Sequencing site investigation before an application

Ground investigation done after the design is frozen tells you what you have got wrong. Done before it, the same money tells you what to draw. On a small residential site the sequence that works is a desk study while the layout is still fluid, intrusive work before the drawings are finalised, and a monitoring programme running in parallel with the application rather than after the decision. That order costs no more and removes the two expensive surprises: a foundation solution nobody budgeted for, and a remediation requirement that changes the plot layout.

The constraint that sets the programme is not the drilling. It is groundwater and ground gas monitoring, which take months and cannot be compressed.

Why investigation before design saves money

Four design decisions depend on ground conditions, and all four are hard to reverse once drawn.

Foundation type and depth, which on shrinkable clay near trees is set by soil shrinkage potential and tree water demand rather than by bearing capacity.

Whether surface water can go to ground, which decides whether the layout needs space for an attenuation tank and an outfall.

Whether a basement or a lower ground floor is achievable, which depends on groundwater level, permeability and the effect on neighbours.

Where clean cover is needed for gardens, which on a site with made ground can dictate levels, retaining structures and the position of soft landscaping.

A scheme drawn without that information will usually need changing, and a change at planning stage is a revised drawing while a change after permission is a section 73 application. On a self-build plot that difference can be three months.

Desk study first, intrusive work second

The desk study is the Phase 1 preliminary risk assessment, and it is the deliverable that should exist before the layout is settled. It is also the least costly part of the work by a wide margin.

A thorough desk study assembles the following.

1. Historic Ordnance Survey mapping across the site's full recorded history, read for former industrial use, infilled ground, ponds, watercourses, railway sidings and tanks.
2. Published geology and any borehole records held in the national archive for the site and its surroundings.
3. A coal mining report where the site lies in a coalfield referral area, and equivalent checks for other mineral workings.
4. Landfill and waste site records, both licensed and historic.
5. Radon potential for the address, and the protective measures that implies.
6. Flood and surface water risk, groundwater source protection zone status and proximity to watercourses.
7. Any previous site investigation held by the authority on earlier applications for the site, which is frequently available and frequently overlooked.

That last item is worth a direct request to the local planning authority before commissioning anything. A neighbouring site investigated five years ago can answer half your questions.

The desk study produces a conceptual site model: sources, pathways and receptors, with the plausible pollutant linkages identified. The intrusive work then tests that model rather than exploring at random, which is why doing it the other way round costs more. The resulting contaminated land assessment follows the staged risk management approach set out in the Environment Agency's guidance, moving from risk assessment to options appraisal to remediation and verification.

Combining geotechnical and contamination sampling

One mobilisation, two purposes, if the scoping is done properly. The requirements differ enough that a combined investigation has to be designed as one rather than ordered as two.

Geotechnical testing wants undisturbed samples for strength and consolidation work, and in situ testing at specified depths. Contamination analysis wants representative samples in specified containers, with limited headspace and short holding times before analysis, and an analytical suite chosen from the conceptual site model rather than a default list.

Three decisions have to be made before the rig arrives. Which exploratory holes carry gas and groundwater monitoring installations, because installing them later means re-drilling. What the analytical suite covers, including asbestos in soil, which is found far more often than it is scoped. And who logs the holes, since a single engineer producing one set of logs avoids the situation where the Geotechnical Assessment and the contamination report describe the same stratum differently.

Infiltration testing belongs in the same visit where a soakaway is contemplated, at the depth the soakaway would occupy, and the results feed the Drainage Strategy. Testing at a convenient depth rather than the design depth wastes the exercise.

Access and neighbour consent for trial pits

On a tight site this is the item that slips.

A rig needs a working width and headroom, and a cable percussion rig needs more of both than people assume. Check overhead cables, gateway widths and the turning space. Where the only access is across a neighbour's land, a licence to enter is a negotiation with its own timetable.

Underground services have to be located before any hole is opened, by utility records and a scanning survey, and a hand dug inspection pit to the first metre is the normal precaution. On a site with an existing building, that often means breaking out slabs.

Where excavation is close to a boundary structure, the Party Wall etc. Act 1996 may require notice, and a trial pit adjacent to a neighbour's foundation is exactly the situation the Act contemplates. Factor the notice period in. Demolition brings its own prior surveys, and the sequencing described in demolition and asbestos surveys before a planning submission runs alongside this work rather than after it.

Seasonal constraints on groundwater data

This is the part that cannot be shortened, and it is the reason to start early.

Groundwater levels in England are generally highest in late winter and early spring and lowest in late summer and autumn. A single reading taken in September tells you very little about the level the design has to work to, and an infiltration test in a dry spell on clay can produce a rate that a wetter month would not support.

Ground gas monitoring needs several visits across a range of conditions, including periods of falling and low atmospheric pressure, because gas emission rates rise as pressure drops. A credible monitoring record therefore spans months rather than weeks, and the number of visits depends on the risk identified.

The practical consequence for a programme is this. Install monitoring wells as early as you can, even before the design is settled, and let the monitoring run while the drawings and the application are prepared. A scheme that submits with monitoring in progress and a clear statement of what remains outstanding is in a far better position than one that has to start monitoring after a condition is imposed.

Where a basement is proposed, add a season. A Basement Impact Assessment needs a defensible groundwater regime, and one round of readings will not provide it.

What to submit and what to hold back

Submit the interpretation, not the raw data alone. A factual report with no interpretive section invites a condition requiring the interpretation you have already paid for.

What belongs with the application: the Phase 1 desk study and conceptual site model, the factual results of any intrusive work completed, an interpretive report with conclusions on geotechnical feasibility and contamination risk, a statement of what further work is needed and when, and the infiltration or percolation results where drainage depends on them.

What is better held back: incomplete monitoring rounds presented as conclusions, draft reports, and a remediation strategy committed to before the ground is understood. Everything submitted goes on the public register and can be read by anyone, including a future purchaser and a neighbour's solicitor, so submit documents you are content to be bound by.

One further point of judgement. A report that identifies a risk and proposes to address it is a stronger submission than one that reports no risk on thin evidence. Officers and their contaminated land colleagues are used to residual uncertainty and have conditions for it. What they object to is a conclusion that the evidence does not support, and once that objection is raised the whole document loses weight.

This is general guidance, not legal advice. Land contamination assessment and remediation reporting require a licensed or competent professional, and an AI-assisted draft is a starting point for that professional rather than a substitute, so take qualified advice before relying on anything described here.

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