Sponge cities in the UK: what the idea now means for planning and drainage

Sponge cities are urban areas built to absorb, store and slowly release rainfall instead of piping it away as fast as possible. The phrase came out of China a decade ago. In Britain it has mostly been discussed as something happening somewhere else, an interesting experiment in Wuhan or Shanghai with no obvious bearing on a planning application in Surrey. That stopped being true on 17 August 2026.

England's new National Planning Policy Framework, in force from that date, requires sustainable drainage designed to the 2025 National Standards, extends the requirement to every development rather than major schemes only, and for the first time gives deculverting national policy status. The vocabulary is different. The substance is sponge-city thinking written into planning policy, and it changes what a surface water drainage strategy for planning has to demonstrate.

Arup's Global Sponge Cities Snapshot puts London at 22% sponginess, against 35% for Auckland. The UK is not starting from a strong position.

What is a sponge city?

A sponge city is an urban area that manages rainfall where it lands, through soil, vegetation and open water, rather than routing it straight into a pipe network. Rain is intercepted, held and released slowly, so less of it reaches the drainage system at once. The features are ordinary rather than exotic, and most of them already appear in UK schemes under a different name.

• Rain gardens and swales. Shallow planted depressions that catch runoff from roofs and roads and let it soak away. Green SuDS and nature-based drainage covers the design detail.
• Permeable paving. Surfacing that lets water through into a sub-base rather than shedding it to a gully.
• Green roofs. Planted roof build-ups that hold rainfall and release it over hours instead of minutes.
• Urban wetlands and ponds. Open water that stores volume and treats it on the way through.
• Street trees. Interception and evapotranspiration, plus whatever soil volume sits around the rootball.
• Deculverted watercourses. Streams brought back to the surface, so an open channel does the work a buried pipe was doing.

The term is usually credited to the Chinese landscape architect Kongjian Yu, and China put it into national policy in 2015 through a pilot programme that set coverage targets for participating cities. What the UK has instead is sustainable drainage systems, and the same instinct to deal with rain where it lands that we set out in source control at plot level. The four pillars of SuDS — quantity, quality, amenity and biodiversity — describe what a sponge city is trying to do at street level.

How spongy is the UK?

Not very. On the only comparative measure anyone has published, London sits near the bottom. Arup's Global Sponge Cities Snapshot used its Terrain AI tool to measure green and blue cover across ten city centres, then adjusted for soil type and vegetation to estimate how much rainfall each could absorb.

London's ground conditions are part of the answer. Clay limits infiltration, which is why so much English drainage design ends as attenuation rather than a soakaway, and why a BRE 365 infiltration rate so often decides the shape of a scheme before anything else does. Green space is only half the story. What the ground underneath will take matters just as much.

The other half is what has been built over. CIWEM's sponge cities briefing notes that urban areas have grown by nearly a third while losing a city-sized amount of green space.

Why the UK is talking about sponge cities now

Because surface water is now the largest source of flood risk in England by property count, and the number moved sharply. The Environment Agency's updated national assessment, published in December 2024, put around 6.3 million properties in England in areas at risk of flooding, of which around 4.6 million are at risk from surface water. That surface water figure was a 43% increase on the previous assessment.

By the middle of the century the Environment Agency expects around one in four properties in England to be in an area at risk of flooding.

Much of that increase came from better data and modelling rather than from new rainfall, but the practical effect is the same: pluvial, or surface water, flooding is now the risk most often decided by what happens at the surface, in the first few minutes of a storm. Rainfall patterns are moving too, as we set out in how increasing rainfall is changing flood risk in the UK.

CIWEM puts the cost of getting this wrong in concrete terms. The 2007 floods killed 13 people, affected 55,000 homes and businesses and caused £4 billion in economic losses. In the fifteen years since the Flood and Water Management Act 2010 reached the statute book, two million homes have been built without climate-resilient drainage.

What the evidence says a sponge city delivers

The case has shifted from hydraulic to economic. That is usually what moves policy. CIWEM's October 2025 briefing sets out both halves: the drainage performs, and it costs less than the concrete alternative once the whole life of the asset is counted.

The UK numbers in the same briefing are the ones worth quoting to a client. If sustainable drainage were required across the Office for Budget Responsibility's forecast of 1.3 million new homes, CIWEM puts the additional welfare value to residents at as much as £219 million a year, which it translates into a £3.1 billion property price premium. At the level of a single house, the research found a home with SuDS carries a £4,001 premium over an otherwise identical property with standard drainage, and that the public would pay an average of £281 more a year to live in one.

A home with sustainable drainage is worth £4,001 more than an identical house on a conventional pipe system.

That last figure matters because the cost objection has always been framed as a developer cost with no return. The evidence says some of it comes back through sales values.

From aspiration to Policy F8: what changed on 17 August 2026

Three things, all of them in the flood risk chapter of the new National Planning Policy Framework. Together they take the parts of sponge-city practice that translate to a development site and make them planning requirements rather than good intentions. Our full account of the August 2026 NPPF changes works through the whole chapter; the drainage points are these.

