Perched Groundwater: How to Tell It From the Water Table
Estimated reading time 15 minutes
Perched groundwater is a body of shallow groundwater held above the main water table by a layer of low-permeability ground, with unsaturated material beneath it. That unsaturated gap is the definition. Water standing in a trial pit on a clay site in February is frequently perched, and a designer who records it as the site’s water table will set the maximum likely groundwater level against the wrong horizon.
A single water level on a trial pit log is not a water table. Where water is struck higher than the level it later settles to, the British Geological Survey reads that as the signature of a perched water table.
What is perched groundwater?
Perched groundwater is groundwater separated from the main body of groundwater beneath it by unsaturated ground. It collects where percolating rainfall meets material it cannot pass through quickly: a clay lens, a silt band, a cemented horizon. The water saturates the ground above it. The layer doing the holding is the perching bed, and the top of the saturated zone resting on it is a perched water table.
The US Geological Survey’s committee on groundwater terminology fixed the wording in 1972, and it has not needed revising since:
Perched ground water is unconfined ground water separated from an underlying body of ground water by an unsaturated zone. Its water table is a perched water table. It is held up by a perching bed whose permeability is so low that water percolating downward through it is not able to bring water in the underlying unsaturated zone above atmospheric pressure.
US Geological Survey · Water-Supply Paper 1988 (1972)
The British Geological Survey puts the same mechanism in UK terms. Perched water tables "occur when there are lenses of impermeable material in a rock that is otherwise permeable", allowing "small localised water tables to develop over the lenses, which can overlay a regional water table in the surrounding rock mass".
Two phrases appear on ground investigation reports and they are not synonyms. A perched body holding enough water to supply from is a perched aquifer; where there is not enough water for that, the term is perched groundwater. Most perched bodies under development sites in England and Wales are the second kind, which is why they are treated as a ground constraint rather than a resource. Perched water therefore sits in the evidence base for a compliant surface water drainage strategy, alongside the infiltration rate and the discharge point.
Why water sits above the main water table
Perched groundwater forms when downward percolation is slowed faster than it is supplied. Rainfall moves down through the ground until it meets material of much lower permeability. If water arrives more quickly than that layer can transmit it, the ground above saturates. Below the perching bed the ground stays unsaturated. Water is still draining through it, just slowly, and that dry gap between the two saturated zones is what distinguishes perched groundwater from the regional water table.
- What makes a bed perch water: a permeability contrast with the material above it, not thickness. An iron pan a few centimetres deep will hold water that a metre of silty sand will not.
- Recharge has to keep up: perched water survives only while water arrives faster than the bed beneath transmits it, which is why many perched bodies are a winter phenomenon.
- It escapes sideways. Perched water drains off the edges of the layer holding it, which is why it shows up as seepage on a cutting face, or as a spring line part-way up a slope where the bed daylights. Intercepting it is what a cut-off drain is for.
- Permanence varies. The USGS separates permanent perched water, maintained by frequent recharge, from temporary perched water that drains away between wet periods.
- Its extent is the layer’s extent: a perched body is only as large as the bed supporting it, so two trial pits forty metres apart can give completely different answers without either being wrong.
How to tell perched groundwater from the water table
Compare the depth at which water is first encountered with the level it settles to afterwards. BGS states the test directly: "Comparisons of water strike to rest water level can be used to identify perched water table (strike > level) or confined conditions (strike < level)." Water struck high that then falls away is perched. Water struck low that then rises is under pressure. The difference is diagnostic.
| What the record shows | What it indicates | What it means for design |
|---|---|---|
| Water struck above the level it later rests at | Perched groundwater | The controlling level is the perched horizon, not the regional water table |
| Water struck below the level it later rests at | Confined conditions | Head sits above the stratum; infiltration into it needs separate assessment |
| Strike and rest level broadly coincide | An unconfined water table at that level | Seasonal monitoring still sets the maximum likely level |
| Strikes at more than one depth | More than one saturated horizon | No single groundwater figure represents the site |
Whether the record can be read at all depends on how the instrument was built. BGS notes that the water level in an observation borehole "will be a composite of all aquifers that are penetrated by the borehole", and that an instrument screened to one horizon with the others cased off is a piezometer rather than a general observation hole. On a site carrying both a perched horizon and a deeper water table the two installations answer different questions, and only one of them answers the question Requirement 1.15 asks.
A standpipe left open across a perched horizon and the ground beneath it drains the perched water and reports a level that exists nowhere on the site.
