Flow hydrograph moving method based on inflow-outflow tests and its application on canal seepage calculation

Abstract: Inflow-outflow test and ponding test are both common ways to measure canal seepage and have their own advantages and disadvantages and respective applicable conditions. Inflow-outflow test is more convenient than ponding test, while less accuracy. Inflow-outflow test is conducted in irrigation period, and its calculation accuracy mainly depends on stability of canal flow. The evidence shows that stable flow volume is extremely rare in the measure data from Shijin Irrigation District. Since canal flow volume is controlled by sluice gate commonly, and calculation accuracy is sensitive to flow changes, inflow-outflow test is difficult to apply widely and effectively. Generally, a series of continuous inflow-outflow tests will get a group of data and draw a flow hydrograph. Theoretically, the 2 hydrographs have the same or close shape and parallel nearly when upstream hydrograph flows down to downstream section, and seepage quantity would be revealed by flow difference of the hydrographs between upstream and downstream section. This paper establishes the flow hydrograph moving (FHM) method, by which seepage quantity can be calculated. The FHM method has 3 factors: hydrograph flowing velocity, length between upstream and downstream section and time consumption, and needs 3 steps: drawing hydrographs (according to measure data for 2 times, calculate and line the 2 quantities with X-axis as time and Y-axis as quantity), moving hydrographs (time consumption of moving hydrograph from upstream to downstream section depends on the maximum hydrograph flowing velocity) and measuring flow difference (the moved hydrograph of one section will parallel with the other one, and measure the difference between the Y values of those 2 parallels). However, the measured flows by instruments contain potential errors, which will lead to the intersecting between the 2 hydrographs. Therefore, the moved upstream hydrograph should parallel the downstream one when the 2nd step is finished, which will decide whether data are effective or not. Basically, 5 hypotheses are providing supports to the FHM method: 1) One measure process is regarded as an average flow characteristic to the whole measuring process; 2) Hydrograph flowing velocity value takes the maximum velocity; 3) Seepage characteristics of canal are homogenized, such as longitudinal slope, roughness and cross-sectional shape; 4) Difference in seepage quantity caused by flow variation and seepage is ignored; 5) Hydrograph shape is linear between 2 measurements. However, the maximum velocity of upstream section will be decreased gradually with flow loss, which will influence time consumption, so the difference of the maximum velocity between upstream and downstream section is calculated and averaged by upstream moving and downstream moving in the 2nd step. The FHM method is applied in Shijin Irrigation District. In the case study, 6 kinds of concrete lining forms are designed and constructed, and inflow-outflow tests and ponding tests are both conducted successively. According to the analysis of the ponding test data by non-linear regression on 6 kinds of concrete lining, seepage characteristics showed exponential function between water depth and seepage quantity, and the correlation coefficients of 6 functions are 0.884 (P<0.1), 0.852 (P<0.1), 0.992 (P<0.001), 0.988 (P<0.0001), 0.964 (P<0.01) and 0.991 (P<0.01), respectively. There are 19 couples of hydrographs matching the parallel conditions in inflow-outflow tests data. Taking exponential function simulation results as benchmark, seepage calculation results by the FHM show that the errors from 12 groups are less than 10 m3/(km·h) (63.16% of total data), 6 groups fall in 10-50 m3/(km·h) (31.58% of total data), and the maximum error is 74.08 m3/(km·h). Furthermore, the linear regression for the FHM and exponential functions both revealed that seepage rate increased with water deepening. The possible cause of errors is potential flow fluctuation between 2 measurements. After all, time interval is almost inevitable between them. Seepage calculation using the FHM is more suitable under canal's measuring conditions and makes the application of inflow-outflow tests more effectively.