Strategic Issues Facing Transportation, Volume 5: Preparing State Transportation Agencies for an Uncertain Energy Future (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

125 Natural gas and liquid petroleum gas have a long history of use as transportation fuels both domestically and inter- nationally. Natural gas is a fossil fuel extracted from onshore or offshore wells, sometimes in conjunction with crude oil. It is composed mostly of methane (70% to 90%) and minor amounts of ethane, propane, butane, and other gases (NRC 2013). The gas must either be pressurized to compressed natu- ral gas or cooled to liquefied natural gas for storage in vehicles. Most light- and medium-duty vehicles (i.e., passenger cars, sport utility vehicles, vans, and small-to-medium trucks) that rely on natural gas use CNG, whereas both CNG and LNG are used for heavy-duty vehicles such as buses and cargo trucks (Yacobucci 2005). Currently less than 3% of the natural gas consumed in the United States is used for transportation (NRC 2013). Until recently, LPG has dominated the alternative-fuels transportation market in the United States. Produced as a by-product of natural gas processing and petroleum refining, LPG is a mixture of various hydrocarbons, including pro- pane, propylene, butane, and butylene. LPG is also referred to as propane or autogas (EERE 2012c). Because LPG is gaseous at ambient temperature and pressure, it must be liquefied before it is injected into a vehicle (Yacobucci 2005). Natural gas is viewed as a potentially attractive alterna- tive to conventional petroleum-fueled vehicles for several reasons. First, recent breakthroughs in natural gas extrac- tion technologies—specifically, horizontal drilling combined with hydraulic fracturing (or fracking)—have made it pos- sible to economically recover abundant domestic natural gas resources (IEA 2012). The resulting lower cost of natural gas should help reduce the energy cost of travel, while the domestic supply is important from the perspective of energy security. In comparison to the combustion of gasoline and diesel, natural gas also promises modest reductions in local air pollutants and greenhouse gas emissions. However, natural gas still results in greenhouse gas emis- sions, both from combustion and from the upstream supply chain (ANL 2012, Alvarez et al. 2012), and the fuel is a non- renewable fossil resource. Some, therefore, view natural gas as an intermediate step in a decades-long transition to cleaner, renewable transportation fuels. Additionally, natural gas has competing uses as a cleaner alternative to coal for power gen- eration, as well as in manufacturing. These considerations, and others discussed later in this appendix, make the prospects for a significant shift to natural gas within the transportation sector somewhat uncertain. Although LPG has been a leader in alternative fossil fuels in the past, the greater supply and reduced prices of natural gas suggest that CNG has greater promise for future success in the light-duty vehicle market. Accordingly, this appendix focuses more attention on natural gas than on LPG. B.1 Production, Distribution, and Refueling There is already significant production of natural gas, world- wide and in the United States, along with a well-developed dis- tribution system with an extensive network of pipelines. Yet there are relatively few stations for dispensing natural gas to vehicles, most of which are dedicated for fleet use. Thus the network of dispensing stations would need to be considerably expanded in order to support a significant level of natural gas adoption among the light-duty vehicle fleet. B.1.1 Natural Gas Reserves The proven reserves of natural gas in the world—the quan- tity that appears to be recoverable in the future under existing economic and operating conditions based on available geo- logical and engineering information—have increased steadily over time. From 1991 to 2011, the world’s proven reserves of natural gas rose from 131 trillion cubic meters (tcm) to 208 tcm. Nearly 80% of this total—almost 160 tcm—is divided in roughly equal shares between the Middle East and Europe A p p e n d i x B Natural Gas and Liquid Petroleum Gas

126 and Eurasia. As of 2011, proven reserves in the United States stood at 8.5 tcm, just over 4% of the world total (BP 2012). Total world production of natural gas in 2011 was 3,276 billion cubic meters (bcm). At this rate, the world’s proven- reserves-to-production ratio—that is, the number of years that it would take to exhaust the proven reserves—is about 64 years. The United States produced about 651 bcm in 2011, corresponding to a proven-reserves-to-production ratio of about 13 years (BP 2012). Estimates suggest, however, that the United States has as much as 50 tcm of technically recover- able natural gas, which includes undiscovered, unproven, and unconventional natural gas resources [NPC 2007, Potential Gas Committee (PGC) 2009]. It is quite possible, therefore, that significant U.S. production levels could extend well beyond the 13-year proven-reserves-to-production time frame. A recent breakthrough in drilling methods—the combina- tion of horizontal drilling and hydraulic fracturing—has sig- nificantly increased the estimates of domestic economically recoverable natural gas reserves. Horizontal drilling allows producers to drill vertically and then horizontally to access multiple permeable zones of vertical geologic faults rich with natural gas. Although more costly than the traditional ver- tical drilling approach, this technique has allowed for the economical extraction of natural gas from sources once con- sidered unrecoverable or uneconomical, such as the Barnett shale regions of Texas, which could supply gas for over 40 years [Texas Comptroller of Public Accounts (TCPA) 2010]. The high price of petroleum in recent years has also contributed to the emergence of horizontal drilling as an economical option. Between 2000 and 2010, according to estimates from the EIA, the U.S. proven reserves of natural gas increased by about 71% (EIA 2012d). The production of natural gas from shale has enabled sig- nificant growth in overall U.S. natural gas production over the last several years. Shale formations trap natural gas in rocks of low permeability and low porosity. Horizontal drill- ing enables fracturing of the shale formations to release the natural gas. The overall environmental impact of shale gas, however, remains unclear. The process of extracting the natu- ral gas can result in localized emissions of harmful air pollut- ants in new areas (Litovitz et al. 2013). On the other hand, the net effects on air quality at the regional level might be positive if the natural gas is substituting