Impact of Renewable Fuels Standard/MTBE Provisions of S. 1766
Assumptions of this Analysis
| Table 1. Comparison of Assumptions for Requested Cases Printer Friendly Version 
|
Besides the addition of MTBE unit conversion, the merchant plant conversion allowance, and biodiesel methodology, the analysis scenarios were developed by defining different scenario assumptions. Table 1 shows a comparison of the basic assumptions underlying the Reference, S. 1766, and RFS/No MTBE Ban Cases.
Key Assumptions of “S. 1766” Case
- Annual RFS requirements as stipulated in Section 818 of S. 1766 (See Table 2). After 2012, the RFS requirement is estimated as the same percentage of highway fuel demand that it represented in 2012.
- An adjustment is made to credit every gallon of cellulose ethanol as 1.5 gallons of renewable fuel as per S. 1766.
- A national ban on MTBE starting in 2006. S. 1766 specifies a ban four years after passage.
- States on the East and West Coasts waive the 2.0 percent (weight) oxygen requirement for RFG, as specified in Senator Murkowski’s letter of February 6. Section 824 of the bill grants governors the authority to waive the oxygen requirement but no “backsliding” on air toxics is allowed. This Case assumes that RFG areas that currently use MTBE blended gasoline but have State legislation that would ban MTBE in the future will waive the oxygen requirement.
- Maintaining toxic emissions benefits. Air toxic emissions from RFG are assumed to be maintained at current levels on average.
- The gasoline forecast is based on the AEO2002 Reference Case.
- Areas that currently use RFG are assumed to continue to use RFG. In April 2001, the governor of New Hampshire requested that the EPA allow the State to opt-out of the RFG program immediately. The next opportunity for States to legally opt-out of the program is January 1, 2004. This analysis reflects no change to RFG areas since no decision has been made in regard to New Hampshire.
- The continuation of the ethanol tax exemption is assumed through 2020.In accordance with the Federal Highway Bill of 1998, the exemption is currently 53 cents per gallon but will be reduced by 1 cent per gallon in 2003 and again in 2005. Legal authority for the tax exemption expires in 2007, but because the exemption has been renewed several times since it was initiated in 1978, this analysis assumes that it will be extended at the 51-cent (nominal) level for 2007 through 2020.
- MTBE use outside the United States is assumed to continue. No ban or reduction in the use of MTBE is assumed for other countries.
- The cost of increased ethanol blending results in an increase in ethanol prices resulting from greater demand. Additional ethanol requirements are assumed to be met with domestic supply. The level of corn-ethanol production required in this Case is associated with production costs based on a U.S. Department of Agriculture paper analyzing a renewable fuel standard.18
- Biodiesel consumption is assumed to grow from an estimated 5.4 million gallons in 2001 to 7.3 million by 2020. This growth path is based on consumption in alternative fuel vehicles and is assumed to be used to meet requirements of EPAct. The economics and penetration of biodiesel are highly uncertain and are discussed above.
- As indicated in S. 1766, merchant MTBE plants are allowed up to $250 million in grants for conversion to other uses between the years 2002 and 2004.
Key Assumptions of “RFS/No MTBE Ban” Case
The RFS/No MTBE Ban Case uses the same set of assumptions as the S. 1766 Case with the exception of the following:
- No national ban on MTBE is assumed.
- The oxygen requirement on RFG is assumed to be repealed nationally, as indicated in the February 6 letter from Senator Murkowski.
- No conversion of MTBE units is assumed because MTBE is not Federally banned.
- No grant for merchant MTBE plant conversion is assumed.
Provisions of S. 1766 Not Explicitly Modeled
Distillation Index
An additional provision of S. 1766 would require a reduction of the maximum distillation index (also called drivability index or DI) from 1250 to 1200 for summertime at the refinery gate. The DI generally ranges between 1100 and 1300 and is calculated as a function of the temperatures at which 10 percent (T10), 50 percent (T50), and 90 percent (T90) of the gasoline vaporizes. The spread between these ignition temperatures enables the gasoline to be effective in both cold start and warm start situations (when the engine has already been heated). Because efficient cold starting is key to minimizing hydrocarbon (HC) and carbon monoxide (CO) emissions, the automotive industry advocated the creation of the current DI maximum, and is now advocating the 1200 maximum to help meet increasingly stringent exhaust emission standards. The reduction of the DI is expected to make it more difficult to produce gasoline that can be blended with ethanol. At the same time the automotive industry is advocating a change to the calculation of DI that would add further difficulty to achieving even the current standard with ethanol-blended gasoline.19
The gasoline representation in the PMM does not currently include a DI parameter. The inclusion of a DI parameter would require a recursive optimization process that could not easily or efficiently be incorporated into the model framework given the time constraints of this study. In addition, the PMM does not differentiate between different grades of gasoline that are critical to DI analysis.
