In this post, I am going to briefly talk about how the transportation energy consumption varies across United States. Much of the energy in that sector is from liquid fuels and the transmission and distribution network of these fuels have enormous land use and transportation implications. If we were to think about changing the fuel mix of the transportation fleet (say for example making hydrogen fuel cells or electricity more prominent), we need to think through its implications.
In the United States (US), the transportation sector consumes about 29% of the total energy in 2017, rising from 23.5% in the 1960s even while the energy efficiency of the economy increased. Much of this energy comes from liquid carbon-based fuels contributing to greenhouse gas emissions and bad air quality. While we know a bit about how much energy is consumed in the transportation sector nationally, we don’t really have a grasp on how various local and state policies (incluidng the design, form and function of cities and regions) are affecting transportation energy consumption.
Three components of energy consumption
The three main components that explain energy consumption in the sector is the total volume of travel, mode split and average energy efficiencies of the fleet mix. Volume of travel depends on the distances between origins and destinations, network effects and trip frequencies. Mode split, the transportation mode that trip uses, is dependent on distances that need to be travelled, the availaibility and ease of mode not just for the particular trip, but for subsequent and prior chained trips and trip purposes (e.g. delivery of goods, shuttling kids). The energy efficiency of the fleet mix depends on adoption rates of newer vehicles, organisational/household acquisition rules and procedures and government policies on pollution control mechanisms, tax subsidies, depreciation rules etc.
In addition to these complicated mechanisms through which we can influence energy consumption, we should also be mindful of the different components of energy consumption. We need to account for transportation consumption both in freight and in person travel. Light duty vehicles with short wheel bases (passenger cars, vans, SUVs etc.) only accounted for 52% of highway transportation energy in 2016 (Bureau of Transportation Statistics 2018 Table 4.06). Assuming other vehicle types are largely associated with non-household travel, relying simply on household travel underestimates the total energy consumption in the system and the impact of urban form.
Sales in Gas stations
Sub national information is not easily forthcoming for transportation sector. Since we can not simply proxy total energy consumption from VMT derived from household travel surveys, we need an alternative strategy. Because most of energy is purchased at retail gas stations in the US, we can study the geogrpahic differences in energy consumption, by studying the variation in gas station sales.
I use the 2012 Economic Census by the US Census Bureau to construct a proxy for transportation energy consumption, by analyzing the sales at gas stations in each county (or equivalent areas) of the United States. The Bureau collects extensive data on businesses every 5 years. I use the data is reported at a county level for the retail sector (North American Industrial Classification System (NAICS) code 44-45), in particular, the total sales receipts from the gas stations (NAICS 447). Due to confidentiality concerns, data for 344 counties are not reported.