MEMORANDUM

Subject:	Illustrative Costs Impact of the Final RFS2 Standards, 2014-2016

From:	Michael Shell and Michael Shelby, U.S. EPA, Office of Transportation and Air Quality

To:		EPA Air Docket EPA-HQ-OAR-2015-0111


	In this memo we provide further description of our assessment of the illustrative costs of the final advanced biofuel and total renewable fuel standards for 2015 and 2016.  We start with an overview of considerations of a baseline from which to assess the impacts, followed by descriptions of the specific costs assessed.  We provide cost estimates for three illustrative scenarios  -  one, if the entire change in the final advanced standards is met with soybean oil biodiesel; two, if the entire change in the final advanced standards is met with Brazilian sugarcane ethanol; and three, if the entire change in the volume of the non-advanced renewable fuel required by the final standards is met with corn ethanol. While a variety of biofuels could help fulfill the advanced standard beyond soybean oil BBD and sugarcane ethanol from Brazil, these two biofuels have been most widely used in the past. We believe these scenarios provide a range of possible costs of meeting the final standards. 

      For this analysis, we estimate the per gallon costs of producing biodiesel, sugarcane ethanol and corn ethanol relative to the petroleum fuel they replace at the wholesale level, then multiply these per gallon costs by the applicable volumes established in this rule for the advanced and total renewable fuel categories. Because we are focusing on the wholesale level in each of the three scenarios, these comparisons do not consider taxes, retail margins, and any other costs or transfers that occur at or after the point of blending (i.e., transfers are payments within society and not additional costs).  Also, we did not attempt to estimate the infrastructure costs (e.g., E85 pump installations) that would be required to meet these RFS standards.  In addition, because more ethanol gallons must be consumed to go the same distance as gasoline and more biomass-based diesel must be consumed to go the same distance as petroleum diesel due to each of the biofuels' lesser energy content, we consider the costs of ethanol and biomass-based diesel on an energy equivalent basis to their petroleum replacements (i.e., per energy equivalent gallon (EEG)).


	1.	Baseline Consideration

 
	A number of different scenarios could be considered the "baseline" for the assessment of the costs of this rule.  For the purposes of showing illustrative overall costs of this rulemaking, we are proposing to use the preceding year's standard as the baseline (e.g., the baseline for the 2016 advanced standard is the applicable 2015 advanced standard), an approach consistent with past practices.  Arguments could be made for using a variety of alternative baselines such as the statutory renewable fuel volumes (e.g., the required volumes in the Energy Independence and Security Act (EISA) of 2007 for 2013, 2014, and 2015) or the volumes of renewable fuels that would be produced in the complete absence of the RFS program (e.g., "no RFS").  However, we believe the difference between each year's standards provides information that is illustrative of the magnitude of the potential overall costs of the standards being finalized today without adding the complexity of trying to project a hypothetical future baseline that would only call into question the usefulness of the cost comparison.  

	Use of alternative baselines other than the 2013, 2014 and 2015 applicable standards is problematic from a practical standpoint.  For example, using the 2013, 2014, and 2015 statutory volumes would require estimating a hypothetical set of cellulosic fuel costs as part of the baseline.  Since liquid cellulosic fuels are not currently being produced in significant quantities, estimating their average costs of production would be highly speculative on EPA's part. Furthermore, trying to estimate the hypothetical mix of fuels that could meet the annual EISA volumes would introduce a level of complexity that would not necessarily provide a more robust estimate of costs.  

	There are similar complexities with trying to project what volumes of renewable fuels would be consumed in the U.S. in the absence of any RFS program, a hypothetical "no RFS" baseline.  One option would be to try to estimate the amount and types of biofuels that would have penetrated the market without the RFS program from its inception in the relevant years of consideration 2013, 2014 and 2015.  Estimating the quantities of renewable fuels that would have been used by market participants absent any RFS requirements from the inception of the program would be an uncertain exercise.  For example, absent the RFS, billions of gallons of ethanol would still be produced for consumption in the U.S. fuel market due to ethanol's value as a low cost octane enhancer. Another factor that would impact consumption absent the RFS is individual state renewable fuel mandates, each of which would have to be assessed to determine this baseline.  Thus, we believe that estimating the quantities of renewable fuels that would be used in the absence of the RFS program, as well as their associated costs, would be highly speculative.  

