Transcribed from the conference paper PDF, with original section headings, bulleted lists, the architecture diagram redrawn as SVG, and the handwritten note from Bob Dickinson to Tony Kuch that was attached to the manuscript. Faithful to the source — including the typographic quirks and the truncated sentence fragments where the original scan had layout artifacts.
The Storm Water Management Model has a rich history of its application to analyze and optimize combined and separate storm and wastewater networks. The past decade (1990's) has seen the emergence of many graphical user interfaces (GUIs) for the Storm Water Management Model (SWMM). The most popular of these GUIs for SWMM are Windows SWMM, MIKE SWMM, PCSWMM, and XP-SWMM. These interfaces have increased modeling productivity, model understanding, model interpretation, and data integration to name just a few of the benefits. However, they do not integrate SWMM to the full scope of the design process that may include activities such as drafting, survey and right-of-way maps, surface modeling, watershed delineation, structures, cost calculations and the production of engineering plans and drawings.
This paper discusses the need for modifications and expansion of the design capabilities of SWMM, the need to create a flexible linkage with SWMM to geographic information systems (GIS) and the benefits of coupling SWMM to a comprehensive civil engineering design package. This paper also presents Visual Hydro as a comprehensive storm and wastewater management model fully integrated with CAiCE and its suite of tools for complete civil engineering design of storm and wastewater systems.
The Storm Water Management Model (SWMM) has been used by engineering design offices for more than 20 years. The model has been successfully applied to a range of water resources projects due to its comprehensive list of modeled processes and robust algorithms. Typical applications have included the analysis of storm water systems, capacity analysis of sanitary sewer networks, the prediction of combined sewer overflows (CSO) and sanitary sewer overflows (SSO), and to a lesser extent the inclusion of water quality to these models. SWMM is a deterministic model requiring physically descriptive data and that the system being modeled be adequately discritized.
To a large extent SWMM has not been used for the design of new storm and wastewater systems. The design of these systems, especially storm water systems is regulated by municipal, regional, and federal regulations as to model selection and methodology. The most popular hydrology methods being rational formula or SCS hydrology driven by IDF curves and design storms. Tabular hand calculations, spreadsheets, and software programs employing inlet ratings, manhole losses, the rational formula, nominal pipe size selection, and HGL prediction are typically used for the hydraulic analysis of storm water systems. Similar tabular methods are employed for wastewater networks that include in addition to the above list dry weather flow (DWF) peaking factors, infiltration and inflows, and future growth scenarios.
There are several reasons the SWMM is not more widely used for the design of storm and wastewater systems. Some of these reasons are financial, political, legal, related to software usability and because of model deficiencies. This paper will present some of these reasons and some software solutions to the barriers that prevent a more widely usage of SWMM.
has been excluded due to several internal and ex[scan break] reasons SWMM does not have a sophisticated pipe Over the past 10 years the usage of SWMM has increased
Visual Hydro with its accompanying modules and productivity tools expands the design, visualization, customization and production capabilities of Visual SWMM (formerly XP-SWMM32). Networks can be digitized on AutoCAD or Microstation background drawings with the network data now stored in an ACCESS database. Visual Basic for Applications (VBA) allows the user to customize the software including the link to ArcView and other GIS and database software. This functionality is common to all Visual Hydro Products. The Visual SWMM module of Visual Hydro allows the comprehensive simulation of stormwater and wastewater networks.
This product is an add-on to CAiCE survey and/or design products. It allows storm, sanitary and water distribution networks to be laid out in plan and profile sheets and to be combined with survey data and terrain modeling. It also contains the analytical capabilities for inlet spread and analysis, flow calculations in roadway cross-sections and links to HEC2 and HECRAS for open channel modeling.
Visual Hydro For Drainage is composed of Visual UDD and the Visual Hydro Manager. Visual UDD is a friendly graphics-based storm water decision support system. It is used to model the hydrologic cycle, including storm water runoff and hydraulic simulation of drainage networks. The Visual Hydro Manager incorporates Visual UDD into the CAiCE suite of civil engineering design and visualization products allowing networks to be created from survey data or loaded from an AutoCAD or Microstation drawing.
