Quantification and Evaluation of Onroad Light Duty Vehicle Tailpipe Emissions

Emissions from onroad vehicles are a major source of air pollution in the United States. Vehicle emission factors can be quantified by model estimates and empirical measurements. Model estimates need to be evaluated based on empirical data for its accuracy and sensitivity on emissions. To get accurate emissions from model estimates, vehicle activity is an important input and needs to be quantified as the performance envelope of acceleration and speed. Because of the large sample size of the empirical data and MOVES computational intensity, and simplified version of MOVES is needed to evaluate the sensitivity of factors that affect emissions, and to couple with traffic simulation model.;The MOter Vehicle Emission Simulator (MOVES) is an onroad mobile source emission model developed by the U.S. Environmental Protection Agency (EPA). EPA has evaluated MOVES based on limited data not used in model development. To evaluation MOVES sensitivity with independent empirical data, emission were measured in-filed for 100 light duty vehicles son multiple routes using a Portable Emissions Measurement System (PEMS). For each vehicle, modal fuel use and emission rates are estimated based on ranges of Vehicle Specific Power (VSP). The VSP modes are weighted by time spent in each mode for multiple real-world driving cycles that represent different mixes of arterial and highway driving for cycle averages speeds from 20 to 65 mph to represent cycle average emissions rates. Empirical onroad emission measurements of 100 vehicle also recorded vehicle type, age, driving cycles, and ambient conditions, which are input to MOVES estimations. Comparing empirical emission factors and MOVES estimates, MOVES estimates of cycle average CO2 emission rates are sensitive to vehicle type, cycle average speed, and road types, but not model year or age, and the trends are consistent with empirical data. MOVES emission rates for NOx, CO, and HC are sensitive to vehicle type, model year, age, and cycle average speed, as are the empirical data.;MOVES can be used to couple with travel demand models (TDMs) and traffic simulation models (TSMs) for the purpose of estimating emissions impacts of possible future changes in road infrastructure, vehicle mix, traffic control measures, and other factors. However, MOVES is computationally intensive, and direct dynamic coupling of MOVES to a TDM or TSM can be impractical. To facilitate the capability to estimate link-based emission factors based on second-by-second vehicle speed trajectories, a simplified version of MOVES is demonstrated here. A Cycle correction factor (CCF) is generated for a selected vehicle type and driving cycle based on distribution of time spent in each of 23 operating mode bins. Operating modes are defined by the instantaneous speed and Vehicle Specific Power (VSP). The emission factors estimated by the simplified model are demonstrated to be sensitive to differences between driving cycles with similar average speeds. The errors of the simplified model cycle average predictions are within +/-1% for 92% of the cases among pollutants, ages, and driving cycles, for passenger cars, passenger trucks, light commercial trucks, single unit short haul trucks, and combination long haul trucks. The application of the simplified model is demonstrated based on empirical driving cycles observed from field measurements.;After the simplified version of MOVES has been updated to taking account corrections of ambient conditions including temperature and humidity, variability in light duty vehicles emissions based on vehicle type, age, driving cycles, and ambient conditions is assessed using the simplified version of MOVES as a high throughput tool (HTT). To evaluate the separate and interactive impact of the sources of variability, case studies with combinations of two vehicle types (passenger car and passenger trucks), two model year (vehicles subject to Tier 1 and Tier 2 emission standards), and three ambient conditions ( low, medium and high temperature) are compared based on 591 empirical driving cycles of each case.;Empirical driving cycles including second by second speed, acceleration, and road grade were measured in RTP, NC area for 100 light duty vehicles on both freeways and nonfreeways. A multivariate rank regression based method (SRRC) is used in the analysis for case study based on empirical vehicle type, age, driving cycles, and ambient conditions for the interactions of factors. Based on the results, emission factors are most sensitive to temperature, cycle average speed, and standard deviation of speed.;The overall contributions of the research is to provide a large sample size of empirical emission data and vehicle activity data to evaluate model estimates, and to develop a simplified emission model that can runs much faster for evaluation the sources of variability such as vehicle type, age, model year, driving cycles, and ambient conditions. (Abstract shortened by UMI.).