The proposed reduction in DI would be expected to add to the costs projected in the S. 1766 and the RFS/No MTBE Ban Cases; however, the magnitude of the impact is uncertain. A report by the National Petroleum Council provided a wide range of cost estimates for reducing DI based on different assumptions.20 The NPC analysis was based on a proposal by the Alliance of Automobile Manufacturers which advocated a 1200 DI maximum at the retail level, in contrast to the current standard of 1250 at the refinery gate. The analysis assumed that gasoline would need to meet an 1100 DI at the refinery gate in order to meet 1200 at retail, and resulted in an estimated cost increase of about 7 cents per gallon. A sensitivity analysis, assuming gasoline measuring 1150 at the refinery gate, resulted in additional production costs of only three-fourths of one cent per gallon, compared to the current 1250 DI. 21
Removal of PSI Allowance for Conventional Gasoline
Section 818 of S. 1766 calls for elimination east of the Mississippi River of the current 1 pound per square inch waiver of the Reid vapor pressure (RVP) limitation for fuel blends containing gasoline and 10 percent ethanol in areas not suffering from high ozone concentration levels. This ethanol waiver was originally allowed to make it easier and more economical to blend ethanol into gasoline, since a 10 percent ethanol blend raises the RVP of gasoline by about one pound. Some environmental organizations advocate elimination of the existing 1-pound RVP waiver, arguing that when hydrocarbon emissions increase, ozone levels increase. PMM does not incorporate lifting the waiver because this change cuts through the middle of a refining region and would require detailed regional estimations. The lifting of the waiver may also result in conventional gasoline blendstock of different qualities on either side of the Mississippi River and may require another category of gasoline blendstock in an already stretched distribution system. This new conventional blendstock could not be modeled within the time constraints of this analysis.
Credit Trading Provision
Section 818 of S. 1766 mentions a credit program but does not outline the structure of the program. A credit program would be expected to provide flexibility to refineries that are unable to meet individual targets; however, it would not be expected to significantly modify the aggregate results of this analysis. Due to the minimal expected impact and the lack of specific information about the structure of the program, credit trading was not modeled for this analysis.
Local Market and Price Volatility Issues Not Analyzed
The current modeling approach does not capture some of the changes refiners will make and some of the market dynamics that could influence prices to consumers and alter competition among refiners. The proposed legislation could have disparate impacts among different types of refineries and in different regions. The three-region notional or “aggregate” refinery approach that is used by the PMM does not capture some of these distinctions, which in turn, can affect prices to end users. In addition, NEMS operates as a long-run equilibrium model which projects the levels of domestic production and imports necessary to meet demand requirements. Since supply and demand are always in balance, this approach cannot provide insights into price volatility, either during transitions or after transitions. For many issues, this modeling approach is adequate to simulate refiners’ behavior and associated price implications to consumers, but under the proposed fuel-specification changes, this approach may not fully capture what refiners’ will actually do, particularly on a regional or local basis.
PADD I (East Coast) is an area that could be affected by the disparate effects proposed regulatory changes may have on refineries. PADD I refineries’ markets are mainly on the East Coast, where a large amount of reformulated gasoline is used. Some of these refineries produce mostly reformulated gasoline; therefore, the toxic content of their current gasoline would be much lower than required gasoline toxic limits. The recent Mobile Source Air Toxics (MSAT) rule locks those refineries into those low toxic baselines. Under an MTBE ban, and even with removal of the oxygen mandate, ethanol will be used because of its octane content and other properties. However, the properties of MTBE and ethanol are different, and under an MTBE ban, to meet toxic limits and maintain competitive costs, these refineries could find it more economic to produce less reformulated gasoline components and more conventional gasoline. In this case, Gulf Coast refineries with higher toxic baselines would need to fill the reformulated gasoline gap. The PMM does not simulate these types of individual refinery circumstances. If the situation described evolves, it would produce higher prices to consumers than the PMM implies, and would produce a change in the competitive position of the refiners that the PMM is not designed to simulate.