Visual Hydro For Drainage combines advanced digital surface modeling capabilities with water resource computations. Surface models constructed from spot elevations and breaklines can be visualized in either wireframe or shaded mode. You can also use these surfaces for delineating watershed boundaries, drawing water flow paths, constructing cross sections for open channels, and defining pipe and manhole elevations.
Another important aspect of Visual Hydro For Drainage is its visualization capabilities. Visual Hydro Manager draws pipe networks to scale as shaded or wireframe 3D objects, integrated with ground surfaces, topography, survey features, and other geometry objects. Conduits are drawn with a variety of shapes and colors, and manholes are shown with the correct dimensions. The hydraulic grade line and energy grade line can be superimposed on existing and proposed surfaces on profile sheets. The graphics engine and user interface make viewing operations such as panning, zooming, drivethroughs, and flyovers simple and intuitive. Visualization makes it easy to spot design flaws such as pipes above the ground surface and conflicts with other utilities.
Visual Hydro uses Microsoft VBA to directly link with a multitude of other Windows applications, including Microsoft Office, ArcView, AutoCAD, and Microstation. With the Visual Hydro GIS Interface, you can capture storm and wastewater networks from GIS databases for design and analysis modeling. Modeling results and changes and additions to the network are updated in the GIS. Survey data, surface models, and network geometry can be imported from Autocad and Microstation, and all graphics can be exported to CAD systems.
You will find that Visual Hydro for Drainage and Stormwater Management is a complete software package that combines an advanced digital terrain manager (Visual DTM) with a comprehensive hydrology/hydraulics package (Visual SWMM). Using this integrated package the 2 dimensional and 3 dimensional data representing the watershed is intimately connected to the hydrology and hydraulics analysis. The drainage network is generated from either survey or geometry chains in Visual Hydro with the generated data consisting of invert elevations, ground elevations, and conduit lengths. Visual Hydro can then show the simulation results as 2 and 3 dimensional profiles embedded in the DTM. It is both a revolutionary and evolutionary new means of designing and analyzing storm and sewer networks.
The features of Visual Hydro® can be grouped into several categories: Decision Support or Intelligent Graphics, Plans and Report Production, Modeling or Engineering Capabilities and Integration with other software such as ArcView. The Visual Hydro® Family of products shares the same graphical features and capabilities and they only differ in the engineering or modeling options. For example, CAiCE for Storm Water Management does not have the water quality capabilities of CAiCE for Storm and Wastewater Management Modeling. Although all of the graphical tools would be common to both, Water Quality modeling would only be available in the latter. This allows Visual Hydro® to be customized to a specific discipline or area of expertise.
Features within the Visual Hydro Manager include the capability to: create drainage networks from CAiCE survey and geometry data, store the node and link elements and their data within a Microsoft Access database and edit the networks and object data within CAiCE. These networks are drawn realistically and can be viewed in 3D with other CAiCE objects such as the DTM surface, survey chains and points, and roadway alignments. The network objects employ the full capabilities of CAiCE's feature tables allowing manholes, ponds, pipes and channels etc. to be displayed with common properties such as color, cell or line weight. The intelligence of an object is intrinsic in the database. The database stores much more than the color attributes, it also stores the network model data.
Many Hydrologic and Hydraulic Decision Support Graphic features exist within Visual Hydro. Figure 4 shows some of the features for hydrology, hydraulics, water quality, analysis, and graphics in the combination of Visual Hydro and its Visual SWMM Module. Some of the more common themes and applications of Visual Hydro are discussed in the following subsections.
This component is used to define the geometric layout of the drainage network without doing any hydrology or hydraulic calculations. Its capabilities include:
For example, the user is able to take data from EPA SWMM and XP-SWMM and define the network geometry for visualization and plotting purposes.
Hydraulics deals with the calculation of water flow characteristics within the drainage pipe system. The hydraulics model using the 1 dimensional dynamic wave equation to calculate the flow through the conduits based on the difference between the upstream and downstream hydraulic grade line, the conduit roughness, and the upstream and downstream cross sectional area difference. The hydraulics features and components include:
Visual Hydro has the ability to define the network geometry and characteristics. It performs hydraulic calculations that enable the user to design and analyze the drainage system. If the hydrology component is not used the amount of water flowing into each inlet would have to be determined elsewhere and provided as input to the system in a user defined inflow hydrograph.
