The proposal seeks the development of a consistent and well-detailed methodology for the determination and modeling of vehicular emissions on different temporal and vehicular scales. Successful development of such a methodology would be a step forward for planners and regulators who often struggle as to which currently available model may most accurately predict local/regional vehicle emissions. The PIs have a well-documented history in the field of on-board vehicular measurements, meshing the fields of environmental and transportation engineering. The project appears well-organized and the PIs have carefully explained their experimental, modeling, and dissemination plan, including details on planned statistical analysis and a fairly in-depth description of the dissemination plans. The listed resources at the home institution seem more than sufficient to accomplish the proposed tasks. The proposed work is well organized and represents a valuable step forward in the ultimate quantification of vehicular emissions. The budget seems appropriate and sufficient. The FUEMS methodologies proposed not only advance the state-of-knowledge, but the very project plan enhances training and learning by the implicit inclusion of a number graduate and undergraduate students in the raw research, and further student exposure by inclusion of breaking data, and the process of data reduction, in various courses. The results are to be further disseminated through webpages, local political organizations, and presentations and publications via professional organizations. The recruitment of students into the program seems to be mostly in-house, but will be advertised through the school's minority program offices.
We have demonstrated a robust methodology for quantifying the real world in-use activity, fuel use and emission rates of light duty vehicles, taking into account a wide range of factors that affect speed, acceleration, and road grade, including road types, traffic controls, speed limits, traffic conditions, and routes. Using Portable Emission Measurement Systems (PEMS), we have collected data for a large number of light duty gasoline vehicles using the same study design and data collection and analysis procedures, demonstrating that the procedures are robust to inter-vehicle variability. We have established a repeatable and reliable set of procedures for reviewing, time-aligning, correcting, and removing inaccurate real-world raw data to produce quality assured data from multiple instruments from which a vehicle emissions "fingerprint" can be developed, in terms of modal emission rates for each of 14 vehicle specific power (VSP) modes. The use of VSP enables cycle average emission rates to be estimated for any driving cycle, not just those cycles measured for a particular vehicle on a given day, route, and traffic condition. The use of a consistent 14 mode VSP-based approach to estimating fuel use and emission rates from real-world data enables comparison of inter-vehicle variability controlled for similar engine load, and for comparisons to be made across vehicles for any user-supplied driving cycle. Conversely, the VSP-based approach allows the observed driving activity for data collection of a given vehicle to be compared to the driving activity for any other vehicle, using consistent metrics, to enable assessment of whether differences in such activity exist, the extent of such differences, and if such differences affect cycle average fuel use and emission rates for a given vehicle. Thus, the VSP-based approach demonstrated here enables assessment of both intervehicle and inter-cycle variability on a consistent basis. From a methodological perspective, in addition to developing improved robust procedures for field data collection and data quality assurance and analysis, this work has addressed two key sources of variability in emissions: road grade and cold starts. We developed and demonstrated a method for quantifying road grade from a moving vehicle that is shown to be both accurate and sufficiently precise for the purpose of estimating VSP and assigning each second of field data to a VSP mode. This method is based on relatively low cost consumer grade GPS receivers with barometric altimeters. By making use of multiple GPS/BA receivers, combining the data from multiple receivers, and analyzing the results statistically, accurate estimates of road grade can be made for roads with between -8% and +8% grade. We also propose a hierarchical approach to road grade estimation in which GPS/BA can supplement DGPS receivers. Although DGPS receivers are more accurate, they are also more sensitive to signal loss and thus not able to provide data with as extensive spatial coverage as GPS/BA. With regard to cold starts, we demonstrated a method for measuring real-world cold starts using PEMS and for quantifying the cold start duration and incremental fuel use and emissions associated with cold starts. As hot stabilized exhaust emission rates are lowered, the contribution of cold starts to trip total emissions may be increasing. The data collected and the methods demonstrated in this project set the stage for more flexible, systematic and rigorous quantification of real-world vehicle activity, energy use, and emissions. The approach is flexible because it can represent any driving cycle by taking into account the "microscale" (1 Hz) effect of speed, acceleration, and grade, along with characteristics of different road types, facilities, traffic control measures, levels of congestion, vehicle type, and emission standards. The approach is systematic in that it is based on a robust set of procedures that can be applied under a wide range of conditions, and it is rigorous in quantifying key factors that affect variability in fuel use and emission rates. This work has also demonstrated how emissions models calibrated based on real world data can be incorporated into traffic simulation models (TSMs), and the need for ongoing improvement of TSMs to better estimate vehicle speed trajectories. Because vehicle emission rates for a given segment of road or a given trip between an origin and destination are sensitive to the distribution of time spent in each VSP mode, it is important to have accurate and realistic estimates of vehicle speed trajectories from a TSM as a condition of being able to estimate realistic emission rates. The value-added of being able to use a TSM is the ability to conduct numerical experiments that are infeasible in the real-world, such as assessment of the effect of a change in road geometry, traffic control, or traffic management on vehicle activity and emissions.
