Estimating greenhouse gas emissions of electric delivery trucks

In this study, we investigate the regional differences in emission benefits of batteryelectric delivery truck electrification. In this regard, we build a simulation frameworkto quantify the regional differences in the use phase emissions across the UnitedStates. A vital part of our framework is the machine learning model to predict theunit energy consumption of a battery electric delivery truck which is based on realworld driving data. Using our framework, we perform two case studies to quantifythe effect of ambient temperature and driving profile on the use phase emissions,respectively. In the first case study, we observe that our machine learning model cancapture the increase in energy consumption at low temperatures quite well, howevermore data is needed to predict high temperature effects. As expected, the emissionsare lower in regions where electricity production is cleaner. In the second case study,we observe that our framework can differentiate between the energy consumptionunder aggressive and gentle driving profiles.