for coal in the generation of electric power, a subject being examined in ongoing research. Shale development has also introduced water quality con- cerns (see, for example, Osborn et al. 2011), which could slow the development of natural gas from shale via hydrau- lic fracturing. Additional studies are currently underway to better understand the water quality implications of hydraulic fracturing. Additional unconventional sources of natural gas that might be exploited in future years include other tight gas and coal bed resources. B.1.2 Natural Gas Production The world’s total production of natural gas has increased steadily from 1,100 bcm in 1970 to almost 3,300 bcm in 2011. Dating to the 1980s, the Russian Federation and the United States have consistently accounted for the largest shares of global production. In 2011, the United States produced 651 bcm of natural gas, or almost 20% of the world total (BP 2012). Recent increases in U.S. production stem from unconventional gas supplies such as shale gas (EIA 2013). The states with the highest share of U.S. proven reserves, ranked in descending order, are Texas, Wyoming, Louisiana, Oklahoma, Colorado, New Mexico, Arkansas, and Pennsylvania. Collectively these accounted for more than 80% of total U.S. proven reserves and close to 79% of U.S. production as of the end of 2010. Texas alone has more than a quarter of the nation’s proven reserves. Additional producers, alphabetically, are Alaska, Alabama, California, Florida, Kansas, Kentucky, Michigan, Mississippi, Montana, New York, North Dakota, Ohio, Utah, Virginia, and West Virginia (EIA 2012c). Many of these states already have an extensive natural gas network of underground pipelines that could support delivery for transportation uses. B.1.3 Natural Gas Consumption As with historical production trends, the world’s consump- tion of natural gas has grown steadily over time. The largest consumers of natural gas, by a sizable margin, are the United States and the Russian Federation. The United States con- sumed over 690 bcm (21% of global consumption) and the Russian Federation consumed 425 bcm (13%) in 2011. With recent gains in U.S. production, however, the United States has now become a net exporter of natural gas (EIA 2013). It still imports a significant volume of natural gas by pipeline from Canada along with more limited quantities via pipeline from Mexico and liquefied natural gas via tanker from Trinidad and Tobago, Qatar, Yemen, Egypt, Peru, Norway, and Nigeria (BP 2012). It also exports natural gas to other countries, how- ever, and export volumes have grown to exceed import volumes as of 2013. Natural gas has many uses. It provides a feedstock for elec- tric power generation; it is used to help produce such products as steel, glass, paper, clothing, and bricks; it serves as a raw material in paints, fertilizers, plastics, antifreeze, dyes, photo- graphic film, medicines, and explosives; it heats homes and fuels stoves, water heaters, clothes dryers, and other home appliances; and it propels a small share of the vehicle fleet. As of 2011, the shares of natural gas consumption in the United States across different sectors were as follows: 31% for elec- tric power, 28% for industrial, 19% for residential, 13% for commercial, 6% for oil and gas industry operations, 3% for pipeline and distribution use, and less than 1% for vehicle fuel (EIA 2012b). In considering the future potential for natural

127 gas as a fuel for transportation, it is thus important to consider the competition from other potential applications of natural gas as well (NRC 2013). B.1.4 Natural Gas and Liquid Petroleum Gas Distribution The pipeline infrastructure for distributing natural gas in the United States is expansive. In areas without gas pipelines, natural gas is liquefied at the wellhead and transported to the market as LNG in insulated containers. LNG is also imported into U.S. coastal states via LNG tanker ships. After being trans- ferred from the ships, LNG can be delivered via pipeline or stored in cryogenic tanks or underground caverns. At fueling stations, LNG can either be injected into natural-gas vehicles equipped with LNG fuel tanks or gasified for vehicles with CNG tanks (NPC 2012). CNG can be delivered to retailers through pipelines and tanker trucks much like gasoline. Large electric- or gas-driven compressors are used to produce higher- pressure CNG for onboard storage from low-pressure pipeline gas. Most fueling stations have their own compressors. LPG is more commonly used for home heating and out- door grilling than for transportation. Currently, most of the LPG pipelines run through the central and southern United States. There are no major LPG pipelines in the western states, so LPG is delivered to these states primarily via rail, truck, and ship transport. It is likely that the supply infrastructure of LPG could expand if demand were to rise (WGA 2008). B.1.5 Refueling Infrastructure In contrast to the extensive distribution network for natu- ral gas, available refueling infrastructure is quite limited. To support significant market adoption of natural gas as a trans- portation fuel, the network of refueling stations would need to be expanded significantly. Because natural gas is delivered to many homes in the United States, the installation of home refueling appliances is also possible. Refueling stations. Compared to more than 150,000 gasoline stations operating in the United States (API 2013), the number of natural gas and LPG stations available to the public is quite limited. As of August 2012, there were 2,654 LPG stations, 59 LNG stations, and 1,107 CNG stations in the country (EERE 2012e), and many of these were for fleet use only and not open to the public. While there are more refuel- ing stations for LPG than for LNG or CNG, their patronage has been declining in recent years, with the total number of LPG stations having peaked in 1998 with over 5,300 (WGA 2008). In contrast, the number of natural gas refueling sta- tions has been on the rise. Texas has the largest share of LPG stations in the coun- try, with 18% of the total, followed by California with 9% and Indiana with 7%. California and New York have the most CNG stations, with 22% and 10% of the nation’s total, respectively. Interestingly, Texas is home to less than 4% of the CNG refueling stations, even though it leads the nation in natural gas production (EERE 2012e). Developing a broader market for natural gas as a transpor- tation fuel, then, will likely require a significant build-out of refueling infrastructure, which is a costly undertaking. The amortized cost of building a new dedicated natural gas refu- eling station can be $0.61 to $0.69 per gge of