SWMM 5 (and ICM SWMM and InfoSWMM) is a link-node model

SWMM 5 (and ICM SWMM and InfoSWMM) is a link-node model

The fact that SWMM 5 (and ICM SWMM and InfoSWMM) is a link-node model, where a significant portion (or even all) of a node's surface area can come from the connecting conduits, has several important implications for how the model behaves, how you should build and interpret models, and the overall accuracy of simulations.

Here are the full implications, broken down into key areas:

1. Representation of Physical Processes:

• Approximation of Storage: In reality, junctions in a drainage network often have some physical volume (e.g., manholes, catch basins). In SWMM 5, unless a node is explicitly defined as a storage node with a depth-area curve, its storage volume is approximated based on the surface area contributed by connected links.
• Simplified Geometry: The link-node representation simplifies the complex geometry of real-world drainage systems. It assumes that storage and flow transitions occur primarily at nodes, with links acting as connectors. This simplification can affect the accuracy of simulations, particularly in situations where the physical volume of junctions is significant compared to the volume of the links.
• Emphasis on Link Hydraulics: The model places a greater emphasis on accurately representing the hydraulic behavior within the links (conduits). The St. Venant equations are solved for the links, and the node behavior is derived from the link hydraulics.

2. Model Building and Parameterization:

• Node Sizing: When creating nodes that are not intended to be storage nodes, you don't need to explicitly define their surface area or volume, unless you use the Ponded Area Parameter of a Node. The model will automatically calculate the area based on connected links. However, this means that the node's storage capacity is directly tied to the dimensions of the connecting links. If the physical volume of the node is large you should use a storage node instead of a junction.
• Storage Nodes: For nodes that represent significant storage elements (e.g., ponds, detention basins), you must define them as storage nodes and provide a depth-area or depth-volume curve. This ensures that the storage capacity is accurately represented and not solely dependent on the link areas.
• Link Dimensions: Accurate representation of link dimensions (diameter, width, height) is crucial, as these dimensions directly influence the calculated node surface area and, consequently, the node's storage behavior.
• Ponded Area Parameter: SWMM 5 allows for a "ponded area" at nodes, representing the area available for temporary surface ponding above the node's maximum depth. This adds some flexibility to model surface storage that might exceed the node's normal capacity based on link areas.

3. Simulation Behavior and Accuracy:

• Numerical Stability: The way link areas are assigned to nodes (as discussed in the link flow classification) is designed to enhance numerical stability during the iterative solution of the St. Venant equations.
• Continuity Errors: Inaccurate link dimensions or neglecting the physical volume of junctions can lead to larger continuity errors, especially in systems with significant storage at junctions.
• Sensitivity to Time Step: The model's behavior can be sensitive to the routing time step, particularly if the node storage is dominated by link area contributions. Smaller time steps might be needed to accurately capture rapid changes in node depth.
• Approximation of Dynamic Effects: While SWMM 5's dynamic wave routing is sophisticated, the simplified representation of node storage can affect the accuracy of simulating highly dynamic events, such as rapid filling and emptying of junctions. 
  
  
  

4. Interpretation of Results:

• Node Depth as a Surrogate: Node depth in SWMM 5 is often used as a surrogate for the water level in a manhole or catch basin. However, it's important to remember that this depth might not perfectly correspond to the physical water level, especially if the node's storage is primarily derived from link areas.
• Storage Node Output: When analyzing storage nodes, pay close attention to the depth-area or depth-volume relationship you've defined. The reported depths and volumes will be based on this relationship, not on the link area contributions.
• Continuity Checks: Always check the continuity error reported by the model. Large errors can indicate problems with the model setup, including issues related to node storage representation.

5. Model Limitations:

• Simplified Representation of Junctions: The model's simplification of junction geometry can be a limitation in cases where the physical volume of junctions has a significant impact on the system's hydraulic behavior.
• Two-Dimensional Flow: SWMM 5 is fundamentally a one-dimensional model. It does not explicitly simulate two-dimensional flow patterns that might occur within junctions or at complex flow transitions. 
  
  
  

Best Practices to Mitigate Limitations:

• Use Storage Nodes: For nodes with significant physical storage, always define them as storage nodes with appropriate depth-area curves.
• Refine Link Discretization: In areas with complex hydraulics or significant storage at junctions, consider using shorter link lengths to better capture the spatial variation in flow and storage.
• Calibrate Carefully: If possible, calibrate your model using observed data. Pay attention to water levels at nodes and adjust storage parameters (e.g., depth-area curves, ponded area) as needed.
• Consider Alternative Modeling Approaches: For situations where the physical volume and geometry of junctions are critical, consider using a two-dimensional or coupled 1D/2D model, which can more realistically represent these features. 
  
  
  

In conclusion, understanding that SWMM 5 is a link-node model with node surface areas often derived from connected links is crucial for building accurate models, interpreting simulation results, and recognizing the model's limitations. While this approach simplifies the representation of drainage networks, it's important to be aware of its implications and take steps to mitigate potential inaccuracies, especially when modeling systems with significant storage at junctions or highly dynamic flow conditions.