Comparative evaluation of canal seepage investigation techniques

Various methods are used to estimate the canal seepage rate such as empirical formulae, analytical or analogue studies and the direct seepage measurement techniques, i.e. seepage meters, ponding tests and inflow-outflow tests. Considering significance of physical measurement techniques both ponding and Mow-outflow methods were used to evaluate the seepage losses. These methods have their merits, demerits and limitations which are well understood. However, while interpreting seepage loss evaluations, there can be independent random errors, like the fluctuations in flow conditions affecting different current meter positions, and systematic errors, like the use of different instruments or different methods to measure different terms in the balance. These errors have to be distinguished. The data from various studies in the region have been reviewed and analyzed. A statistical analysis of a number of replications of ponding and inflow-outflow test data of the seepage estimation work on various canals under the Fordwah Eastern Sadiqia (South) project, Bahawaluagar, Pakistan has been done to show the accuracy of results achieved through both the methods. It has been shown that the ponding method is more accurate than the inflow-outflow method as variability of various statistical parameters from the ponding tests is comparatively less. It is proved that inflow-outflow tests should not be used for seepage investigations where seepage rates are fairly small and there is no possibility to use sufficiently long reaches. (C) 2003 Elsevier B.V. All rights reserved.