CNG, corre- sponding to about 125 cubic feet of natural gas (NPC 2012). The estimated amortized cost of adding integrated, modular CNG refueling infrastructure is higher, at $1.09 to $1.15 per gge, since the annual dispensing capacity is lower than with dedicated stations (NPC 2012). Home refueling. Because natural gas is already delivered by pipeline to many homes in the United States, home refu- eling stations are also possible. This can take the form of a garage-mounted appliance that allows a vehicle to connect to the existing natural gas line and refuel overnight. Currently this appliance can be purchased for about $4,500, with an additional cost for installation (O’Dell 2011). Rebates and tax incentives exist to help offset this cost. B.2 Vehicle Technologies As of 2011, there were approximately 253 million vehicles operating in the United States. Most (about 234 million) were light-duty vehicles—cars, minivans, pickup trucks, and SUVs (BTS 2013). Approximately one million light-duty AFVs were in use during the same year. This count includes vehicles run- ning on LPG, natural gas, E85 (an 85% ethanol blend; the count does not include the share of flex-fuel vehicles believed to run mainly on gasoline instead of E85), grid electricity, and hydrogen. About 77,000, or 8% of the AFVs, were LPG vehicles, and another 66,000, or 7%, were CNG vehicles (EIA 2012a). A significant share of both LPG vehicles and CNG vehicles are fleet vehicles. As shown in Table B.1, private fleets and municipal governments are the predominant users of CNG and LPG as transportation fuels (EIA 2012a). In response to incentives and regulations in the federal Energy Policy Act, federal and state agencies have invested considerably in CNG vehicles and LPG vehicles. About 90% of the federally owned fleet of CNG vehicles is made up of light-duty vehicles pur- chased to fulfill EPA requirements (Yacobucci 2005). Manufacturers of NG- and LPG-powered vehicles. CNG and LPG vehicles can be supplied to the public through two routes: either an OEM directly assembles or manufactures the vehicle, or a converter modifies a gasoline or diesel vehicle to operate on CNG or LPG. In 1992, only two OEM light- and medium-duty natural gas-powered vehicles were available in the U.S. market. The

128 separate fueling systems, enabling the vehicle to operate on either fuel (EERE 2012f). LPG vehicles also operate similarly to gasoline-powered, spark-ignition vehicles. Because LPG is stored in the vehi- cle as a lower-pressure liquid, a regulator first vaporizes a controlled amount of LPG. The vapor is passed into a mixer where it is combined with filtered air. This mixture is then introduced into the combustion chamber, where it is spark- ignited and burned to produce power. In the last 15 years, LPG injection engines, which allow the injection of LPG into the combustion chamber as a liquid, have also been devel- oped [Energy Technology System Analysis Program (ETSAP) 2009]. In the United States, onboard CNG storage is typically accommodated with a tank pressurized at between 2,000 psi and 4,000 psi. The CNG tank on the Honda Civic GX holds 8 gge of CNG at 3,600 psi, and the tank reduces available trunk space by half compared to a gasoline-powered Civic. Higher- pressure tanks of up to 10,000 psi are also possible, reducing the amount of space required to store a given amount of fuel but also increasing the cost and energy required to compress the gas (NRC 2013). LPG, which is gaseous at ambient tem- perature and pressure, is compressed to a liquid and stored in vehicle tanks at around 200 psi. Vehicle conversion technology. The conversion of con- ventional vehicles to run on natural gas or LPG is an option for increasing the number of CNG and LPG vehicles on the road. Most gasoline engines can be converted to operate on natural gas with some modifications to the fuel system, including a new fuel tank, new fuel lines, and changes to the electronic control unit (Yacobucci 2008). Because natural gas and LPG engines typically use spark ignition, it is generally easier to convert a gasoline engine, which also relies on spark ignition, to operate on natural gas or LPG. Light- and medium-duty gasoline vehicles can be retrofit- ted with a CNG or LPG fuel system using converter kits made by small-volume manufacturers, or SVMs (NGVA 2010). number of models offered by OEMs subsequently increased, peaking at 18 models in 2002, and then declined rapidly. In the latter part of the first decade of this century, Honda, with its Civic GX, was the only OEM to offer a natural gas- powered light-duty vehicle. Spurred by the enhanced supply and reduced price of natural gas enabled by horizontal drill- ing technologies, however, an increasing number of OEMs have once again begun to offer natural gas models. As of 2012, five manufacturers—Chevrolet, Ford, GMC, Honda, and the Vehicle Production Group—were offering light-duty CNG vehicles, and Ford was marketing a bi-fueled truck capable of running on natural gas (EERE 2012a). OEM LPG vehicle production began in 1997 with three different models of Ford trucks. In 2001 and 2002, Chev- rolet and GMC also contributed to the mix of light- and medium-duty LPG vehicles available to the public but have since discontinued their productions, leaving Ford as the lone producer of OEM LPG vehicles. Following model year 2009, Ford terminated its LPG vehicle production as well (EERE 2012a). Buyers can now only acquire an LPG vehicle through conversions. In some cases, buyers can place an order with an automaker, which contracts with a vehicle converter to modify the standard OEM vehicle into an LPG vehicle. In addition to CNG and LPG vehicles relying solely on an internal combustion engine, natural gas–electric hybrid and LPG-electric hybrid vehicles have also been developed. In the United States, only heavy-duty models (i.e., buses and trucks) of these hybrids are available. Although not available in the United States, Hyundai (2009) has released the first light-duty LPG-electric hybrid vehicle—the Hyundai Elantra LPi hybrid. NG and LPG vehicle technology. CNG light-duty vehicles inject natural gas into spark-ignition engines, much like gas- oline engines do. Vehicles can be dedicated—that is, designed to rely solely on natural gas—or provide a bi-fuel configura- tion. The latter includes both a CNG and gasoline tank and User Group (2011) CNG LPG Federal agencies 4,548 75 State agencies 3,733 1,626 Electric power providers 2,313 289 Natural gas fuel providers 1,739 12 Propane fuel providers 16 1,129 Transit agencies 336 101 Other private fleets and municipal governments 53,295 73,415 Total 65,980 76,647 Source: EIA 2012a. Table B.1. Fleet users of light-duty CNG and LPG vehicles.

129 (EERE 2013a). Absent federal tax credits, it is more challeng- ing to recoup the price premium associated with natural gas vehicles through fuel-cost savings over time. B.3.2 Energy Cost of Travel While natural gas and LPG vehicles entail significant price premiums, they also provide operational savings over time. When compared on an energy-equivalent basis, CNG vehicles achieve roughly the same fuel economy as gasoline- powered vehicles. The base-model 2012 Honda Civic Sedan, for example, has an EPA rating of 39 mpg on the highway, while the 2012 Honda Civic natural gas model is rated at 38 mpg (Honda 2012). Yet natural gas is much cheaper than gasoline. As of July 2013, the average U.S. retail price for gasoline was $3.65 per gallon, while the average retail price for CNG was $2.14 per gge (EERE 2013b). Even with this differential, however, the high premium cost associated with CNG vehicles leads to long payback times for the average driver (MIT 2011). LPG is also less expensive than gasoline, costing an aver- age of $2.73 per gallon as of July 2013 (EERE 2013b). Yet LPG has a lower Btu (British thermal unit) rating than gasoline, potentially leading to lower fuel economy. Rather than signifi- cant fuel-cost savings, lower maintenance costs are perhaps the most important reason for LPG’s success in the market for high-mileage vehicles. LPG offers a combination of high octane, low carbon, and low oil-contamination characteristics that result in longer engine life in comparison to conventional gasoline engines. Because the fuel’s mixture of propane and air is completely gaseous, the cold-start problems associated with liquid fuels are also reduced (EERE 2012b). Looking forward, changes in the relative prices for gaso- line, natural gas, and LPG are likely to have a significant influence on the degree to which the latter two emerge as competitive alternatives within the light-duty vehicle fleet. The expectation of natural gas remaining reasonably in- expensive, for instance, could induce auto manufacturers to develop more CNG models in the coming years. As a point of reference, the EIA, in its reference-case scenario in the most recent Annual Energy Outlook, projects that the price of gaso- line will increase at an annual rate of 0.8% through 2040, that the price of CNG will increase at an annual rate of 0.9%, and that the price of LPG will increase at an annual rate of 0.4% (EIA 2013). Such long-term forecasts, however, are subject to great uncertainty. B.3.3 Operational Performance Comparing same-size engines, CNG engines generate slightly less power than gasoline engines, leading to slower acceleration and less power climbing hills (Yacobucci 2008, SVM engine conversion systems may be installed into a new or used vehicle. System installations are usually handled by the SVMs themselves or their qualified system retrofitters. From the 1970s to 1990s, unregulated conversion kits were available from dozens of manufacturers. Since the 1990s, all gaseous fuel engine systems on the market have been engi- neered and tested to comply with the EPA emissions stan- dards. Hence, all conversion kits are now tightly regulated. B.3 Cost and Performance Natural gas offers less-expensive per-mile energy costs compared to conventional vehicles and can potentially reduce greenhouse gases and local air pollutant emissions. Natural gas and LPG vehicles currently entail a significant price pre- mium, however, and natural gas faces challenges with respect to vehicle range and, to a lesser degree, power. B.3.1 Vehicle Cost The current cost differential between a natural gas or LPG vehicle and an otherwise comparable conventional vehicle is significant. The suggested retail price for Honda’s 2012 Civic Sedan, for example, starts at $15,955, while the suggested retail price for its Civic natural gas model begins at $26,305 (Honda 2012). The only LPG vehicles available recently have been light- and medium-duty Ford and GMC trucks retro- fitted with a conversion kit (EERE 2010). The cost of LPG conversion ranges from $4,000 to $12,000, depending on the make and model of vehicle (EERE 2012d). The conver- sion cost for a CNG vehicle is also significant; NGVA (2011) estimates the cost of converting a new light-duty vehicle to natural gas at between $12,000 and $18,000 with installation. Vehicle manufacturers do expect that the incremental cost for CNG vehicles should decline as sales volume increases, enabling greater economies of scale, though they could remain moderately more expensive than conventional vehicles. An important factor in the additional cost of a CNG vehicle is the compressed storage tank. The cost of an onboard steel CNG storage tank capable of storing 8 gge of CNG is about $1,250, and a carbon fiber tank can be more than $3,100 (NPC 2012). Some buyers of CNG vehicles may also choose to install a home refueling station, which can add another $4,000 to $5,000 to the cost of owning and operating a CNG vehicle. To partially offset such costs, the federal government and many states have offered tax credits or other incentives for natural gas and LPG vehicles. Consumers who purchased a CNG vehicle before the end of 2010, for example, could qual- ify for a federal tax credit of up to $4,000 (fueleconomy.gov 2012), and consumers who installed a home CNG refueling appliance by the end of 2013 qualified for a tax credit of up to $1,000 under the Alternative Fuel Infrastructure Tax Credit

130 mum efficiency, and maintain durability (M. J. Bradley and Associates 2005). B.3.4 Greenhouse Gas Emissions Compared to gasoline vehicles of comparable fuel effi- ciency, CNG vehicles offer about a 14% reduction in well-to- wheel carbon dioxide emissions (ANL 2012). A major concern, however, is that natural gas vehicles or upstream distribution infrastructure could leak methane, itself a greenhouse gas. Methane has a 34-times greater climate-warming impact than carbon dioxide, so even modest levels of leakage in the natural gas supply chain could undermine the GHG benefits of CNG in comparison to petroleum (Alvarez et al. 2012). To examine the potential climate implications of natural gas and other alternative-fuel vehicles in relation to conventional vehicles, the research team conducted an exercise relying on emissions factors from ANL’s GREET model to compare the performance of alternate fuel types and vehicle technologies. For each of the fuel types included, the team included GREET’s specifications for a generic mid-size passenger vehicle in 2010 along with two hypothetical configurations for the same size vehicle in the 2050 time frame. Both of the hypothetical future models are broadly consistent with expectations or forecasts reported in the literature, but one embeds modest base-case assumptions, while the other is more optimistic. Many of the same technologies that will improve ICE vehicles will also improve the fuel economy of CNG vehicles. To provide a bounding estimate for future improvements, the researchers used the maximum mpg considered by the National Petro- leum Council (NPC 2012) as an assumption for the 2050 base case. For the 2050 optimistic case, the researchers assumed a percentage-based improvement similar to that for the ICE optimistic case described in Appendix A. Table B.2 lists