Development of an Activity-based Windowing Approach to Evaluate Real-World NOx Emissions from Modern Medium and Heavy-Duty Diesel Trucks

The introduction of in-use emissions regulations by the United States Environmental Protection Agency (U.S.EPA) requires medium-duty (MD) and heavy-duty (HD) engine manufacturers to demonstrate emissions compliance during in-fleet operation. In the United States (U.S.), the Not-to-Exceed (NTE) method is used to evaluate real-world emissions compliance from on-highway MD and HD trucks. Regulatory agencies, engine manufacturers and research entities have identified that the NTE method incorporates numerous exclusions and evaluates emissions compliance only under selective operating conditions that are favorable for the selective catalytic reduction (SCR) system to reduce oxides of nitrogen (NOx) emissions efficiently. Such operation is typically encountered only by vocations that experience sustained highway driving operation, which is not entirely representative of actual highly diverse real-world operation experienced by the engine/aftertreatment system. Evaluation of real-world driving emissions (RDE) plays a critical role in monitoring and ensuring the performance of emissions control systems. Portable emissions measurement system (PEMS) serves as a robust tool to assess emissions levels during real-world operation. However, utilization of PEMS for large-scale deployment is time-consuming, labor-intensive, and expensive. As a vision of potential elements for a next-tier of in-use NOx monitoring systems, there is an actively growing research and regulatory interest to evaluate the feasibility of using existing on-boardNOx sensors for HD on-board NOx compliance. However, research studies have highlighted that NOx sensor measurements are also subjected to cross-sensitivity from other species in the exhaust stream. The global objective of the study was to develop an alternative approach that attempts to bridge the gap between current in-use certification procedures and highly diverse real-world operation for evaluation of in-use NOx emissions. The study outlines a vehicle activity-based windowing (ABW) approach that provides an event-based bifurcation of the engine and aftertreatment operational conditions. The thermal boundary bin exhibits bin boundary conditions favorable for SCR catalytic activity. Results of the study show that the 90th percentile of ABW bin-1a bsNOx emissions (i.e., on average of the individual vehicle datasets) was below the current NTE NOx limit. In terms of data usage, the ABW approach provides a unique opportunity of utilizing ~95% and 83% (Phase-1 and Phase-2) of test activity acquired from valid ABW trips. In comparison, the current NTE approach evaluated over a diverse in-use test activity collected as part of the HDIUT program exhibits utilization of only a sparse amount (i.e., less than 10%) of in-use test activity for emissions compliance evaluation. In light of using existing on-board NOx sensors for the screening of in-fleet activity, the study evaluates measurement thresholds of NOx sensors under real-world operating conditions. In the absence of a substantial amount of ammonia (NH3), it was observed that the average measurement deviation was within ±10% for NOx concentration levels between 10 ppm and 200 ppm. However, statistical principal component analysis (PCA) indicates a hypothetic relation between NOx sensor measurements and rapid changes in water (H2O) concentrations