Lead local flood authorities ask for the distinction in writing. Responding on an outline application at Halterworth in February 2024, Hampshire County Council as lead local flood authority recorded that "a perched water table was recorded at 3-4m below ground level in all three boreholes", and required that "monitoring wells should be installed so that the perched water table can be measured without it draining through the less permeable soils".
None of this is separate from what a ground investigation for drainage and flood risk is for, from when infiltration testing is required, or from reading a BGS borehole log before a number from it reaches a design.
Where perched groundwater forms in UK ground
Perched groundwater occurs wherever permeable material overlies something that sheds water, which covers a great deal of English and Welsh geology. BGS sheet explanations record it by name across the London Basin and the Chilterns, and it is routine in glacial till. Published mapping rarely shows it. The layers responsible are too thin and too local to appear on it.
- Glacial till: sand and gravel beds within boulder clay hold perched water, which BGS describes as having "limited storage indicative of their isolated nature".
- The Lambeth Group and Reading Formation: interbedded sands and clays that BGS records as producing "a multi-layered aquifer and perched water tables", with springs where the sands meet clay.
- London Clay and the Claygate Member: BGS lists a perched water table among the formation’s standard ground conditions.
- The Bagshot Formation: low-permeability clay layers within it create perched water that BGS notes "may impair slope stability".
- Clay-with-flints over Chalk: an aquitard. It reduces infiltration into the Chalk beneath and pushes run-off to its margins.
- Made ground: fill of unknown composition and thickness, recorded on the geological map only where someone happened to record it, which can both create a perched horizon and sit on top of one already there.
Desk study will not flag any of it. BGS says so in the user guide to its own Infiltration SuDS Map: the dataset "does not consider the presence of perched water tables, which may form above layers of low permeability material, causing a water strike above the level of the regional water table". Perched water tables sit in that guide’s explicit list of exclusions, alongside contaminated land and unrecorded made ground.
A BGS research borehole near Boyton End, on the Essex–Suffolk border, shows the size of the error that follows. The site had been chosen because surrounding records "mostly showed a sequence of homogeneous boulder clay resting directly on Chalk". Then a saturated silt band turned up between 4.75 and 5.30 metres, perched water stood at 4.5 metres below ground while the Chalk water table sat at 26.8 metres, and the perched horizon nobody had expected changed how the borehole itself had to be built. Twenty-two metres separated the two levels. The mapping had called that ground uniform. Unda’s ground conditions and geology checker will tell you whether infiltration is worth pursuing at all; no desk tool will tell you whether a lens two metres down holds water in February, which is where a measured drainage strategy starts.
What perched groundwater means for infiltration and the 1 metre rule
The 2025 National Standards for Sustainable Drainage Systems require that "the base of infiltration drainage features shall not be within 1m of the maximum likely groundwater level on the site". Nothing in the Standards excludes perched groundwater from that definition. A perched horizon two metres down, above a proposed soakaway, is groundwater for the purposes of Requirement 1.15, and a design set against a regional water table twenty metres below it will not comply. The shallow horizon governs.
- Establish which water has been measured, by comparing every strike against its rest level before any figure reaches the design.
- Install to the right horizon, screening instruments so a perched level can be read without draining it into the ground beneath.
- Monitor through the wet season. Requirement 3.11 asks for groundwater levels "confirmed across the site through site investigation during high groundwater level periods".
- Set the maximum likely groundwater level against whichever horizon controls the feature, which on a perched site is the shallow one.
- Test infiltration in the material that will receive the water, not in a permeable lens inside it.
CIRIA’s SuDS Manual guards the last of those in its infiltration assessment checklist, which asks the assessor to confirm a test result "is likely to be representative of the wider ground mass (eg the test has not been undertaken in a limited extent of sand within a mass of clay)". A perched water table and an unrepresentative BRE 365 infiltration test are frequently the same lens of ground measured twice: the sand that took water quickly in the test is the sand that held the perched water in the first place, and neither result describes the clay around it.
The word itself appears in none of the national documents. "Perched" is absent from the 2025 National Standards, from the August 2026 NPPF, and from Planning Practice Guidance on flood risk. One council has written down what that silence means:
The national standards for sustainable drainage design defines groundwater as water which is below the surface of the ground in the saturation zone, and in direct contact with the ground or sub-soil. No distinction is made between the aquifer and perched groundwater.