the assumed on-road (as opposed to EPA-rated) fuel economy for the current and future conventional and CNG vehicle specifi- cations used in the analysis. Based on these assumed specifications and GREET’s emis- sions factors, Figure B.1 graphs the GHG emissions perfor- mance for current and future conventional and CNG vehicles in grams of CO2-equivalent per mile. Note that the estimates are for well-to-wheels emissions, including extraction, process- ing, distribution, and combustion of the fuel within the vehicle. NRC 2013). LPG engines, on the other hand, have shown improved power performance. The recent introduction of sequential port-injection provides a great improvement over the older air-valve LPG fuel system. This product allows the engine control system designed for gasoline to distrib- ute LPG into an engine more precisely and uniformly. The result is an LPG engine with better power and fuel economy (WGA 2008). Although natural gas’s octane is higher than that of gaso- line (130 compared to 87 for unleaded gasoline), its storage density, even compressed, is much less than that of gaso- line. As a result, a CNG fuel tank needs to be much larger than a gasoline tank to accommodate the same range (EERE 2013c). In practice, natural gas vehicles may be configured with somewhat larger tanks and in turn offer modest reduc- tions in range. The tank in the 2012 Honda Civic NG, as already mentioned, displaces half of the trunk space and pro- vides for a range of 192 miles (EPA city rating) to 304 miles (EPA highway rating; NRC 2013). Note that EPA ratings are generally overoptimistic with respect to mileage in real- world driving conditions, so the actual range may be even less. Due to the limited range, CNG vehicles have often been used for applications in which vehicles are operated in local- ized areas and can easily return to a central base to refuel. The range of LPG vehicles is greater, at 300 to 400 miles (Yacobucci 2005). The performance of natural gas engines is sensitive to the specific composition of the natural gas. The fuel quality stan- dard for natural gas vehicles is to include a minimum of 95% methane. At the time of extraction, natural gas contains between 70% and 90% methane. The processed natural gas in pipelines typically consists of 85% to 99% methane, varying seasonally. Natural gas suppliers could upgrade the composition of the gas for transportation sector needs, but the extra purification could increase fuel price. Variations in fuel composition can upset the air-fuel ratio in the engine, leading to higher emissions and lower efficiency. Significant amounts of heavier petroleum gases, such as propane and butane, can increase the tendency for engine knocking, leading to loss of engine power and pro- gressive engine damage. In addition to the composition of the hydrocarbons, the water content, sulfur content, and residual compressor oil content influence the degree to which CNG engines are able to produce low emissions, operate at maxi- Vehicle Technology Current MPG 2050 MPG (Base) 2050 MPG (Optimistic) Gasoline ICE 24.8 72 91 CNG (gge) 25.6 69 93 Source: Computations by authors based on data from ANL (2012), NPC (2012), and NRC (2013). Table B.2. Assumptions in emissions comparisons for CNG vehicles.

131 B.4.2 Factors Affecting Market Prospects The remainder of this section discusses some of the main obstacles that CNG and LPG will need to overcome in order to achieve significantly greater market share in the light- duty fleet. Availability of refueling stations. The current lack of refu- eling stations stands as a major barrier to wide adoption of LPG and CNG vehicles. As noted earlier, there were fewer than 2,700 LPG stations in the country as of 2012, just over 1,100 CNG stations, and fewer than 60 LNG stations (EERE 2012e). While the number of LPG stations exceeds the number of refueling stations for any other type of alternative fuel, it still represents less than 2% of the more than 150,000 stations for conventional gasoline and diesel (API 2013). Further, many of the existing natural gas and LPG refueling stations are for dedicated fleets and are not accessible to the general public. The provision of CNG refueling infrastructure is particu- larly challenging financially given that CNG stations can cost three to four times as much as conventional gasoline stations (WGA 2008). The National Petroleum Council estimates that a dedicated CNG station costs about $1.5 million, while a modular CNG station connected to an existing gasoline station would cost around $400,000 (NPC 2012). Absent additional investment, however, the low availability of refu- eling stations dampens demand for vehicle adoption, and low vehicle market share in turn discourages investment in additional refueling stations. In light of this problem, natural gas and LPG are perhaps most feasible, at least in the near term, for vehicles that are centrally garaged, are operated in a well-defined geographical area, and can be fueled at a single location or at a limited network of stations. The most cost-effective markets for CNG and LPG vehicles are in high fuel-use fleets such as transit, airport, taxi, shuttle, munici- pal, refuse, ports, and delivery and distribution. Long-haul trucks, which can refuel at stations located strategically along B.3.5 Local Air Pollutant Emissions All internal combustion engines, including those operating on CNG and LPG, emit ozone-forming compounds such as NOx and non-methane VOCs—termed “criteria pollutants” and regulated by the EPA. LPG and natural gas vehicles, how- ever, generally produce less air pollutants than gasoline vehi- cles when compared on a well-to-wheels basis. CNG vehicles, for example, emit 45% less VOCs, 12% less PM2.5, and 27% less NOx emissions than otherwise comparable gasoline- fueled ICE vehicles (ANL 2012). B.4 Market Prospects The potential advantages of LPG in comparison to gaso- line- or diesel-fueled vehicles are rather modest, with little evidence to suggest that the market share for LPG is likely to expand significantly in the coming decades. In contrast, CNG appears to hold greater promise due to its lower cost and moderate emissions improvements. For CNG to succeed, however, it will need to overcome a number of challenges, such as the lack of adequate refueling infrastructure, the cur- rent high price premiums for CNG vehicles, and competing uses for natural gas in other sectors. B.4.1 Future Market Projections In the reference case for its Annual Energy Outlook, EIA (2013) projects that CNG and LPG vehicle sales will expand modestly in the coming decades, rising to 32,000 and 56,000 vehicles per year, respectively, by 2040. This would represent less than a half of a percent of all light-duty annual vehicle sales. In contrast to EIA’s rather tepid forecast, OEMs—encouraged by dramatic declines in the cost of natural gas—have in recent years begun to introduce new CNG models. If this trend holds, CNG adoption might easily exceed EIA’s current projections. Source: Computations by authors based on data from ANL (2012), NPC (2012), and NRC (2013). 0 100 200 300 400 500 CNG Gasoline ICE Grams CO2 Equivalent per Mile of Travel (well to wheels) 2010 2050 Base 2050 Opmistic Figure B.1. GHG reduction prospects for CNG in 2050.

