The International Network of Irrigation Testing Laboratories (INITL) undertook a sprinkler intercomparison testing exercise to generate data for an objective comparison of the performances of the different facilities and identify opportunities for further improvements. Three impact sprinklers were tested in four laboratories in accordance with established standards. The plots of flow rate-pressure profiles were found to be consistent, and correlated to those obtained from previous studies. Although there were slight variations in the measured flow rate data, the mean flow rates at each pressure level were the same for the three sprinklers, and the shapes of the radial distribution profiles were similar. The deviations of reconstituted flow rates (from the measured) of at least two tests in each facility were found to exceed the recommended limit. The sprinkler software developed by INITL was found to have a good correlation with a related commercial software program.

The International Network of Irrigation Testing Laboratories (INITL) undertook a laboratory inter-comparison testing exercise of three sets of drippers in the period 2013 and 2014. The four testing facilities that participated in this exercise are based in following countries: Australia, Brazil, France and China. The objective of the testing programme was to compare results from different independent testing facilities to enable individual laboratories identify potential opportunities for improvement of their performances. This would also facilitate the INITL to make proposals for harmonisation of the testing methods. The maximum coefficient of variation, cᶹ, at the manufacturer's recommended operating pressure of 100 kPa, was found to be 3.76%, which was significantly smaller than the 7% recommended by ISO 9261 [5] as the maximum allowable variation of the flow rate of the test sample. The emitter exponent was determined to be approximately 0.5 which is consistent with results obtained from past studies. At the operating pressure of 100 kPa, it was found that although the average flow rates from the participating laboratories were similar, there was a difference in the dispersion of data. Datasets for the 4 Lh‾¹ dripper model fitted a normal distribution model, while on the other hand, some data sets for the 2 and 8 Lh‾¹ dripper models were not normally distributed. A higher dispersion of measurements can be interpreted as a higher instability of the testing conditions. The variance for the 2Lh‾¹ dripper model was found to be homogeneous, while non-homogeneous for the other two models. The latter implies that at least one of the laboratories presented an uncertainty of measurement significantly different from the others. The measurement uncertainty undertaken in this study demonstrated that there were opportunities to improve the measurement process. Recommendations and suggestions for harmonisation of test procedures and improvements in individual laboratories are also identified in this paper.