All the efforts were made to minimize the errors in measurements due to equipment, humans, flow-fluctuations. Evaporation losses were measured at the site using class-A evaporation pan (IRI 2019). If seepage losses are small, evaporation and rainfall must also be considered even though these factors generally have no significant effect on seepage loss (Kraatz 1977). In the present study, rainfall was observed/measured by installing standard rain gauges at the test sites. Well-calibrated current meters were used to measure the discharge. The two-point and six-tenths depth methods are two recommended methods for the observation of velocities in channels (IRI 2019). The two-point method is recommended for large canals, where the depth of water is more than 0.75 m. It consists of measuring the velocities at 0.2 and then at 0.8 of the depth of water from the water surface. Then the average of these two observations is adopted as the mean velocity in a vertical. The accuracy obtained with this method is high. In situations where the two-point method is not applicable or where the depth of water is less than 0.75 m, the six-tenths method is used. This method consists of measuring the velocity at 0.6 of the depth from the water surface and adopting this velocity as the mean velocity in that vertical. This procedure gives satisfactory results (Kraatz 1977). The water gauge in the channel was kept constant to minimize the fluctuations in the discharge of the channel during test duration. Discharges of all outlets were measured with similar accuracy to keep minimum errors. All measurements were repeated three times to get an average value to be used in the further calculation for seepage measurements (IRI 2019). A summary of inflow–outflow tests performed is given in Table 3. As evident from Table 3, a total of 10 tests were performed on eight channels. The total length of these eight channels is 692,709 ft (211 km) out of which 224,200 ft (68 km) length has been tested which becomes 32% of the total length. One test was performed on the unlined channel (Khadir Disty), which was being lined at the time of field testing. The range of test reaches is 4–9 km with an average value of 7 km of test reach. It is a significant length of test reach and is generally not available in the literature (Shah 2019). Each test was repeated to get an average and more accurate results.