Rain gardens and the highway adoption problem

A rain garden inside the public highway is a maintenance liability that nobody has agreed to take on, and that is the real reason schemes lose them at technical approval rather than at planning. The drainage case is usually sound. The question the highway authority asks is who cuts the planting, who lifts the silt out of the forebay and who pays for it in year fifteen, and a layout that cannot answer that tends to be redrawn with a conventional gully and a tank.

Get the adoption conversation started before the street geometry is fixed. Once the carriageway width, the parking bays and the plot boundaries are set, there is nowhere left to put a feature that has to sit outside the adopted limits.

How a rain garden works in a street

A street rain garden is a shallow, planted depression receiving runoff from the carriageway or footway through one or more kerb inlets. Water ponds for a short period, filters through an engineered soil, and then either infiltrates to the ground or collects in a perforated underdrain and discharges at a controlled rate. CIRIA C753, the SuDS Manual, calls the family of features bioretention and sets out the layer build-up, the loading ratios and the drawdown expectations.

The attraction for a designer is that it takes runoff out of the pipe network at source, delivers water quality treatment, and does it in space that already exists between parking bays and crossings. The attraction for the authority writing the local plan policy is the same, which is why so many design codes now show them. The gap between the code and what gets built is the adoption problem.

Depth is the first constraint. Ponding depth in a trafficked street is normally kept shallow for safety and to limit the kerb reveal, which means the feature has to be wide or long rather than deep, and that competes directly with parking. On a sloping street, check-dams or terracing are needed, and the same constraints that apply to runoff on a steeply sloping site apply at street scale.

Why highway authorities resist adoption

Section 38 of the Highways Act 1980 lets a highway authority agree to adopt a street constructed by a developer. Nothing in that section obliges it to adopt any particular feature, and a planted bioretention basin is not a conventional element of highway construction. Several objections come up repeatedly, and they are reasonable ones.

• The authority's maintenance budget and its direct labour organisation are set up for gully emptying, patching and gritting, not horticulture.
• Inspection regimes and condition standards for bioretention are not written into the authority's own design guide, so there is nothing to inspect against.
• Liability for a feature that fails and floods a property is unattractive where the failure mode is poor maintenance.
• Commuted sums are often calculated on the authority's standard rates, which may not cover a specialised regime over a design life of thirty years or more.
• Statutory undertakers need access to apparatus under the highway, and a 1 metre deep filter media trench complicates every future excavation.

Some authorities have adopted street SuDS and published standard details for them. Many have not. The only way to know is to ask the authority that will sign the section 38 agreement, in writing, early.

Kerb inlets, forebays and silt

Silt is what actually kills these features. Road runoff carries sediment, tyre wear and grit from winter treatment, and a filter surface that blinds over stops infiltrating. The design response is to intercept sediment where it can be lifted out at low cost.

A forebay, meaning a small hard-surfaced or gravel settlement area immediately inside the inlet, does most of that work. It needs to be reachable from the carriageway by a suction unit without closing a running lane, which affects where the feature can sit relative to parking bays. Kerb inlets need a throat wide enough to pass leaf litter and a sacrificial grating arrangement that can be cleared by hand.

Then specify the failure case rather than the design case. A partially blocked inlet is the normal operating condition in autumn, not an exception, and the exercise of modelling a partially blocked flow control is worth repeating for the inlet itself. Show the overland flow route that applies when the inlet blocks, and show that it runs along the kerb line to the next gully rather than across a footway and into a doorway.

Winter salting and the planting

Highway authorities salt carriageways through the winter, and that chloride load arrives in the rain garden in the first flush of every thaw. Planting specified for a garden setting does not survive it.

The specification has to work for three conditions at once: short-duration inundation, summer drought in a free-draining filter media, and intermittent salinity. Sedges and rushes, ornamental grasses, salt-tolerant shrubs such as dogwood and willow, and herbaceous species chosen for coastal planting all have a place. Avoid anything that needs consistently moist soil, and avoid a monoculture, because a single pest or a single bad winter removes the whole feature. Our note on planting specifications for swales and wetland margins covers the same ground for off-carriageway features.

Two practical additions. Keep a gravel or stone mulch over the filter surface in the splash zone, so that salt crust and grit can be lifted without disturbing the media. And write the establishment period into the planning condition or the section 38 agreement, with a replacement obligation, because year one and year two failures are normal and somebody has to replant.

Who maintains it when adoption is refused

Where the highway authority will not adopt, the feature has to sit outside the adopted highway on land with an identified owner and an enforceable maintenance obligation. The usual routes are:

1. Place the rain garden in a landscape strip outside the highway boundary, maintained by a management company funded through estate charges, with the obligation secured by a section 106 agreement.
2. Transfer the feature to the sewerage undertaker under a section 104 adoption agreement, which the sewerage sector guidance now contemplates for some SuDS components, subject to the undertaker's own design requirements.
3. Retain it in a parcel transferred to the parish or town council or a land trust, with a commuted sum.
4. Keep it within curtilages as a private feature on each plot, which works for roof runoff and not for carriageway runoff.

Each route changes the layout. Option one needs a strip of land and a visibility check so that planting does not grow into a splay. Option two needs the undertaker's agreement at the design stage rather than at the connection application. Option three needs a willing recipient, which is rarer than it sounds.

Note that Schedule 3 to the Flood and Water Management Act 2010, which would create a SuDS approving body and change the right to connect, has still not been brought into force in England as at October 2026. Announcements have been made about implementation, and until it takes effect adoption remains a matter of agreement rather than a statutory route, so check the current position before relying on it.

Settling the position before the layout is fixed

The sequence that avoids a redesign:

1. At pre-application, ask the highway authority whether it adopts bioretention within the highway and ask for the standard detail if it does.
2. Ask the lead local flood authority, as the statutory consultee on surface water for major development, what it expects on maintenance and whole-life costing.
3. Ask the sewerage undertaker whether it would consider adoption and on what conditions.
4. Record all three answers and carry them into the Drainage Strategy as the basis for the chosen arrangement, with the maintenance schedule, the responsible party and the access arrangements set out.
5. Only then fix the street cross-section.

That last point is the one people skip. A 5.5 metre carriageway with parallel bays and a 2 metre service strip leaves no room for a feature that turns out to need its own land, and by the time technical approval raises it the plot layout is in a sale agreement.

Write the maintenance plan as if the person reading it in year twenty has never seen a rain garden, because they probably have not. Then put the drawdown time, the design storm, the sediment removal frequency and the planting schedule in one table that can be handed over with the street.

For the wider evidence base on a scheme like this, enter your postcode on PlanScope to see which reports the site is likely to need. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.