Energy Losses Assessment of Smallholder Farmers’ Surface Water Irrigation Pumps in South and Southeast Asia Using Entropy Generation Principle

Abstract

One of the most serious problems among smallholder farmers in South and Southeast Asia associated with the use of a surface water irrigation pump is low engine performance. The main cause of this low performance is the decrease in the flow field energy conversion mechanism caused by irreversible processes. The energy conversion theory suggests that pump efficiency is maximum when the loss is minimum. Whatever the origin of the losses, the deterioration in engine performance is due to a deterioration in the reversibility of the pump system. In this study, the pump is classified as the propeller impeller (PI), the improved axial or typical impeller (TI), and the conical hollow-shaped impeller (CI). Entropy production is applied to the pump on design improvement and loss sources location and mechanisms. The entropy production consists of viscous dissipation and turbulent dissipation. In this study, the pump design improvement of various designs based on entropy production has been studied in detail to predict energy loss in areas such as the inlet section, impeller, or discharge pipe. With the entropy generation, the optimum efficiency of different pump designs CI, PI, and TI were determined. The results showed that in all designs, more than 63% of the total entropy generation came from turbulent distribution. The flow in the pumps was analyzed in detail in comparison with entropy generation. It was found that the entropy generation rate increased in the secondary flow direction and was consistent with free-stream velocity. The PI design at the inlet pipe should be modified for reducing flow separation and entropy generation. All design impellers showed high energy losses, especially near the hub and tip along the leading edge and trailing edge. Therefore, it is possible to determine which features of the flow and entropy generation are relevant to the pump improvement.