Introduction
Methodology
Transportation Fuel Consumption
Vehicle Sales
Fuel Economy
Why Fuel Consumption is Higher than Expected: "Shortfall"
Effects
"On the Road" Fuel Efficiency Effects
Stock Turnover Effects
Horsepower (HP)/Performance Effects
Size Class Consumer Purchase Shifting Effects
AFV Sales Effects
VMT Effects
Quantification and Distribution of "Shortfall" Effects
The Office of Transportation Technologies (OTT), Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy (DOE), requested that the Energy Information Administration (EIA) analyze the impacts on petroleum prices, demand, and refinery operations of an increased demand for diesel fuel stemming from greater penetration of diesel-fueled engines in the light-duty vehicle (LDV) fleet of the U.S. transportation sector, compared with the Annual Energy Outlook 1998 reference case. This request was made as part of EE's "Quality Metrics" initiative, which is designed to collect a wide range of data and information required for the Government and Performance Results Act of 1993, the National Performance Review's Performance Agreements with the President, and Executive Order 12862 on setting Customer Service Standards. OTT also wanted the impacts to respond to inquiries from the White House, Congress, and other entities of DOE.
The specific cases that OTT requested EIA to analyze were as follows:
(1) Advanced diesel technology begins penetrating the market in 2003,(1) increasing to 10 percent of new LDV sales by 2010, and constant thereafter.
(2) Advanced diesel technology begins penetrating the market in 2003, increasing to 20 percent of new LDV sales by 2010, and constant thereafter.
(3) Advanced diesel technology begins penetrating the market in 2003, increasing to 30 percent of new LDV sales by 2010, and constant thereafter.
OTT also requested that the advanced diesel cases include a direct injection diesel technology with 50 percent higher fuel efficiency than an equivalent conventional gasoline engine. According to the Fuel Economy Guide,(2) current turbo direct injection diesel technology achieves approximately 63 percent higher fuel economy than a gasoline engine in the Volkswagen Jetta model, indicating that such high efficiency diesel engines are already commercially available.
By comparison, in the Annual Energy Outlook 1998 (AEO98) reference case, the reference case for this report, penetration of diesel engines in 2010 is only about 0.3 percent of new sales of LDVs. Thus the requested cases represent a significant increase in the penetration of diesels in the LDV fleet beginning in 2003.
This analysis was conducted using EIA's National Energy Modeling System (NEMS). NEMS is an integrated model that represents the supply, conversion, and end-use demand sectors in domestic energy markets. It also contains macroeconomic and international modules to incorporate the effects of economic factors and world oil markets. By balancing energy supply and demand, NEMS projects production, imports, consumption, and prices of energy through 2020. The transportation and industrial demand modules, the Petroleum Market Module, and the macroeconomic and international oil modules of NEMS were used in the preparation of this report.(3) The demand modules represent the consumption of energy to meet end-use sector requirements based on underlying factors governing the demand for energy, such as prices, gross domestic product (GDP), demographic factors, and the costs and performance characteristics of energy-consuming equipment (including automobiles). The demand modules provide regional results at the Census Division level. The PMM represents the conversion of crude oil to petroleum products and their distribution to end-use sectors, based on the costs and technical characteristics of refinery operations, the costs of distribution, and the demand for products, subject to refineries' engineering constraints. The PMM is a three-region model, consisting of the East Coast, the central United States (including the Gulf Coast), and California. The PMM provides prices of all petroleum products to the demand modules. Both the macroeconomic and international modules were also run for this analysis in order to incorporate feedback effects of increased diesel fuel penetration to economic growth and world oil prices. In addition, the industrial demand model was also run to determine the change in the mix of industrial fuel demand as a result of changing prices for distillate fuel.
In order to analyze the impacts of increased diesel fuel penetration, the following cases were developed:
Reference Case. The Reference Case in this report is the same as that prepared for the AEO98. The results in this case are based on the expected continuation of existing laws, regulations, and policies. This case serves as the comparison case for analyzing the impacts of increased diesel penetration beginning in 2003.