132 most frequently requested, high-sales fleet vehicles. This in turn limits the variety of vehicle models available to the public (WGA 2008). The year before the new EPA regula- tion addendum took effect, about 14,000 alternative-fuel conversions were accomplished. In the year the new policy was enacted, the number of conversions of all alternative-fuel vehicles dropped to 8,500 and then steadily declined in sub- sequent years to around 1,000 conversions per year (Mokh- tarian and Cao 2004). Consequently, LPG vehicles lost more than half of their market share between 1995 and the middle of the first decade of the 21st century. The decline in LPG vehicle ownership may be due to the lack of vehicle choice, but it is also possible that the limited availability of options reflects market rejection of LPG vehicles. In the state of California, CARB certification in addition to EPA certification is required for all conversions, increas- ing the financial cost for conversion manufacturers and for buyers. Most fuel system providers are relatively small and cannot finance both the EPA and CARB certifications. Most companies, therefore, choose to obtain only the EPA certi- fication to gain market access in 49 states, foregoing CARB certification and the California market. Hence, adoption of LPG and CNG vehicles through conversions in the state of California is a difficult strategy (WGA 2008). It is still possible that more OEMs could begin to offer CNG and LPG models. As described earlier, OEMs began to offer natural gas models in the early 1990s and LPG models in the late 1990s. Market offerings increased through the early 2000s but then entered a period of decline. In just the past few years, however, stimulated by rising oil prices as well as the increased supply and decreased price of natural gas result- ing from horizontal drilling techniques, the number of OEM CNG models, at least, appears to be on the rise once again, as shown in Figure B.2. This could reflect an expectation on the Interstate highway system, are another promising market for natural gas (WGA 2008). One option for expanding refueling infrastructure is to build partnerships between large anchor fleets and fuel providers that construct new fueling stations to service the anchor fleets. Many natural gas providers will build a station at no cost to the customer in return for a minimum fuel volume contract. A commitment of 250,000 gallons per year is generally adequate for fuel providers to construct a new refueling station. The sta- tion can be opened to the public, thereby increasing the avail- able refueling infrastructure to local, smaller fleets and private vehicle owners. The development of public infrastructure may encourage greater adoption of natural gas vehicles and LPG vehicles, especially by smaller fleets that cannot afford their own refueling station (WGA 2008). Cost and availability of CNG and LPG vehicle models. The currently limited number of OEM CNG and LPG models, the relatively limited number of conventional vehicle models that can be converted to CNG or LPG, and the high price pre- mium associated with both new vehicles and conversion kits have the effect of limiting adoption of CNG and LPG. From the 1970s to 1990s, universal conversion kits allowed the conversion of a greater selection of vehicles at a lower cost than today. In 1997, an addendum to the Clean Air Act raised the testing requirements for CNG and LPG conver- sions and rendered universal kits impermissible. The tighter regulations, while protecting the environment, also made con- versions more difficult and costly. Under the more-stringent regulations, conversion kits for different engine families must be tested and certified separately. For instance, a conversion company must design and manufacture a unique conversion kit to certify the 2008 model year 4.6L V8 Ford engine family. Due to the high investment cost, small-volume manu- facturers that develop conversions focus their efforts on the Source: EERE (2012a). 0 2 4 6 8 10 12 14 16 18 20 1995 2000 2005 2010 Av ai la bl e O EM M od el s CNG LPG Figure B.2. Available OEM CNG and LPG light-duty models, 1992–2012.