Visual Hydro allows the user to define the input necessary for many hydrology methods. It performs the computations for the rational method and exports the results as a triangular user defined hydrograph. It is capable of:
Hydrology is the conversion of rainfall data into water runoff, and calculation of the volume of water flowing into the drainage system. The hydrology components includes:
Visual Hydro can generate cross sections to be used in the dynamic modeling of water flow in open channels, streams, ditches, and rivers. These channels are simulated in the hydraulics layer of the Visual SWMM module. The data is exported as either a HEC-2 or HEC-RAS cross section using the command File=>Export Translators=>to Hec=>X Section to HEC-2.
The spread analysis table is used for computation of how far the water flowing on a road (in a storm) will spread. Computations take into account the spacing and type of drainage inlets along the road, longitudinal and transverse slope of the road and properties of the area contributing (such as the total area, the imperviousness and the time of concentration) to the stormwater reaching each inlet. This analysis is required for urban stormwater drainage design and enforces the requirements for the maximum submerged area as well as the maximum depth of water flowing on a road in a design storm.
Visual Hydro can be used to design sizes for culverts based on inflow and outflow information and culvert characteristics (size, shape, entrance and exit loss, number of barrels etc.). The flow in the culvert is estimated using the 1D dynamic wave equation.
The calculation and modeling of stand-alone retention ponds and basins, mostly for site design applications is done in Visual Hydro. The hydrology data input component can optionally be used to provide the input data for the volume of water to be handled.
Handwritten note attached to the manuscript, addressed to co-author Anthony Kuch.
The Toronto paper made a case that the EPA Storm Water Management Model had a 20-year track record as a research and analysis tool, but had not been adopted for the everyday design of storm and wastewater systems. The authors argued the gap was not in the science of SWMM. It was in the surrounding workflow: drafting, survey, right-of-way, surface modeling, watershed delineation, structures, cost calculations, and plan production.
Visual Hydro, built on the CAiCE civil engineering platform, was offered as the bridge. The Visual SWMM module gave you the EPA SWMM engine. The surrounding CAiCE environment gave you survey data, terrain modeling, AutoCAD and MicroStation interoperability, an Access database, and VBA customization. The promise was a single environment where the same network you analyzed was the network you drew, surveyed, and submitted.
Two things in the paper aged well. First, the diagnosis that SWMM was not failing on science but on workflow integration. Second, the prediction that a CAD-native modeling environment was the right home for design work. Visual Hydro did not survive the 2002 Autodesk acquisition of CAiCE, but the architectural argument it made is essentially what ICM plus Forma is doing today, and what BIM-for-water continues to push toward.
The paper grouped the barriers into five categories. They have not entirely disappeared, but each has shifted.
| Barrier | What the paper said (circa 2000) | Status in 2026 |
|---|---|---|
| Financial | Commercial SWMM GUIs cost more than the spreadsheet-and-Rational-Formula approach most municipalities already paid for. | EPA SWMM5 is free. PCSWMM, ICM, InfoSWMM, XPSWMM, Civil 3D Storm and Sanitary, and InfoDrainage all compete. Cost is no longer the gate. |
| Political / Regulatory | Many agencies prescribed Rational Formula or SCS in regulation. Dynamic wave routing was not in the rule book. | Many large municipalities now require dynamic wave for capital projects. Smaller jurisdictions still default to Rational and SCS. |
| Legal | Engineers signed and sealed using methods their state board recognized. SWMM dynamic wave was new territory. | Largely resolved. SWMM5 is the EPA reference implementation and accepted nationwide. |
| Usability | EPA SWMM was a Fortran card-image program. The interfaces of the era helped, but design-grade workflows (inlet sizing, profile sheets, plan production) were missing. | SWMM5 ships with a polished GUI. ICM and Civil 3D extend it to plan and profile. Vibe-coded browser apps fill smaller niches. |
| Model deficiencies | SWMM lacked a real pipe-sizing routine. EXTRAN handled hydraulics but not design iteration. | SWMM5 itself still does not size pipes; the design loop lives in the GUI layer. ICM, XPSWMM, and Civil 3D add it. EPA SWMM5+ is exploring direct design routines. |
The paper noted that the cost of a commercial SWMM seat in the late 1990s was a serious barrier for small-to-mid municipalities. Free Fortran source from the EPA helped, but the supported, GUI-driven products carried real license fees. Twenty-five years later the calculus has inverted: the engine is free, the GUIs span free to premium, and the marginal cost of trying a model on a project is essentially zero.
