Identifying the effect of vehicle operating history on vehicle running emissions

This research aims to determine the overall effects that a vehicle's short-term operating history has on its subsequent emissions, and how Vehicle Specific Power (VSP)-based vehicle emission models could be improved to account for these effects. Current VSP-based emission models, such as the U.S. EPA's MOtor Vehicle Emission Simulator (MOVES), only predict emissions based on instantaneous vehicle activity and the corresponding VSP value; the past short-term vehicle operational activity is not taken into account (e.g., the last 10-20 s of operation). For example, instantaneous vehicle emissions could be affected by a hard acceleration vs. a deceleration event at that particular point in time. This paper attempts to determine the accuracy of VSP-based emission models, which may suffer due to the fact that the history effects are being overlooked. A number of experiments were carried out in order to determine the anomalies resulting from instantaneous estimation as opposed to taking short-term vehicle operating history into account. These experiments compare model estimates with actual emission measurements. A quantitative analysis shows that the higher power operating modes (such as modes 33, 35, 37, 38, and 40 in MOVES) had the greatest variability - sometimes in the range of 60-100% - due to the effects that vehicle operating history has on carbon monoxide (CO). Hydrocarbons (HC) in higher power operating modes also vary 40-60% depending on the driving cycle. For lower power operating modes (e.g.. MOVES modes 1-30), the uncertainty for all pollutants was significantly less. It was also established that the carbon dioxide (CO2) and nitrogen oxide (NOx) estimations conducted by MOVES were least affected by the vehicle operational history effects compared with other emissions. As such, MOVES emission results are more accurate for mild to normal driving cycles, but there is greater uncertainty for higher power driving cycles. (C) 2012 Elsevier Ltd. All rights reserved.