	The choice of a baseline will have a directional implication for the costs assessed.  For example, reducing the standards from a high renewable fuel volume baseline (e.g., the EISA 2013, 2014, or 2015 volumes) would show cost savings (assuming a gallon of renewable fuel is more costly than its petroleum-based fuel equivalent).  Conversely, increasing the volumes from a low renewable fuel volume baseline (e.g., the "no RFS" volumes) would show cost increases as a result of the 2014, 2015, and 2016 standards (assuming a gallon of renewable fuel is more costly than its petroleum-based fuel equivalent).  For the purpose of providing illustrative costs, we have opted for a more straightforward and real world baseline assumption of the applicable standard for the previous year as a baseline to assess the costs. In evaluating the costs for 2014, we would use the actual applicable standards for 2013 as the baseline.  In evaluating the costs for 2015 and 2016, we would use the applicable standards for 2014 and 2015, respectively, as the baseline.  

	 The 2014 standards were not finalized in 2014 so it is difficult to estimate what their costs may have been. Market participants may have anticipated that a final 2014 standard would require higher levels of biofuels than the market would provide in the absence of the standard, which would contribute to the positive RIN prices witnessed in 2014. In contrast, the 2014 standards being finalized in this rulemaking represent reductions in both the advanced and conventional volumes compared to the 2013 standards, suggesting cost savings for this final 2014 rule. Finally, the 2014 standards being finalized in this rulemaking are based on actual production levels in 2014, suggesting that the 2014 standards we are proposing have no costs. Given the disputable foundation of this historical dynamic, we have not attempted to estimate the costs of the 2014 standards. Therefore, we only evaluate costs for the 2015 and 2016 advanced standards and total (i.e., non-advanced) standards.
 

2.	Cost of Displacing Petroleum-Based Diesel with Soybean-Based Biodiesel
      

	Consistent with our previous work in this area, EPA's estimate is a "bottom-up" engineering cost analysis that quantifies the costs of producing a gallon of soybean-based biodiesel and then compares that cost to the energy-equivalent gallon of petroleum-based diesel.  

	Similar to the analysis EPA undertook for the 2013 annual standard, we have made the simplifying assumption that the marginal biodiesel feedstock used to meet the higher total advanced biofuel standard will come from soybean oil.  EPA recognizes that a variety of different feedstocks in addition to soybean oil could be used to make additional biodiesel.  These include other oils (e.g., canola oil) as well as waste oils, fats, and greases.  However, projecting the proportions of different feedstocks that may be used to make extra biodiesel as a result of the standard and the different feedstocks' costs would be speculative.  Based on historical data, 52 percent of the feedstock used for U.S. biodiesel production in 2015 (through August) was from soybean oil, the largest source of feedstock used to make biodiesel.  Other feedstocks such as oils derived from wastes are generally less costly so they will be preferentially used regardless of the volume set.  However, there are not enough oils derived from wastes to fully supply the advanced standards we are proposing. Thus, we believe it is appropriate to make the simplifying assumption that the marginal feedstock used to produce biodiesel to meet the total advanced biofuel standard will be from soybean oil.  

	We base our estimates of soybean oil feedstock costs in 2015 and 2016 on the latest United States Department of Agriculture/World Agricultural Supply and Demand Estimate (USDA-WASDE) forecasts.  Soybean oil feedstock costs represent the majority of the overall costs of biodiesel, usually somewhere between 70 and 90 percent of overall costs.  In addition to soybean oil feedstock cost, we assume variable production costs to estimate the total per gallon cost.  We use two sources to represent a range of potential variable costs.  For the higher cost estimate, we used per gallon cost estimates from the University of Missouri, Food and Agricultural Policy Research Institute (FAPRI-Missouri) March 2015 U.S. Baseline Briefing Book.  They estimate variable biomass-based biodiesel production costs of $0.47/gallon, net of the coproduct (i.e., glycerin) value for both 2015 and 2016.  FAPRI-Missouri also include an average net operating return to produce biodiesel of $0.10 per gallon in 2015 and $0.39 per gallon in 2016.  For our lower cost estimate, we used a forecast of 2014 average biodiesel variable cost projections of $0.50/gallon from the National Biodiesel Board.  We then compare these ranges to estimates of the average price of wholesale petroleum-based diesel in 2015 and 2016.  We assume that the wholesale price of petroleum-based diesel represents its costs.  EPA obtained the projected average wholesale petroleum-based diesel 2015, and 2016 cost estimates from the U.S. Department of Energy, Energy Information Administration's (EIA) Short Term Energy Outlook (STEO) Report.  The STEO price is $1.71 in 2015 and $1.82 in 2016.