Visual SWMM was the rebranded XPSWMM32 product line moved into the CAiCE environment. It ran from VBA code inside a CAD engine, with model data stored in a Microsoft Access database. The computational engine was still SWMM — the Fortran RUNOFF, TRANSPORT, and EXTRAN blocks — but the surrounding workflow was now drafting-aware.
Architecture diagram from the original paper, redrawn for clarity.
Visual Hydro for Drainage paired Visual UDD — a hydrology and hydraulic decision support tool — with the CAiCE digital terrain manager. Surfaces built from spot elevations and breaklines drove watershed delineation, pipe inverts, and channel cross sections. Networks rendered as shaded 3D objects superimposed on the DTM, with HGL and EGL drawn on profile sheets against existing and proposed grade. The paper's explicit pitch was that 3D visualization made design flaws — pipes daylighting above ground, utility conflicts — immediately visible.
The author list reflects a specific moment. Bob had been SVP at XP Software 1992-1999, where he designed the .xp and .xpx file formats and the XP visual engine that wrapped SWMM3 and SWMM4. He then joined CAiCE Software as lead developer of the Visual suite: Visual SWMM, Visual Hydro, Visual Culvert, Visual Inlets. The Toronto paper sits at the end of that arc. Two years after publication, Autodesk would acquire CAiCE for its road and highway tools, and the water products would be discontinued. The same architectural argument the paper makes — that water modeling belongs inside the design environment — would not get another serious commercial push for almost two decades.
The original paper enumerated each Visual Hydro module. Here is the same list with the SWMM5 equivalent or successor for each capability.
| Visual Hydro module | What it did | SWMM5 / modern equivalent |
|---|---|---|
| Hydro Mapping & Plotting | Defined inlet and manhole locations from survey and geometry chains. Pipe sizes and characteristics. 2D and 3D visualization. Spreadsheet editing. Plotting on plan, profile, and cross-section drawings. | SWMM5 GUI handles the model side; CAD and GIS integration moves to ArcGIS Pro, ICM, Civil 3D, or InfoDrainage. The Access database has been replaced by section-based .inp files and modern geodatabases. |
| Hydraulics | 1D dynamic wave routing through conduits. HGL over time. Interconnected ponds and open channels. Surcharge-based pipe sizing. | SWMM5 FLOW_ROUTING DYNWAVE in [OPTIONS]. The St. Venant solver was rewritten in C as dynwave.c, descended from the original EXTRAN Fortran. See the Dynamic Wave Lab tab. |
| Hydrology data input | IDF.TBL user-editable rainfall tables. Design-storm selection. Horton and Green-Ampt infiltration parameters. Watershed boundary tools. | SWMM5 has [RAINGAGES], [TIMESERIES], [INFILTRATION] with Horton, Modified Horton, Green-Ampt, Modified Green-Ampt, and Curve Number options. Watershed delineation lives in GIS or terrain tools, not in the engine. |
| Hydrology calculations | EPA Runoff method, SCS, Rational, Santa Barbara Urban Hydrograph. Watershed modeling from DTM. Runoff to inlets. | SWMM5 retains the nonlinear-reservoir Runoff method as its native approach. Rational and SCS workflows continue in design tools alongside SWMM5 for analysis. |
| Open channel / river analysis | Cross-section generation, dynamic routing in the EXTRAN Hydraulics layer, export to HEC-2 and HEC-RAS. | SWMM5 [TRANSECTS] + dynamic wave handles natural channels. HEC-RAS remains the standard for detailed riverine work; coupling is now done via 1D-2D in ICM or via file exchange. |
| Inlet spread analysis | HEC-22 style: spread on the road as a function of inlet spacing, longitudinal and transverse slope, contributing area, imperviousness, time of concentration. | SWMM5 added [STREETS], [INLETS], and [INLET_USAGE] sections in 5.2, putting HEC-22 inlet hydraulics directly in the engine. See the SWMM5 Theory Update tab. |
| Culvert and weir design | Inflow/outflow culvert sizing with entrance and exit losses, barrel count, 1D dynamic wave estimation. | SWMM5 supports culverts via [XSECTIONS] with culvert codes and an HDS-5 inlet-control check. Weirs and orifices have native sections with coefficients matching the design literature. |
| Retention / detention pond design | Stand-alone pond sizing for site design, optionally driven by Visual Hydro hydrology output. | SWMM5 [STORAGE] nodes with functional, tabular, or shape-based stage-storage curves; [OUTLETS] for rating-curve discharge. LID controls extend this to bioretention and infiltration ponds. |
The Toronto paper documents the world as it was when SWMM4 was still the production engine and EXTRAN was still Fortran. The complete rewrite to SWMM5 in C, started around 2002 and released in 2005, changed almost every aspect of how the engine works. Here is what would have to be added to the paper today.
