Safety · South Africa

Dam safety evaluation in South Africa

A dam safety flood evaluation in South Africa classifies the dam by wall height and downstream hazard, takes the recommended design discharge (RDD) and safety evaluation discharge (SED) for that category from the SANCOLD guidelines, routes both floods through the reservoir and checks the freeboard. Studio's Dam Safety Evaluation does each step in one workspace and shows the working behind every number.

What the evaluation covers

A dam with a safety risk in South Africa, one with a wall higher than 5 m that can store more than 50 000 m³, falls under the Dam Safety Regulations (GN R139 of 2012). Its flood evaluation follows the SANCOLD guidelines in four steps: classify the dam, set the two design floods, route them through the reservoir, and check that the freeboard covers the surcharge plus wind and waves. The Studio tool runs for South African dams and works through those four steps in order.

Step 1: classification

Size comes from the maximum wall height (SANCOLD 1991, Table 2.1). Hazard comes from the potential consequences of failure: no expected loss of life for low, up to 10 lives for significant and more than 10 for high, together with economic, social and environmental loss. The two combine in Table 2.3:

Size (wall height)Low hazardSignificant hazardHigh hazard
Small (> 5 m and < 12 m)IIIII
Medium (12 m to < 30 m)IIIIIII
Large (≥ 30 m)IIIIIIIII

A large dam is Category III whatever is downstream, because of the volume it can release. You can override the size or category where the Dam Safety Office has classified the dam differently, and the override is recorded.

Step 2: the RDD and SED

The recommended design discharge is a return-period flood set by size and hazard (SANCOLD 1991, as reproduced in SANCOLD 2011 Vol. II, Table 3.2):

SizeLow hazardSignificant hazardHigh hazard
SmallQ₂₀ to Q₅₀Q₁₀₀Q₁₀₀
MediumQ₁₀₀Q₁₀₀Q₂₀₀
LargeQ₂₀₀Q₂₀₀Q₂₀₀

For a small low-hazard dam the tool takes the upper end of the range, Q₅₀.

On the PMF basis the safety evaluation discharge is a fraction of the probable maximum flood. The fractions are those of the SANCOLD (1986) interim guidelines; the SANCOLD (1991) guidelines require the PMF itself where the fraction is 1.0 and scale it down for the other dams (Görgens et al. 2007, Table 5.2):

QSED=k QPMFQ_{SED} = k \, Q_{PMF}
SizeLow hazardSignificant hazardHigh hazard
Small0.20.50.7
Medium0.50.71.0
Large0.71.01.0

On the RMF basis, SANCOLD (1991) sets the SED as the regional maximum flood of the site’s K-region, or of the region one step lower (RMF−) or higher (RMF+) (Görgens et al. 2007, Table 5.1):

SizeLow hazardSignificant hazardHigh hazard
SmallRMF−RMF−RMF
MediumRMF−RMFRMF+
LargeRMFRMF+RMF+

The step goes to the neighbouring region of Kovács’s (1988) regionalisation (2.8, 3.4, 4.0, 4.6, 5.0, 5.2, 5.4, 5.6), and the tool evaluates its RMF with the Kovács (1988) equations at the catchment area. A medium high-hazard dam on an 85 km² catchment in region 5.0 has an RMF of 922.0 m³/s and an SED of 1 203.6 m³/s, the RMF of region 5.2.

You choose which basis the SED uses. For the RMF basis you enter or link a saved RMF calculation; where the table steps the K-region, the tool also needs the catchment area and the site’s Kovács (1988) K-region, and warns until it has them. The RDD peak can be typed in or linked from a saved flood calculation: Rational Method, regional maximum flood, QRT, flood frequency analysis or TR-55. The PMF can be entered from a separate rainfall-runoff study, or estimated with the Rational formula from a PMP depth that you type or link from a saved probable maximum precipitation lookup:

QPMF=C⋅(P/Tc)⋅A3.6Q_{PMF} = \frac{C \cdot (P / T_c) \cdot A}{3.6}
SymbolMeaningUnits
QSEDQ_{SED}Safety evaluation dischargem³/s
kkSANCOLD SED factor for the dam’s size and hazard–
QPMFQ_{PMF}Probable maximum flood peakm³/s
CCRunoff coefficient, typically 0.8 to 1.0 for a PMF–
PPPMP depth for a storm lasting TcT_cmm
TcT_cTime of concentrationh
AACatchment areakm²

PP must be the PMP for a storm lasting TcT_c. The national grid gives 1-day depths, so convert first with a PMP depth-duration relation for the site: dividing a 1-day depth by a TcT_c of a few hours overstates the PMF by the ratio of the 1-day depth to the TcT_c-duration depth. The tool does not make that conversion, and it warns when a linked or imported 1-day PMP meets a TcT_c under 24 hours.

