Laboratory experiments on the removal of soil plugs during soil piping and internal erosion

Plugging of soil pipes can be detrimental to hillslope stability. A soil pipe commonly becomes plugged through internal erosion when the pipe wall collapses. When a pipe becomes plugged, a pressure buildup occurs upslope from the plug. This pressure may be enough to remove the plug, or the pressure may continue to build in the soil matrix, which may lead to hillslope failures. Based on field observations, it is known that both processes occur, but limited data exist to understand if and when plug removal or pressure buildup occurs or how to model these processes. This study involved wellcontrolled laboratory experiments to determine instantaneous pressure buildup behind soil plugs and conditions for which an idealized plug in a soil pipe will be removed. Laboratory experiments were conducted with a smooth or roughened 100 cm long clear polyvinyl chloride (PVC) pipe. A pipe plug (3 or 6 cm) was established 90 cm along the pipe length. Triplicate experiments were conducted with two pipe diameters, two soil types (sand and sandy loam), two plug lengths, three pipe roughness values, various packing densities, and with both dynamic and constant pressure heads. Digital pressure gauges were installed to monitor pressures in the pipe both before and after the plug. The upslope pressure and the length of time that the plug withstood the pressure before removal were recorded. Regardless of pressurized time, all plugs were initially mobilized as intact plugs and then in many cases quickly disaggregated. Some sandy loam plugs withstood pressures of 100 cm; more cohesive plugs could withstand much higher pressures. In addition, pressurized times exceeded 1000 s for some plug conditions even with short (3 to 6 cm) plugs. Therefore, hillslopes upslope of a plugged soil pipe may experience considerable pressure buildup for extended periods of time. Plug physical characteristics are important; for example, the plug's bulk density had a positive exponential relationship with the pressurized time. Soil water dynamics inside the plug will need to be considered, as well as the hydraulic boundary providing inflow to the soil pipe. These experiments provided insight into the dynamics of plugged soil pipes with the eventual goal of assisting in being able to better predict internal erosion and hillslope failure. Future experiments are needed across a wider range of soil types.