Increased Feedstock Imports Push Soy from Biofuels

By Scott Gerlt • ASA Chief Economist

Growth in biofuel use of soybean oil has been a bright spot for soybean farmers. Over 20 new or expanding crush plants have been announced since the start of the renewable diesel boom to provide increased soybean oil to the sector.

The heightening demand for soybeans to feed the crush plants pulls up the soybean prices that farmers receive. Since the beginning of 2023, imports of competing feedstocks have increased significantly, which pushes soybeans out of biofuel use at the margin. This article examines the relationship between feedstock imports and soybean oil use within the Renewable Fuel Standard.

The RFS Wrinkle

Understanding how feedstock imports affect soybeans requires some background on the Renewable Fuel Standard. The current form of the RFS, established through the Energy Independence and Security Act, was enacted into law in 2007[i]. Blending levels of biofuels into the domestic fuel supply were historically set by the legislation. The Environmental Protection Agency had authority to alter the statutory blending levels and was tasked with administering the program. Starting in 2023, the RFS no longer provided targeted blending levels to EPA, and the agency was provided the authority to set them as it deemed fit. EPA has currently set blending levels through 2025.

The required blending levels in the RFS are called Renewable Volume Obligations (RVOs). EPA determines the total gallons to consume and divides that number by expected domestic fuel consumption. The resulting percentage is the RVO that is passed on to fuel refiners and importers. These parties must then demonstrate compliance by turning over Renewable Identification Numbers (RINs) to EPA.

Every gallon of biofuel produced in the United States bound to be used in the RFS is assigned RIN(s). When the biofuel is blended with petroleum, that RIN is detached from the biofuel. The RIN can then be turned in to EPA to demonstrate RVO compliance, or it can be sold to another party to allow that party to meet their obligations. This system allows refiners to achieve aggregate industry compliance while obtaining flexibility in their individual biofuel blending levels.

Figure 1

Source: Environmental Protection Agency
https://www.epa.gov/renewable-fuel-standard-program/renewable-identification-numbers-rins-under-renewable-fuel-standard

While this oversimplified summary of the RFS is still quite complex, it is necessary to understand the biofuel market. Biodiesel and renewable diesel (biomass-based diesel or BBD) provide a public environmental good that is unpriced by the market (positive externality). Since both soy-based biofuels cost more to produce than a corresponding gallon of petroleum fuel, and the biofuel producer is not compensated for the positive externality, the RVOs essentially provide a minimum and maximum blending rate for biomass-based diesel in the United States[ii].

These blending rate parameters have a practical implication for BBD: The RVOs set a fixed consumption level in the United States. As a result, the RIN represents the marginal difference between the cost of producing BBD (including any other incentives that may be provided) and the price of petroleum diesel. In effect, within a current RVO, decreasing feedstock costs do not change the biofuel consumption level, only the value of the RIN, which falls by the amount of the decrease in feedstock costs[iii]. This is due to a rightward shift in the demand curve. As long as the slope of the demand curve does not change, the producer profits of the industry will remain unchanged under economic theory due to the change in the RIN price.

Expectations vs. Reality

EPA set the current RVOs in 2023 for BBD using assumptions about the availability of feedstocks to produce the fuels. The agency estimated domestic feedstock availability and net imports and BBD imports to set the blending levels. Importantly, EPA did not set RVOs based upon biofuel sector production capacity but rather on the amount of feedstocks that could be used at the production facilities.

EPA released the 2023 to 2025 RVOs in June 2023 after a long rule-making process. Imports of biofuels and feedstocks were already starting climb at that point but had not yet peaked (Figure 2). Canola received a pathway to be used in renewable diesel under the RFS in 2023, which increased imports. Used cooking oil imports started arriving from China. At the same time, biodiesel imports from the European Union and Canada started increasing significantly in 2023.

Figure 2

The impact of the imports is dependent upon the assumptions EPA made when setting the RVOs. Figure 3shows the difference between EPA's assumptions regarding BBD net imports and production by feedstock and actuals. Only the first five months of 2024 are available. The EPA assumption for BBD production from soybean oil and fats, oils and greases (FOG) (which include UCO and tallow) was close to reality in 2023, but that pattern has changed in 2024. If the current pace continues, biofuels produced from FOG will, by the end of the year, exceed EPA's assumptions by 500 million gallons for 2024. Meantime, soy will be more than 100 million gallons below the assumption.

Figure 3

It is worth noting in Figure 3 that the actual numbers are for only production while EPA assumptions are for production and net imports. EPA does not break out whether imports are via feedstocks or finished biofuels. The agency assumed that these together would total 649 million gallons in 2023, 166 million gallons in 2024 and zero gallons in 2025[iv]. The gallons would be equally split between soybean oil, canola oil and FOG each year.

The reality has been quite different than these expectations. Through the first seven months of 2024, the U.S. is on pace for net imports of renewable diesel at 441 million gallons and 337 million gallons for biodiesel. All the renewable diesel gallons are from one producer[v] that primarily uses FOG as a feedstock[vi]. As a result, most of these 441 million renewable diesel gallons should be added to the 2024 FOG data in the chart. The biodiesel imports are primarily coming from Canada and the EU, and they are likely produced from mostly a mix of canola and soybean oil.

At the end of 2024, the Biodiesel Blenders Tax Credit (BTC) of $1 per gallon will expire. Starting in 2025, the Clean Fuel Production Credit or 45Z will take its place, which will not be available for imported biofuels. As a result, imports of finished biofuels likely will decline in 2025 and likely will partially substitute for feedstock imports. Some of the increase in 2024 biofuel imports may be an effort to capture the BTC before expiration.

