Experimental investigation of a novel thermoelectric generator design for exhaust waste heat recovery in a gas-fueled SI engine

In this paper, a series of tests are performed on the exhaust waste heat recovery using a thermoelectric generator (TEG) in a propane-fueled spark-ignition (SI) engine. A three-layer TEG is designed, consisting of two engine coolant exchangers (Ec_hex) positioned on both surfaces of the rectangle exhaust heat exchanger (Ex_hex). Unlike the literature papers, a novel design is created by placing copper serpentine pipes in each Ec_hex to allow propane to pass through the TEG. The reason for this is both to ensure the evaporation of propane and to increase the & UDelta;T temperature differences by increasing the cold surface activity of TEG. A total of 12 pieces of thermo-electric modules (TEMs), 56x56 mm in size, are positioned between the rectangular Ex_hex and each Ec_hex, and a total of 24 pieces of TEMs are used in TEG. Experimental studies were performed for eight different engine speeds in the range of 1500-5000 rpm for TEGs with and without a propane inlet. In addition, computational fluid dynamic (CFD) analyses are performed to visualize both the cold and hot surface temperature contours of the TEMs. The results of the paper illustrate that the DC electrical power of the TEG with propane is higher in the range of 11.5-12.1% compared to the TEG without propane in the 1500-5000 rpm range of the engine. At 4500 rpm engine speed, the propane input TEG produced a maximum of 90.2 W of DC electrical power and a 3.02% energy conversion efficiency. In addition, there is a good correlation between the experimental and numerical findings, varying between 3% and 15% with the engine speed.

Automated summary

This paper presents a series of tests on the utilization of a thermoelectric generator (TEG) to recover waste heat from a propane-fueled spark-ignition engine. The experimental and numerical results demonstrate that the TEG with propane achieves a higher DC electrical power and there is good correlation between the experimental and numerical findings.