Influence of Different Natural Gas Blends on the Regulated Emissions, Particle Number and Size Distribution Emissions from a Refuse Hauler Truck 2012-01-1583

Natural gas is a potential alternative to conventional liquid fuels for use in automotive internal combustion engines. The primary goal of this study is to understand how gas composition changes might impact the performance or emissions of a natural gas vehicle or engine. For this study, a waste hauler truck equipped with a 2001 Cummins 8.3L C Gas Plus lean burn spark-ignited engine and an oxidation catalyst was operated on the William H. Martin Refuse Truck Cycle (RTC). This cycle was developed to simulate waste hauler operation and consists of a transport segment, a curbside pickup segment, and a compaction segment. The vehicle was tested on 7 different gases including two gases representative of Texas and Rocky Mountain Pipeline Gases, a Peruvian LNG, a Middle East LNG-Untreated with high Wobbe number (above 1400), two gases representative of those located within the state that have low methane number (MN) as well as varying hydrocarbon compositions, and a CNG blend produced from an LNG fuel tank with similar properties to the first two gases. The experimental results indicate the fuel composition had a noticeable effect on fuel economy and carbon dioxide (CO₂) emissions of the waste hauler, with the high energy content and higher hydrocarbons gases exhibiting higher fuel economy and CO₂ emissions. Emissions of nitrogen oxides (NO

) were also influenced by fuel composition, and increased for gases with higher levels of heavier hydrocarbons. Total hydrocarbon (THC), nonmethane hydrocarbon (NMHC), and methane (CH₄) emissions were affected by the concentration of higher hydrocarbon molecules in the gases, with gases with higher levels of CH₄ showing high THC and CH₄ emissions and lower NMHC emissions. Decreases in particulate matter mass, particle number, and, in some cases, CO emissions were also found for the gases with more heavy hydrocarbons, while fuel quality had a minimal impact on particle size distributions.