Description of Study Area

Figure 2.1 illustrates the study area, including the Low Impact Developments (LIDs) and the detention pond. For this tutorial, the GeoSWMM storage unit representing the detention pond (originally created in Tutorial 03), SU1 has been resized to reflect a smaller Water Quality Capture Volume (WQCV). This adjustment accounts for the reduced runoff resulting from the implementation of LIDs.

As discussed in Tutorial 05, infiltration trenches were found to be the most effective LID type for decreasing the volume of water requiring treatment in the pond’s WQCV. As such, because the infiltration trenches can hold and treat 19,950.48 ft3 (see Table 2.5), that portion of runoff volume no longer needs to be handled by the regional detention pond. So, instead of the pond needing to handle the full 33,968 ft3 (as originally designed in Tutorial 03), we can subtract the amount managed by the trenches. That leaves the pond with a smaller volume requirement of 14,017.52 ft3 (33,968 - 19,950.48). Table 2.2 lists the model object quantity summary.

Along with the detention pond, the pond’s outlet structures were also redesigned to control this new WQCV and meet the general design criteria introduced in Tutorial 03 (e.g. 40 hr drawdown time for the WQCV and peak runoff of the 2-, 10- and 100-yr storms). Table 2.3 compares the storage curve of the pond designed in Tutorial 03 without LIDs and the pond designed for this tutorial with LIDs. Table 2.4 does the same for the dimensions and inverts of the orifices and weirs that comprise the pond’s outlet structure.

Table 2.2 : Network Object Summary

Table 2.3 : Storage Curve for the old and re-designed pond

Table 2.4 : Properties of the detention pond’s old and redesigned outlet structure

Table 2.5 : Infiltration Trench Volume calculation from Tutorial 05

Water Quality Treatment in LIDs

As described in Tutorial 05, the filter strip and infiltration trench LIDs are modeled as subcatchments in order to represent the combined effects of infiltration and storage on storm runoff. In this tutorial, we'll also look at how these LIDs help lower pollutant levels in the runoff they manage.

Since there are no universally accepted models that accurately simulate pollutant removal by LIDs, the most practical approach is to use average removal rates for specific pollutants based on field data from published literature.

GeoSWMM allows users to assign a constant BMP Removal Efficiency for each pollutant, based on the land use. At every time step in the model, the pollutant load from a specific land use is reduced by the percentage specified by the user. This reduction also applies to any runoff flowing into that subcatchment from upstream areas.

In this case, the LIDs added in Tutorial 05 receive runoff from upstream but don’t produce any pollutants themselves. To manage this limitation, in Tutorial_06_Initial.gdb, we’ll create a new land use category called “LID.” This category will be assigned a specific BMP Removal Efficiency for TSS, which will track pollutant reduction more effectively.

Water Quality Treatment in Detention Ponds

Detention ponds are modeled as storage unit nodes within GeoSWMM. By adding Treatment Functions to the storage node’s properties, the model can reduce the pollutant concentrations in the pond’s outflow. This tutorial uses an empirical exponential decay function to model solids removal through gravity settling within a pond. For controlling the WQCV event, the pond fills relatively quickly over a period of 2 hours and then drains slowly over an extended period of 40 hours during which solids removal occurs. At some interval t during this drain time, and assuming homogenous concentration, the fraction of particles with a settling velocity ui that are removed would be ui t/d where d is the water depth. Summing over all particle settling velocities leads to the following expression for the change in TSS concentration C during a time step t:

Where, C_t is the total concentration of the TSS particles at time tand f_i is the fraction of particles with settling velocity u_i. Because is generally not known, it can be replaced with a fitting parameter k and in the limit Equation 2.1 becomes:

Note that k has units of velocity (length/time) and can be thought of as a representative settling velocity for the particles that make up the total suspended solids in solution.

Integrating Equation (2.2) between times t and t + Δt and assuming there is some residual amount of suspended solids C*that is non-settleable leads to the following treatment function for TSS in the pond:

Equation 2.3 is applied at each time step of the simulation to update the pond’s TSS concentration based on the current concentration and water depth

Since there are no universally accepted models that accurately simulate pollutant removal by LIDs, the most practical approach is to use average removal rates for specific pollutants based on field data from published literature.

Simulating Treatment within a Conveyance Network

GeoSWMM can apply water quality treatment at any node of a drainage system’s conveyance network. Treatment for a node is defined by opening its Property Editor and selecting the ellipsis button next to the Treatment. This brings up a Treatment Expression dialog box in which the user can define a treatment function for each pollutant that passes through the node (Shown in Figure 2.2).

The treatment function for a given pollutant can have one of the following forms:

R = f (P,R_P,V)

C = f (P,R_P,V)

where R is the fractional removal, C is the outlet concentration, P is one or more concentrations given by the pollutant names (e.g., TSS), R_P is one or more pollutant removals (e.g., R_TSS), and V is one or more of the following process variables: FLOW (flow rate into the node), DEPTH (water depth above node invert), HRT (hydraulic residence time), DT (routing time step) and AREA (node surface area). Some examples of treatment expressions are:

R =

With a fractional removal expression, the new concentration at the node, C, is defined as C_in(1-R) where C_in is the inflow concentration to the node. Also, when a concentration P appears in an expression applied to a non-storage node, it is the same as C_in for the node whereas for a storage node it is the current concentration C in the storage unit.