Price volatility, another area of concern to consumers, could arise under some of the proposed fuel regulatory changes. NEMS is an equilibrium model and, by its nature, does not deal with the imbalances in supply and demand that cause price spikes. The regulatory changes being considered have supply implications that could create price volatility during transition periods. For example, both MTBE bans and DI reduction result in a loss of gasoline production capability that, without further refinery investment, could create price pressures as MTBE and DI changes take effect.
For example, there is the loss of MTBE volumes, which ethanol would only partially replace. Also, in order to use ethanol in place of MTBE, some refineries might have to reduce the “light”, low-boiling-point gasoline components now being used that have higher RVP in order to keep the ethanol-blended gasoline within RVP limits. Lowering the DI, on the other hand, can cause refiners to reduce the “heavy”, high-boiling-point components now being blended into gasoline. Removing these components reduces the amount of gasoline that can be produced without making further investments. Furthermore, imported sources of reformulated gasoline, which are important to the East Coast, are likely to diminish if ethanol-blended reformulated blendstock for oxygenate blending (RBOB) or other formulations not needed abroad are required in the United States. During the transition to a Federal MTBE ban or DI reduction, such supply reductions could add to price volatility. There may not be the time and resources to adjust capacity when engineering and construction resources are likely to be strained meeting the low sulfur fuel requirements alone.
Even after the transition, the East Coast could be subject to more volatility than it has seen historically if it uses ethanol-blended RFG extensively. S. 1766 effectively requires that another major gasoline component be shipped some distance into the Northeast and be stored separately in a distribution system that is working near capacity limits. Whether this increases the potential for volatility requires a more detailed analysis of the distribution system and its potential expansion path along with the implications of increased number of products.
In summary, to comprehensively analyze all of the regional price implications and changes in competitive positions potentially stemming from the proposed legislation requires additional modeling approaches needing more time than was available for this report. The price and volatility implications could be significant, but the magnitude is unknown.
Results of the S. 1766 Case
 Figure Data |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Although S. 1766 includes a year-by-year schedule for the RFS, EIA's renewables projections for the S. 1766 Case do not coincide with the levels specified in the schedule for a number of reasons (Figure 3). In 2006 when the Federal MTBE ban is assumed to occur, ethanol blending increases significantly, exceeding the RFS target by more than 1 billion gallons. The higher level of ethanol blending than the RFS target specified in S. 1766 disappears by 2010 due to increasing RFS targets. Beginning in 2010, the level of ethanol blending is determined predominantly by the RFS schedule rather than the assumed MTBE ban. However, the provision for a cellulose ethanol credit (allowing each gallon of cellulose ethanol to count as 1.5 gallons of renewable fuels) is projected to reduce the amount of renewable fuels required in the RFS schedule by about 10 million gallons in 2003 growing to 130 million gallons by 2012, and about 370 million gallons by 2020 (Table 2).
The RFS requirements can be met either with ethanol or biodiesel. In this analysis, the RFS is projected to be met predominantly with ethanol which represents at least 99.5 percent of total renewables in all years. Biodiesel is assumed to grow from 5.4 million gallons in 2001 to 7.3 million gallons in 2020, reflecting EIA’s estimate of the amount of biodiesel that will be used to fulfill EPAct requirements. As discussed in the Methodology Section, a higher level of biodiesel penetration may occur if it becomes widely used as a lubricating agent for ultra-low-sulfur diesel. Because of the uncertainty in the biodiesel market, "high biodiesel penetration" estimates were developed which could result in 139 million gallons of additional biodiesel and less ethanol used to meet the RFS requirement for 2006, 347 million gallons for 2012, and 625 million gallons for 2020. Given the higher biodiesel penetration assumption, biodiesel would represent 3.5 percent of total renewables in 2006, 7.1 percent in 2012, and 11.9 percent in 2020 in the S. 1766 Case.
S. 1766 is projected to yield carbon emissions from petroleum in the transportation sector which are 4.6 million metric tons lower than those in the Reference Case in 2006, and 7.2 million metric tons lower in 2020. These changes represent a decline of between 0.8 and 1.0 percent of carbon emissions from petroleum used for tranportation from Reference Case levels.
