Cities were drained, and now they are being un-drained

There is a settled way of talking about urban drainage. It is worth asking how much of it survives contact with the detail.

The old strategy was speed

Twentieth-century urban drainage aimed to collect rainfall and move it into a pipe network and out of the city as quickly as possible. That works well until the receiving watercourse cannot take the volume, at which point the problem has simply been relocated downstream. It also removes water from the ground where it fell, which contributes to dry soils, dying street trees and subsidence.

The alternative approach reverses the objective: slow water down, hold it locally, and let some of it soak in.

Sustainable drainage is a set of ordinary things

Permeable paving, rain gardens, swales, detention basins and green roofs each hold a small volume and delay its release. Individually they are minor; distributed across a catchment the cumulative effect on peak flow is substantial.

They also filter pollutants, which addresses a water quality problem that pipes make worse. They work on frequent moderate rainfall and are overwhelmed by the rare extreme event, so the honest claim is that they reduce how often a system floods rather than that they prevent the flood anybody remembers.

Combined sewers are the awkward legacy

Many older cities have sewers carrying both foul water and rainfall in the same pipe. When rainfall exceeds capacity, the system discharges the mixture into rivers to avoid backing up into buildings. Separating these networks is enormously expensive, which is why reducing the rainfall entering them at source is the practical route.

Separated systems leak the problem back anyway, since misconnected plumbing sends washing machines and bathrooms into surface water drains that discharge straight to a watercourse, and that is a fault found house by house with no efficient way to detect it.

Retrofit is the hard part

New development can be required to manage runoff on site, and many jurisdictions now do exactly that. The existing city is where most of the impervious surface is, and retrofitting it requires touching every street.

Measured properly, programmes that attach drainage retrofit to routine street renewal make progress; standalone programmes generally do not. A large share of the impervious surface is privately owned as well — roofs, yards, driveways, car parks — which no street programme can reach, so the available levers there are charging for surface water by area, grants for removing hard surfacing, and requiring permission before a front garden becomes a parking space.

The co-benefits are what fund it

Rain gardens and street trees deliver shade, biodiversity, air quality and amenity alongside drainage. Schemes justified on multiple benefits attract funding that drainage alone does not. This is the practical argument for designing them as public realm improvements rather than as engineering works.

At street level, the argument carries a trap, because a scheme sold on five benefits gets assessed by five departments against five different tests, and the ones that actually get built are generally those where a single budget holder was willing to pay for the whole thing and treat the rest as free.

Local statute governs most of this, and it varies street by street.

Adoption and maintenance decide whether it lasts

A rain garden needs weeding, silt clearance and replanting, and a permeable surface clogs if nobody sweeps it, so these assets fail gradually through neglect rather than visibly through breakage. Many schemes stall on the question of which authority adopts them afterwards, because a feature that is simultaneously drainage and landscaping falls between two budgets and two departments. Where the adoption route is agreed before construction the schemes survive; where it is left to be settled later, the maintenance quietly never happens and the failure is blamed on the technique.

Designers also have to plan for exceedance — the route water takes once every system is full — and deliberately steering that flow across car parks and along roads is what keeps it out of habitable rooms.

The takeaway

Stop moving water away quickly. Almost everything else follows from that reversal.

Measure the walk, not the map.