Over the next several decades, countries transitioning from gasoline and diesel to alternative fuels will need to invest heavily in new refueling station infrastructures. Refueling stations will need to be located in places convenient for early adopters of alternative-fuel vehicles (AFVs). What constitutes "convenience", however, has been a matter of debate. In particular, two groups of optimization models used to deploy new stations have taken divergent views of convenience. One group aims to minimize distance traveled from homes to stations, while the other locates stations "on the way" -- on the routes people regularly travel from their origins to their destinations. Which approach best fits the actual behavior of AFV drivers is unknown, however. For insights into refueling behavior, researchers have looked to two studies from 1983 and 1984. Given the paucity of AFVs on the road at that time, these studies used diesel drivers as a proxy for the refueling patterns that early AFV drivers might exhibit given the relative scarcity of alternative-fuel stations. Patterns of refueling stations for contemporary alternate-fuel drivers are much sparser than the diesel-fuel networks of the 1980s, so new models are needed that are better suited to current conditions. This research project will survey drivers about their refueling patterns and behaviors to better understand the assumptions underlying the deployment of new stations. Drivers will be interviewed at stations while they refuel. The survey will ask for the geographic details of their current trip. Socioeconomic and demographic data will be gathered and drivers will be asked whether they detoured from their desired route, why they chose this station, and other questions. To investigate the effects of station site and situation, a number of AFV stations and a control group of gasoline stations will be sampled to represent a variety of geographic settings, including urban, suburban, interstate highways, and arterial streets. Geographic information system-based analyses will be used to measure distances from home, deviations from shortest paths, and geographic characteristics of station-service areas.
This project will improve understanding of evolving travel behavior for which past studies do not provide sufficient information or insights. Research conducted nearly 30 years ago will be updated and expanded to examine behaviors of actual alternative-fuel vehicle drivers. The project will contribute to the literatures on detouring behavior, distance decay, retail service areas, and adoption of AFVs, and it provide a much-needed empirical foundation to theoretical models for optimal station location. The project will contribute to wider-ranging efforts to combat global climate change and local air pollution by helping to overcome the foremost barrier to the use of alternate-fuel vehicles, namely the lack of a refueling infrastructure. Understanding where and when drivers refuel will enable industry to locate stations more efficiently. The U.S. will benefit by saving money on stations, increasing adoption of AFVs, reducing dependency on foreign oil, and developing green jobs and clean industries in the U.S.
Over the next several decades, countries transitioning from gasoline and diesel to alternative fuels will need to invest heavily in new refueling stations. It is especially important that stations be located conveniently for early adopters of alt-fuel vehicles (AFVs). What constitutes "convenience," however, is a matter of debate, with little evidence to back up the assumptions. In particular, two classes of optimization models for deploying new stations take divergent views of convenience. One group minimizes distance traveled from homes to stations, while the other locates stations "on the way" from origins to destinations. The objective of the research, therefore, was to survey drivers about their refueling patterns to better understand the assumptions underlying infrastructure planning. Our study was designed to determine what drivers consider convenient based on their revealed rather than stated preference, i.e., not by what they say they want but by what they actually did. We interviewed over 500 drivers while they refueled at 10 stations in Greater Los Angeles, and asked, among other questions, where their trip started, where they are going, and where they live. We studied 5 compressed natural gas (CNG) stations and 5 nearby gasoline stations as a control group. GIS was used to estimate their routes driven, distances from home, and deviations from shortest paths to refuel (Figure 1). We published two refereed papers in respected scientific journals, and are nearly finished a third. We presented our research at six conferences, five universities, and two infrastructure companies. We confirmed our most important hypothesis that, when AFV drivers are faced with a choice between a station closest to home vs. a station most on their way (shortest detour), they more frequently refuel along their way (Figure 2). We isolated those drivers who did not have a station that satisfied both criteria. Drivers facing such a choice chose the station farther from home but more on their way by more than a 10:1 margin (Table 1). Deeper analysis of drivers who chose a station neither closest to home nor most on their way showed that they more often chose the station with the second or third shortest detour rather than the second or third closest to home (Table 2). This result has important implications for planning infrastructure networks. Our research showed that actual AFV drivers, faced with a sparse refueling network of stations mostly located at fleet depots (a very common strategy in the early stages of commercialization), are willing to refuel far from home and mid-route if it’s not too far out of their way. This means that for the early stage of commercialization, stations should be placed at major funnel points of our highway systems so they can be more or less on the way for as many trips as possible. This is a counter-intuitive result for many people because, given the widespread gasoline station network today, it's almost always possible to find a gasoline station that is both close to home and on the way. This will not be possible in the early stages when there will be only a few stations to serve an entire metropolitan region (e.g., 7 CNG stations for 4 million people in Greater Phoenix in 2014). With very few stations, most drivers will not have a station close to home—but many could have one on their way if the same number of stations are located near the key highway interchanges through which hundreds of thousands of drivers may go every day. A second paper summarized the differences between the refueling behavior of CNG and gasoline drivers, attributable mainly to how drivers adapt to the lack of stations and shorter vehicle driving range. CNG drivers refuel farther from home, with fuller tanks, and more often mid-route. They refuel more on work trips and detour 5-6 minutes from shortest paths, compared with 1-2 minutes for gasoline drivers (Figure 2). A third paper being finalized studies the behavior of commercial AFV drivers. While many drivers of buses, shuttle vans, taxis, delivery or repair vehicles, and trucks have a fuel station at their home fleet base, they also refuel at other fleets’ bases, and smaller companies usually don’t a station at their facility. Little is published about this unstudied aspect of early-stage AFV commercialization, yet is of great importance to supplying the base demand for the early-stage stations. The lack of a refueling infrastructure has consistently been identified as the largest barrier to the transition to alternative fuels. It is half of a classic chicken and egg dilemma facing car manufacturers and energy companies. Understanding where drivers refuel enables industry to locate stations more efficiently. The US will benefit by saving money on stations, increasing adoption of AFVs, reducing our dependency on foreign oil, improving local air quality, slowing global warming, and developing green jobs here in the US.