EPA SWMM5 replaced the 80-column card-image input with a section-based ASCII .inp file. Each section is bracketed by [SECTIONNAME] and contains whitespace-delimited records. Section order is irrelevant — the parser reads everything before simulating. There are now 40+ sections covering project setup, hydrology, hydraulics, loading, water quality, controls, and visualization.
The XPX file the paper described was Bob's interchange format at XP Software — intentionally simple, [NODE] and [LINK] sections that survive in spirit in SWMM5's bracket-section syntax. The card-image .xp format with its GROUP + CARD + Index structure defined in Swmfield.c is no longer the input path, but legacy XP files still exist and Bob has published a parser for them.
SWMM4 did not have native RDII. Sanitary modelers calibrated external sources and inflow time series. SWMM5 added the [HYDROGRAPHS] and [RDII] sections, implementing the three-triangle RTK unit hydrograph (R·T·K for each of fast, medium, and slow response) convolved with rainfall. Bob coded rdii.c during the EPA SWMM5 CRADA at CDM Smith. See the RDII / RTK Lab tab for an interactive triangle builder.
Low Impact Development controls — bioretention cells, green roofs, infiltration trenches, permeable pavement, rain gardens, rain barrels, rooftop disconnection, vegetative swales — were not in any version of SWMM4. They appeared as [LID_CONTROLS] and [LID_USAGE] in SWMM 5.0 and have expanded since. Each control is a stack of layers (surface, soil, pavement, storage, drain, drainmat) with explicit water-balance physics.
SWMM4 offered Horton and Green-Ampt. SWMM5 added Modified Horton (recovery of infiltration capacity during dry periods), Modified Green-Ampt (better behavior at small rainfall intensities), and SCS Curve Number. The choice is set in [OPTIONS] under INFILTRATION. See the Infiltration Lab tab.
Visual Inlets and Visual Hydro had to compute spread and inlet interception externally and feed inflows into SWMM as boundary conditions. SWMM 5.2 added [STREETS], [INLETS], and [INLET_USAGE] — gutter geometry, HEC-22 grate, curb, slotted, and combination inlets, and the capture-bypass logic at every junction. The paper's inlet-spread analysis is now an engine feature.
EXTRAN's explicit half-step scheme was replaced in SWMM5 with an iterative Newton-Raphson solution at each link, with surface area weighting at junctions for stability. The variable timestep was reworked. Backwater, flow reversal, and supercritical-to-subcritical transitions are handled more cleanly. The lineage is direct — EXTRAN Fortran descended into dynwave.c — but the numerics were rebuilt.
Two-zone aquifer modeling with lateral and deep groundwater flow; snowmelt with separate snowpack accounting per subcatchment subarea; pollutant buildup and washoff with explicit functions; rule-based [CONTROLS] that read state variables and modify links during simulation. The paper had hooks for water quality through WASP5; SWMM5 internalized most of it.
The SWMM5+ Technical Advisory Committee at CIMM is actively developing the next-generation engine, with parallel computation, improved force-main physics, and direct design routines. SWMM2D and OpenSWMM 5.3/6.0 are extending into 2D overland routing. Caleb Buahin's HydroCouple architecture is moving toward standardized coupling with receiving-water and groundwater models. The Toronto paper's instinct that SWMM needed to be embedded in a richer environment is still driving the roadmap.