Step 3: routing

Both floods are routed through the reservoir from full supply level with level-pool (modified Puls) routing, using the elevation-volume curve and the spillway rating: an uncontrolled ogee or a broad-crested weir. Without a linked or pasted hydrograph, the inflow is a triangle with time to peak tp=tb/2.67t_p = t_b / 2.67. The result gives the peak outflow, the maximum water level and its margin to the non-overspill crest (NOC) for each flood, and the spillway capacity at the NOC. The same storage-indication method is explained on the detention pond calculator page.

Step 4: freeboard

The freeboard check follows the SANCOLD (2011) freeboard guidelines. Wind set-up uses the Zuiderzee form,

S=U2F62 000 dS = \frac{U^2 F}{62\,000 \, d}

with UU the over-water wind speed in km/h, FF the fetch in km and dd the average depth in m. The significant wave height comes from the fetch and wind speed, and the run-up on the upstream face depends on the slope and its protection through a roughness factor:

Upstream slope protectionRoughness factor γf\gamma_f
Smooth concrete1.00
Rough concrete0.90
Exposed soil0.90
Grass0.85
Rip-rap, single layer0.55
Rip-rap, double layer0.50

The components are added in six combination scenarios. Category I uses scenarios 1 and 4; Categories II and III use 2, 5 and 6, which cover the RDD with wind and waves, the RDD with gate failure, and the SED surcharge with a landslide wave. The required freeboard is the largest scenario total (or the simplified minimum for the dam type, if larger), compared with the provided freeboard, NOC minus FSL.

Worked example

A hypothetical 28 m earthfill water-supply dam in the KZN Midlands has a significant hazard rating (fewer than 10 lives at risk). Its catchment is 85 km². The full supply level is 669.0 m a.s.l. with 24 million m³ in store, the NOC is at 673.0 m, and a 60 m uncontrolled ogee spillway (CdC_d = 2.18) has its crest at FSL. A flood study gives a 1:100-year peak of 285 m³/s and a PMF study gives 1 133 m³/s. Both hydrographs are triangles with a 16 h base, routed at 10-minute steps. The 1:100-year wind is 30 m/s over a 2.5 km fetch with an average depth of 8 m, and the 1:2.5 upstream face has double-layer rip-rap.

StepResult
Size, hazard, categoryMedium, significant, Category II
RDDQ₁₀₀ = 285 m³/s
SED0.7 × 1 133.3 = 793.3 m³/s
RDD routedOutflow 161.5 m³/s, level 670.14 m (1.14 m above the spillway crest)
SED routedOutflow 528.3 m³/s, level 671.43 m (2.43 m above the spillway crest)
Spillway capacity at NOC1 129.9 m³/s
Wind set-up, wave height, run-up0.08 m, 1.31 m, 1.31 m (over-water wind 36 m/s)
Governing scenarioRDD surcharge + run-up + set-up = 2.53 m
SED surcharge scenario2.43 m
Freeboard required / provided2.53 m / 4.00 m, sufficient

The reservoir attenuates the RDD by 43 % and the SED by 33 %. The governing case is the design flood with wind and waves, not the extreme flood, and the dam has 1.47 m of freeboard to spare. As cross-checks at the same site, the 1-day PMP from the national grid (Johnson and Smithers 2020) is 929 mm, and the regional maximum flood for 85 km² in K-region 5.0 is 922 m³/s on the Du Plessis and Masule (2023) curve.