133 Unresolved environmental concerns about fracking. Technological and economical improvements in horizon- tal drilling with hydraulic fracturing to further expand the economically recoverable supply of natural gas could rein- force the cost advantages of natural gas over petroleum. It is unclear, however, whether the environmental impacts of this extraction technique could constrain its future applica- tion. As previously discussed, there are remaining concerns about the effects on surrounding groundwater supplies (e.g., Osborn et al., 2011). While some experts are cautious in their views about the long-term production potential from uncon- ventional sources, others believe that as much as 64% of U.S. natural gas production could come from unconventional sources by 2020 (Vidas and Hugman 2008). Competing uses for natural gas. As noted earlier, natural gas is already used in many other applications, such as power generation, industrial processes, and residential heating and cooking. Depending on available supply, its use in transpor- tation could hinge on its relative value across these different possible uses (NRC 2013). Potential for biomethane. Gathering natural gas from landfills, animal waste, sewage, and other renewable resources could provide an additional supply of natural gas and remove a greenhouse gas (methane) that otherwise would have been released into the atmosphere. The crude natural gas collected from such sources, typically referred to as biogas, can be puri- fied and refined into biomethane to produce fuel for vehicles. The production technology to collect, process, and distribute biomethane is currently more established in Europe and else- where than in the United States. Iceland operates 100% of its natural gas vehicle fleet using biomethane, and Switzerland and Sweden power more than 50% of their natural gas vehi- cles with biomethane (WGA 2008). In Austria, the number of biomethane production plants has grown from approxi- mately 50 in 1999 to 350 in 2008 (Baumgartner, Kupusovic, and Blattner 2010). As part of an economic recovery package, the European Commission invested five billion euros in the European Green Cars initiative, and biomethane for trans- portation was one of the main research topics (EC, undated). Only a few projects to process biogas from landfills for trans- portation currently exist in the United States, mostly in Cali- fornia (Lear 2008). As the European examples show, lack of technical capabilities is not a barrier to the increased pro- duction of renewable natural gas. The Department of Energy estimates that the United States could produce enough bio- methane to replace 10 billion gallons of gasoline annually, given supportive economic and regulatory conditions, from such sources as landfills, animal waste processing, and sewage (WGA 2008). Potential for methanol. Another technical option receiv- ing some attention is to convert natural gas into methanol, which can then be used to power existing flex-fuel vehicles the part of auto manufacturers that the current low cost of natural gas will persist, leading to increased adoption pros- pects and in turn setting the stage for even more OEM models being offered in the future. Performance limitations. CNG vehicles face a range limi- tations in comparison to conventional vehicles, and this could deter some consumers from choosing natural gas. With CNG having lower energy density than gasoline, the vehicles must carry larger fuel tanks, in turn reducing cargo space. These problems could be addressed in the future by producing hybrid-electric versions of the vehicles, which would allow for greater range as well as a smaller fuel tank (WGA 2008). To remedy the short range of CNG vehicles and the safety concerns associated with highly pressurized CNG fuel tanks, the U.S. Department of Energy set a goal to develop a tech- nology that could store as much natural gas at the lower pressure of 500 psi as can be stored in a conventional CNG tank at 3,600 psi. This goal was achieved in 2007 by a team of researchers that developed carbon briquettes with nanopores capable of storing natural gas (NSF 2007). Once integrated with vehicle manufacturing, such technology could provide safer, lighter, and smaller fuel tanks. Other novel materials capable of holding an even greater density of natural gas and extending the range of CNG vehicles are also possible in the future (NRC 2013). Another performance issue for CNG vehicles is power. The acceleration power of the Honda Civic GX, for example, is less than that of its gasoline counterpart due to lower horse- power (113 hp versus 140 hp) and greater weight (2,910 lbs versus 2,652 lbs). The heavier weight of the GX results from the larger and heavier CNG fuel tank (Alapati 2010). Here again it is possible that this issue could be at least partially addressed through the addition of hybrid-electric vehicle technology. Perceived safety concerns. LPG and CNG are safer to use than gasoline in some respects. The ignition temperature of LPG is higher than that of gasoline, and the concentra- tions for both LPG and natural gas have to be greater than for gasoline before the fuels ignite. The dissipation rate of natural gas is also faster than that of gasoline. Still, public perceptions of the hazards of natural gas and LPG tanks and refueling stations could hinder broader adoption of these technologies. The most significant safety difference between these alternative fuels and gasoline is that the latter is stored at refueling stations and on vehicles as a liquid at atmospheric temperature and pressure. In contrast, LPG and CNG are stored under pressure. CNG, stored under high pressure at 2,000 psi or more on light-duty vehicles, presents the greatest concerns. Valve or tank failure at this pressure could result in an explosion, while a gas leak in a confined space, such as a garage, could lead to an accumulation of gas at a dangerous level of concentration (WGA 2008).

