Flow variation in irrigation system components

Abstract

This research evaluated flow characteristics in three irrigation system components: Venturi injector, chemigation oil drop generator, and pressurized landscape bubbler. Foliar chemicals are often applied as oils through irrigation systems in order to prevent wash off from leaves. The system application uniformity and efficiency depends on oil drop size distribution. The primary parameters that influence oil drop size are water flow rate and eddy shear stress, interfacial tension and oil viscosity. In this laboratory experiment, the oil-based chemical drop diameter distributions were evaluated over a range of these physical parameters. Drop sizes were in the range of an equation used to predict drop size in turbulent flow; however, there was very little change in drop size with change in physical parameters. The injection flow rate for four chemicals (CAN17, UAN32, soybean oil, and OrchexRTM) and water was measured over a range of pressure differentials between the upstream and downstream side of the Venturi, and over a range of chemical temperatures. The injection flow rate for water, with low viscosity, did not change significantly with temperature. However, the injection rate for the four chemicals was correlated with temperature and viscosity. If the chemical tank temperature variation is 20°C during the day, then the injection flow rate variation would be in the range of 50% for soybean oil, 30% for OrchexRTM, 10% for UAN32, and 5% for CAN17. Five commercial bubbler with flow rate screw adjustments were evaluated: RainbirdRTM, ToroRTM, IrritrolRTM , HunterRTM, and LegoRTM. The coefficient of variation at full flow for the five brands ranged from 8.2% to 20.8%. Bubbler flow rates were measured at 140 and 210 kPa over a range of number of screw turns. Equations were developed for each bubbler based on the emitter flow equation, Q = KHx. Because K and x varied with number of screw turns, equations were developed for K and x for each bubbler based on regression of flow data. Correlation between calculated flow rates with the equations and measured flow rates was in the range of 80% to 90%.