Leaky Dams and Flood Risk: What They Do, Where They Work, and What They Can’t Fix

Anyone who has watched the latest series of Clarkson's Farm will have seen Jeremy Clarkson wading into a stream at Diddly Squat to build what he gleefully describes, after a bit of reading on the government's website, as "large, leaky, wooden dams", partly because the government will help pay for them. It is played for laughs, as most things at Diddly Squat are. But behind the comedy sits a real, well-tested piece of flood management.

Leaky woody dams are permeable timber structures built in sequence across small streams to slow the flow, hold water back briefly and lower the flood peak downstream. In England they are a funded capital item under Countryside Stewardship, at £461.39 for a small dam and £764.42 for a large one, and the evidence behind them is real but bounded. They take the edge off the floods that happen often and do progressively less as events get more extreme.

Leaky dams are not a gimmick. They are one of the most widely trialled natural flood management techniques in the UK, they have a measurable effect on flood risk in the right setting, and they are being built at scale from the North York Moors to the Welsh valleys. They are also routinely misunderstood, including, occasionally, on television. Here is what a leaky woody dam actually does, where it works, what the grant pays and the limits worth being honest about.

At Pickering in North Yorkshire, 167 leaky woody dams and 187 heather bale dams helped cut the town's annual flood risk from around 25% to less than 4%.

What is a leaky woody dam?

A leaky woody dam, also called a leaky dam or a leaky barrier, is a partial obstruction placed across a small stream or ditch, usually built from logs, branches and woody material. Unlike a conventional dam, it is designed to leak. Water passes through and around it most of the time, so the channel keeps flowing and fish and sediment can move through. It only does its real work when flows rise.

The idea is to mimic what a fallen tree, or a beaver, would do naturally: hold the channel back a little, raise water levels behind the structure during heavy rain, and spill some of that water out sideways onto the surrounding land. Leaky dams are rarely built singly. They are installed in sequences of tens or even hundreds along a watercourse, so their individual effects add up across a catchment.

The split between a small and a large leaky woody dam is a funding distinction rather than a hydraulic one. Countryside Stewardship separates them by channel width: item RP32 covers streams between 1m and 2.99m wide, and RP33 covers channels between 3m and 5m. Anything wider sits outside both.

How leaky dams reduce flood risk

The principle behind every leaky dam is the same one that underpins almost all good drainage thinking: slow the water down. A flood happens when too much water arrives in one place too quickly. Anything that delays and spreads that arrival reduces the peak flow.

A leaky dam does this in a few ways. It provides temporary attenuation behind the structure during a storm, holding back volume that would otherwise rush downstream. It pushes higher flows out of the channel and onto the floodplain, where the water sits, soaks in and is released slowly over the following days. And across a whole series of barriers, it desynchronises the catchment, so water from different tributaries no longer arrives at the vulnerable town or junction all at once. That lowers and flattens the flood peak.

This is the same "slow the flow" logic that drives source control in sustainable drainage, sits at the top of the SuDS hierarchy, and underpins the broader idea of spongy landscapes that hold rainfall in the land rather than shedding it. A leaky dam is simply that principle applied to a stream channel. None of it changes what a developer has to prove about their own site, though: catchment work sits alongside a flood risk assessment for planning, never instead of one.

Do leaky woody dams actually work?

Yes, within limits, and the evidence is now reasonably good rather than merely hopeful.

The best-known UK example is Slowing the Flow at Pickering in North Yorkshire, a Defra-funded pilot led by Forest Research after the 2007 floods. The scheme placed 167 leaky woody dams and 187 heather bale dams in the streams above the town, alongside a flood storage bund and new woodland. Together these measures helped cut Pickering's annual flood risk from around a 25% chance in any given year to less than 4%. It remains the reference case for what coordinated natural flood management can achieve.

Controlled research backs this up. A study by Cardiff University and the University of Worcester monitoring leaky barriers on a small Shropshire river found the structures could store enough water during a major event like Storm Dennis to fill more than four Olympic swimming pools. Upland modelling studies typically find leaky dams shave around 10% off the flood peak for smaller, more frequent storm events.

Two more recent pieces of work sharpen the picture. A three-year before-and-after monitoring study on three streams in Coverdale, in the Yorkshire Dales, is one of the few UK assessments built on measured stage data rather than modelling alone. And in January 2026 researchers at Loughborough University reported that once sediment movement is written into the models, leaky dams can cut peak flows by up to a third in storm events, while stressing that the structures are dynamic, reshape as the river reshapes around them, and need monitoring rather than a one-off build.

