Hydraulic & Hydrologic Modeling
A practical guide for water engineers: what each type of model does, when a project actually needs one, which free tools exist, and the QA checks that separate a defensible model from a colorful guess. Plus the roughness and runoff coefficients you’ll look up constantly.
Hydraulic vs hydrologic — the one-paragraph version
Hydrology answers “how much water shows up, and when” — rainfall becomes runoff. Hydraulics answers “what does the water do once it’s moving” — how it flows through pipes, channels, rivers, and structures. They chain together: a hydrologic model produces inflow hydrographs, and a hydraulic model routes those flows through the system. If someone says “we modeled the watershed,” they usually mean hydrology; “we modeled the storm drain system” means hydraulics.
| Hydrologic modeling | Hydraulic modeling | |
|---|---|---|
| Question | How much runoff from this rainfall? | Will this pipe/channel contain the flow? |
| Typical methods | Rational method, TR-55 / Curve Number, unit hydrographs | Manning’s equation, gradually-varied flow, full dynamic routing |
| Common software | HEC-HMS, SWMM (runoff), TR-55 worksheets | HEC-RAS, SWMM (routing), EPANET (pressurized networks) |
| Typical deliverable | Peak flow or hydrograph at each inlet | Water surface profiles, pipe capacity, flood extents |
Does this project need a model, or will hand calcs do?
- Hand calcs are enough when: a single pipe or channel, steady uniform flow, small simple watershed (rational method), or a quick feasibility check. Use the Manning calculator and rational method calculator.
- Build a model when: flows change over time (hydrographs), backwater or tailwater effects matter, networks with branches and loops, storage/detention routing, or the regulator explicitly requires it.
- The gray zone: a “simple” project with one tricky element (a culvert under tailwater, a looped water network) often justifies modeling that one element rather than the whole system.
Free software worth knowing
All of these are free, maintained by US federal agencies, and widely accepted by regulators:
| Software | By | Does |
|---|---|---|
| EPANET | EPA | Pressurized water distribution networks — pressures, flows, water age, chlorine residual |
| SWMM (Storm Water Management Model) | EPA | Urban rainfall-runoff and stormwater routing — the workhorse for storm drain design |
| HEC-RAS | US Army Corps of Engineers | River and channel hydraulics, floodplains, bridges/culverts, dam breach |
| HEC-HMS | US Army Corps of Engineers | Watershed hydrology — rainfall to runoff hydrographs |
Commercial options (WaterCAD, SewerCAD, InfoWorks, PCSWMM) add GUIs, CAD/GIS integration, and support contracts. Pick what your client and reviewing agency accept — a perfect SWMM model is useless if the reviewer requires HEC-RAS.
Model QA checklist
Run through this before any model result leaves your desk:
1. Mass balance. Inflow ≈ outflow ± storage change. If the model creates or destroys water, nothing else matters.
2. Boundary conditions documented. Every model edge needs a defensible assumption (normal depth? known WSE? rating curve?). Undocumented boundaries are the #1 source of wrong models.
3. Roughness justified. Manning’s n is the most-abused input in hydraulics. Every n value should trace to a table or field observation — see the lookup below.
4. Calibration vs validation. Tune the model on one storm/event, then test it on a different one before trusting it. A model that only reproduces its calibration data proves nothing.
5. Sensitivity check. Vary the top 2–3 uncertain inputs ±20%. If the design decision flips, you need better data, not a prettier model.
6. Physical sanity. Velocities in a plausible range? Flow directions correct? No water flowing uphill? Check the animation/profile, not just the summary table.
7. Datum and units. Confirm vertical datum (NAVD88 vs NGVD29 still bites people) and unit consistency — especially in models converted from imperial templates.
Before you build: data checklist
- Topography / DEM or survey cross-sections at the right spacing
- Rainfall data (design storm distribution + depth, or continuous record)
- Land use / imperviousness for runoff computation
- Existing infrastructure: pipe sizes, inverts, rim elevations (field-verify critical ones)
- Downstream boundary: tailwater elevation, tide, or receiving water rating
- Calibration data: at least one observed event with rainfall + flow/stage
Missing calibration data doesn’t always stop a project, but it changes what you can claim — an uncalibrated model is a planning-level estimate, and the report should say so.
Lookup: Manning’s n (typical values)
| Surface / material | n |
|---|---|
| Concrete pipe / finished concrete channel | 0.013 |
| PVC / HDPE pipe | 0.009–0.011 |
| Corrugated metal pipe | 0.022–0.026 |
| Brick / vitrified clay | 0.014–0.015 |
| Earth channel, clean | 0.022 |
| Earth channel, weedy | 0.030 |
| Gravel / riprap channel | 0.035–0.040 |
| Natural stream, clean banks | 0.030 |
| Floodplain, light brush | 0.050 |
| Floodplain, dense trees | 0.100–0.150 |
Typical values compiled from standard references (Chow, USACE, FHWA). Field conditions vary — adjust based on inspection.
Lookup: Rational method runoff coefficients (C)
| Surface | C |
|---|---|
| Asphalt / concrete pavement, roofs | 0.90–0.95 |
| Gravel roads / shoulders | 0.40–0.60 |
| Lawns, sandy soil, flat | 0.05–0.10 |
| Lawns, clay soil, steep | 0.25–0.35 |
| Residential (1/4-acre lots) | 0.40–0.50 |
| Residential (1-acre lots) | 0.30–0.40 |
| Commercial / industrial | 0.70–0.90 |
| Parks / cemeteries | 0.10–0.25 |
| Woodland, flat | 0.05–0.10 |
Typical ranges; check local stormwater manual — many jurisdictions publish their own C tables that govern.
Related calculators on this site
Manning’s Equation
Channel and pipe capacity, normal depth — the hand-calc behind every hydraulic model.
Weir & Orifice Flow
Outlet structures and control devices in detention and distribution.
Glossary
Calibration — adjusting model parameters (usually roughness, losses) so output matches observed data.
Validation — testing the calibrated model against independent data it wasn’t tuned on.
Boundary condition — the assumed behavior at the model’s edges (inflow hydrograph upstream, water level downstream).
Hydrograph — flow vs time at a point; the standard hand-off from hydrology to hydraulics.
Backwater — upstream water-level rise caused by a downstream constriction or high tailwater.
Normal depth — the depth where gravity and friction balance in uniform flow (Manning’s solution).
Critical depth — the depth of minimum energy for a given flow; controls behavior at drops and constrictions.
Water age — time since water entered a distribution system; EPANET tracks it for water-quality analysis.
This page is an introductory guide, not a substitute for software documentation or engineering judgment. Verify coefficients against your local design manuals.