Equestrian flood risk and drainage: the planning rules for stables, field shelters and ménages

Planning permission for stables is needed far more often than owners expect, and the flood risk and drainage requirements that come with it are the part almost nobody writes about. Horses are kept on low-lying grazing land: the cheap, wet, unarable ground that floods. So equestrian applications land in Flood Zone 3 at a rate that would alarm most housebuilders. This guide covers when planning permission for stables, field shelters and ménages is required in England, when a flood risk assessment is triggered, and what a drainage strategy for an equestrian yard actually has to show.

It also answers the question the internet does not: how a riding arena should be treated in a surface water calculation. There is no published guidance on that anywhere in the UK, and the assumption most reports make is wrong.

Britain has around 850,000 horses across roughly 331,000 horse-owning households, and a sector worth about £5 billion. There is no register of arenas, no count of equestrian premises at flood risk, and no British Standard for how an arena is built.

Do you need planning permission for stables?

You almost certainly do. Grazing horses on land is agriculture and needs no permission. Keeping them, which means feeding, rugging, schooling and sub-dividing into paddocks, is a material change of use that does. Any stable block, tack room or hay barn built for that use is operational development in its own right, and agricultural permitted development rights do not reach it.

The distinction comes from section 336 of the Town and Country Planning Act 1990, which defines agriculture to include "the use of land as grazing land" and "the breeding and keeping of livestock". Livestock is then narrowed to creatures kept "for the production of food, wool, skins or fur, or for the purpose of its use in the farming of land". A riding horse is none of those. So grass eaten in a field is agriculture; a horse kept, exercised and trained is not. Courts have read it that way since Sykes v Secretary of State for the Environment in 1981.

• Supplementary feeding. Hard feed and hay replacing grass as the main food source.
• Sub-division. The field split into small individual paddocks rather than run as one.
• Schooling equipment. Jumps, poles, a lunge ring or a marked-out riding area left in place.
• Buildings and structures. Stables, a field shelter, a tack store, a horsebox hardstanding.
• Commercial activity. Livery, a riding school, breaking or training for third parties.
• Stocking density. The New Forest National Park Authority uses less than 0.5 hectares per horse as its working guideline.

Part 6 of the General Permitted Development Order, the agricultural right that lets a farmer put up a building on a five-hectare unit, is unavailable. It excludes any "building, structure or works not designed for agricultural purposes" and requires land in agricultural use for a trade or business. Class Q converts a barn to a house, not to stables. Class R can move an agricultural building into a flexible commercial use, and since May 2024 that list includes Class E and outdoor sport and recreation, but nobody has yet tested it on a livery yard and we would not rely on it without a lawful development certificate.

Private stables and livery yards sit in no use class at all. The Use Classes Order does not mention horses, stables, riding or equestrian anywhere, so they are sui generis and every change into or out of them needs permission. That makes an equestrian scheme one of the few rural proposals where the planning application, the flood risk assessment for planning and the drainage strategy all have to be built from first principles rather than lifted from a precedent.

Field shelters: when a shelter on skids still needs permission

A genuinely mobile field shelter can escape planning control, but two separate tests have to be passed and most shelters fail the second. The first asks whether the shelter is a building operation, judged on size, permanence and physical attachment. The second asks whether putting it there changes the use of the land, and a shelter that exists to serve kept horses usually does.

The chattel test comes from rating and planning case law, and inspectors apply it in exactly those three words: size, degree of physical attachment, and permanence. A shelter that is genuinely moved around a field on skids, sits on no base, is connected to nothing and does not stay put may be a chattel rather than a building. One craned into position on a hardcore pad and left there for five years is a building, whatever the brochure says.

Even a genuine chattel does not help if the field's lawful use is agricultural grazing: an enforcement notice at a South Downs National Park site in 2024 required the removal of two stables and one field shelter as operational development facilitating an unauthorised change of use to equestrian.

