Flow hydraulics and sediment transport in pervious rockfill detention dams

There is growing interest in using pervious rockfill detention dams to control and mitigate floods in Iran. This kind of dam is constructed of rock particles without any core inside and shell on the upstream face. Rockfill detention dams function in such a way that the peak point of outflow hydrograph will be smaller than the peak point of the inflow hydrograph. During the passing of a flood the flow through a rockfill dam is mostly laden with sediment. If sediments pass through, the dam will be safe, function well and downstream scouring will be low. Otherwise, gradually sediment particles will settle in pore spaces of dam and partial clogging occurs. In such cases the flood may overflow the dam, erode the downstream bed, bank and the dam itself and finally may cause damages. Therefore it is necessary to investigate the various aspects of sediment laden flows through the pervious rockfill dam to achieve a safe hydraulic design. Critical hydraulic gradient, sediment transport, water discharge,rating and water surface profile equations are the main components of hydraulic design of flow through pervious rockfill dam. Some semi empirical researches have been conducted in the past to study the problems but the results have indicated limitations and these results cannot be generalized. In this respect scale effect,exclusion of some effective parameters and the formulation of basic equations are the subjects that have to be considered and revised. Regarding these limitations rock particle sizes, sediment particle sizes, dimensions of laboratory rockfill dams and magnitude of Reynolds number which were used in the previous experimental works were small and stand in the lower limits of those are being used in real field condition. In addition by Appling a fixed bed slope flume and sediment free flow in previous researches the effect of bed slope changes and sediment laden (two phase) flow has been excluded. Moreover the formulations of some equations such as critical hydraulic gradient and sediment transport through pervious rockfill dam indicate some ambiguous parameters (critical water discharge and hydraulic gradient) which are undeterminable in real field condition. The main objective of this research is to investigate the non-cohesive suspended sediment laden flow through a pervious rockfill detention dam. From a review of various formulated equations the Sakthivadivel formula (critical hydraulic gradient in laminar flow), dimensional analysis and the pipe theory including Darcy-Wiesbach and continuity equations have been incorporated to develop related formulas. To calibrate and validate the equations formulated, 180 laboratory tests have been conducted. Statistical analysis (by means of SPSS software) and the resulting correlation coefficient (R2) and mean square error (mse) were used as criteria for assessment. Five equations have been improved and developed for the prediction of the critical hydraulic gradient, friction coefficient, rating curve, water surface profile and the sediment transport rate in pervious rockfill dam to overcome the limitations of previous research results. Validation of the above mentioned equations showed good agreement to the real laboratory data with the mean square errors of 4.46E-5, 0.29, 7.7E-6, 5.19E-4 and 6.4E-4 respectively. The improved critical hydraulic gradient was calibrated and validated using 36 laboratory test results with a correlation coefficient of 0.92 and a mean square error (mse) of 4.46E-05. Comparing with two equations proposed by earlier researchers (mse, .0003 and 0.029) it showed a better agreement to the observed critical hydraulic gradient. The friction coefficient-Reynolds number equation developed as a basic equation in the pipe theory was calibrated, validated and compared with three equations proposed by previous researchers using 46 laboratory test results. The results showed a correlation coefficient of 0.74 and a mean square error of 0.29 while previous equations have mean square errors of 0.66, 113 and 197. This indicates a better agreement of the predicted friction coefficient (with the improved equation) with the laboratory derived friction coefficient. The rating and water surface profile equations developed and validated using 34 laboratory test results have mean square errors of respectively 7.7E-06 for the rating equation and less than 5.19E-04 for all 34 water surface profiles. These equations when compared with two formulas proposed by previous researchers showed better agreements to the observed upstream water depths and also observed water surface profiles. A sediment transport equation has also been formulated, calibrated,validated using 144 test results and compared with the three equations presented by other researchers. Calibration and validation of the dimensionless sediment transport equation indicated a correlation coefficient (R2) of 0.933 and mean square error of 6.4E-06 respectively. The comparison indicated the mean square errors of 6.4E-4,7.1E-03, 2.3E-02 and 0.041 respectively for the formula developed and the other formulas, thus suggesting a better agreement for the equation proposed in this study to the laboratory measured sediment transport. In conclusion, the equations developed and validated in this study showed more accurate predictions and calculations of hydraulic behaviors of sediment laden flow through rockfill dam than previously presented formulas in this field. These improvements can enable engineers to design safer pervious rockfill detention dams.

Preview PDF FK 2011 126R.pdf Download (945kB) | Preview

Actions (login required)

View Item