That last line matters. The honest reading of the evidence, set out in more detail in a closer look at the Environment Agency's natural flood management research and in the EA's own Working with Natural Processes evidence directory, is that leaky dams are most effective against the floods that happen often, and progressively less effective as events get more extreme. They take the edge off the regular soakings, not the once-in-a-generation deluge.

Where leaky dams work, and what they can't fix

Leaky dams are not a universal answer, and treating them as one is the quickest way to be disappointed.

They work best on small watercourses with narrow channels, where there is room beside the channel for water to spill into and a town or asset downstream worth protecting. That usually means steep upland streams, but not always: the size of the channel and the space around it matter more than the altitude. They work poorly, or not at all, on large lowland rivers where fluvial flood risk is driven by a far bigger catchment, in deep engineered channels, or where there is simply nowhere for displaced water to go. In the wrong place a leaky barrier can move a problem rather than solve it. That is exactly why the work needs designing, not simply doing.

• Scale. A handful of dams on one farm makes little difference, and meaningful flood reduction needs many structures across a catchment, which means cooperation between landowners.
• Extreme events. As storms get rarer and larger, the proportional benefit falls away, so leaky dams should sit alongside engineered defences and property-level resilience, not replace them.
• Your own site. A leaky dam upstream does not remove the need for a site-specific flood risk assessment or a compliant drainage strategy.

Natural flood management reduces risk across a catchment. It does not exempt a site from demonstrating its own safety.

A lowland example: Debenham after Storm Babet

Debenham, a village in the clay headwaters of the River Deben in mid Suffolk, shows the technique at work well away from the hills. On 20 October 2023, during Storm Babet, more than 70 homes and businesses there flooded internally. Three tributaries of the Deben flow into the village, and water came off the surrounding fields so fast that properties flooded before the river gauge reached its warning trigger. Suffolk County Council's Section 19 investigation found that the rain had overwhelmed the natural flow routes, on clay soils that shed water before it can soak in. It recommended slowing the flow in the fields above the village.

That work was finished in October 2026. At Crows Hall a timber leaky dam now sits across a field-edge ditch, where it catches debris and holds water back during a storm. At Elm Farm an attenuation feature on arable land stores water in heavy rain and releases it at a controlled rate. The two sites cost £122,219, paid from the Anglian Eastern Regional Flood and Coastal Committee's Local Levy, with the initial designs funded through Defra's Water Environment Improvement Fund.

Suffolk County Council expects the works to "significantly reduce flood risk to people and property", and says they cannot guarantee complete protection.

A ditch on a lowland farm can host a leaky dam as readily as a moorland gully, provided the channel is small and the water has somewhere harmless to go. Debenham had also tried natural flood management before. A £200,000 scheme of ponds, bunds and storage areas was delivered in the late 2010s. The village flooded anyway when 68mm of rain fell in 21 hours at the nearby Earl Soham gauge. No analysis of how those earlier features performed has been published, but the outcome matches the evidence above. Small measures take the edge off frequent floods, and an extreme storm overwhelms them. That is why the new works sit alongside property flood resilience, watercourse clearance and a revised flood warning service, not in place of them. The same pattern runs through the other UK natural flood management schemes.

For anyone planning to build in a village like Debenham, upstream works change the catchment but not the planning test. A site still has to be shown safe for its lifetime with climate change allowances applied. A published Section 19 report on the village is also evidence the lead local flood authority will expect a flood risk assessment for planning to have read.

A volunteer-built example: Bloxham after Storm Bert

Bloxham, a village of around 3,400 people south-west of Banbury in north Oxfordshire, shows the same technique arriving by a different route: built by residents rather than commissioned by a council. On 24 November 2024, during Storm Bert, the Bloxham Brook peaked at 2.2m on the village gauge while surface water poured off the mudstone and clay slopes of Hobb Hill. Oxfordshire County Council's Section 19 investigation recorded up to 35 properties flooded internally across seven hotspots, from a mix of river and surface water flooding. The parish council puts the number of properties damaged at 62 or more, and BBC News reported an estimated £2 million of damage, with some households out of their homes for up to 18 months.

The Section 19 report recommended talking to landowners about natural flood management upstream of the village. A flood working group set up by the parish council took that forward with Thames21, which mapped the options in an opportunity study for Oxfordshire County Council. On 29 and 30 October 2026, around a dozen volunteers each day, led by Thames21, built six leaky woody dams in The Slade, a nature reserve on the edge of the village, three on each branch of the stream that runs through it. A tree contractor felled and cut the timber on the reserve a week earlier, Cherwell District Council issued the watercourse consent, and a £9,000 grant from Oxfordshire County Council paid for the work.

The parish council expects the six dams to have a "beneficial but incremental impact" on slowing water into the village in a flood.

