An optimization routine on a prediction and selection model for the turbine operation of centrifugal pumps

The paper presents an experimentally validated optimization routine for the turbine-mode operation of radial flow centrifugal pumps. The optimization routine outlined here is designed to be used with prediction (predicting turbine mode characteristics of a pump) and selection (selecting the most appropriate pump for turbine-mode operation) models. The optimization routine improves upon previous uncertainties in prediction, especially in the low specific speed range.The optimization routine is evaluated experimentally for three pumps with specific speeds of 18.2 rpm, 19.7 rpm and 44.7 rpm, and a significant improvement in the accuracy of the turbine predictions with the errors for all the three pumps falling within the ±4% acceptance bands in the full load operating region is found.It is also shown how the optimization routine validates an approach to selection and prediction based on model experiments and classical principles of applied turbomachinery (specific speed-specific diameter or the Cordier/Balje plots). Such an approach is shown to be the most economic in terms of pump mode input variables.The paper recommends the extensive use of the optimization routine in micro hydro and other energy recovery projects involving pumps as turbines and the creation of a database of accurate field results that can be used to improve the routine further.