On-road Testing and Characterization of Fuel Economy of Light-Duty Vehicles

The potential discrepancy between the fuel economy shown on new vehicle labels and that achieved by consumers has been receiving increased attention of late. EPA has not modified its labeling procedures since 1985. It is likely possible that driving patterns in the U.S. have changed since that time. One possible modification to the labeling procedures is to incorporate the fuel economy measured over the emission certification tests not currently used in deriving the fuel economy label (i.e., the US06 high speed and aggressive driving test, the SC03 air conditioning test and the cold temperature test). This paper focuses on the US06 cycle and the possible incorporation of aggressive driving into the fuel economy label. As part of its development of the successor to the MOBILE emissions model, the Motor Vehicle Emission Modeling System (MOVES), EPA has developed a physically-based model of emissions and fuel consumption which accounts for different driving patterns. This model could be used to analyze surveys of U.S. driving behavior, as well as that represented by available driving cycles, and derive a weighting of the driving cycles which best represents current on-road driving behavior. In 2001, the U.S. Environmental Protection Agency (EPA) conducted a pilot study of real-world emissions and fuel economy using a SEMTECH-G Portable Emissions Measurement System (PEMS). Second-by-second measurements of emissions were obtained from 15 vehicles. Various other operating parameters were also recorded. Fuel economy can be calculated from carbonaceous emissions. Most of these vehicles were also tested on a dynamometer over the FTP and US06 cycles. Since the data set is limited in the number of tests conducted, this analysis is best seen as a pilot study. In this paper, we apply the MOVES fuel consumption model to the driving patterns measured in the 2001 PEMS study and the FTP, HFET and US06 cycles. We derive weights for these three cycles which best represent the driving activity of each vehicle. We then model the on-road fuel economy based on fuel economies measured over the three driving cycles and compare them to the measured on-road fuel economies.