Leaky sewer systems: Influence of soil properties and sewer failure characteristics on the shape of contaminant plumes

At the current time, cities harbor more than 4 billion inhabitants. According to the United Nations’ projections, an increment by 2.5 billion is expected until mid-century. This will create enormous stresses to the water resource management in urban regions, including detrimental impacts on both groundwater quantity and quality. For instance, leakages from aging urban sewer network systems may lead to uncontrolled recharge and contamination. Sewer-borne contaminants cover a broad bandwidth of substances including pathogenic microbiota, nutrients and emerging contaminants. These substances may be highly persistent and accumulate in the subsurface over time. This, in turn, may pose a long-term threat to urban ecosystems. Hence, understanding the spatiotemporal distribution of sewer-borne plumes within the subsurface is of strategic importance. Sewer failures may include, among others, pipe blockades, local collapses and smaller cracks, as well as leaking joints between pipe segments. The intensity of sewer exfiltration to the soil and eventually to the aquifers depends on a variety of influencing factors, including pipe diameter and failure type as well as pipe burial depth and distance to the groundwater table. In this context, this study’s specific aim is to investigate the effect of selected vadose zone and aquifer properties and of failure characteristics on the final shape of sewer contaminant plumes to eventually delineate groundwater contamination characteristics solely from sewer network properties.

Results from two numerical studies, employing the HYDRUS 3D software code for variably saturated flow and transport simulation, will be presented. First, a small-scale principal model setup with a single pipe defect was designed to investigate the effect of soil type, colmation layer properties, pipe water level, defect shape and natural groundwater recharge on the shape of the plume in the vadose zone and at the aquifer table. Hereby, the simulations included both constant and varying pipe water levels. To define a de-facto worst-case scenario, continuous water injection as well as conservative transport (i.e., no decay or sorption) were assumed for most simulation runs. Besides the pipe water level, the intensity of precipitation was found to be a major influencing factor on the contaminant plume dimensions. In a second step, an intermediate-scale model involving a long pipe was conducted to further investigate overlay effects of multiple contaminant plumes. Here, multiple defects were positioned along the pipe in various distances, starting from a quasi-continuous line source and ending at a rather broad interval. It was found that the plume shapes on larger scale were very similar for most defect positionings, if the averaged injection rate remains the same. The direction of groundwater flow was altered in addition to the variation of the defects’ positions. Here, the contaminant plumes became slightly skewed.

The presentation will also give a short outlook to future works which will include simulations on city district scale employing HPC-capable codes such as ParSWMS, ParFlow and/or OpenGeoSys, and a comparison to simplified modeling approaches.