Innovative Approach to Quantify RDII Sources Using SWMM Groundwater Module

Hazem Gheith, ARCADIS-US Inc, Columbus, Ohio, USA


An innovative approach was employed to generate the Rain Driven Inflow and Infiltration (RDII) using the groundwater module of the Storm Water Management Model (SWMM) by splitting the service area into sub-catchment features that correspond to the various RDII sources. Each RDII source was set to contribute to a subsurface aquifer. These aquifers represented the different manmade trenches within the service area. The hydrologic cycle, which included the surface runoff, evapotranspiration, surface infiltration, deep percolation and RDII processes, was appropriately configured for each RDII source using SWMM’s runoff, aquifer and groundwater modules. Hydrologic parameters for each RDII source were selected to generate RDII hydrographs that match the three RDII stages observed in the flow meter readings. In most old houses, there are no sump pumps (usage started during the 1960’s) to drain the foundation and water from downspouts splashes at a high intensity from the roof drains onto the house perimeter. Additionally, soil media around the house perimeter is typically a granular material at the foundation level topped with soil layers that are in a highly disturbed condition due to excessive roots and planting activities. These conditions cause the surface runoff to percolate at a high speed to the foundation. Also, due to construction practices in older houses, percolated flows could enter the sanitary system from the foundation drains, either through a direct connection or through the loose connections between the 4-inch house connection pipe and the 6-inch lateral service connection. Accordingly, a fast RDII stage would be expected from the house foundation. This condition is represented in SWMM by using high infiltration capacity in the buffer areas around the house perimeter and a low soil conductivity slope to accelerate the percolation process.

Lateral pipes are usually placed under a pervious surface at the front or backyard of the house and extended towards the main collector sewer. In addition to subsurface waters from the foundation drains, surface water over the lawn area could also percolate into the lateral pipe trenches. Lateral pipes are prone to high defect conditions due to soil movements and root intrusion. Also, as part of construction practices, the 6-inch lateral pipe is connected to the main collector sewer by punching a hole in the main sewer. This connection is typically performed poorly, causing a leakage path for the lateral trench subsurface water to enter the main sewer. Main sewer trenches could also collect water if placed under a previous surface or under cracked impervious pavements. Defects in the main sewer and the manhole structures can cause the trenched waters to enter the main sewer as RDII at a slower pace than the foundation drain. This configuration is represented in SWMM by assuming a less infiltration capacity than values assumed in the house perimeter. In addition to the fast RDII intrusion into the lateral pipes and main sewers from trenched waters, a slower RDII hydrograph is observed as the water from the natural, more condensed soil material outside the trenches starts to seep into the manmade trenches. This slow RDII seepage stage could extend for more than a month in clay soils.

The proposed approach was applied to over 100 meter basins that have long-term flow monitoring records spanning several years. Each meter had an average annual of 40 measurable RDII response events. Proposed approach predicted high accuracy between computed hydrographs and monitored data including the impact of back-to-back storms in the long term simulation. Calibration accuracy for predicted peak RDII flows and RDII volume was mostly within 15% from the measured peak flows and volumes. Collection systems models prepared with the proposed approach would facilitate a more accurate flow prediction in long term simulations using available climate and rainfall records.

Permanent link: