Modeling Sanitary Sewer Groundwater Inflow Rehabilitation Effectiveness in EPA-SWMM5 Using a Two Aquifer Approach

The Broadview neighborhood in the northwest corner of Seattle, Washington, experiences frequent wet weather sanitary sewer overflows (SSOs) into the public rights-of-way and backups into private residences. Several sewer studies initiated by Seattle Public Utilities (SPU) have indicated groundwater inflow to the sanitary sewer system as the leading source of wet weather flows. Through a business case evaluation process, SPU selected flood grouting (Sanipor) as the preferred method of sewer rehabilitation for a pilot project. The business case, as well as a description of the technology and its application, can be found in a 2013 WERF paper “Flood Grouting for Infiltration Reduction on Private Side Sewers.”

Hydrologic and hydraulic modeling in the EPA-SWMM5 platform was used to calculate the effectiveness of the pilot project at reducing groundwater inflow. Flow monitoring deployed at the outlet of the rehabilitated project basin and at the outlet of the downstream non-rehabilitated basin allowed for hydrologic calibration of the basin wide model to pre- and post-rehabilitation conditions. Hydrologic calibration required a two aquifer approach to capture both a rapid response groundwater inflow signature as well as a long term seasonal signature. Capturing both signatures in the hydrologic model allowed for a more precise hydrologic calibration and a more accurate estimation of event infiltration volumes, especially in late wet season storms where the long term aquifer is more active. At the time of model calibration, very little was known about the hydrogeologic conditions of the area, which necessitated an estimation of how to set up the two aquifer approach. The existence of two aquifers was an unproven hypothesis. A subsequent hydrogeologic exploration of the area has proven the existence of a shallow unconfined aquifer as well as a deeper confined aquifer in the area. This exploration has confirmed the modeling approach and informed the conclusions that came from the modeling effort.

Recalibration of both the rehabilitated project basin as well as the non- rehabilitated downstream basin for post-rehabilitation conditions was required as the flow monitoring indicated a transfer of groundwater downstream once the upper project basin was rendered less porous. Long term simulations of the project basin model indicated the reduction in peak 5-minute groundwater inflow for a 10-year storm to be 33% and the reduction in average annual total groundwater inflow to be 69%. However, 73% of the groundwater removed from the rehabilitated project basin flowed downgradient and infiltrated into the non-rehabilitated lower basin in the form of the long term groundwater response. This paper will discuss the two aquifer groundwater modeling approach, the flood grouting effectiveness, the unintended consequences of a less porous sewer system, and how modeling in the EPA-SWMM5 platform informed the project’s conclusions.