Small Refineries, Big Consequences: What Exemptions Mean for Biofuel Demand

By Scott Gerlt, PhD • ASA Chief Economist

Small refinery exemptions can undercut headline biofuel blending numbers with much of the burden falling on biomass-based diesel.

Since the enactment of the Renewable Fuel Standard in 2005, no single issue deriving from the law generates as much controversy as Small Refinery Exemptions. SREs continue to be litigated and debated 20 years later and have even led to recent attempts to revise the RFS to deal with the issue. Although not well understood outside biofuel circles, SREs heavily affect biofuel demand and subsequently agricultural prices. EPA's current methods for dealing with the issue have largely removed their negative effects from agriculture, but there is no guarantee this treatment will continue in the future.

History

Under the RFS, petroleum refineries and importers are obligated parties that must demonstrate that biofuels have been blended into petroleum fuels at the percents prescribed by EPA. Each gallon of biofuel produced or imported is associated with Renewable Identification Numbers. When the biofuel is blended with petroleum, the RINs can be detached from the biofuel and turned into EPA to show the proper amount of biofuel was blended. Alternatively, an obligated party can make up a shortfall in its own blending by purchasing RINs from other parties that have excess.

The RFS intended to ease small refineries (defined as processing less than 75,000 barrels per day) into compliance with the program by exempting all small refineries from RFS compliance from 2006 until 2011[1]. Beginning in 2011, individual refineries could petition EPA for an extension by demonstrating that RFS compliance would cause "disproportionate economic hardship."

A detailed history of SREs would fill a dissertation, but Figure 1 shows the historical exemption levels compared to the overall biofuel blending mandate. An easy way to think of this is the percent of the mandate waived through SREs. SREs peaked in 2017 at over 9% of the blending obligation after relatively low levels in the previous years. Starting in 2020, SREs were generally between four and five percent of obligations before shooting up to the second highest level in recent history in 2025.

SREs Explained

Understanding how SREs impact biofuel demand can best be demonstrated by a simple example with manufactured numbers. Suppose gasoline and diesel on-road consumption for the U.S. totals 200 billion gallons per year. Also suppose that EPA wants 20 billion gallons of biofuels blended into the fuel supply. In this case, they would set the blending rate at 10%.

Now suppose that after the EPA sets the blend rate at 10% that they exempt refineries responsible for 40 billion gallons of fuel from compliance. This would reduce the total volume of the obligated parties to 160 billion gallons. As a result, only 16 billion gallons of biofuels would be consumed. The exemptions cost 4 billion gallons of biofuel demand.

EPA does have one tool available to them to address this situation. They can increase the blending rate on the remaining obligated parties to make up for the exemptions. This is called reallocation. In our example, EPA would divide the 20 billion gallons of intended biofuel use across the 160 billion gallons of remaining obligated fuel. The exempted parties would blend no biofuel while the remaining parties would blend at a rate of 12.5%. The total biofuel consumption would remain at 20 billion gallons, but some parties now blend nothing and others blend more. If exemptions are granted, this option allows for biofuels to avoid negative impacts.

While this simple example demonstrates the overall impact on biofuels, it excludes some important nuances of the RFS. There are multiple blending rate targets in the RFS that each have their own biofuel eligibility. In essence, there isn't one "biofuel" but many with differing rules about which blending rates they can help fill. This detail leaves biomass-based diesel (biodiesel, renewable diesel and sustainable aviation fuel) uniquely exposed to SREs.

Understanding this issue involves a brief explanation of the biofuel categories in the RFS. For simplicity, this explanation will omit cellulosic biofuels, which are a very small part of the program. The remaining categories are:

• Biomass-based diesel (D4): Biodiesel, renewable diesel and sustainable aviation fuel that has at least a 50% reduction in greenhouse gas emissions.
• Advanced biofuels (D5): Non-biomass-based diesel biofuels with at least a 50% GHG reduction.
• Conventional (D6): Biofuels with a GHG reduction between 20% and 50%. Note that the RFS specifies that this is the only category that corn ethanol may participate in.

EPA sets targets and corresponding blending levels for each of these categories. For the categories listed, fuels with a lower number D code may be used for the blending targets corresponding to higher number D codes. For example, biomass-based diesel (D4) RINs can be used for advanced (D5) or conventional (D6) blending requirements. The opposite is not true, and D6 RINs cannot be used for any other obligations.

The economics of the program design require that biofuels generating higher number D codes must be cheaper to produce than those with lower D codes. Otherwise, biomass-based diesel would simply replace ethanol production in the conventional category. In effect, there is no ethanol mandate in the RFS but only a category it can fill at a lower cost up to blending limits.