Ashford Borough Council, whose 2014 supplementary planning document remains the most useful English guidance on horse development, puts it plainly. The question "rests largely upon the degree of permanence", there have been "conflicting appeal decisions", and factors such as whether a base has been laid or the structure is fixed to the ground all count. If the shelter is really a permanent building in a fixed spot, apply. If you want certainty either way, a certificate of lawfulness costs half a planning fee, and it is a great deal cheaper than a retrospective application after an enforcement officer has been.

A ménage is an engineering operation, not a surface

A ménage needs planning permission in all but the most rudimentary cases. Building one involves excavation, levelling, a stone sub-base, a drainage system and usually kickboards and fencing, all of which are engineering operations and development in their own right. The surface is the last 100 millimetres of a structure, and it is the structure the planning system is looking at.

Floodlighting is a separate battle and usually a losing one in open countryside. The New Forest National Park Authority states that "requests for lighting for manèges will be resisted", and Ashford treats high-level floodlights on columns as "normally unacceptable". Both documents also press applicants to site the arena tight against existing buildings rather than out in a field, and to choose ground flat enough that re-grading is unnecessary.

That paragraph is, as far as we can find, the only place an English planning authority has written down the two flood mechanisms an arena actually creates. It is a decade old and it is more useful than anything the national guidance says. The second mechanism it names, displacement of flood water by a raised structure, is the same problem that drives compensatory flood storage on larger schemes.

When does an equestrian scheme need a flood risk assessment?

Policy F4 of the August 2026 National Planning Policy Framework requires a site-specific flood risk assessment for all development in Flood Zones 2, 3a and 3b, with no size threshold and no small-building exemption. In Flood Zone 1 it is triggered by a site of one hectare or more, land with critical drainage problems, or land shown at risk of flooding from any source now or in future. Our guide to planning for flood risk under the NPPF sets out how the chapter fits together, and how to do a flood risk assessment covers the method itself.

Two of those Zone 1 triggers catch equestrian schemes constantly. The one-hectare threshold applies to the site, not the building, so drawing the red line around the paddocks as well as the yard takes most holdings past it. And the "any source" trigger, widened in the August 2026 rewrite, sweeps in every low-lying rural site the Flood Map for Planning shows at surface water risk. If you are unsure where a site sits, our flood risk map by postcode is the quickest first check.

The Environment Agency must be consulted on a new stable block in Flood Zone 3, because a new building is neither a householder development nor an extension under 250 square metres, and therefore falls outside the minor-development band that standing advice covers, and our explainer on planning with EA standing advice sets out where the line falls. Anything within 20 metres of a main river triggers consultation in any flood zone. If the two documents sound like the same thing, they are not: our guide to whether you need an FRA, a drainage strategy or both sets out where one stops and the other starts.

How is a stable block classified for flood risk?

Nowhere in national policy, which is the honest answer. The word "equestrian" does not appear in the August 2026 NPPF, and neither does "horse" or "stable". Annex F, Table 2 has no entry for stables, field shelters or arenas. The classification is argued by analogy in every application, and it is worth pinning down with the case officer before the report is written.

In practice the analogy runs two ways and both are defensible. Stables, field shelters, tack and feed stores, hay barns and indoor schools are treated as less vulnerable, either by reading across from "land and buildings used for agriculture and forestry" or by landing squarely in the "assembly and leisure" and "storage" entries of the same class. Paddocks, grazing and an outdoor arena have a strong case for water-compatible development, under the entry for "amenity open space, nature conservation and biodiversity, outdoor sports and recreation".

• Less vulnerable in Zone 3a. Acceptable under Annex F, Table 3, with no exception test. This is what makes equestrian development viable on floodplain land at all.
• Less vulnerable in Zone 3b. Should be refused. A stable block on the functional floodplain is a policy refusal, not a design problem.
• Water-compatible in Zone 3b. Acceptable, but only with no net loss of floodplain storage, no impedance of flows and no increase in risk elsewhere.
• A groom's flat changes everything. Residential accommodation is more vulnerable, the exception test applies in Zone 3a, and where a scheme mixes vulnerabilities the highest one governs unless the authority agrees to assess it in component parts.