The paper's Hydraulics section described 1D dynamic wave routing as the calculation of conduit flow from upstream and downstream HGL difference, roughness, and area. This widget shows that same calculation for a single circular pipe at quasi-steady state, with Manning's equation as the closure that SWMM5's dynwave.c uses inside its iterative loop.
The paper mentioned Horton and Green-Ampt as the two infiltration options. SWMM5 added Modified Horton, Modified Green-Ampt, and SCS Curve Number. Toggle the method and watch the cumulative loss curve change.
Rainfall-Derived Infiltration and Inflow was not in the engine when the Toronto paper was written. SWMM5 added it via the RTK method: three triangular unit hydrographs (fast, medium, slow) each parameterized by R (fraction of rainfall that becomes RDII), T (time to peak in hours), and K (recession ratio). The three triangles are convolved with rainfall and summed.
The Toronto paper is a snapshot from one specific year on a much longer timeline. Here is where it sits.
| Era | What was happening | Where this paper fits |
|---|---|---|
| 1971–1973 | EPA contracts the original SWMM development. Metcalf & Eddy, University of Florida, and Water Resources Engineers build the first version. Fortran on mainframes, punch-card input. | Pre-paper. Wayne Huber at UF leads the academic side; Bob Dickinson starts coding SWMM3 at UF as an undergraduate. |
| 1974–1988 | SWMM3 and SWMM4 mature. Huber, Dickinson, Barnwell, and Branch publish the SWMM4 User's Manual in 1988 (850+ citations to date). EXTRAN block becomes the standard dynamic wave routine. | The base SWMM science the paper builds on. |
| 1992–1999 | XP Software (Australia) wraps SWMM3/4 in a graphical environment. Bob, as SVP, designs the .xp and .xpx file formats. XPSWMM becomes the dominant commercial SWMM platform internationally. | The XPX file format the paper diagrams (Visual Hydro → XPX → Visual SWMM Interface → .DAT) is Bob's design from XP Software. |
| ~2000 — this paper | Bob moves to CAiCE Software. Visual SWMM is the rebranded XPSWMM32. Visual Hydro, Visual Culvert, Visual Inlets fill out the drainage design suite. All four run from VBA inside AutoCAD or MicroStation. | This paper. |
| 2002 | Autodesk acquires CAiCE for its road and highway tools. An Autodesk VP famously tells Bob in Clearwater that he “doesn't understand the role of rainfall in road design.” The water tools are discontinued. | Two years after this paper, the products it describes are gone from the market. |
| 2002–2005 | EPA funds a complete C rewrite of SWMM. Lew Rossman is lead author; Bob codes the RTK/RDII unit hydrograph (rdii.c) under the CRADA at CDM Smith. SWMM5 ships in 2005. |
Almost all the “new” features in the SWMM5 Theory Update tab originate here. |
| 2006–2021 | Innovyze builds InfoSWMM and InfoSewer as ArcMap extensions wrapping SWMM5. Bob is Product Sector Leader for 18 years. PCSWMM, MIKE+, and ICM (originally Wallingford / HR Wallingford) compete. | The commercial SWMM landscape consolidates around four or five products, all built on the SWMM5 engine. |
| 2021 | Autodesk acquires Innovyze. Bob joins Autodesk as a Water Technologist. The arc closes: code from the CAiCE era is back inside Autodesk, via a different acquisition path two decades later. | The paper's premise — that water belongs inside the civil design environment — finally has serious corporate backing. |
| 2022–2026 | SWMM 5.2 adds streets and HEC-22 inlets natively. The SWMM5+ TAC at CIMM coordinates the next major engine release. SWMM2D and OpenSWMM extend toward 2D and coupled simulation. The March 2026 ArcMap discontinuation forces large-scale migration from InfoSWMM and InfoSewer to ICM. | Current era. |
The single most durable claim in the Toronto paper is that SWMM's adoption gap was a workflow problem, not a science problem. Twenty-five years later that is still the right diagnosis. The engine is free, the science is settled enough for production use, and the remaining friction is in moving between the design environment, the model environment, and the GIS environment cleanly. ICM plus Forma, Civil 3D Storm and Sanitary, and the vibe-coded apps that fill the gaps between the heavyweight tools are all chasing the same goal the paper described.