	The final 2014 standard is being set at the actual level of advanced biofuels produced in 2014, 2.67 billion gallons. The total advanced biofuel volumes are being finalized for 2015 at 2.88 billion gallons and 3.60 billion gallons in 2016.  

	Consistent with the approach we used for the 2013 BBD standard, EPA estimates that increasing volumes of soybean oil for biodiesel will create upward pressure on the price of soybean oil.  Based on work from Iowa State University, we assume each additional 200 million gallons of soybean oil used to produce extra biodiesel results in an increase in soybean oil costs of 3.2 cents per pound.  However, given the current low-price or glutted condition of commodity markets in general, including soybean oil markets, it is also possible that changes in biomass-based diesel volumes would have minimal impacts on prices in the near-term, before supply adjusts. Therefore, in our low price estimate, we assume there is no upward price pressure on soybean oil prices from greater biomass-based diesel volumes. Applying the cost impact assumptions to the projected range of USDA-WASDE soybean oil prices results in soybean oil price estimates at 32 to 34 cents per pound for 2015, and 28 to 40 cents per pound for 2016.  Considering non-feedstock variable production cost estimates from FAPRI-Missouri and the National Biodiesel Board, and low and high WASDE soybean oil estimates, the resulting biodiesel costs range from $3.16 to $3.42 per gallon in 2015 and $2.82 to $4.28 per gallon in 2016 (on a diesel gallon equivalent basis).

	Comparing the petroleum-based diesel price estimate and multiplying by the volume of fuel displaced on an energy equivalent basis, results in an overall annual cost of $203 to $240 million in 2015 and $480 to $1,182 million in 2016. Table 1 below shows estimates of per gallon increases on an energy equivalent basis as well as total annual costs for 2015 and 2016 of using greater amounts of biodiesel to meet the total advanced biofuel standard.  

	These cost estimates assume that the increase in the total advanced biofuel standard is met with soybean-based biodiesel, which is a simplifying assumption used for illustrative purposes.  In all likelihood a variety of different sources of feedstocks and renewable fuels (e.g., biodiesel from other feedstocks, sugarcane ethanol, sorghum ethanol, CNG/LNG from biogas) are expected to be produced to meet the 2015 and 2016 total advanced biofuel standard. It should be noted that soybean oil costs and petroleum-based diesel costs could vary depending upon evolving supply and demand factors in their respective markets that are not reflected in current USDA/EIA price estimates.

                                    Table 1
Illustrative Per Energy Equivalent Gallon Cost Difference and Total Cost Estimates of Meeting the Advanced Standard in 2015 and 2016 Using Biomass-Based Diesel to Displace Diesel
                                       
                                       
                                     2015
                                     2016
                                       
                                      Low
                                     High
                                      Low
                                     High
Per Gallon Differences in Cost ($/gallon)
                                     1.45
                                     1.71
                                     1.00
                                     2.46
Total Annual Costs (millions of dollars)
                                      203
                                      240
                                      480
                                     1,182



3.	Cost of Displacing Petroleum-Based Gasoline with Sugarcane Ethanol

	While BBD can be used to meet the total advanced biofuel standard, other advanced renewable fuels can also be used to comply.  One possible option is the use of sugarcane ethanol, the majority of which is imported to the U.S., principally from Brazil.  In this section, we provide illustrative estimates of what the potential costs might be for a scenario in which we assume that all additional volumes used to meet the 2015 and 2016 total advanced biofuel standards above the previous year's total advanced biofuel standard are met with imported Brazilian sugarcane ethanol.  We compare the cost of sugarcane ethanol and gasoline at the wholesale stage, since that is when the two are blended together and represents the approximate costs to society absent transfer payments.  On this basis, EPA estimates the costs of producing and transporting a gallon of sugarcane ethanol to the blender in the U.S.  