An appeal at Baildon in August 2025 dismissed a change of use from stable and store to holiday accommodation because the building sat in Flood Zone 3b. Raised floor levels, resilient construction and an evacuation plan could not cure a Table 3 refusal.

The flood risk vulnerability classification is worth reading in full before you commit to a position, because the compatibility table does the work in every one of these cases. The same reasoning drives the much tougher treatment of permanent residential caravans, which sit a whole band higher.

The sequential test trap: your change of use is exempt, your stable block is not

Policy F5 exempts changes of use from the sequential test. It does not exempt buildings. So an applicant who obtains a change of use from agriculture to the keeping of horses without a sequential test, and then applies for the stables a few months later, is caught by a test the first application never mentioned. This is the commonest way an equestrian scheme comes unstuck.

The exemptions in F5(2)(c) are householder development, small non-residential extensions under 250 square metres, and changes of use other than to caravan, camping, chalet or mobile home sites. A new standalone stable block is not an extension, and it is not a change of use. It is operational development, and the test bites. The same carve-out logic, read the other way, is why campsites and glamping sites lose the exemption their neighbours keep.

1. Show no built development is at risk. Under F5(2)(b)(i), if no building, access route or escape route and no land raising sits in an at-risk area, the test does not apply. On a large holding this is usually the cheapest answer: relocate the yard to the Zone 1 corner rather than rebuild it where it stands.
2. Check the access route as well as the building. An appeal at Ellesmere in June 2025 failed precisely here. The caravans and buildings were in Flood Zone 1, but the only access ran through Zones 2 and 3, and the inspector held the sequential conclusion to be "an incorrect assumption".
3. Argue a proportionate area of search. Policy F5(1) limits the search to "the anticipated catchment of the development in terms of its likely occupiers or users", and the September 2025 planning practice guidance added an express proportionality steer. A private stable serves the horses on one specific holding, so its functional catchment is a parish, not a district.
4. Turn the functional tie into a suitability failure. A site in another settlement that cannot serve the grazing land the stables exist to support is not suitable, regardless of who owns it. That answers the case officer's standard objection that the alternative site need not be in your client's ownership.

Interrogate the council's own strategic flood risk assessment before building a search from scratch. On one recent Severn Levels scheme the decisive evidence was the authority's own Level 2 SFRA, which recorded that every potential development site in the village was flood-affected and there were no reasonably available sites at lower risk. A sequential test report for planning built on the council's evidence is far harder to argue with than one built on estate agents' listings.

Floor levels, safe access and the horsebox nobody plans for

Policy F7 requires development to be safe for its lifetime, with safe access and escape routes where needed and residual risk managed under an agreed emergency plan. For a stable block that is harder than it sounds, because horses need level access at grade. You cannot put a loose box on stilts, so equestrian schemes are made acceptable by warning time and resilience rather than by dry design.

The Environment Agency recommends floor levels at least 600 millimetres above the estimated flood level, reducible to 300 millimetres where there is high certainty about that level, with flood-resistant materials to the same height. That margin is the freeboard, and it is the number a case officer will look for first. On tidal sites the practical compromise is usually a floor set 300 millimetres above surrounding ground, a mezzanine refuge above the design breach level for people, resilient construction to absorb the residual depth, and an operational commitment that the yard closes when a flood warning is issued. Which climate change allowance the levels are set against matters as much as the freeboard itself.

• A horsebox route, not a car route. Evacuating eight horses means a trailer or lorry on firm, wide, non-inundated ground, and it takes hours rather than minutes.
• Somewhere to go. World Horse Welfare advises arranging evacuation premises in advance and moving feed and bedding to higher ground before the water arrives.
• A realistic lead time. Tidal flooding is forecast and the cycle gives roughly six hours between peak and trough; a flashy ordinary watercourse gives no such warning.
• An honest limit. World Horse Welfare is blunt that horses stranded in deep or fast-moving water need Fire and Rescue with specialist training and equipment. There is no published depth at which a horse becomes unsafe to lead out, and inventing one would be worse than saying so.
• Raised electrics and services as well as a raised floor. Sockets, consumer units and pumps above the design level, and non-return valves on drainage outlets.

