Misfire Detection for Spark Ignition Engines Based Upon Cycle-by-Cycle Exhaust Temperature Sensing

Misfiring is a serious issue for large spark-ignited industrial natural gas engines because it increases fuel consumption and emissions and can ultimately lead to engine damage. Continuous misfiring of an engine cylinder can often be diagnosed based upon vibration measurements or conventional exhaust temperature measurements. The aim of the research described in this paper was to develop a means of detecting not only frequent misfiring of a cylinder, but also isolated, individual misfires. This would make it possible to detect problems early in the progression from sporadic misfire events to the onset of continuous misfiring. The approach used in the study is based upon the sensing of cycle-by-cycle fluctuations in the exhaust temperature from each cylinder. Proprietary circuitry and signal processing techniques are used to extract high frequency temperature information from durable sheathed thermocouples similar to those commonly used for these types of engines. These signal fluctuations can be used to detect momentary drops in exhaust temperature that occur as unburned mixture from a misfire exits the cylinder. The system was developed and validated through engine dynamometer tests using an automotive engine equipped with laboratory-grade cylinder pressure sensors for reference measurements. Further testing was conducted using a medium speed stationary natural gas engine. In both cases, the system was shown to be able to detect and count individual and consecutive misfire events.