Workflows · South Africa

Design flood estimation in South Africa

Design flood estimation in South Africa means running several of the methods in the SANRAL Drainage Manual on one catchment and choosing a design value from the spread. Studio's Design Flood Estimation does this from a single outlet point: it delineates the catchment on a hydrologically corrected Copernicus GLO-30 DEM, derives the parameters and runs up to eleven methods for the 1:2 to 1:200-year floods.

What the tool calculates

You place an outlet on the map. Studio delineates the catchment on a hydrologically corrected Copernicus GLO-30 (30 m) DEM, traces the longest watercourse and derives the parameters the South African methods need: area, the 10–85 slope, time of concentration, mean annual precipitation, land cover, soils, veld zone, SDF basin, RMF K-region and the NaFSAT 2.0 clusters. It then pulls the design rainfall and runs every method you ticked for the return periods you chose, from 1:2 to 1:200 years. The Studio tool runs for sites in South Africa only, because its rainfall grids and regional classifications stop at the border.

Before anything is charged, you see the delineated catchment and the outlet the engine actually used, and you confirm it. That step exists because a pin a short way off the channel, or on a tributary within about 150 m of a confluence, can return the wrong river’s catchment.

Most of the deterministic methods share the Rational form, with the SDF making the runoff coefficient depend on return period:

QT=CT IT A3.6CT=C2+YT2.33(C100−C2)100Q_T = \frac{C_T \, I_T \, A}{3.6} \qquad C_T = \frac{C_2 + \dfrac{Y_T}{2.33}\left(C_{100} - C_2\right)}{100}
SymbolMeaningUnits
QTQ_TPeak discharge for return period TTm³/s
CTC_TRunoff coefficient for return period TT–
ITI_TDesign rainfall intensity at duration TcT_cmm/h
AACatchment areakm²
C2C_2, C100C_{100}SDF basin calibration coefficients%
YTY_TReturn-period factor, Drainage Manual Table 3.13 (0.84 at 1:5, 1.28 at 1:10, 2.05 at 1:50, 2.33 at 1:100)–

The methods in one run

MethodFamilyOutputSource
Rational, three rainfall alternativesDeterministicPeak, triangular hydrographDrainage Manual §3.5.1
SCS-SADeterministicPeak, hydrographSchmidt & Schulze (1987)
Standard Design FloodDeterministic, regionally calibratedPeak, triangular hydrograph, and a Van Bladeren (2005) adjusted peak where a sub-catchment matchesAlexander (2002)
Unit HydrographSynthetic unit hydrographPeak, hydrographHRU Report 1/72
Direct Runoff Hydrograph (DRH)Storm-loss rainfall excess routed through a single linear reservoir (Muskingum, X = 0); no unit hydrographPeak, routed hydrographBauer & Midgley (1974); DWS flood frequency guide (2025) §4.4
RMFEnvelopeOne maximum flood, no return periodKovács (1988), Du Plessis & Masule (2023)
MIPI 7.1EmpiricalPeak per return period, 1:10 to 1:100 onlyDrainage Manual §3.6.3
CAPAEmpiricalPeak per return periodMcPherson (1983); DWS flood frequency guide (2025) §5.4
Index Flood, PRMRegional statisticalPeak per return periodCalitz & Smithers (2020), NaFSAT 2.0
QRTRegional statisticalPeak per return periodCalitz (2020), NaFSAT 2.0

The Rational run uses all three of the Drainage Manual’s rainfall sources: the Midgley & Pitman depth-duration-frequency curves (Alternative 1), TR102 station depths with the modified Hershfield equation (Alternative 2) and the RLMA&SI design-rainfall grid (Alternative 3). The runoff coefficient, area and TcT_c are shared, so the spread between the three is the uncertainty that comes from the rainfall database alone.

Catchment-area limits

The method list is gated by area applicability ranges, judged against a preliminary catchment area. Above a hard upper bound the method is disabled. Below a lower bound it stays available and is flagged as indicative. The Rational, SCS-SA and Unit Hydrograph ranges are those of SANRAL Drainage Manual (6th ed.) Table 3.2. Table 3.2 sets no area limit on the SDF or the empirical methods; the 10 – 40 000 km² SDF range is the one Studio applies.

MethodRecommended catchment areaUpper bound
Rational< 15 km² (advisory, DM §3.5.1.1)Flagged, still runs
Alternative Rational< 15 km² (advisory, same as Rational)Flagged, still runs
SCS-SA≲ 30 km² (advisory, Schmidt & Schulze 1987)Flagged, still runs
SDF10 – 40 000 km²Hard
Unit Hydrograph15 – 5 000 km²Hard
DRH15 – 5 000 km²Hard
RMF, MIPI 7.1, CAPAAny sizeNone
Index FloodAny size (advisory)None
PRM≲ 50 000 km² (advisory, NaFSAT)Flagged, still runs
QRT10 – 10 000 km² (advisory, NaFSAT)Flagged, still runs

The Drainage Manual also says MIPI 7.1 is suited to rural catchments larger than 100 km², so treat it as a check below that.

Why several methods are run and compared

No single South African method is calibrated for every catchment, and each family fails differently. The Rational and SCS-SA results depend on a runoff coefficient or curve number you can argue about. The SDF and unit hydrograph carry regional calibrations from older flood records. Index Flood, PRM and QRT were fitted to gauged rural floods and never see land cover. The RMF is an envelope above the largest recorded floods, so it bounds the others rather than competing with them.

When families built on different data agree, that agreement is the evidence for your design value. When they disagree, the pattern usually points at something specific about the catchment: urban cover, a dam upstream, dolomite, or an outlet near a cluster boundary. Studio puts every method on one table and one chart so the disagreement is visible before you choose.

Worked example

One road job, three outlets

A batch holds a 6.2 km² culvert catchment, a 26.4 km² stream crossing and a 7 200 km² river bridge, with every method ticked. The applicability check returns:

Method6.2 km²26.4 km²7 200 km²
RationalRunsFlagged above 15 km²Flagged
SCS-SARunsRunsFlagged above 30 km²
SDFFlagged below 10 km²RunsRuns
Unit Hydrograph, DRHFlagged below 15 km²RunsNot run (above 5 000 km²)
QRTFlagged below 10 km²RunsRuns
RMF, MIPI 7.1, CAPA, Index Flood, PRMRunsRunsRuns

Each site runs the subset its own area allows, so the river bridge does not stop the culvert from getting a unit hydrograph check.

The regional cross-check on the 26.4 km² catchment

This catchment lies east of Pretoria. The delineation gave AA = 26.45 km², MAP = 725 mm and TcT_c = 2.14 h, the 1:10-year 24-hour point rainfall is 97.8 mm, and the centroid sits 4.20 decimal degrees from the coast in NaFSAT cluster 31. Running the regional methods on those inputs gives:

Method1:10 (m³/s)1:50 (m³/s)1:100 (m³/s)
PRM34.673.497.6
Index Flood19.337.847.6
QRT20.424.024.2
RMF 2023 (K = 4.8, envelope)314314314

Index Flood and QRT agree at 1:10, then the three methods spread to a factor of four at 1:100. The land cover from the delineation is about 68 % urban, and none of the three regional methods can see that. Here the design value has to lean on the Rational and SCS-SA results, which do respond to impervious cover, with the regional methods as a rural baseline and the RMF as a ceiling.

Inputs and where they come from

You supply the outlet, the methods and the return periods. Outlets can be clicked on the map, pasted as coordinates, uploaded as a point layer (GeoJSON, Shapefile, KML, KMZ or GeoPackage) or taken from a saved watershed delineation. Everything else is derived, and an expert can override the parts that matter most: the Kovács TcT_c correction, the rainfall region for Alternative 1, the SCS-SA storm type, lag and rainfall duration, the RMF K-value, the NaFSAT cluster, the QRT DR10% and the coast distance.

Limitations and when to use something else

  • The catchment comes from a 30 m DEM and cannot be edited in the tool. Where piped urban drainage decides the boundary, draw your own catchment and use the Rational Method calculator.
  • The regional statistical methods were calibrated on rural gauged catchments. Urban cover, large dams and dolomite are outside their calibration.
  • Where a reliable gauge record exists at or near the site, at-site flood frequency analysis should carry more weight than any regional method. The stream gauge finder shows what records exist.
  • The RMF is an envelope and not a design flood. For dams, take the RMF and the PMF into a dam safety evaluation.
  • Hydrographs come from the Rational, SDF, SCS-SA, unit hydrograph and DRH methods. The empirical and regional methods return peaks only, so a storage or routing study needs one of the hydrograph methods, or the unit hydrograph generator on your own inputs.

What the full Studio tool adds

A submission can hold up to 200 outlets, with a bulk discount from 10 sites. Each site gets its own result page, and a batch page compares the peak flows by site with a CSV export. The results page puts every method on one peak-flow table and chart with a flood frequency curve, plots the hydrographs for each return period, and lets you hide a method from the view or set it aside with a written reason in the engineer’s review. Every finished estimate downloads as a Word report and an Excel workbook, with the catchment and watercourse as GeoJSON or Shapefile.

The price depends on the delineated catchment area and is shown on the confirmation step before you submit. You can save a half-finished request as a draft, and resubmit a finished one with changed parameters. The Index Flood method, the probabilistic rational method, the QRT calculator and the regional maximum flood also exist as standalone calculators for catchments you have delineated elsewhere.