Most equestrian flood risk assessments we read are residential templates with the word "stables" substituted. One recent report specified flood-resilient kitchen units, a raised gas meter and solid hardwood stairs for a single-storey timber stable block with none of those things, then advised occupants to take refuge on the first floor of a building 3.6 metres to the ridge. It is a small thing, but it tells a case officer nobody looked at the site. It is also the sort of detail that gets an application held at validation.

Your ménage is not a soakaway

A riding arena is a positively drained, piped structure that discharges at a point. It is not an infiltrating surface, and treating it as permeable ground in a runoff calculation is the most commonly unexamined assumption in equestrian drainage. A standard 20 by 40 metre arena is 800 square metres of engineered surface, comparable to eight or ten house roofs, and it is routinely excluded from the calculation altogether. The reasoning that catches out paved-over gardens applies here at ten times the scale.

The build-up is consistent across the trade, because there is no British Standard and no CIRIA guidance for arena construction. Manufacturers' technical literature is the only body of specification that exists. Strip to firm subsoil; a woven geotextile on the compacted formation; 150 to 200 millimetres of clean angular stone, typically 40 to 75 millimetres and free of fines; a non-woven separation membrane of at least 300 grams per square metre; then around 100 millimetres of silica sand laid in two lifts, sometimes with rubber or fibre on top. Drainage is a herringbone: a 100 to 150 millimetre spine across the centre with 75 to 100 millimetre perforated laterals entering at 45 degrees at five-metre centres, backfilled with clean washed gravel. The sports-turf literature, which is properly engineered, puts comparable laterals at five to ten metre centres and falls in the region of 1 in 100, and is a better analogue than anything in the equestrian trade press.

Every published arena specification assumes a positive piped outfall to a ditch, watercourse or soakaway, and the lower woven membrane is there precisely to stop water reaching the subsoil.

Some arenas are sealed outright. Ebb-and-flow and capillary systems are built as watertight basins on welded liners by design, and at least one English planning submission specifies a heat-welded 650 gram membrane battened up the perimeter boards. In those cases nothing infiltrates at all. The arena is a 100 per cent impermeable catchment with the stone voids as its only storage, and every millimetre of rain that does not evaporate leaves through a pipe.

The performance also decays. The geotextile is the known weak point, and the trade markets membrane-free variants specifically because the fabric "can quickly become blocked or damaged over time". Sand degrades mechanically under hoof and harrow, generating fines that migrate down and blind it, and manure, hay and leaves worked into the surface accelerate it. Nobody has published a permeability curve for an arena at five or ten years old, which means any calculation crediting the surface with infiltration is crediting a design assumption with no maintenance-adjusted evidence behind it. The same argument is settled ground for permeable paving, where clogging and maintenance are an expected part of the submission.

The defensible position is to treat the arena as a positively drained sub-surface filter discharging at a point, evidence the sub-base voids as attenuation storage rather than assuming them, and design the outfall properly. At roughly 30 per cent voids, 175 millimetres of clean stone under a 20 by 60 metre arena gives something in the order of 63 cubic metres of storage, which is a real and useful number but one that has to be demonstrated rather than asserted. If the outfall goes to a ditch you will also need ordinary watercourse consent from the lead local flood authority, and a flood risk activity permit if it is a main river. If the ditch runs along your own boundary, the riparian responsibilities for keeping it clear are yours as well.

What an equestrian drainage strategy actually has to show

Policy F8 applies to any proposal that "could affect drainage on or around the development site", proportionate to its scale, with no floor area or site threshold at all. The only threshold in the policy is that lead local flood authority advice is required for major development, which almost no equestrian scheme is. So the duty applies in full while the consultation that would test it does not. What the authority expects to see is set out in our guide to the drainage information needed for planning permission.

