Introduction

The purpose of this page is to provide a comparison between EPA-SWMM Street Inlets and HEC-RAS Inlets. A simple EPA-SWMM model including 4 Streets with Inlets was recreated in HEC-RAS, simulations were run and results were compared. This page explores the differences in how inlets are handled between HEC-RAS and EPA-SWMM, provides some insight into parameterizing inlets in HEC-RAS, and includes discussion on comparisons of results. The EPA-SWMM and HEC-RAS models for this comparison are available for download at the link below.

The example dataset is provided for demonstration purposes only and should not be used for other purposes. Some of the data within the dataset may have been altered specifically for the purpose of testing and demonstration and do not reflect actual conditions at the location. The example model has been simplified where possible for storage size and computation runtime purposes, and is not reflective of detailed study models. 

Download: Street Inlets Comparison.zip

Geometry Setup

The EPA-SWMM dataset was derived from the example datasets that are included in the EPA-SWMM setup package. The model includes 4 full Streets that share the same cross-sectional shape, each with 1 combination-style Inlet on either side of the street. A secondary drainage system is attached to the Street Inlets. The EPA-SWMM model geometry is shown below:

The dimensions of the combination inlets defined are shown in the diagram below. In addition to these dimensions, the "P-50" bar was defined for the grate. 

The most common application of pipe network Top and Side Inlets in HEC-RAS is to use them in combination with a 2D Area to which the inlets are connected. For modeling flow and inlets on a street in HEC-RAS, the underlying terrain should be a fine enough resolution to capture the capture the crown and cross slope of the street and the 2D mesh faces should be adjusted to capture the shape of the street. In this example a street terrain with identical dimensions to the EPA-SWMM Street was created by defining cross-sections and creating a terrain from them as described here. However, this step is not required for field datasets where the terrain model defines the street already. A cross-section of the street terrain is shown below.

A 2D mesh was created in HEC-RAS 2025 using quadrilateral regions, and imported into HEC-RAS 7.0.1 for the street surface. The mesh was created ensuring the cells faces were aligned perpendicular to the street flow, and the crown of the street was captured. with a cell face.

Next a Pipe Network was drawn in HEC-RAS Mapper with identical attributes to the EPA-SWMM model. A trunk line was drawn down the center of the street, and laterals were drawn to inlets at the curb on either side of the street.

The lateral lines connecting street inlets in this dataset were modeled using the Instantaneous Modeling Approach for conduits. This option allows flow into the inlets on the laterals to be instantaneously transferred to the receiving trunk node without performing the hydraulics or including the conduits in the pipe computational mesh. However, the hydraulic properties are computed for the laterals, and lumped into the receiving node cell. The benefits of doing this are faster computation times and a more stable model.

To create the combination inlet in HEC-RAS, a Top Inlet was first defined to represent the grate portion of the combination inlet. HEC-RAS computations for inlets most closely follow that of on-sag inlets from Federal Highways' HEC-22 (Hydraulic Engineering Circular 22) guidance - weir flow is computed for an inlet until it is appropriate to transition to orifice flow. HEC-RAS computes the flow from user entered weir length, orifice area, weir coefficient, and orifice coefficient attributes in combination with the computed head at the inlet. Having these attributes broken out individually gives the user some flexibility in deciding how the inlet will perform in a given setting. For instance, to make the inlet perform more like an on-grade inlet the weir length can be reduced to just the frontal length and a lower weir coefficient can be set.

The attributes for the top inlet were set to match the dimension defined in the EPA-SWMM model. For an on-sag inlet the Weir Length was set to the perimeter of the grate inlet minus the curb side (L+2W). The Orifice Area  was set to the area of the grate inlet multiplied by the Open Area Ratio of 0.9 for a "P-50" bar as provided in HEC-22. All other parameters were kept default.

Then a Side Inlet was defined to represent the curb inlet portion of the combination inlet. The side inlet was set to a Box shape with dimension that match the curb inlet size. All other parameters were kept default.

Reference Lines were added to the 2D mesh in HEC-RAS at stationing that matches the locations of the nodes between the EPA-SWMM Streets. The purpose of this is to provide a comparison of bypass flow in the streets after each set of inlets. The orange reference lines are shown and labeled below for each of the streets.

Boundary Conditions

The boundary conditions for this test are very simple. The 2D mesh has an external upstream flow boundary providing inflows to the street, and a downstream normal depth boundary allows the bypassing street flow to leave the model domain. The pipe network only receives flow from the inlets connected to the 2D mesh, and allows flow to leave the model domain with a stage boundary set low enough to mimic a free outfall. 

Results and Discussion

The EPA-SWMM model was run with inlets set to the on-sag position, and compared to the HEC-RAS model results setup as described above. Inlet capture and bypass flows are compared between the two models below. 

The 'Street1' node and Reference Line show the full boundary condition flow in each of the models because no inlets have been intersected yet. Note the SWMM and RAS flows are on top of each other in the plot below. 

The 'Street 2' node and Reference Line show the flow that bypassed the inlets at node 'J1' of the pipe network. As shown in the plot, about 1.2 cfs less flow (≈19%) is bypassing the inlets in the HEC-RAS model.

So that means the inlets at 'J1' are capturing more flow in the HEC-RAS model. Curiously, looking at the inlet flows and inlet depths at 'J1', the ponding depths are consistently higher in the EPA-SWMM results (0.43 ft vs 0.35 ft peak) which is likely due in part to the differences in surface routing (2D vs 1D Street) between the two models. Though EPA-SWMM shows a higher head on the inlet, HEC-RAS shows higher inlet capture flows demonstrating the HEC-RAS inlets are more efficient. This observation is confirmed when looking at the computed inlet rating curves for EPA-SWMM, HEC-RAS, and standard weir flow. 

In the inlet curve above you can see that for a given head, the EPA-SWMM inlet captures less flow than the HEC-RAS inlet. This difference is because in EPA-SWMM, the depth at the inlet is reduced to an effective-depth which accounts for the cross-slope of the street and the inlet depression geometry. On the other hand, HEC-RAS is simply using the depth at the inlet to compute flow, which is why the HEC-RAS results are right on top of the standard weir flow rating. 

The 'Street3' node and Reference Line show that the street flows that bypassed the inlet at node 'J2' are now nearly equal, and the inlet flow magnitudes at 'J2' are now reversed - EPA-SWMM captures more flow at 'J2'. This is because more street flow reaches 'J2' in the EPA-SWMM model because it captured less at 'J1' upstream. Similar to 'J1', the ponded depth at 'J2' is higher in EPA-SWMM (0.29 ft vs 0.16 ft) and this difference is the reason driving more inlet flow at 'J2' in the EPA-SWMM model. 

This same trend of inlet and bypass flows continue at inlets J3 and J4. Looking at the inlet / bypass flows for the whole system (in the plot below) tells the story: The more efficient inlets in HEC-RAS front load the capture at 'J1', but EPA-SWMM inlets quickly catch up due to more bypass flow and higher inlet ponding depths.

Conclusion 

A simple street inlet capture dataset was created in HEC-RAS and EPA-SWMM, simulations were run, and model results were compared. In comparing the results it was found that HEC-RAS inlets are more efficient than EPA-SWMM inlets because HEC-RAS uses the true head at the inlet to compute the inlet flow while EPA-SWMM uses a reduced effective depth at the inlet as recommended by HEC-22 (Hydraulic Engineering Circular 22). The rating curve plot below shows this difference, with HEC-RAS following the standard weir flow equation, and EPA-SWMM requiring a higher head for a given flow. 

Despite this per-inlet efficiency difference, the total flow captured across the four inlets ends up nearly the same in both models - the more efficient HEC-RAS inlets front-load capture at 'J1', while EPA-SWMM has deeper ponding and more bypass to the downstream inlets. In addition, the ponding depths in the EPA-SWMM model were noticeably higher. While a part of this was due to the difference in inlet flow computations, the street flow computations played a part as well. For the same street geometry without inlets, EPA-SWMM streets still produce higher water surface elevations in the street as shown in the plot below. Though the inlet depth differences are in the range of hundredths of feet, they still had an impact on flow capture and bypass results.