Technology developments in directional drilling and hydraulic fracturing have led to increased natural gas reserves. Development of these unconventional resources is an energy intensive process. Prime-movers of unconventional well development were previously identified to be over-the-road trucks, drilling engines, and hydraulic stimulation engines. Diesel engines dominate these markets but industry is attempting to cut costs by using dual fuel and dedicated natural gas engines. On-road engines are subject to the transient FTP cycle for certification and off-road engines are subject to the 5-mode ISO 8178 D-2 cycle. It is well known that in-use activity can differ from certification activity. Significant in-use activity data for each prime-mover were collected and a Markov-Chain Monte-Carlo Simulation with a genetic algorithm was used to develop test cycles for each.
The developed test cycles allowed for operation of a smaller yet similar engine within a controlled laboratory environment. Laboratory tests utilized a Cummins 8.9L ISL-G to analyze the emissions of new cycles compared to certification cycles and to examine the effects of fuel quality on emissions. The ISL-G is a spark-ignited engine used for heavy-duty trucks and could see market penetration in fleets serving the well development industry. It is similar in technology to the Waukesha LI7044, which is used in drilling operations — both employ air fuel ratio control and three-way catalysts. For the case of “pump” quality fuel, compressed natural gas was used. The developed OTR truck cycle produced higher brake-specific emissions of CO2, CO, NOx, and lower HC emissions compared to the FTP. The drilling and fracturing cycles tended to have lower CO2 and HC emissions but higher CO emissions when compared to the D-2 cycle. Two additional fuel blends were used on the new cycles and represented blends with higher ethane and propane fractions — which are common to shale gases that could fuel prime-movers in the future. The minimum recommended methane number for this engine was 75 and additional fuel blends had methane numbers of 75.5 (propane blend) and 75.3 (ethane blend). As expected, CO2 emissions increased with increased alkane concentration, while opposite trends were shown for THC and CH4. NOx emissions also tended to decrease with higher ethane and propane blends, across all cycles. For all cycles and fuels, HC emissions were predominately CH4 - 94–97%. Variations in activity and the effects of different fuels should be addressed when estimating emissions since using standard certification or emissions factors may not be representative of in-use emissions.