10-Percent Case. This case assumes that all aspects of the reference case are in effect, with the exception that penetration of new sales of LDVs by diesel-fueled engines are forced to 10 percent of all sales by 2010 (ramping up from reference case values starting in 2003). In particular, the availability of imports of refined petroleum products, the configuration of refineries, and the quality of diesel fuel produced by domestic refiners are all assumed to remain the same. Light-duty vehicle sales are set to achieve the assumed diesel penetration levels, and AFV sales are held the same as the AEO98 reference case level (8 percent by 2020). Because the purpose of these model runs was to assess the impacts of various levels of diesel vehicle sales penetration, diesel vehicle attributes were not altered, with the exception of increasing the diesel fuel economy level to 50 percent above comparable conventional gasoline fuel economy levels, as requested by OTT.
20-, 30-Percent Cases. Each of these cases was exactly the same as above, except that diesel-fueled sales of LDVs were ramped to 20 and 30 percent, respectively, by 2010. Again, all other aspects of the transportation and refinery model assumptions were unchanged, with the exception of the fuel economy level for diesel.
Low-Sulfur Diesel Case. In this case, all of the assumptions of the 30-Percent case were combined with the assumption that the maximum sulfur content for diesel fuel for LDVs would be 50 parts per million (ppm), compared to 500 ppm in the other cases. This case was run to show the impact of requiring a low-sulfur diesel fuel to enable the use of catalytic converters to reduce emissions of oxides of nitrogen that would result from higher diesel fuel demand. Additionally, product imports for 50 ppm diesel were not made available, while imports of 500 ppm diesel and other distillate were made available at the same price and volume as in the 30 percent case. Imported ultra-low sulfur diesel was not incorporated because there is currently no reliable data upon which to make any assumptions concerning the price at which the imported product would be available. Therefore, the prices for ultra-low sulfur diesel reflect the marginal cost at the average U.S. refiner absent any international ultra-low sulfur diesel supply effect.
Total transportation sector fuel consumption is lower with increasing penetration levels of diesel vehicle sales across the cases. Table 1 indicates that total transportation fuel consumption is as much as 0.34 million barrels per day oil equivalent (mmbdoe)(4) lower for the 30 percent diesel sales share case by 2020 compared with the reference case, with smaller differences for the 10 and 20 percent cases. Gasoline consumption is as much as 2.0 mmbdoe lower than the reference case by 2020 (Figures 1A and 1B). Gasoline consumption is displaced with distillate fuel use, which is 1.55 mmbdoe higher than the reference case in 2020 in the 30 percent case. Additional displacement of gasoline consumption results from slightly higher alternative-fuel consumption of about 0.06 mmbdoe in 2020 in the 30-percent case (Table 1). Total petroleum use in the transportation sector is as much as 0.43 mmbdoe lower by 2020 at 30 percent penetration levels for diesel, compared with reference case levels. The difference between total petroleum demand and total transportation demand is due to the increase in AFV consumption, together with the increase in the associated losses for electricity-related consumption.
Total transportation consumption is lower than in the reference case because of the higher efficiency of diesel-fueled engines. In the reference case, the overall efficiency of the LDV fleet is approximately 20.7 miles per gallon (mpg) in 2010, rising to 21.2 mpg by 2020. In the diesel penetration cases, because of the assumption that diesel engines are 50 percent more efficient than conventional gasoline engines, the 2020 fleet efficiency rises to as high as 23.0 mpg in the 30-percent penetration case. New gasoline-powered cars average 29.0 mpg in 2020 in the 30-percent case, with new diesel-powered cars averaging 44.1 mpg. For light-duty trucks, the corresponding efficiencies are 20.5 mpg for gasoline-powered engines, and 30.3 for diesel-fueled new trucks.
The mix of fuel consumption in 2020 results in a reduction in gasoline's share of total transportation fuel, from 54 percent in the reference case, to as low as 43 percent in the 30-percent penetration case. At the same time, the use of distillate fuel in transportation is higher, with the share going from 18 percent in the reference case to as high as 27 percent in the 30-percent penetration case by 2020. Although petroleum consumption is lower in all cases, petroleum as a percent of total transportation sector fuel use remains relatively unchanged at approximately 94 percent in 2020.