	We use the sugarcane ethanol free on board (FOB) price at Santos, the port of San Paulo, Brazil, from which sugarcane ethanol is typically shipped to the U.S.  Using daily price data from the Oil Price Information Service (OPIS) from October 2014 to October 2015, we estimate a range of prices for Brazilian ethanol exported to the U.S. Since OPIS does not project future sugarcane ethanol prices from Brazil, and a bottom-up approach to estimate sugarcane ethanol prices from sugar futures prices is uncertain, we make the simplifying assumption that the observed high and low estimates from January 1, 2014 to the present are reasonable high and low price estimates for 2015 and 2016 as well. We assume that the reported price of sugarcane ethanol represents both the production and transportation costs of delivering sugarcane ethanol to the port at Santos. Using data collected by Argus in a daily Argus Americas Biofuels report, we assume that it costs an additional $0.19/gallon to transport the sugarcane ethanol from Santos to the U.S. where the ethanol can be blended with gasoline.  Based on the range of OPIS prices, the delivered price of sugarcane ethanol to the U.S. ranges from $1.53 to $2.31 per gallon.  For comparison purposes, we used the average wholesale price of gasoline for 2015 and 2016 from the STEO, which is $1.70 per gallon and $1.67 per gallon, respectively.  

	Taking the difference between the sugarcane ethanol and the wholesale gasoline price estimates, and multiplying that by the volume of petroleum displaced on an energy equivalent basis, results in an overall estimated annual cost of $186 to $431 million for 2015 and $656 to $1,493 million for 2016.  Table 2 below shows estimates of per gallon increases on an energy equivalent basis as well as total annual costs for 2015 and 2016 of using greater amounts of sugarcane ethanol to meet the total advanced biofuel standard.  These cost estimates assume that the increase in the advanced biofuel standard is met with sugarcane ethanol, which is a simplifying assumption used for illustrative purposes.  In all likelihood a variety of different sources of feedstocks and renewable fuels (e.g., biodiesel, sorghum ethanol, CNG/LNG from biogas) are expected to be produced to meet the increase in the 2014, 2015, and 2016 advanced biofuel standards.  It should be noted that sugarcane ethanol costs and wholesale gasoline costs could vary depending upon evolving supply and demand factors in their respective markets that are not reflected in current OPIS/EIA price estimates.


                                    Table 2
Illustrative Per Energy Equivalent Gallon Cost Difference and Total Cost Estimates of Meeting the Advanced Standard in 2015 and 2016 Using Sugarcane Ethanol to Displace Gasoline
                                       
                                       
                                       
                                     2015
                                     2016
                                       
                                      Low
                                     High
                                      Low
                                     High
Per Gallon Difference in Cost ($/gallon)
                                     0.89
                                     2.05
                                     0.91
                                     2.07
Total Annual Costs (millions of dollars)
                                      186
                                      431
                                      656
                                     1,493



4.	Cost of Displacing Petroleum-Based Gasoline with Corn Ethanol 


	 In this section, we assess the difference in cost associated with a change in the implied volumes available for conventional (i.e., non-advanced) biofuels for 2015 and 2016.  We provide illustrative estimates of what the potential costs might be in a scenario where corn ethanol is used to meet the entire conventional renewable fuel volumes.

	The implied 2014 volume allowance for conventional renewable fuel is 13.61 billion gallons, 14.05 billion gallons in 2015, and 14.50 billion gallons in 2016. If corn ethanol is used to meet the difference between the implied 2014 to 2015 and 2015 to 2016 conventional renewable fuel volumes, an increase in 440 million gallons of corn ethanol would be required in 2015 and 450 million gallons in 2016.  In Table 3, we show estimates of differences in costs on a per gallon energy equivalent basis between corn ethanol and wholesale gasoline as well as total annual costs of the conventional biofuel standard for 2015 and 2016. 

