Investigation of multiple lift-tube air-lift pumps: Experimental evaluation and predictive modelling

Due to limited pumping capability the commercial use of the single lift-tube air-lift pump has been confined to very little applications. To overcome the limited pumping capability of single lift-tube air-lift pump, multiple lift-tubes were examined as a way to improve the pumping capacity. Experiments were conducted on air-lift pumps having 2 and 3 lift-tubes. Lift-tubes having the inner diameter of 10 mm and length of 500 mm were used in the experiment. The working fluids used in the experiment were water and air. Tests were performed at three different submergence ratios of 0.3, 0.5 and 0.7. It was found that using multiple lift-tubes is a viable solution as it improves the pumping capacity of the air-lift pump. However, the flow rate and the flow regime of each lift-tube in a multiple lift-tube configuration were observed to be non-uniform. Moreover, due to non-uniform flow regime analytical modelling of a multiple lift-tube air-lift pump showed an inability to accurately predict the pumping characteristics. To overcome this, a novel artificial neural network (ANN) technique was used to predict their pumping capacity. The results showed that the ANN was capable of accurately predicting the performance of the multiple-lift tube bubble pump.