1. Sustainable drainage must be designed to the National Standards. Policy F8 requires drainage designed in accordance with the 2025 National Standards for Sustainable Drainage Systems, which had been guidance rather than policy since June 2025. The standards cover runoff destination, everyday and extreme rainfall, water quality, amenity, biodiversity and whole-life maintenance.
2. It applies to every development, major or not. The design and maintenance requirements now reach past the major-development threshold. A small site can no longer treat drainage as a question for somebody else's application.
3. Culverting is discouraged and deculverting is favoured. National planning policy now discourages enclosing an existing watercourse without compelling justification and supports removing culverts and restoring natural channels. This is the first time deculverting has carried that status.

One thing did not change. Schedule 3 of the Flood and Water Management Act 2010 remains uncommenced in England, so there is no separate SuDS Approving Body here and the route still runs through planning. Wales commenced it in 2019 and has had a SuDS Approving Body ever since. Asked about it in January 2026, the government said only that a final decision on commencing Schedule 3 "will be made in due course", according to the House of Commons Library briefing.

The rural half: spongy landscapes and the catchment

A sponge city stops at the urban edge. Most of the water does not. CIWEM published a companion report on spongy landscapes in March 2026, applying the same principle to farmed land: soils, vegetation, wetlands, ponds, ditches and floodplains that absorb, slow and filter rainfall rather than shedding it.

Soil does most of the work. Healthy soil holds pore spaces that let water soak in and move through the ground; compacted or degraded soil sheds it. That makes cover crops, herbal leys, reduced tillage and managed grassland flood-risk interventions as well as agricultural ones, a link we have looked at separately in how soils influence flooding and resilience.

• North East Cotswolds. Modelling suggests land management changes across the catchment could store around 2 million cubic metres of additional water and reduce flood peaks by about 21% for a 1 in 2 year event.
• Spains Hall Estate, Essex. A whole farm reservoir approach could cut peak flood flows by 15% to 30% and provide around 5,942 megalitres of storage, roughly 12% of the planned Fens Reservoir.
• Dry weather too. The same Spains Hall modelling indicates downstream summer flows could rise by 5% to 10%, because water held in soils and landscape features is released gradually.

These are the measures covered by natural flood management more generally, from leaky dams to floodplain reconnection, and by research on sediment ponds and flood control. The direction of policy is the same: England's first Land Use Framework treats flood risk as a land use constraint rather than only a site-level planning matter, and CIWEM, whose role in environmental management and planning sits behind both reports, has been arguing the catchment case for years.

What a sponge city cannot fix

Sponginess changes the timing and the volume of runoff. It does not remove flood risk. The benefit is real at the frequent end of the range and fades as the event gets larger, which is the part most coverage of the idea leaves out and the part a planning officer will ask about.

• The effect is bounded. The Environment Agency's 2024 Working with Natural Processes evidence directory reviewed around 800 studies across 17 measures and found combined measures can cut flood flows by roughly half and lower peak levels by 10 to 30 centimetres downstream in a one-in-25-year storm, while stating plainly that natural flood management cannot guarantee a level of protection on its own. Our read of that evidence goes through it measure by measure.
• The evidence is lopsided. POSTnote 768, published in May 2026, found the research base weighted roughly ten to one towards potential benefits over potential disbenefits, which is why borrowed performance figures need site-specific justification. We covered that evidence gap when it landed.
• Ground conditions bite. Where clay dominates, infiltration is limited and the scheme becomes a storage problem instead. Sponginess is not evenly available.
• Retrofit is expensive. Sheffield's Grey to Green converted 1.3 kilometres of inner-city street, at £3.6 million for the first phase between 2014 and 2016 and £6.3 million for the second between 2019 and 2022.
• Nobody has solved maintenance. CIWEM's briefing names the missing piece: there is still no clear mechanism for adopting these features and maintaining them, and Schedule 3 was the mechanism that was meant to fix it.

What it means for a development site

Since 17 August 2026 the position for a planning application in England is straightforward, whatever the wider debate about sponge cities. Every development needs sustainable drainage designed to the National Standards with arrangements for maintenance across the lifetime of the development, major schemes need Lead Local Flood Authority input, and enclosing an existing watercourse needs compelling justification.

• A site-specific flood risk assessment where the site triggers one. Our flood risk assessment for planning applications covers the sequential and exception tests as well as the assessment itself.
• A surface water drainage strategy that works down the discharge hierarchy and shows why each step was rejected before the next was reached.
• Infiltration testing where soakaways or permeable surfaces are proposed, because the standards will not accept an assumed rate.
• Maintenance and adoption arrangements, named and funded, rather than deferred to a condition.
• A route to discharge the conditions once permission is granted, which our discharge of drainage conditions service handles.

One caution worth stating plainly. Upstream sponginess, whether a catchment scheme, a neighbouring green roof or a council retrofit programme, does not reduce your own obligations. No Lead Local Flood Authority will accept a modelled catchment benefit in place of site-specific evidence, and the drainage strategy for your planning application still has to stand on its own.

If you are working out what Policy F8 means for a live application, talk to our drainage engineers about your surface water strategy and we will tell you what the scheme needs.

Frequently asked questions