	To calculate the cost of corn ethanol, we use a bottom-up engineering cost framework starting with USDA-WASDE estimated corn prices for 2015 and 2016.  We then estimate the 2015 and 2016 marginal production costs of corn ethanol using data from the University of Missouri, FAPRI March 2015 U.S. Baseline Briefing Book.  There are many factors that influence the price of corn ethanol besides the RFS program. The price of corn is one of the primary drivers of the corn ethanol price, and corn is a commodity demanded for other end uses in addition to ethanol including food and livestock feed.  Changes in demand for these other end uses will impact corn prices which will in turn affect corn ethanol prices. Also, corn ethanol is demanded internationally so changes in international demand will have impacts on domestic prices independent of the RFS program. 

	Using historical relationships developed by Iowa State, we assume each increase of one billion gallons of corn ethanol results in an increase in corn costs of $0.21/bushel.  Using this relationship, we estimate that corn prices will increase by roughly $0.09 bushel in 2015 and $0.18 per bushel in 2016 (approximately $0.03 and $0.06 per gallon of corn ethanol for 2015 and 2016, respectively).  Based upon this methodology, EPA estimates that a gallon of corn ethanol will cost $1.77 per gallon ($2.66 per gallon on a gasoline gallon equivalent basis (GGE)) in 2015 and $1.79 and $2.00 per gallon ($2.68 and $3.00 per GGE) in 2016.  To estimate the premium per gallon against wholesale gasoline, we estimated the price of a gallon of wholesale gasoline using the latest EIA STEO estimates.  As mentioned previously, STEO estimates an average wholesale gasoline price of $1.70 per gallon for 2015 and $1.67 per gallon for 2016.  We assume that the wholesale price of gasoline represents the costs of producing gasoline.   

      Taking the difference between the corn ethanol and the wholesale gasoline price estimates, and multiplying that by the volume of petroleum displaced on an energy equivalent basis, results in an overall estimated annual cost of $424 million for 2015 and $453 to $597 million for 2016.  Table 3 below shows estimates of per gallon increases on an energy equivalent basis as well as total annual costs for 2015 and 2016 of using greater amounts of corn ethanol to meet the entire conventional renewable fuel volumes.  It is worth noting that corn ethanol costs and wholesale gasoline costs could vary depending on evolving supply and demand factors not reflected in current USDA/EIA price estimates.

                                    Table 3
Illustrative Per Energy Equivalent Gallon Cost Difference and Total Cost Estimates of 
Meeting the Change in the Total Standard for 2015 and 2016 Using Corn Ethanol to Displace Gasoline
                                       

                                     2015
                                     2016

                                       
                                      Low
                                     High
Per Gallon Difference in Cost ($/gallon)
                                     0.96
                                     1.01
                                     1.33
Total Annual Costs (millions of dollars)
                                      424
                                      453
                                      597


	The short time frame provided for the annual renewable fuel rule process does not allow sufficient time for EPA to conduct a comprehensive analysis of the benefits of the 2015 and 2016 standards and the statute does not require it. Moreover, as discussed in the proposed rule establishing the 1.28 billion gallon requirement for BBD in 2013, the costs and benefits of the RFS program as a whole are best assessed when the program is fully mature in 2022.  We continue to believe that this is the case, as the annual standard-setting process encourages consideration of the program on a piecemeal (i.e., year to year) basis, which may not reflect the long-term economic and environmental effects of the program.  Therefore, for the purpose of this annual rulemaking, we have not quantified benefits for the 2015 and 2016 final standards.  We do not have a quantified estimate of the GHG impacts for the single year (e.g., 2015, 2016).  When the RFS is fully phased in, the program will result in considerable volumes of advanced and cellulosic biofuels that will reduce GHG emissions in comparison to the fossil fuels which they will replace. EPA estimated greenhouse gas, energy security and air quality impacts and benefits in the 2010 RFS Final Rule.  

