Infiltration testing that fails: what to do next

A failed soakaway test does not stop a scheme, but it does change the drainage strategy, the layout and sometimes the number of plots. If your soakaway test failed, planning-wise you now have to move down the discharge hierarchy and evidence why infiltration was ruled out. Authorities and lead local flood authorities accept that. What they will not accept is a strategy that quietly drops the failed test out of the report and assumes a sewer connection.

Three things need doing in order: understand why it failed, decide whether to retest, then redesign around the answer.

What a failed infiltration test looks like

BRE Digest 365 is the method almost every authority expects. A trial pit is dug to the proposed base level of the soakaway, filled with water three times in succession, and the time for the level to fall from 75 per cent to 25 per cent of the effective depth is recorded each time. The infiltration rate follows from the geometry and the slowest of the three results is used for design.

Failure shows up in four ways. The rate comes out too low to be usable, commonly below about a millionth of a metre per second in clay soils. The pit does not empty at all between fills. Water arrives from the sides rather than draining away, which means you have hit groundwater. Or the three results diverge so widely that no design value can be defended.

A rate that is merely modest is not a failure. Low rates can work with a larger footprint, more surface area relative to volume, or a series of shallow features rather than one deep one. Check the arithmetic before you abandon infiltration, because a 50 per cent larger soakaway usually costs less than a sewer connection.

Common causes: groundwater, clay, made ground

Groundwater is the most frequent reason, and the fix is geometry. A soakaway needs unsaturated ground beneath it, and the usual working rule is at least a metre between the base of the feature and the highest expected water table. Where the test was done in August, the winter water table may be a metre higher, so a pit that looked dry can be submerged in February. Standpipes monitored over a winter period are the only reliable evidence.

Clay is the second. London Clay, Oxford Clay, Mercia Mudstone and glacial till all give rates that rule out infiltration for anything but the smallest feature. That is a geological fact rather than a testing failure, and the desk study should have predicted it. Look at the geological mapping before you dig.

Made ground is the awkward one. Fill on a brownfield site can be highly permeable in places and impermeable in others, which produces divergent results across a site. It also raises a separate objection: infiltrating clean surface water through contaminated fill can mobilise pollutants into groundwater, and in a source protection zone the Environment Agency will resist infiltration regardless of the measured rate. The groundwater flooding evidence you gather for the flood risk work usually answers half of this question already.

Whether to retest deeper or elsewhere

Retesting is worth doing in three situations and a waste of money in the others.

Retest at a different location where the geology varies across the site, the first pit was placed for access convenience rather than where the feature will go, or the mapping suggests a permeable band under part of the site. Retest at a different depth where a shallower horizon of sand or gravel exists above the clay, because a shallow, wide feature in that horizon can work when a deep one cannot. Retest in a different season where the failure was caused by a high water table and you intend to argue a seasonal case, which requires monitoring rather than a single repeat.

Do not retest simply hoping for a better number in the same ground. And whatever you do, record both sets of results. A report showing one favourable test, where the site investigation log reveals two earlier failures, destroys the credibility of the whole submission.

Moving down the discharge hierarchy

The hierarchy is settled: infiltration first, then discharge to a surface water body, then to a surface water sewer, and only as a last resort to a combined sewer. Each step down needs the step above ruled out with evidence.

Watercourse discharge gets skipped too readily. A ditch 40 m away across a neighbour's field is a real option if an easement can be agreed, and consent from the lead local flood authority for a new outfall into an ordinary watercourse is a normal process rather than an obstacle.

Justifying a sewer connection to the authority

The right to connect surface water to a public sewer under section 106 of the Water Industry Act 1991 exists, but the undertaker can refuse or restrict a connection on capacity grounds, and it will normally impose a maximum discharge rate. That rate is what drives everything else in the design.

Your justification needs four items, and it should be short. The infiltration test results with the location, depth, date and method. The geological and groundwater reasoning for why infiltration is not viable. Written evidence that a watercourse discharge was investigated, with the reason it was rejected. And the undertaker's pre-development enquiry response giving the connection point and rate.

Put all four in the Drainage Strategy as an appendix and summarise them in one paragraph in the main text. Officers and flood authority reviewers look for that paragraph first.

Storage implications of losing infiltration

This is where the failure costs money. Infiltration removes water from the system. A restricted connection only delays it, so you have to store the difference.

An illustrative example. A 0.4 ha site of nine houses would have discharged to soakaways with no attenuation. The test fails, and the undertaker permits a connection at 5 litres per second. Attenuating the 1 in 100 year event with a climate change allowance now needs a storage volume of a few hundred cubic metres, which at 1 m effective depth in a geocellular tank is a footprint comparable to two plots. Options are a basin in open space, tanked storage under the access road, oversized pipes, permeable paving sub-base storage, or a combination. Every figure here is illustrative and the volume comes from the hydraulic calculation for your own site.

Three practical consequences follow. Layout has to change, because storage needs a place and a level. The finished floor level strategy may change, since storage under a road needs cover and gradients. And the maintenance and adoption position changes, since tanked storage under a private drive stays private and needs a management arrangement.

Deal with this at layout stage. Retrofitting several hundred cubic metres of storage into an approved layout is the sort of change that becomes a fresh application.

Reporting the failure honestly in the strategy

Write the failure up as a finding, with the raw data in an appendix: pit locations on a plan, dimensions, depths, the three fill and fall records, the calculated rates, photographs and the date. Then state the design decision that follows and the consent route you are pursuing.

Two habits make reviewers trust the document. Give the groundwater evidence its own short section, including any monitoring period and the highest level recorded. And say what happens if the assumption proves wrong at construction, because a retest at formation level is normal practice and the verification stage will expect it.

For a single plot the same logic applies at a smaller scale, and the drainage evidence for one new dwelling is a useful comparison for how much detail a minor application actually needs. On flood zone sites the strategy also has to sit consistently with the Flood Risk Assessment, because a contradiction between the two documents is the fastest way to a request for further information.

Enter your postcode on PlanScope to see which drainage and flood documents your site is likely to need before the next round of testing.

This is general guidance, not legal advice. Check the position with your local planning authority, the lead local flood authority or a qualified professional before relying on it.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.