FAQ

Common questions

Which flood method should I use for a SANRAL road project?
The SANRAL Drainage Manual (6th ed., 2013) does not name one method for every catchment. Table 3.2 gives each method an area range, and the usual practice is to run every method that suits the catchment and adopt a value with a written reason. The Standard Design Flood is normally expected on SANRAL work inside its range. Confirm the required methods and return periods in the project's drainage brief before you start.
What is the standard design flood (SDF)?
The SDF is Alexander's (2002) regional method for South Africa. The country is divided into 29 SDF basins, each with a calibrated 1:2-year and 1:100-year runoff coefficient. The coefficient for any return period is interpolated between them with the return-period factor, and the peak follows from the Rational form with design rainfall at Tc. Table 3.2 of the Drainage Manual sets no area limit on the SDF; Studio's Design Flood Estimation offers it for 10 to 40 000 km² and flags catchments below 10 km².
How do you estimate peak flow for an ungauged catchment in South Africa?
Without a gauge you combine methods that do not share inputs: rainfall-runoff methods such as the Rational, SCS-SA, SDF and unit hydrograph, regional statistical methods such as Index Flood, PRM and QRT calibrated on gauged floods, and the RMF envelope as an upper check. Where the families agree the value is well supported. Where they disagree, find out which assumption the catchment breaks before choosing.
Why do the methods give such different answers?
They rest on different data. The Rational and SCS-SA methods respond to land cover and soils, the regional methods respond only to area, rainfall and distance to the coast, and the empirical methods were fitted to older records. Urbanisation, dams and dolomite all push the families apart. The Rational result alone can differ by rainfall database, which is why Studio reports all three Drainage Manual rainfall alternatives.
Can I run a design flood estimate outside South Africa?
Not with this tool. The terrain model, design rainfall grids, SDF basins, veld zones and regional clusters it uses all cover South Africa only. For other countries use Watershed Delineation for the catchment, Design Rainfall where a national dataset is supported, and the Rational Method or TR-55 calculators for the peak flow.
What do I get when the estimate is finished?
A results page with the catchment, longest watercourse, land cover, design rainfall and every method's parameters, a peak-flow table and chart for each return period, and hydrographs where the method produces one. You can download a Word report, an Excel workbook, and the catchment and watercourse as GeoJSON or Shapefile.
Full documentation: Design Flood Estimation
Every input, option and equation in the Studio tool, with its limitations.
Run it from your AI agent
This engine is open to Claude, ChatGPT, Cursor and other MCP clients as quote_flood_estimate, submit_flood_estimate, submit_flood_batch and get_flood_request. The agent passes the inputs, the engine does the calculation, and the record is saved to your project for you to review.

References

  1. SANRAL (2013) Drainage Manual, 6th ed. South African National Roads Agency SOC Ltd, Pretoria. Chapter 3, Flood Hydrology.
  2. Alexander, W.J.R. (2002) The Standard Design Flood. Journal of the South African Institution of Civil Engineering, 44(1), 26–30.
  3. Smithers, J.C. (2012) Methods for design flood estimation in South Africa. Water SA, 38(4), 633–646.
  4. Hydrological Research Unit (1972) Design flood determination in South Africa. HRU Report 1/72, University of the Witwatersrand, Johannesburg.
  5. Schmidt, E.J. and Schulze, R.E. (1987) Flood volume and peak discharge from small catchments in Southern Africa, based on the SCS technique. WRC Report TT 31/87, Water Research Commission, Pretoria.
  6. Du Plessis, J.A. and Masule, S. (2023) Revised Regional Maximum Flood (RMF) method and regionalisation. Journal of the South African Institution of Civil Engineering, 65(3), 2–19.
  7. Calitz, J.P. and Smithers, J.C. (2020) Development and Assessment of Regionalised Approaches to Design Flood Estimation in South Africa. WRC Report No. 2748/1/20, Water Research Commission, Pretoria.
  8. Calitz, J.P. (2020) Development and assessment of regionalised approaches to design flood estimation in South Africa. PhD thesis, University of KwaZulu-Natal, Pietermaritzburg.
  9. McPherson, D.R. (1983) Comparison of mean annual flood peaks calculated by various methods. Technical Report TR119, South African National Hydrology Symposium Proceedings. Department of Environment Affairs and WRC, Pretoria.
  10. Department of Water and Sanitation (2025) A Guide to Flood Frequency Estimation Methods as Applied in the DWS, 11th ed. (version 2). DWS, Pretoria.

Last reviewed by the Delta Hydro engineering team.

Keep the whole study in one place

Delta Hydro Studio puts this tool alongside catchment, rainfall, runoff and hydraulics tools, with projects, team sharing and Word and Excel reports. Run them by hand, or let your AI agent run them for you to review.