134 EERE. 2012d. Propane Vehicle Conversions. Alternative Fuels Data Cen- ter. http://www.afdc.energy.gov/vehicles/propane_conversions. html (accessed September 21, 2012). EERE. 2012e. Alternative Fueling Station Total Counts by State. Alterna- tive Fuels Data Center. http://www.afdc.energy.gov/fuels/stations_ counts.html (accessed September 20, 2012). EERE. 2012f. Natural Gas. Alternative Fuels Data Center. http://www. afdc.energy.gov/vehicles/natural_gas.html (accessed September 21, 2012). EERE. 2013a. Federal Laws and Incentives. Alternative Fuels Data Center. http://www.afdc.energy.gov/laws/ (accessed September 11, 2013). EERE. 2013b. Fuel Prices. http://www.afdc.energy.gov/fuels/prices.html (accessed September 12, 2013). EERE. 2013c. Natural Gas Benefits and Consideration. Alternative Fuels Data Center. http://www.afdc.energy.gov/fuels/natural_gas_ benefits.html (accessed September 12, 2013). EIA. 2012a. Renewable and Alternative Fuels: Alternative Fuel Vehicle Data. http://www.eia.gov/renewable/afv/users.cfm (accessed Sep- tember 12, 2013). EIA. 2012b. Natural Gas Explained: Use of Natural Gas. http://www.eia. gov/energyexplained/index.cfm?page=natural_gas_use (accessed September 20, 2012). EIA. 2012c. U.S. Crude Oil, Natural Gas, and NG Liquids Proved Reserves: With Data for 2010. EIA. 2012d. Natural Gas: U.S. Dry Natural Gas Proved Reserves. http:// www.eia.gov/dnav/ng/hist/rngr11nus_1a.htm (accessed September 20, 2012). EIA. 2013. Annual Energy Outlook 2013. ETSAP. 2009. Liquid Petroleum Gas and Natural Gas Internal Combustion Engines. Fueleconomy.gov. 2012. Federal Tax Credits for Alternative Fuel Vehicles. http://www.fueleconomy.gov/feg/tax_afv.shtml (accessed Septem- ber 21, 2012). Honda. 2012. To Each Their Own Civic. http://automobiles.honda. com/civic/ (accessed September 21, 2012). Hyundai. 2009. “Hyundai Elantra LPi Hybrid Official Press Release” IEA. 2012. World Energy Outlook 2012. Lear, J. 2008 (November 20). “NGV Technology Efforts: Biogas for Transportation.” Presented at the Natural Gas Vehicle Technology Forum, Downey, California. Litovitz, A., A. Curtright, S. Abramzon, N. Burger, and C. Samaras. 2013. “Estimation of Regional Air-Quality Damages from Marcellus Shale Natural Gas Extraction in Pennsylvania.” Environmental Research Letters, 8 (1): 014017. MIT. 2011. The Future of Natural Gas. MIT Energy Initiative. M. J. Bradley and Associates. 2005. Natural Gas as a Transportation Fuel: Best Practices for Achieving Optimal Emissions Reduction. Northeast States Center for a Clean Air Future and International Council on Clean Transportation. Mokhtarian, P. L. and X. Cao. 2004. The Future Demand for Alternative Fuel Passenger Vehicles: A Diffusion of Innovation Approach. California Department of Transportation. NGVA. 2010. Guide to Available Natural Gas Vehicles and Engines. NGVA. 2011. Fact Sheet: Converting Light-Duty Vehicles to Natural Gas. NPC. 2007. Meeting the Challenges of the Nation’s Growing Natural Gas Demand. NPC. 2012. NPC Future Transportation Fuels Study: Advancing Technology for America’s Transportation Future. NRC. 2013. Transitions to Alternative Vehicles and Fuels. The National Academies Press, Washington, D.C. with a modest investment in an air/fuel mixture control and an alcohol sensor. The additional vehicle expense is esti- mated at between $100 to $200. The equivalent of a gallon of gasoline in methanol can be produced for about $2.00 when natural gas costs $8 per million Btus (MIT 2011). Natural gas is currently priced well below that figure, suggesting that methanol could be a cost-effective alternative to gaso- line. Methanol can also be blended with gasoline for use in vehicles, similarly to the way ethanol is blended with gaso- line. Despite the favorable economics, methanol demand has remained low in recent years, in part due to the phaseout of methyl-tertiary butylether (MTBE), a fuel additive aimed at increasing fuel octane levels that was linked to groundwater contamination issues. MTBE is derived from methanol and butane. Methanol also has about half the energy content of traditional gasoline, so the range of a tank of 100% methanol fuel is only about half that obtained from a similarly sized tank of gasoline. Nevertheless, it has been suggested that methanol represents a promising alternative fuel with the potential to fuel a large portion of the U.S. transportation fleet (MIT 2011). References Alapati, J. 2010 (January 18). “Car Review: Honda Civic GX.” Reporter mag.com. Alvarez, R. A., S. W. Pacala, J. J. Winebrake, W. L. Chameides, and S. P. Hamburg. 2012. “Greater Focus Needed on Methane Leakage from Natural Gas Infrastructure.” Proceedings of the National Academy of Sciences, 109 (17): 6435–6440. ANL. 2012. The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET), Fuel Cycle Model version 1 2012. U.S. Department of Energy. API. 2013. Oil & Natural Gas Overview: Service Station FAQs. http://www. api.org/oil-and-natural-gas-overview/consumer-information/ service-station-faqs (accessed July 20, 2013). Baumgartner, B., M. Kupusovic, and H. T. Blattner. 2010. National Report on Current Status of Biogas/Biomethane Production – Austria. Intelligent Energy - Europe Programme. BP. 2012. BP Statistical Review of World Energy 2012. BTS. 2013. Table 1-11: Number of U.S. Aircraft, Vehicles, Vessels, and Other Conveyances. National Transportation Statistics. http:// www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/publications/ national_transportation_statistics/html/table_01_11.html (accessed September 6, 2013). EC. Undated. Green Cars. http://ec.europa.eu/research/industrial_ technologies/green-cars_en.html (accessed September 25, 2012). EERE. 2010. Clean Cities 2010 Vehicle Buyer’s Guide. EERE. 2012a. Light-Duty Vehicle Search. Alternative Fuels Data Cen- ter. http://www.afdc.energy.gov/vehicles/search/light/ (accessed September 21, 2012). EERE. 2012b. Propane Benefits and Considerations. http://www.afdc. energy.gov/fuels/propane_benefits.html (accessed September 21, 2012). EERE. 2012c. Propane Fuel Basics. Alternative Fuels Data Center. http://www.afdc.energy.gov/fuels/propane_basics.html (accessed September 21, 2012).

135 Vidas, H. and B. Hugman. 2008. Availability, Economics, and Production Potential of North American Unconventional Natural Gas Supplies. The INGGA Foundation. WGA. 2008. Transportation Fuels for the Future: Natural Gas and Propane. Yacobucci, B. D. 2005. Alternative Transportation Fuels and Vehicles: Energy, Environment, and Development Issues. Congressional Research Service. Yacobucci, B. D. 2008. Natural Gas Passenger Vehicles: Availability, Cost, and Performance. Congressional Research Service. NSF. 2007. “From Farm Waste to Fuel Tanks.” Press release. O’Dell, J. 2011. “BRC Fuel Maker Again Selling Phill Home CNG Fuel Station.” http://www.edmunds.com/autoobserver-archive/2011/03/ brc-fuelmaker-again-selling-phill-home-cng-fuel-station.html. Osborn, S. G., A. Vengosh, N. R. Warner, and R. B. Jackson. 2011. “Methane Contamination of Drinking Water Accompanying Gas- Well Drilling and Hydraulic Fracturing.” Proceedings of the National Academy of Sciences, 108 (20): 8172–8176. PGC. 2009. Potential Supply of Natural Gas in the United States. TCPA. 2008. Chapter 5: Natural Gas. In The Energy Report 2008.