In general, net petroleum imports are projected to be about one percent lower than in the Reference Case. Net petroleum imports are 156,000 barrels per day below Reference Case levels in 2006 (decline of 1.2 percent), and 227,000 barrels per day lower in 2020 (a decline of 1.4 percent). The lower import projections translate into a reduction in the import share of petroleum consumption of between 0.4 and 0.7 percent. The net import reductions may be overestimated in this case because it does not reflect any net increase in energy consumption required for the additional renewables/feedstock production. The inability to capture imports of unfinished products, including iso-octane and alkylates that may come into the market may also contribute to an over-estimation of the decline in imports. On the other hand, the net import projections do not account for the possible increase in MTBE exports from the United States to other countries that might occur if domestic use of MTBE is banned.
 Figure Data |
 Figure Data |
 Figure Data |
Prior to 2006, projected average national prices of all gasoline and of RFG are not significantly different from the Reference Case. After the Federal MTBE ban is assumed to become effective in 2006, the national average price of all gasoline is projected to be about 4 cents per gallon higher (Figure 4), and the national average RFG price between 9.0 and 10.5 cents per gallon higher relative to the Reference Case (Figure 5). The higher projected prices reflect the loss of volume, oxygen, and octane associated with banning MTBE. Ethanol cannot fully compensate for all of these characteristics and is more expensive than MTBE.
Despite the $250 million allowance for merchant plants to convert MTBE units to other uses provided for in S. 1766, this analysis indicates that conversion may be uneconomic for “on-purpose” plants that do not produce their own feedstocks because the feedstock costs may be prohibitive.
Results of the RFS/No MTBE Ban Case
The RFS/No MTBE Ban Case reflects no Federal MTBE ban, and, therefore, no associated increase in ethanol blending requirements. As a result, the projected level of renewables is effectively set by the RFS schedule. The cellulose ethanol credit and results in actual renewables consumption that is below the RFS target levels (Figure 6). After 2012, the RFS target is determined as the percentage of total highway demand that renewables achieved in 2012. By 2020, total renewables consumption in this case is projected to be 40 million gallons per year higher than in the S. 1766 Case, because the relatively high gasoline prices associated with S. 1766 have a slight dampening effect on gasoline demand which in turn reduces blending.
Projections of carbon emissions are lower than the Reference Case but not as low as those in the S. 1766 Case. As a percentage of carbon emissions from petroleum transportation fuels in the Reference Case, this Case results in declines ranging from 0.2 to 0.7 percent between 2006 and 2020, compared with declines in the S. 1766 Case ranging from 0.8 to 1.1 percent.
The projected reduction in net petroleum imports is smaller than in the S. 1766 Case. A reduction in petroleum imports of 61,000 barrels per day is projected for 2006, and 189,000 barrels per day for 2020, compared to S. 1766 Case reductions of 156,000 barrels per day in 2006, and 227,000 barrels per day in 2020. Most of this difference can be attributed to MTBE imports which are allowed in this case but are not allowed in S. 1766. As mentioned above, net imports in the S. 1766 Case may not fully capture the potential for imports of unfinished petroleum products such as iso-octane and alkylate. Neither this case nor the S. 1766 Case account for the additional energy that may be used to produce the additional renewables and their feedstocks.
Projected prices in the RFS/No MTBE Ban Case are well below the S. 1766 projections because no Federal MTBE ban is assumed. In the absence of an MTBE ban more ethanol is available to be blended into conventional gasoline, instead of being pulled into RFG blending to help replace MTBE. Beginning in 2006, projected RFG prices in the RFS/No MTBE Ban Case rise gradually to about 1 cent per gallon higher than the Reference Case by 2012, where they remain through 2020. The impact on the price of all gasoline remains below one-half cent per gallon compared to the Reference Case through 2020.
The relatively minor change in national average prices projected in this case may not translate into minor changes in localized markets. Although the aggregate nature of the PMM precludes an analysis of local markets, information about the shift in ethanol blending from RFG toward conventional gasoline would point to possible price increases in conventional gasoline markets that do not currently blend ethanol. After 2012 when the growth of the RFS targets slows, ethanol blending into conventional gasoline is projected to increase up to 2.8 billion gallons per year over Reference Case levels without a Federal MTBE ban, compared to an increase of no more than 0.5 billion gallons per year in the S. 1766 Case which assumes a Federal ban. This higher level of ethanol blending into conventional gasoline occurs because MTBE is still blended into RFG, leaving most of the RFS requirement for ethanol to be absorbed into conventional gasoline markets.