That gap is why so many equestrian drainage submissions consist of the SuDS hierarchy recited in six bullet points and a sentence saying the change in runoff will be negligible. It is also why they get conditioned. A surface water drainage strategy for planning has to name the impermeable areas, derive an allowable discharge rate, size the storage and say where the water goes, all of it against the national SuDS standards (the non-statutory Defra standards published in 2025) and, on anything with a water quality dimension, the four pillars of SuDS rather than quantity alone.

Two real contrasts show how completely the ground conditions govern the answer. On a Weald Clay site, infiltration was ruled out and a 338 square metre roof area produced a greenfield QBar of 0.2 litres per second, requiring 26.6 cubic metres of cellular storage under the yard and a Hydro-Brake with a 24 millimetre orifice. On a chalk site, BRE 365 infiltration testing across two trial pits and six cycles returned a lowest rate of 0.34 metres per hour, the IH124 greenfield rate rounded to zero litres per second, and the whole scheme became two soakaways with no outfall at all.

Two practical lessons sit in that table. First, on free-draining chalk the greenfield rate rounds to nothing, so there is no legitimate discharge rate to design to and infiltration stops being optional. Second, at equestrian scale the allowable rate is well under a litre per second, which forces a flow control with an orifice the width of a pencil, on a site generating hay chaff, shavings and hoof-borne grit. That is why the six-monthly flow control inspection in the maintenance table is the most important line in the whole report, and why a vortex control beats a plain orifice plate every time.

A geology map is not a soakage test. The SuDS hierarchy requires infiltration to be pursued to the maximum extent practical, and ruling it out on a British Geological Survey sheet invites the lead local flood authority to ask for evidence, so knowing when infiltration testing is required is worth an hour of anyone's time. Trial pits are cheap next to a crate tank, and on a sandy or chalky site they can delete the attenuation structure entirely. Our BRE 365 infiltration rate calculator and the ground conditions for a soakaway checker will both give you a first view before anyone digs, and the rest of our drainage calculators will size the pipes and storage once you have a rate. Where the water table is the constraint rather than the geology, a groundwater flood risk assessment is the document that settles it.

Where does the muck heap runoff go?

Not into your surface water system. Muck heap leachate entering a ditch or stream is an unpermitted water discharge activity under regulation 12 of the Environmental Permitting Regulations 2016 and an offence under regulation 38. Yard wash-down water is not domestic sewage, so the general binding rules for small sewage discharges are unavailable to it. Both have to be contained.

A horse produces 24 litres of manure a day under the Nitrate Pollution Prevention Regulations 2015, carrying 58 grams of nitrogen and 56 grams of phosphate, or about 21 kilogrammes of nitrogen a year. With bedding in, the real heap is two and a half to three and a half times that volume. Nordic research on horse paddocks found leaching of 1.2 kilogrammes of phosphorus and 95 kilogrammes of nitrogen per hectare per year, against 0.5 and 19 on reference agricultural soil, rising to 7.4 and 484 on the excretion area of a sandy paddock.

• The Farming Rules for Water name horses expressly. The 2018 regulations define livestock to include horses, so a livery yard or riding school run as a trade or business is bound by them. They apply only to land used for agriculture "for the purposes of a trade or business", which appears to leave the private owner outside them altogether.
• Keep the heap 10 metres from water and 50 from a borehole. Do not store manure within 10 metres of inland freshwaters or coastal waters, or within 50 metres of a spring, well or borehole. Both distances are statutory, under regulation 9 of the 2018 rules and regulation 23 of the nitrate regulations.
• SSAFO does not apply. The 2010 silage and slurry regulations define livestock as animals kept for food or wool, which excludes horses, so the engineered containment regime that governs a dairy yard does not govern a livery yard.
• Nitrate rules do apply. The 170 kilogramme per hectare whole-holding nitrogen limit is reached at roughly eight horses per hectare, and hardstanding is expressly excluded from the area calculation, so building a large arena on a small holding tightens the limit.
• Temporary field heaps have their own siting rules, including a prohibition on any field liable to flooding or waterlogging.