Astute readers may notice that the totals for 2024 in Figure 3 do not add to zero as might be implied by a fixed demand for the biofuels. Several factors can contribute to this. First, if total fuel supplied for domestic consumption is different than projected while setting the RVOs, required volumes will differ from intentions given that the RVOs are percentages. Second, there are multiple categories of the RFS that BBD can fill. BBD will often backfill some portion of the RVOs that are left unfilled by ethanol. Yet, according to recent estimates for 2024, the difference between EPA's volumetric standards and the actual number of RINs projected to be unfilled by ethanol in the applied standards is negligible[vii]. Volume obligations can also be deferred in the RFS. About 2 billion RINs were deferred in 2023. Additionally, up to 20% of RINs can be carried over to the next year. While these can allow discrepancies in individual years, the long-run totals cannot deviate due to carryovers.

Soybeans Short-changed?

As aforementioned, the RIN price should equal the marginal difference between the cost of producing the biofuel minus any other revenues generated by the biofuel. Historically, the marginal gallon of BBD was produced from soybeans. Through 2023, the amount of soybean oil used for BBD production exceeded FOG use. The pace for 2024 will likely reverse that, making fats, oils and greases the marginal gallon.

Figure 4 reflects this reality. Gross margins for producing renewable diesel from soybean oil were about $1 per gallon from mid-2022 through mid-2023. Two important events happened after that point. First, EPA's RVOs came in lower than expected by markets, which dropped the value of the marginal feedstock at the time (soybean oil).

Figure 4

Second, imports of waste feedstocks increased, pushing down the demand for soybean oil in biofuels. As a result, the gross margin for producing renewable diesel from soy dropped from about $1 per gallon to $0.50 per gallon. As UCO is now becoming the marginal gallon, the returns from using it are now approaching the $1-per-gallon gross margin that soy once held.

In essence, traditional biodiesel facilities and renewable diesel plants that use soybean oil must operate at much lower long-term profitability after the surge in feedstock imports. While many factors affect margins, this demonstrates evidence in support of the economic theory outlined above.

Import Implications

Setting all the complex RFS and economic theory aside, what do imports mean in practical terms for soybeans?

Table 1 breaks down the numbers for 2024 if the current pace of imports and production are maintained for the rest of the year. Domestic production of BBD from fats, oils and greases would amount to 513 million gallons above what EPA assumed (including EPA's projection of net imports of BBD).

This is equivalent to the soybean oil from 354 million bushels of soybeans, which equates to 6.7 million acres of soybeans. The FOG use in BBD in excess of EPA's projection for 2024 is equivalent to the soybean oil from 10 crush plants (based on the average crush plant size in the United States). There are 63 operational domestic crush plants.

Only four states are projected to produce more than 354 million bushels of soy in 2024 (Illinois, Iowa, Minnesota, Indiana). If renewable diesel net imports are added to the mix, the FOG BBD supply for 2024 approaches almost 1 billion gallons in excess of EPA's expectations. This is equivalent to the soybean oil from 659 million bushels of soy. Only one state (Illinois) is expected to produce more soybeans than that in 2024. This amount is equivalent to the soybean oil from 12.4 million acres and 18 crush plants.

Challenges Ahead

Given the way the RFS works, under the current RVOs, the increase in FOG-based biofuels can only happen at the expense of biofuels produced from other feedstocks, particularly soy. The change from the BTC to the 45Z next year should help with renewable diesel imports but will likely encourage even more imports of UCO and tallow feedstocks.

Additionally, California is proposing to restrict soy and canola feedstocks in their Low Carbon Fuel Standard Program and require traceability for those feedstocks within a few years[viii]. Furthermore, the U.S. just permitted market access for Brazilian UCO[ix]. These factors further encourage domestic production with foreign feedstocks.

One potential solution to this development is for the RVOs to be increased enough to allow room for all feedstocks in the RFS. However, this could not happen until 2026 at the earliest. It would also require foreknowledge by EPA of foreign UCO and tallow supplies. Given the opaqueness of these markets, it is unlikely that the agency would do a better job estimating that than was done under the current rule.

Likewise, updating soybean's carbon intensity scores to reflect current science in the biofuel programs would also help reduce the surge of foreign feedstocks. This is generally a long-term process, and California has been unwilling to open up the most outdated portions of the scoring.

Under the current policy framework, foreign fats, oils and greases have become the marginal feedstock used to produce biomass-based diesel. Given the structure of the RFS, this displaces soybeans in the marketplace for the current RVO and will reduce profitability for plants that can't make the feedstock switch.

Economic theory shows that under a given RVO, RINs adjust based upon feedstock costs keeping aggregate biofuel industry returns approximately unchanged. This is born out by observing gross margins between different feedstocks. The result is that, without a structural change, an appreciable amount of domestic soybean oil production expansion that was intended for biofuels will now be left out of the programs.

[i] https://www.epa.gov/renewable-fuel-standard-program/renewable-fuel-standard-rfs2-final-rule
[ii] In reality, obligations can be deferred and limited amounts of RINs can be retained across calendar years. However, in the long-run these effects will wash out.
[iii] This assumes that the cost of producing BBD never falls below the price of petroleum diesel.
[iv] https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P1017OW2.pdf
[v] https://www.eia.gov/todayinenergy/detail.php?id=62704
[vi] https://www.neste.com/renewable-raw-materials
[vii] EPA was projecting that 7.585 billion RINs of the standards would be unfilled by ethanol for 2024. Current Bloomberg estimates put this number at 7.61 billion RINs based on current EIA projections.
[viii] https://ww2.arb.ca.gov/sites/default/files/barcu/regact/2024/lcfs2024/15day_notice.pdf
[ix] https://www.gov.br/agricultura/en/news/brazil-celebrates-second-market-access-in-the-united-states-for-agricultural-products-this-year