only elevates domestic feedstock preference. Imports in the chart are either biofuel imports based on the feedstock or domestically produced biofuels from imported feedstocks. FOG stands for fats, oils and greases and includes used cooking oil and tallow. EPA expects that imports will continue to supply the bulk of FOG and canola oil.
Also worth noting in Figure 2 is the growth in domestic soybean oil production. EPA used estimates of increasing crush expansion to arrive at the numbers. Note that the growth in soybean oil is expected to be used in renewable diesel as opposed to biodiesel. The imports of soybean oil are primarily through finished biofuels as opposed to raw feedstock.
The RFS granted EPA the authority to exempt refineries with capacity of less than 75,000 barrels per day from the RVO requirements. These small refinery exemptions were last granted in 2017 when over 16 billion gallons of gasoline and diesel were exempted affecting over 1.7 billion RINs. There are currently 189 outstanding petitions from 2016 to 2025 that EPA on which EPA must make a decision.
If the average size of the petitioners is equal to those that received the SREs in 2017, the result would be 89.2 billion exempted fuel gallons affecting 12.6 billion RINs. Given that EPA’s entire RVO proposal for 2026 is 24.02 billion RINs, any decisions that EPA makes granting the backlog of petitions will have major effects on biofuel markets.
The details of compliance for any granting of historical SREs can become quite complicated depending on whether the petitioner has already complied with the RVO for that year or has deferred their obligations. One option that EPA has for SREs going forward is to reallocate the exempted gallons among the remaining obligated parties.
EPA has committed to this option for any SREs granted in 2026 and 2027. EPA estimates up to 18 billion gallons of gasoline and diesel could be exempted in 2026 and 2027 through SREs. However, due to reallocation, any SREs granted for those years would not change the volumes of biofuels obligated to be blended.
EPA’s treatment of historical SREs is one of the biggest X factors regarding how the agency will handle the RVOs. Yet, several others present significant uncertainty.
In practice, EPA sets the RVOs based on RINs instead of gallons. The RINs are divided by the sum of gasoline and diesel consumption. The resulting rates are multiplied by actual throughput for a refiner to determine their RIN compliance obligations.
Prior to this proposal, D4 RINs either produced 1.5 RINs per gallon (biodiesel) or 1.7 RINs per gallon (renewable diesel). This proposal increases this range from .75 RINs per gallon (imported biodiesel) to 1.6 RINs per gallon (fully domestic renewable diesel). EPA assumed a conversion factor of 1.6 in the past RVO to convert between RINs and gallons for BBD. While the methodology changed slightly in the current proposal, EPA assumes an overall conversion factor just under 1.3 in this RVO.
The actual split between imports and domestic supply will have a large effect on actual gallons of consumption since compliance is in RINs. For instance, if the proposed D4 mandate were met with a 50/50 combination of biodiesel and renewable diesel, only 4.59 billion gallons would be needed to comply instead of 5.61. Adding in the backfill in the D5 and D6 would raise that total back up to 5.61, coincidentally. On the other hand, if more of the RVO is met through imports than assumed by EPA, the actual gallons of consumption would be larger than the baseline level.
Another uncertainty with actual compliance is the difference between predicted gasoline and diesel production in 2026 and 2027 versus actual production. EPA uses EIA’s Annual Energy Outlook to predict these values. If fuel consumed is less than predicted, the actual biofuel utilized will be lower than targeted in the RVO. The opposite also holds true. Given historical discrepancy, EPA started inflating the AEO numbers in the 2023 RVO to try and correct for historical differences.
For the current proposal, EPA was relying on the AEO released in 2023. A 2025 update has been released, and EPA plans to use the new version in the final rule. The AEO 2025 has 7% higher gasoline and diesel consumption than the AEO 2023. By itself, this will reduce the actual blending requirements as the fewer RINs per gallon of consumed fuel would be needed to meet the targets.
The update from the AEO 2023 to AEO 2025 will have a large impact in another area. To project domestic ethanol consumption for the next two years, EPA estimated the distribution of E0, E15 and E85 ethanol blends (the number corresponds to the percent of ethanol in the blend). They assumed the rest of the gasoline pool in AEO 2023 was blended at the standard E10 level. This led EPA to 13.8 billion gallons of ethanol consumption in 2026 and 2027. EPA assumed that BBD would fill the rest of the 15-billion-gallon conventional gap.
The AEO 2025 projects about 6 billion gallons more gasoline consumption than in the AEO 2023. Using EPA’s methodology, this results in 600 more gallons of ethanol consumption over the next two years between EPA’s final RVO rule and the current proposal.
Ethanol RINs are generally cheaper to produce than BBD RINs. For this reason, the switch from the AEO 2023 to the AEO 2025 would result in 600 more gallons of assumed ethanol consumption and 600 fewer gallons of BBD consumption in the final rule if the RVOs remain unchanged. If EPA wishes to keep BBD consumption unchanged, the agency would need to increase the RVOs to account for the larger projected ethanol volume. In other words, the update to the newest AEO, which EPA has pledged to use, would allow for higher RVOs without changing any assumptions about BBD availability.
The RVO proposal for the next two years presents an opportunity for soybeans. Biofuel consumption of soybean oil has been extremely low in 2025 due in large part to RVOs that were set too low. This has hurt biofuel producers, soybean crush plants and farmers. The proposal should help recharge the industry. In fact, soybean oil prices increased by the 6% limit on daily market movement and soybean prices increased by 3% on the day of the announcement.
The RVOs that account for increased biofuel capacity and feedstock availability should allow the industry to run more closely to capacity and new soybean crush plants to be better utilized. The change in RIN generation to differentiate between domestic and foreign would move domestic soybean oil away from being a residual feedstock.
This proposal is just that—a proposal. While it signals EPA’s desire to better support biofuels, it is not a regulation. EPA will hold hearings and accept written comments to gather feedback on the proposal. The agency will likely incorporate feedback into the final rule, which they plan to release by Oct. 31. The final version will have a significant impact on the soybean industry over the next few years.
