Date of Graduation


Document Type


Degree Type



Statler College of Engineering and Mineral Resources


Mechanical and Aerospace Engineering

Committee Chair

Mridul Gautam.


A study was conducted to evaluate and correct possible inaccuracies in in-use emissions measurements, which were made with Portable Emissions Measurement Systems (PEMS). West Virginia University developed a PEMS termed the Mobile Emissions Measurement System (MEMS) for measurement of heavy duty on road diesel emissions and reporting them in brake specific units (g/bhp-hr). In particular, the study focused on inaccuracies in the engine Electronic Control Unit (ECU) broadcast data logged by the Dearborn Protocol Adapter (DPA) and the Data Link Monitor (DLM) software. Data for this study were collected both in the West Virginia University's Engine Emission Research Laboratory with engines running FTP cycles on a dynamometer, and during actual on-road tests conducted with the MEMS. The dynamometer test data show substantial and consistent inaccuracies of time stamps associated with the messages acquired with the USB DPA adapters while the DPA adapters with serial communication proved to be accurate. These time stamp inaccuracies associated with the USB DPA adapters result in erroneous evaluation of brake specific on-road emissions due to misalignment between the emission data and the power values acquired from the ECU. This time stamp error results in the USB DPA/DLM collected data file to be shorter, in terms of time, than the actual test. However, this time "shrinkage" can be readily corrected by linearly stretching the time array associated with the DLM collected data. The factor for the correcting stretch can be determined by aligning and matching the erroneous data with some independent but relevant measurement, such as CO2 mass emissions rate. This method for correcting ECU data reduced the average error in brake specific emissions from 6.24 percent to 0.13 percent, when compared to brake specific emissions reduced with serial DPA/DLM data.