Transportation sector carbon emissions (Figure 2) are slightly lower across the cases, by as much as 13 million metric tons (mmt) compared with the reference case by 2020, in the 30-percent penetration case. This represents less than 2 percent of the sector's reference case emissions. Refinery fuel consumption (included in the industrial sector) is also up to 0.227 mmbdoe lower in 2020 compared with the reference case, resulting in total carbon reductions--including those in the transportation sector--of as much as 20 mmt in 2020. Refinery fuel consumption falls because of the reduced still gas consumption needed to meet the revised slate of petroleum product demands.
Table 1 includes the macroeconomic impacts due to the assumptions of increased diesel fuel penetration. In general, there is a net economic benefit due to the increased penetration of diesel-fueled LDVs, because world oil prices and other petroleum prices are slightly lower as a result of the decreased overall demand for petroleum products. Real GDP is less than 0.1 percent higher than the reference case ($8.5 billion in real 1992 dollars) in the 30 percent case by 2020. Real disposable personal income (DPI) also changes very little compared with the reference case in 2020, with an additional $11.3 billion (0.14 percent of DPI) in the 30 percent case. It should be noted, however, that this study does not take into account all of the effects that might arise from a higher level of disel demand, such as the impacts on suppliers of equipment to refineries, the feedback effects due to trade, or the tax revenue consequences of lower petroleum prices. It is not clear whether the ultimate impact would be negative or positive for the economy; however, the first-order impacts of lower petroleum prices would be beneficial.
Total vehicle sales in 2020 are slightly higher due to the higher DPI levels. Approximately 105,000 additional units are sold in the 30 percent case. Gasoline vehicle sales (Table 2) in 2020 are lower as a result of the displacement by diesel vehicles. In the 10 percent case gasoline vehicle sales are almost 1.5 million units lower than in the reference case, or approximately 720,000 cars and 763,000 light trucks by 2020. More drastic reductions in gasoline vehicle sales occur in the 20 and 30 percent cases with reductions of over 3.0 million units (1.57 million cars and 1.47 million light trucks) and almost 4.6 million units (2.41 million cars and 2.17 million light trucks), in the two cases respectively. The loss in gasoline vehicle sales is more than offset by the increase in diesel light-duty vehicle sales. Almost 4.6 million (2.52 million cars and 2.05 million light trucks) more diesel vehicles are sold in the 30 percent diesel case, compared with the reference case.
Diesel fuel efficiencies are assumed to be 50 percent higher than gasoline vehicles in all years (Table 3). When combined with the additional diesel vehicle sales, new car fuel economy in as much as 4.26 mpg higher in 2020 in the 30-percent case, compared with the reference case. New light truck fuel economy is 3.11 mpg above the reference case by 2020 in the 30-percent case. However, because of the slow turnover of the entire stock of vehicles, the improvement in the fleet is much less dramatic. By 2020, average fleet car economy is 1.72 mpg above the reference case in the 30-percent penetration case, with the corresponding truck fleet economy 1.73 mpg higher.
"Shortfall" refers to the fact that the total reduction in consumption as a result of the increased penetration of new diesel-fired vehicles, together with the improved efficiency of those engines, is not as great as would be expected(5) from a first-order calculation of the impacts. Even assuming reference case VMT and other factors, the factors discussed below mitigate the reduction. Although new car mpg and new light truck mpg are higher than in the reference case, these new fuel economy improvements from higher diesel sales penetration levels do not result in identically equivalent improvements in light-duty vehicle "on the road" stock mpg. Several factors that result in this "shortfall" effect can be traced back to lower gasoline prices, higher DPI, and slow turnover of the vehicle stock. In addition, because VMT is a function of the cost per mile of driving, higher efficiencies and lower fuel prices increase VMT in the diesel penetration cases, further mitigating the benefits of the high-efficiency diesel engines. Each of these factors is discussed in turn.
The net effect of the "shortfall" phenomena results in "on the road" stock mpg that is higher than the reference case, but not as high as new vehicle efficiencies. On the road stock efficiencies are only as much as 1.77 mpg higher for the 30-percent penetration case by 2020, compared to the reference case. This represents less than half of the corresponding new vehicle fuel economy improvement by 2020.
Within each of the three diesel cases (Table 4), new car and new light truck efficiency is "degradated" to reach the "on the road" new efficiency, which is lower than the rated efficiency by approximately 16 percent for cars and 21 percent for light trucks. These efficiency losses reflect the "degradation factor", which accounts for the difference between EPA rated new fuel economy and actual "on the road" efficiency.(6) Degradation factor components consist of elements such as the ratio of city to highway travel, road congestion, and average highway speed.
