Preliminary studies of development of clogging prediction method for stormwater infiltration systems
Stormwater infiltration practices are widely used for management of urban stormwater runoff. They help to restore the pre-urban environment by directing stormwater into the ground, and are also a cost effective and efficient way of controlling stormwater. Several researchers (Lindsey et al., 1992) have conducted field studies of these systems, to evaluate their long-term performance, and have found that the biggest threat to these systems is the possible failure of the system due to clogging.
A study has thus been undertaken to develop a fundamental understanding of the clogging processes of stormwater infiltration systems, by conducting one and two dimensional laboratory (1D Rig and 2D-Rig) experiments, and based on this knowledge, to develop a new, reliable and robust method for sediment transport and clogging prediction.
A vertical column (1D-Rig) was constructed from 20cm diameter mountable segments. One half of the column was filled by crushed stones (infiltration medium) and the other half by soil, in order to study sediment transport in the vertical direction. The 2D-Rig, which is 3x2x0.35m in size with removable Perspex panels, was used to construct a section of a typical infiltration trench and was used to study clogging in two dimensional flow regimes.
The 1D and 2D rigs have been operated in a similar way. Synthetic stormwater was prepared in the separate tanks by mixing sediments collected from a stormwater retarding basin (i.e. < 300 ) with tap water, to maintain constant total sediment concentration. Tipping bucket rain gauges were used to monitor the system outflow. Water pressure in the system and the soil was monitored by piezometers and pressure sensors. Tensiometers were also used in the 2D-Rig to monitor the moisture content in the soil around the infiltration system and hence to evaluate negative pressure in unsaturated soils. During these experiments, water samples were collected from the crushed stone layer and from the inflow in order to monitor the total sediment concentration and its particle size distribution. Experiments were undertaken for three hydrologic regimes; (1) constant water levels, (2) constant flow rates, and (3) continuous wetting and drying conditions.
Each experiment was stopped when the outflow had decreased to 10% of the original flow. The segments/panels of the rigs were carefully dismantled, and sediments trapped along the column were measured by weighing the content of each segment separately. So far, seven 1D-Rig experiments and one 2D-Rig experiment have been carried out and the details of these experiments are presented in this paper.
It was clearly observed that the hydraulic conductivity of the infiltration systems decreased with time due to clogging of the system in all the above experiments. However, it was not possible to clearly see a clogging layer at the soil/stone interface. To study the position of the clogging layer, the recorded hydraulic heads were analysed. In all experiments recorded it was found that the hydraulic head in the stones did not change during the course of the experiment, while all heads measured in soils gradually declined. The hydraulic conductivity of the soil near the interface was also analysed and the results showed the decline of the hydraulic conductivity at the interface, while the experiment progressed. Therefore, it is apparent that the growth of the clogging layer is primarily at the soil/crushed stones interface.
The particle size distribution and TSS experimental data are analyzed to determine distribution of the sediment within the system (ie. the percentage that gets trapped around the dominant water level as well as the drop in the sediment size below this level). The distribution of the sediment across the filter media is therefore assessed as well as the hydraulic resistance of the clogging layer, for all experiments carried out so far.
A physically based model is currently being developed using these results, to model physical clogging for traditional stormwater infiltration systems constructed with crushed stone overlaid by soil. For modeling of the hydraulic conductivity of the clogging layer the initial approach will be to test using an established method (Reddi et al., 2000), developed for sand filters based on the Kozeny model (Leonard, 1962). This paper will present an attempt to test this modeling approach using the results collected during 1-D experiments. A detailed description of the methodology adopted and the preliminary results of the 2D-Rig experiment will be also presented.
Leonard A.G. (1962), "Engineering properties of soils." Foundation Engineering, McCraw-Hill, New York, 107-139.
Lindsey G., Roberts L. and Page W. (1992), Inspection and Maintenance of Infiltration Facilities, Journal of Soil and Water Conservation, 47(6), 481-486.
Reddi L.N., Ming X., Hajra M.G. and Lee I.M. (2000), Permeability reduction of soil filters due to physical clogging, Journal of Geotechnical and Geoenvironmental Engineering, 126(3), 236-246.