The Thames21 study shows what incremental means in numbers. It identified field-corner bunds, pond desilting and more leaky dams on five landholdings, covering about a quarter of the 7.18 km² catchment above the village, with around 3,535 m³ of temporary storage between them. That is roughly 1.3% of the runoff from a storm the size of Storm Bert, rising to about 3.5% for a more typical significant storm. The Slade dams sit outside that package. None of the farmland measures has been built yet, and the working group's November 2026 update says the next phase depends on finding a project manager and funding. The pattern matches the national evidence above: small measures help most with frequent floods and are overwhelmed by the largest.

Bloxham also puts a figure on what a community scheme costs. The grant paid for Thames21's design and supervision and for the tree contractor, not just timber; spread across six dams, that is about £1,500 each, against £461.39 for a small dam under RP32. Consent still came before construction, even for volunteers working on a nature reserve. For development, nothing changes: a site in Bloxham still needs a flood risk assessment for planning that shows it is safe on its own terms, with climate change allowances applied, and the published Section 19 report is local evidence the lead local flood authority will expect that assessment to have read.

For a fuller view of where these techniques sit in the toolbox, and the genuine evidence gaps that remain, the article on nature-based flood and drought resilience is a useful companion, as is the overview of natural flood management schemes more broadly.

How much does the government pay for a leaky woody dam?

Countryside Stewardship pays £461.39 for each small leaky woody dam and £764.42 for each large one, under capital items RP32 and RP33. Both are available through Capital Grants 2026, which opened in July 2026 with £225 million behind it, and through Countryside Stewardship Higher Tier. Both also need written support from a Catchment Sensitive Farming adviser before you apply.

Clarkson was right about one thing: the government will contribute to the cost. What the television version leaves out is how much of the money is governed by conditions that have to be met before a single log goes in the water, as the GOV.UK guidance on RP32 and RP33 sets out.

• Catchment Sensitive Farming. RP32 and RP33 are among 24 capital items that need written support from a CSF adviser following an advisory consultation.
• A £25,000 ceiling. Natural flood management is one of six item groups in Capital Grants 2026, and that group is capped at £25,000 per application.
• One application a year. The Rural Payments Agency accepts one eligible Capital Grants application per calendar year for each Single Business Identifier you manage.
• Nothing paid retrospectively. The grant cannot pay for capital works done, or materials bought, before your agreement starts.
• No fixed deadline. The round closes when the available funding is allocated, with notice given on Defra's Farming blog.
• Targeting still applies. The items are aimed at catchments with permeable soils where groundwater recharge is the goal, or areas targeted for flood risk and water resource benefit.

One point of confusion is worth clearing up, because the schemes have moved around a good deal. The leaky dam grant is not a Sustainable Farming Incentive item. SFI26 reopened in 2026, with a first window running to the end of August and a second opening in September, but the money for leaky woody dams sits in Countryside Stewardship, not there. Funding is best treated as a helpful contribution towards a scheme that already makes sense, rather than a reason to build in the first place.

Larger, council-led schemes are usually paid for differently. Debenham's two 2026 sites were funded from Local Levy, not from RP32 or RP33. Local Levy is money the lead local flood authorities in a region contribute and their Regional Flood and Coastal Committee allocates to local priorities.

What consent do you need before building one?

Placing a structure in a watercourse usually requires consent before any work begins. On an ordinary watercourse that means ordinary watercourse consent from the lead local flood authority; on a main river it means an Environment Agency flood risk activity permit. Owning land next to a watercourse also carries riparian responsibilities that do not go away because the structure is well-intentioned.

The grant conditions say the same thing in their own words. Both RP32 and RP33 warn that you may need prior consent from the Environment Agency, your lead local flood authority or an internal drainage board, and the Catchment Sensitive Farming consultation is usually where that surfaces. As Clarkson also discovered, you generally cannot just start digging. Building first and asking later is how a well-intentioned project becomes an enforcement problem.

What it means for landowners and developers

For a farmer or estate, leaky dams are a low-cost way to hold water in the landscape, reduce downstream flooding, settle sediment and support biodiversity, provided they are placed where they will help and consented properly. For developers, the more important takeaway is the opposite of reassurance: upstream natural flood management is part of the catchment picture that assessments are expected to account for, but it never substitutes for getting your own site right. Headline reduction figures from a named scheme cannot be borrowed for an unrelated site, and climate change allowances still apply in full whatever is happening upstream.

If your scheme sits below a catchment where this kind of work is happening, or you need to demonstrate how surface water and river flows behave around your site, that is squarely a question for a flood risk assessment, a surface water drainage strategy and, where the data is coarse, site-specific flood modelling. Unda can help with all three.

Frequently asked questions