An important factor in the "shortfall" effect of stock efficiency relative to new vehicle fuel efficiencies is the slow turnover in the vehicle stock. To illustrate this phenomenon, Table 4 displays harmonically sales weighted average efficiencies for combined new light-duty vehicles for each of the three diesel cases. Comparisons between the combined "on the road" new mpg and "on the road" stock mpg can be attributed to the slow turnover in the stock, because older vehicles have a lower efficiency than that assumed for new vehicles. Table 4 shows the new fuel efficiencies for 2020 only, but prior to 2020 the new vehicle fuel efficiencies are lower (because gasoline-powered engine efficiencies are lower than diesel-powered, and because new cars improve in efficiency every model year), and these vehicles leave the stock very slowly. Even after 10 years, 75 percent of all cars, and 81 percent of all light trucks purchased in a given year are still on the road.(7) Although the assumptions for the runs included reaching diesel sales penetration levels of 10, 20, and 30 percent of total light-duty vehicle sales by 2010, these same levels are not attained in the vehicle stock. As shown in Table 5 and Figure 3, by 2020 diesel vehicle stocks are as much as 24 percent of total vehicle stock in the 30-percent case, compared with less than 0.5 percent in the reference case.
With rising income levels from the positive macroeconomic feedback effects and falling gasoline prices, horsepower (HP) or performance demanded is above the reference case in all three cases (Table 6). Consumers choose higher HP within each size class for both cars and light trucks. Automobile HP is up to 11.12 HP higher than the reference case across the cases by 2020.
Similarly, light truck horsepower exceeds the reference case by as much as 11.0 HP for the 30-percent case in 2020. This reflects consumers' desire for more performance-oriented vehicles when purchasing power increases, either because the costs of the vehicles are reduced, or there is more disposable income available. Since efficiency is inversely related to HP, this effect causes lower efficiency relative to the reference case.
Higher income levels and lower gasoline prices also lead to consumer purchase shifts away from small vehicles and toward larger vehicles (Table 6). By 2020 consumers have shifted away from purchasing compacts and instead purchase more mid-size and large size cars. Less purchase shifting across size classes occurs within light trucks, where consumers purchase more small sport utility vehicles rather than small vans.
To summarize the above two effects, both performance and size class shifts cause new car and new light truck fuel economy to be lower than would be the case without these effects. All new car and light truck fuel efficiencies in 2020 are actually lower relative to reference case efficiency, by approximately 0.1 to 0.94 mpg (Table 3). Table 7 also illustrates the effects of performance and size class shifts on new fuel efficiencies for gasoline and diesel vehicles. Given the reference case new fuel economy levels for cars and light trucks in 2020, and combining this with the assumption in all three cases that diesel vehicles are 50 percent higher in new vehicle fuel economy, we show an expected diesel efficiency both with and without shifts in purchases to larger, higher performance vehicles. In each case, the actual new fuel efficiency for diesel vehicles is lower than the expected efficiency. A portion of the difference can be attributed to a lower actual efficiency for both gasoline and diesel (as a result of the increase in HP and size class shifting effects)
Although almost all new vehicle fuel efficiencies in the diesel penetration cases are lower than the reference case, due to rising HP, AFV fuel efficiencies are not as adversely affected as gasoline and diesel vehicles, because AFVs are inherently more efficient than vehicles powered by traditional fuels. The net result is a relative advantage in both fuel efficiency and vehicle range (a function of fuel efficiency, holding the fuel tank size constant), translating into higher AFV sales. AFV sales in 2020 are 182,000, 161,000, and 110,000 vehicles higher than the reference case total of 1.16 million vehicles, in the 10-, 20-, and 30-percent diesel penetration cases, respectively. Fuel efficiency gains cannot offset the rising HP effects across the cases, however, because AFV fuel prices are relatively equal across the diesel cases while income levels continue to spur HP demands. Therefore, AFV sales are lower the higher the diesel penetration (beyond the 10-percent penetration level), as the advantages of the fuel price effects on higher fuel efficiency are overcome by the countervailing force of the rise in HP. The total "shortfall" effect from higher AFV sales results in rising alternative-fuel consumption of as much as 0.709 mmbdoe in 2020 in the 10- and 20-percent cases compared with the reference case (Table 8).