The Environment Agency can serve an anti-pollution works notice under section 161A of the Water Resources Act 1991 where polluting matter "is likely to enter" controlled waters, before any pollution has actually occurred.

The engineering answer is the one agriculture worked out decades ago and the British Horse Society repeats: keep clean roof water separate from dirty yard water. Clean rainwater from roofs is expressly exempt from permitting and can go to a soakaway or watercourse, but the moment it mixes with yard water it becomes polluting matter needing containment. On a typical yard the roofs dwarf the dirty footprint, so separation can cut the contained volume by an order of magnitude. It is the cheapest compliance measure available, and it is the same logic that makes the first flush the dirtiest few millimetres of any rainfall event.

The cleanest arrangement we have seen reclassifies the problem out of the drainage system entirely. Wash-down bay flows and runoff from the manure trailer hardstanding both go to a sealed cesspool and are tankered off, while the utility room goes to a separate foul drainage strategy designed to Building Regulations Part H. Note that CIRIA C753's Simple Index Approach, which every drainage strategy cites, has no agricultural, livestock or equestrian land-use category at all, and indexes only suspended solids, metals and hydrocarbons, none of which is the pollutant of concern on a horse yard. Applying it to a stable yard produces a technically compliant answer to the wrong question.

What the LPA will condition, and how to discharge it

Expect a pre-commencement condition requiring full details of foul and surface water drainage, a timetable for implementation and a maintenance and management plan for the lifetime of the development. Equestrian conditions increasingly specify permeability testing to BRE 365 by name, and require "all necessary permits, consents and permissions" to be secured before the details are approved, as one Wiltshire equestrian appeal in August 2025 shows.

1. Get the ground data first. Trial pits and BRE 365 testing, ideally in winter when groundwater is at its highest, plus a topographical survey tied to the national grid. Both are cheap and both are the evidence everything else rests on.
2. Fix the contributing areas honestly. Roofs, yard aprons, the horsebox hardstanding, the access track and the arena, with a stated position on the arena's permeability rather than a blanket sentence about permeable construction.
3. Size the system and evidence the storage. Greenfield rates from IH124, storage sized to the 1 in 100 year event with the appropriate climate change uplift, source control at plot level where it helps, and a flow control that will not block. Many authorities want it summarised on a SuDS proforma.
4. Deal with the dirty water separately. Muck store, wash bay and yard drainage kept out of the surface water system and out of the treatment train.
5. Secure the consents. Ordinary watercourse consent for a ditch outfall, a flood risk activity permit within 8 metres of a non-tidal main river or 16 metres of a tidal one, and a trade effluent consent if wash-down water goes to a public sewer. In a drained fenland or levels catchment the consenting body may be an internal drainage board rather than the county.
6. Write the maintenance plan as if you meant it. On a yard producing hay chaff and shavings, three-monthly inspection of gutters and catchpits and six-monthly inspection of the flow control are not boilerplate.

A drainage strategy that never engages the lead local flood authority is a strategy that has not yet been tested. If yours has already been refused or objected to, our guide to resolving an LLFA objection sets out the usual causes, an Environment Agency objection follows a different route again, and discharging drainage planning conditions covers the process from condition to sign-off.

Equestrian schemes are small, and they are treated as small, which is exactly why they attract templated reports that fall over at validation. The two things that decide most of them are settled before a word is written: whether the buildings can sit outside the flood risk, and whether anyone has taken an honest position on what the arena does with rainfall. Get those right and the rest is arithmetic.

Unda prepares flood risk assessments and drainage strategies for equestrian schemes across England and Wales, from a single field shelter to a competition yard. If you have a stables or ménage application in the pipeline, talk to our chartered flood risk consultants about a flood risk assessment for your planning application. We quote within the hour, on 01293 214444 or enquiries@unda.co.uk.

Equestrian flood risk and drainage: frequently asked questions