Figure 1 shows how this works in practice. EPA set the conventional (D6) target at 15 billion RINs for 2026. However, the agency only expects ethanol to fill 14.27 billion RINs[2] of this. The rest is backfilled by BBD. The advanced category is similar with BBD filling 380 million RINs. When an SRE is granted, each of the categories (D codes) is proportionally reduced with the biofuel on top of the stack taking the loss of demand hit first. In other words, the bar shrinks pushing out what was on top of the stack. As a result, BBD doesn't just lose what is in the D4 category but loses across all three categories until it is pushed out. For this reason, BBD generally takes the bulk of SRE losses. Empirical analysis supports this conclusion.

Set 2 Changed SRE Treatment

EPA sought to proactively address the SRE issue in the RVOs for 2026 and 2027 that were released in March 2026. When the blending levels were released, there were over 180 outstanding SRE petitions that could exempt about 12 billion RINs. RINs must be used in the year of generation or the following year, and EPA expected almost no outstanding RIN obligations for pre-2024 compliance years. EPA made decisions on 2023 and 2024 SREs but did not yet know the final level for 2025. Instead, for 2025 EPA used a three-year average for exempted gallons to estimate the expected SREs. The agency then reallocated 70% of the RINs provided by SREs onto the volume obligations for 2026 and 2027.

EPA assumed that 7.55 billion gallons of gasoline and diesel would be exempted in 2025 with SREs. EPA carried that assumption of exempted gallons forward for 2026 and 2027. Given the percentage standards before reallocation, this results in 1,121 RINs in 2026 and 1,141 RINs in 2027. EPA fully reallocated these amounts into the RVO for 2026 and 2027. The agency stated they intend to proactively reallocate all SREs in future set rules but did not intend to reopen Set 2 to address any differences between their assumed SRE levels and actual SRE levels.

The stated intention to not recalculate the percentage standards in the middle of the set rule provided assurance to markets that they could proceed without worrying about changes midstream. However, it did not account for other factors that did change midstream. These factors include "more small refineries requesting exemptions and changes in financial circumstances."

Table 2 shows the actual granting of SREs. Calendar year 2024 ended up almost 200 million RINs above expected levels. The big change was in 2025 where almost 800 million more RINs were granted in SREs than initially estimated. Instead of 7.55 billion gallons of exempted fuel, 13.40 billion gallons were exempted. The volume of exempted RINs will likely further increase in 2026 and 2027 above what was planned for due to higher percent obligations on those years (Table 2).

EPA has committed to reallocating 100% of the additional 2025 SRE RINs above the originally estimated 990 million through a supplemental rule to Set 2. If EPA keeps the same formulas originally used in Set 2, the reallocated RINs will increase the percent standards for 2027 while 2026 standards remain unchanged. However, the likely higher SRE volumes for compliance years 2026 and 2027 will likely also need to be addressed, either in a supplemental rule or Set 3.

Not only do SREs cause market disruption, but rumors of SREs can as well. Before news broke on August 20^th that the administration was considering granting more SREs, the D6 RIN was $2.25 according to Argus. Within a week, it dropped to $1.78 and has now sits at about $2.00. The simple uncertainty over a period of slightly more than a week wiped out value for biofuel producers and subsequently farmers.

EPA's plan to proactively reallocate 100% of SREs gives as much certainty as possible if SREs are to be granted. While granting more SREs for 2025 than expected has caused disruption, the declaration of intent to reallocate all of the additional exempted volumes calmed markets. Continuation of this calm is dependent on quickly moving through a supplemental rulemaking process to incorporate the reallocation into the percent standards. While the SREs are finalized, the reallocation is not. EPA will also need to consider how to handle likely increased SREs for 2026 and 2027 as well.

Small refinery exemptions have been controversial for many years as they historically gutted the biofuel blending. Even 20 years after the RFS was passed, the transitory mechanism still causes volatility and uncertainty in biofuel markets. While headline RVO numbers often receive much press, without knowing actual SREs and their treatment the RVO numbers by themselves merely reflect a starting point.

[1]For 2011 and 2012, 14 small refineries were granted an extension of the initial exemption for the year. Another 10 that did not receive an automatic extension were granted exemptions based on petitions for 2011. That number changed to 9 in 2012.

[2] For ethanol, one gallon is equivalent to one RIN. Biodiesel and renewable diesel generate more RINs per gallon (1.5 and 1.6 to 1.7, respectively) due to higher energy content.