Arun District Council · Groundwater monitoring guidance note, 24 October 2025
The absence of the word is not an exemption. Perched water is caught by the general definition of groundwater, so the 1 metre separation applies to it exactly as it applies to an aquifer, and the same goes for the pollution controls that bite inside a groundwater source protection zone. Arun adds that a high water table is common across its district "whether this is perched groundwater, the aquifer or springs", which on a clay site usually means the shallow horizon governs.
Is that water in your trial pit the water table?
Unda’s chartered consultants interpret groundwater records, scope the monitoring and set a design groundwater level that will stand up to the LLFA. Quote within 60 minutes.
Get a drainage strategy quoteWhen perched groundwater decides a planning application
Unresolved perched water causes more delay than refusal, but it does reach appeal. At Angmering in West Sussex, an inspector examining a drainage scheme for a development already partly built found the groundwater evidence unsatisfactory. The appellant’s own case "refer[red] to perched groundwater affecting pre-development groundwater investigations", while the appellant’s witness suggested "there was confusion in earlier reports" on the point.
The conclusion turned on the same 1 metre separation:
In light of the issues experienced during groundwater investigations, which included possible surface water flooding and perched or passing through water affecting boreholes, I am reluctant to accept winter groundwater levels would not affect the Drainage Scheme without further investigations/monitoring.
Appeal decision APP/C3810/W/24/3349316, Angmering · 31 March 2025
Note what the inspector did not find. The scheme was never shown to have failed, and no finding was made that it would flood anything; the evidence did not allow a conclusion either way, and on the discharge of a drainage condition that is enough to leave it undischarged. Perched groundwater seldom destroys a drainage strategy outright. What it does instead is remove an authority’s confidence in the groundwater level, and an authority that cannot rely on that number will not accept the infiltration design resting on it, however competent the rest of the submission.
What settles it is evidence: the right instrument in the right stratum, read across a winter, and a design groundwater level that states on its face which horizon it refers to. That is the substance of a groundwater flood risk assessment and of what a planning application has to submit on groundwater, and it costs less than the redesign that follows an objection. On one Unda scheme groundwater modelling reshaped the drainage for 130 homes where soakaways failed, which is the same problem solved before determination rather than after it.
Frequently asked questions
Is perched groundwater the same as a perched aquifer?
Not quite, though site reports use the terms interchangeably. The distinction is yield: a perched body holding enough water to develop a supply from is a perched aquifer, and where there is not enough water for that, the correct term is perched groundwater. Most perched bodies on development sites in England and Wales fall into the second category.
Does perched groundwater show on any national map?
No. BGS states that its Infiltration SuDS Map "does not consider the presence of perched water tables", and lists them among the dataset’s exclusions. Published geological mapping is drawn at 1:50,000 with a minimum mappable thickness, so the thin lenses that cause perching fall below the resolution of the map. The same limitation applies to the national water table and groundwater flood datasets. Perched water is found by investigation.
Can a soakaway be built above perched groundwater?
Sometimes, but the perched level becomes the level the design has to clear. Requirement 1.15 asks for 1 metre between the base of the feature and the maximum likely groundwater level, and a perched horizon above the soakaway is that level. The options are to raise the feature, move it beyond the extent of the perching bed, or manage the runoff another way down the SuDS hierarchy.
Does perched groundwater count as groundwater flooding?
It sits outside the usual classification. BGS identifies three main UK settings for significant groundwater flooding: unconfined major aquifers such as the Chalk, shallow unconsolidated sedimentary aquifers, and groundwater rebound in urban centres. Perched water is not among them. It is a ground conditions and drainage design problem, though perched water reaching the surface still produces wet ground and seepage.
Can perched groundwater disappear between one site visit and the next?
Yes, which is why dry-season investigation is unreliable on clay sites. Temporary perched water, in the USGS’s terms, lasts only while recharge continues, then drains over the edge of or through the perching bed. A trial pit dug in August, on ground that holds perched water every winter, can be completely dry. Timing decides the answer.
About the author. Jackie is a co-founder and Director of Unda with 30+ years in flood risk, and sits on CIWEM’s South Eastern Branch committee. Unda has been trading since 2014, is a CIWEM Business Partner with CIWEM member and chartered (C.WEM MCIWEM) consultants, and has delivered 5,000+ flood risk assessments and drainage strategies across England and Wales.
Jackie Stone · MSci, BSc (Hons), DIC, CIWEM Environmental Partner
Latest news
One of our experienced Flood Risk Consultants will get back to you within 60 minutes