WATER-BASED RANKINE-CYCLE WASTE HEAT RECOVERY SYSTEMS FOR ENGINES: CHALLENGES AND OPPORTUNITIES

Much of the fuel energy in an internal combustion engine is lost as heat, mainly through hot exhaust gas. The high energy losses, and high temperatures of the exhaust gas, provide favorable conditions for applying a waste-heat recovery system. Among the available options, systems based on the Rankine cycle show the highest potential in terms of reducing fuel consumption. Water or water-based mixtures have several advantages over organic fluids as working fluids for such applications of the Rankine cycle, in terms of cost, thermal stability, safety and complexity of the system. They also have several disadvantages, including possible freezing for pure water, high boiling temperature and high heat of vaporization. Hence, higher temperatures and amounts of waste heat are needed for reliable operation of the system. However, few experimental investigations have addressed the practical challenges associated with water and their effects on the performance and operation of a system in a driving cycle. This paper presents results of experiments with a full-scale system for recovering waste heat from the exhaust gas recirculation (EGR) of a 12.8 L heavy-duty Diesel engine on a test bench. The working fluid used in the experiments was deionized water and a 2-cylinder piston expander served as an expansion device. The engine was kept in standard configuration, except for minor modifications required to implement the heat-recovery system. The prototype EGR boiler was designed to fit in the space initially designated for the production EGR cooler. The assembly was operated in the operating points of the European Stationary Cycle (ESC). The results show that the trade-off between boiling pressure, sufficient superheating of the water and cooling of the EGR caused by the pinch-point in the boiler poses a major challenge when using water as a fluid. Low flow rates at low load points were challenging for boiler stability. During operation, the blow-by of working fluid into the lubrication system of the expander and vice versa was also problematic. Special steam-engine oil with high viscosity and good water separation capability was used to weaken this effect. The Rankine cycle-based test system attained a thermal efficiency of 10% with EGR as the only heat source. Results, major constraints and possible means to improve the system when using water as a working fluid are presented here. Simulation models developed for the EGR boiler and the piston expander supported this effort