Integrated dual-cycle energy recovery using thermoelectric conversion and an organic Rankine bottoming cycle

Abstract

Hot engine exhaust represents a resource that is often rejected to the environment without further utilization. This resource is most prevalent in the transportation sector, but stationary engine—generator systems also typically do not utilize this resource. Engine exhaust is considered high-grade heat and can potentially be utilized by various approaches to produce electricity or to drive heating and cooling systems. One idea for this application is to combine an organic Rankine cycle and thermoelectric conversion. This approach is being developed to more fully utilize the thermal energy contained in hot exhaust streams. The model developed here is composed of a high-temperature heat exchanger which extracts thermal energy for driving the thermoelectric conversion elements and a closely integrated bottoming cycle to capture the large amount of remaining thermal energy in the exhaust stream. Many interacting parameters that define combined system operation are employed in the model to determine the overall system performance including output power, efficiency, and total energy utilization factors. In addition, the model identifies a maximum power operating point for the combined system. The model can identify the optimal amount of heat to remove from the exhaust flow to drive the thermoelectric elements for maximizing the combined cycle output. The model has been developed such that heat exchanger UA H (heat transfer coefficient multiplied by heat transfer area) values, thermal resistances, and the thermoelectric figure of merit ( ZT) can be investigated in the context of system operation. The model also has the ability to simultaneously determine the effect of each cycle design parameter on the performance of the overall system, thus giving the ability to utilize as much waste heat as possible. Results of exercising the model give system performance and inter-relationships between various design parameters as they affect overall performance.