Non-parametric statistical evaluation of biodiesel emissions from transit buses

Thumbnail Image
Mudgal, Abhisek
Major Professor
Abhisek Mudgal
Committee Member
Journal Title
Journal ISSN
Volume Title
Research Projects
Organizational Units
Journal Issue
Is Version Of
Civil, Construction, and Environmental Engineering

Research on biodiesel emissions has been triggered by depleting fossil fuel resources and environmental protection concerns. However, vehicular emissions are inadequately understood and quantified because of large variations in individual vehicle emissions with changing operating conditions, engines and fuels. More research is needed to evaluate biodiesel emissions especially from heavy duty vehicles which include transit buses. Past research findings have been contradictory mostly in case of NOx emissions. These make it essential to carry out further research especially with the help of on-road measurement devices which can capture the real-time variation in operating conditions, unlike dynamometers and remote sensing devices.

Emissions data were collected using a portable emission measurement system (PEMS) from three transit buses fueled with regular diesel (B0), 10% biodiesel (B10) and 20% biodiesel (B20). At an interval of one second, NOx, HC, CO, CO2 and particulate matter (PM) were recorded along with speed, acceleration, other engine parameters, and number of passengers for all the nine bus-fuel combinations. Emissions were found to exhibit auto-correlation and non-normal distributions, which necessitated a binning-based approach and the use of non-parametric statistics respectively for data analysis. Emission rates were not proportional to percentage of biodiesel. This was also seen when the same batch of biodiesel was tested using a dynamometer. Therefore, B10 and B20 were evaluated separately. The commonly used VSP formula was modified to account for passenger weight and load imparted by the use of air-conditioning. Emissions from each fuel were binned by speed and vehicle specific power (VSP). Emissions from each fuel were grouped by VSP into three bins.

Emissions varied monotonically with VSP. Further, no significant change in result was obtained upon using the new formula. Statistical tests were performed to compare emissions from B10 and B20 to B0. Evaluation of B10 revealed that NOx, HC, CO, and CO2 emission rates decreased for Tier-1 buses. For Tier-2 bus, NOx, HC, CO2 and PM emission rates increased while CO emission rates decreased. With B20, HC and PM emission rates decreased for all the buses. NOx and CO, CO2 results were contradictory. Decrease in PM emissions is very significant particularly for heavy duty vehicles in terms of freight demand. Decrease in HC is not significant for diesel engines. Likewise, inconsistency in CO emissions is also immaterial while inconsistency in NOx emissions supports previous researches.

Thu Jan 01 00:00:00 UTC 2009