Sizing Stormwater Infrastructure

Sizing stormwater infrastructure is a calculation with four inputs and one very common omission — the route water takes when the storm is bigger than the pipe.

Sizing Stormwater Infrastructure — practical guidance from the team running stormwater management on the Garden Route. Below is how the calculation actually works, what a return period means, and why the minor and major systems have to be designed together.

The calculation, in plain terms

Peak flow from a catchment is estimated from three things: how much of the rain runs off rather than soaking in, how intense the rain is for the storm you are designing against, and how large the contributing area is. Multiply them and you have a flow rate, which sizes the pipe or channel.

The runoff coefficient represents how hard the surfaces are. A roof or a paved yard sheds almost everything; grass and garden absorb a substantial proportion. Development increases the coefficient, which is why it increases runoff.

Rainfall intensity comes from published data for the region and depends on both the return period you are designing for and the storm duration — and shorter storms are more intense. The relevant duration is the time of concentration: how long water takes to travel from the furthest point of the catchment to the point you are sizing. That is why a small steep catchment can produce a higher peak flow than a larger flat one.

Area is the contributing catchment, which is not the same as the site boundary. Water arriving from upslope land you do not own still has to be dealt with, and on Outeniqua footslope sites that external catchment can dwarf the site itself.

Return periods, and choosing one

A return period describes probability, not frequency. A 1-in-20 year storm has a five percent chance of being equalled or exceeded in any given year — it does not arrive once every twenty years, and two can happen in consecutive winters.

Different elements get different return periods, and the logic is consequence-based.

Minor system — road channels, catchpits, erf drainage, residential piped reticulation — is designed for a shorter return period. When it is exceeded, water flows overland, which is inconvenient rather than dangerous.

Major elements — culverts under roads, structures where failure means washout or property damage — are designed for a longer return period, because the consequence of exceedance is severe.

Both get designed together, which is the part most often skipped.

And there is a climate argument for headroom. Intensity statistics are derived from historical records, and recent years in many regions have delivered events above what those records suggested. Designing to the bare minimum is a choice; designing with some capacity in hand costs relatively little at construction and a great deal to retrofit.

The overland flow path

This is the single most important concept in stormwater design and the one most often absent from residential drawings.

When a storm exceeds the piped system’s capacity, the water does not vanish. It flows overland, following the topography. That route exists whether or not anybody designed it.

Good design chooses it deliberately — down a road reserve, along a designed swale, through an open channel, across a park — and makes sure it discharges somewhere that can take it. Bad design ignores it, and the water finds its own route, which is frequently through a garage, a lower-lying erf or a neighbour’s living room.

Practically, this means checking levels. Is there a continuous overland route from every low point to a legitimate discharge? Are building thresholds above the level water would reach along that route? Is a driveway falling toward a garage door the low point of a catchment?

It also means kerbs, thresholds and finished floor levels are part of the stormwater design, not just architectural details. A threshold 150 mm above the surrounding paving is a stormwater measure.

The elements, and how they get sized

Gutters and downpipes — sized on roof area and rainfall intensity. Undersized gutters overtop in exactly the storms that matter and send water down the wall.

Surface channels and kerbing — sized on the flow they collect and the gradient available. Flat channels silt up; steep ones outrun their outlets.

Catchpits and gullies — the entry points, sized on inlet capacity rather than pipe capacity. A large pipe fed by an undersized or blocked grating is limited by the grating. Silt sumps only work if somebody empties them.

Piped reticulation — sized on the design flow and laid to a fall that keeps velocity high enough to be self-cleansing but not so high that it scours. Access at every change of direction.

Culverts — sized for the major storm, with headwalls, and scour protection at both ends. The outlet is where they most often fail, because concentrated discharge onto unprotected ground cuts a channel in one storm and then works backwards.

Attenuation — storage sized on the difference between pre- and post-development runoff, with a restricted outlet that enforces the release rate. The outlet is the control; storage without it is just a pond.

Frequently asked questions

What does a 1-in-50 year storm mean?

That it has a two percent chance of being equalled or exceeded in any given year — not that it happens once every fifty years. Two can occur in consecutive years without anything being wrong with the statistics. Return periods describe probability, and they are chosen based on the consequence of exceedance: minor systems get shorter periods, culverts and structures where failure means real damage get longer ones.

How are stormwater pipes sized?

From an estimated peak flow, which comes from the contributing catchment area, how hard its surfaces are, and the rainfall intensity for the chosen design storm. Intensity depends on the storm duration, and the relevant duration is the time it takes water to travel from the furthest point of the catchment to the point being sized — which is why a small steep catchment can generate a higher peak flow than a larger flat one.

What is the overland flow path and why does it matter?

It is the route water takes once a storm exceeds the piped system. It exists whether or not it was designed — the only question is whether it runs down a road reserve or a swale, or through somebody’s garage. Designing the pipes and ignoring the overland route is the most common conceptual failure in residential stormwater, and it is what turns an unusual storm into damage.

Does water from neighbouring land count?

Yes, and it is frequently ignored. The contributing catchment is not the same as the site boundary — water arriving from upslope land you do not own still has to be dealt with. On Outeniqua footslope sites that external catchment can be substantially larger than the site itself, and cut-off drains upslope are then part of the design rather than an optional extra.

Why do catchpits block so often?

Because a catchpit is the entry point to the system and it collects whatever the surface delivers — leaves, silt, litter. The silt sump is designed to trap that material, and it only works while there is space in it. A catchpit that has not been emptied is effectively a solid lid, and no amount of pipe capacity downstream compensates. On the Garden Route, where rain falls all year, they fill continuously rather than seasonally.

Should I design with extra capacity?

There is a reasonable case for it. Rainfall intensity figures come from historical records, and recent years have in many places delivered events above what those records implied. Adding capacity at construction is comparatively cheap; retrofitting it into a completed development is expensive and disruptive. The judgement is about consequence — headroom matters most on the elements where exceedance causes real damage.

Related reading

• Stormwater design for Garden Route sites
• Building a road: the layerwork stack
• Stormwater management in George
• Drainage systems in George