VMT rises slightly across the cases and is higher relative to the reference case with higher diesel penetration levels, because of the macroeconomic feedback effects of higher income levels and lower gasoline prices. Lower gasoline prices reduce the cost of driving, which increases VMT. By 2020 (Table 9) total VMT is higher by up to 25.53 billion miles (0.78 percent of total VMT) in the high penetration case, compared with the reference case.
In order to provide an estimate of the quantitative effects of each "shortfall" phenomenon, a partial derivative methodology was used. Sequential runs of the model were made, changing one "shortfall" variable at a time. For example, one run was made in which VMT was held at reference case levels, with all other assumptions consistent with the 30-percent penetration case.
In this run, total consumption of diesel fuel was .065 mmbdoe (138 trillion Btu) lower than in the 30-percent case. This represents the "shortfall" effect due to increasing VMT. The net effect of each of the variables upon fuel consumption was then calculated relative to the reference case in the year 2020, and compared to "expected" savings of 1920 trillion Btu. The results are shown in Table 10 and Figure 4. This methodology is only an approximation, in that it ignores the effects of combining variables, and it assumes that there are no cross product effects. Since the individual contributions do not sum to the total "shortfall" they have been normalized in Table 10 on a proportional basis.
The largest "shortfall" effect, 41 percent, was attributed to the rise in horsepower that occurred in response to the lower gasoline price and the secondary macro-economic feedback effect of slightly higher income. Slow turnover in the stock, which effectively yielded only a 24 percent diesel stock penetration despite a 30 percent diesel sales penetration by 2020, contributed 37 percent of the total "shortfall" effect. VMT "shortfall" effects, which amounted to 18 percent of the total shortfalls, occurred from the decline in gasoline prices making the cost of driving per mile decline resulting in more driving.
The last column in Table 10 reflects a scenario in which manufacturers hold HP and size class shifts at base case levels. Figure 2 carbon levels for the 30 percent diesel case would decline by an additional 6.7 mmt carbon, bringing the total carbon savings to 19.7 mmt of carbon from the trnsportation sector, and 27 mmt from all sectors.
1Although the original OTT
request was to assume that penetration begins in 2000, during discussions with OTT staff
it was decided to delay the onset of additional penetration until 2003, due to the short
lead time.
2U.S. Department of Energy, Model Year 1998 Fuel Economy
Guide, DOE/EE-0136, U.S. Government Printing Office, October 1997, (Washington D.C.)
3The remaining NEMS modules were not included because of
expectations that results from those modules would not differ significantly from the
reference case as a result of increased diesel penetration.
4Quantities in the text have been converted from quadrillion
Btu, as shown in the tables, to million barrels per day oil equivalent, using the
conversion factor 5.8 million Btu/barrel.
5"Expected" fuel savings were calculated as
follows: Reference case LDV fuel consumption in the transportation sector in 2020 is 19.20
quads. If there were a full 30 percent penetration in the LDV stock (as opposed to LDV
sales) by diesel vehicles in 2020, they would represent 30 percent of that consumption, or
5.76 quads. Since by assumption diesel engines are 50 percent more efficient than gasoline
engines (or 1.5 times as efficient), consumption would be only two-thirds (the inverse of
1.5) the consumption in the reference case, or 3.84 quads. The difference between the
reference case and diesel-adjusted consumption is therefore 1.92 quads, which represents
"expected" savings in Table 10.
6Decision Analysis Corporation of Virginia, Fuel
Efficiency Degradation Factors, prepared for the Energy Information Administration,
Final Report, Subtask 1, August 3, 1992.
7U.S. Department of Energy, prepared by Oak Ridge National
Laboratory, Transportation Energy Databook: Edition #17, ORNL-6919, pg. 3-9 and 3-10,
August 1997, (Oak Ridge, Tennessee).
To Refinery Operations
Back to Service Report Homepage
File last modified: February 2, 1999
Contact the National
Energy Information Center for questions at (202) 586-8800
If you having technical problems with this site, please contact the EIA Webmaster at wmaster@eia.doe.gov