Limitations

  • The tool does not model a dam break or map inundation. The hazard rating comes from your own consequence assessment, which the tool records.
  • The PMF has to come from a rainfall-runoff study or a justified Rational estimate on a storm-duration PMP. The SED is only as good as that number.
  • Routing is level-pool from FSL. Long reservoirs with a sloping water surface, gated operating rules and multiple outlets need a fuller reservoir model.
  • The default triangular inflow fixes the volume through its base time. Peak and volume both decide the routed level, so use a real hydrograph where you have one.
  • The SANCOLD tables are applied as tabulated: the 1986 PMF fractions and the 1991 RMF K-region steps. Evaluations saved before 8 October 2026 used 1.1 × PMF for large high-hazard dams and RMF multipliers of 0.4 to 1.7; reopening one recomputes it. Climate change allowances are a separate decision.
  • The classification must be agreed with the DWS Dam Safety Office, and the report signed by an Approved Professional Person.

What the full Studio tool adds

You can start from a dam’s record in the DWS register of dams, draw the embankment section, and switch the flow, routing, freeboard and consequences modules on or off. For Category II and III dams the workspace adds the crest survey and monitoring records the regulations ask for. A Method tab shows every equation with this dam’s values substituted, and the evaluation saves to your projects and exports as a Word report and an Excel workbook. Spillway ratings can be checked in detail with the weir flow calculator, and catchment design floods come from a design flood estimation.

FAQ

Common questions

Which dams need a safety evaluation in South Africa?
Under the National Water Act and the Dam Safety Regulations (GN R139 of 2012), a dam with a safety risk is one with a wall higher than 5 m that can hold more than 50 000 m³. Such dams are classified by size and hazard, and the owner must have them evaluated periodically by an Approved Professional Person. The flood evaluation is one part of that safety evaluation.
What is the difference between the RDD and the SED?
The recommended design discharge is the flood the spillway must pass with the dam undamaged and the normal freeboard intact, set as a return-period flood such as the 1:100 or 1:200-year. The safety evaluation discharge is the extreme flood the dam must survive without failing, even if the spillway and parapet are stressed. SANCOLD sets it as a fraction of the probable maximum flood, with the regional maximum flood as the other basis.
How is a dam's SANCOLD category decided?
Size comes from the maximum wall height: small is more than 5 m and less than 12 m, medium is 12 m to less than 30 m, and large is 30 m or more. Hazard comes from the potential consequences of failure downstream, rated low, significant or high. The two combine in a matrix to give Category I, II or III, which sets the flood criteria.
What freeboard does a dam need?
The freeboard is the height of the non-overspill crest above full supply level. It must cover the flood surcharge plus wind set-up and wave run-up in the governing combination for the dam's category, as set out in the SANCOLD (2011) freeboard guidelines. The required value depends on fetch, design wind speed, upstream slope and protection, so it is calculated, not read from a table.
Does the tool replace an Approved Professional Person?
No. It does the classification, flood criteria, routing and freeboard arithmetic and shows the working, but the classification must be agreed with the DWS Dam Safety Office and the safety evaluation report must be signed by an Approved Professional Person. Treat the output as the calculation file behind that report.
Full documentation: Dam Safety Evaluation
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 evaluate_dam_safety. The agent passes the inputs, the engine does the calculation, and the record is saved to your project for you to review.

References

  1. SANCOLD (1991) Guidelines on Safety in Relation to Floods. Report No. 4, South African National Committee on Large Dams, Pretoria.
  2. SANCOLD (1986) Interim Guidelines on Safety in Relation to Floods. Report No. 1, South African National Committee on Large Dams, Pretoria.
  3. SANCOLD (2011) Guidelines on Freeboard for Dams, Volume II. South African National Committee on Large Dams, Pretoria.
  4. Kovács, Z. (1988) Regional Maximum Flood Peaks in Southern Africa. Technical Report TR 137, Department of Water Affairs, Pretoria.
  5. Görgens, A., Lyons, S., Hayes, L., Makhabane, M. and Maluleke, D. (2007) Modernised South African Design Flood Practice in the Context of Dam Safety. WRC Report No. 1420/2/07, Water Research Commission, Pretoria.
  6. Department of Water Affairs (2012) Regulations regarding the safety of dams in terms of section 123(1) of the National Water Act, 1998. Government Notice R139, 24 February 2012.
  7. 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.
  8. Johnson, K.A. and Smithers, J.C. (2020) Updating the estimation of 1-day probable maximum precipitation in South Africa. Journal of Hydrology: Regional Studies, 32, 100736.

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.