Compact, real-time exhaust gas recirculation rate sensor for use in natural gas combustion engine control

Natural gas fueled engines offer a 25% reduction in greenhouse gas emissions compared to diesel. However, achieving similar fuel efficiency to diesel while also minimizing harmful emissions is a challenge. Emissions and efficiency targets can be simultaneously achieved for natural gas by using a three way catalyst and exhaust gas recirculation (EGR), but this requires that the air-fuel ratio and EGR rate be measured and precisely controlled to avoid knocking combustion while maximizing efficiency. Such on-engine control can be achieved with high speed, precise, and robust measurements of EGR rate at a size and cost level compatible with onboard control systems. In this work, we present a compact, real-time laser absorption spectroscopy-based EGR sensor with potential for low-cost, on-engine control. We develop a new, computationally efficient data fitting algorithm called Integral Hashing to enable fully autonomous, low latency, real-time operation of the laser sensor at 954Hz. The complete sensor, including lasers and drivers, data acquisition, and data processing is contained in a single 25 x 25 x 8cm enclosure supplied with 15W of power. We validate the sensor to within 2% EGR rate against an extractive non-dispersive infrared (NDIR) gas sensor on a 250kW 6 cylinder natural gas engine by simultaneously measuring a set of steady state EGR rates between 0% and 25%. Additionally, we measure transient EGR rate fluctuations that are too fast to detect with the NDIR measurement. The sensor system presented here could be used to precisely control the EGR rate, improving the efficiency, emissions, and overall adoption of natural gas engines.