Accredited laboratories for soil and water analysis

A contamination report stands or falls on where the numbers came from. If the soil results in a Phase 2 investigation were produced by a laboratory without the right accreditation for those determinands, the local planning authority's contaminated land officer can reject the report without engaging with a single conclusion in it. That happens more often than developers expect, and the remedy is another round of sampling on a site that has usually already been fenced and cleared.

The accreditation question has a short answer and a long one. The short answer: soil and water analysis relied on for planning should be carried out by a laboratory accredited to BS EN ISO/IEC 17025 by the United Kingdom Accreditation Service, and for most contaminated land work the Environment Agency's Monitoring Certification Scheme accreditation is expected on top of that. The long answer is about which determinands are covered, which are not, and how you tell.

Why accreditation decides whether the data counts

Planning decisions on contaminated land rest on a risk assessment, and a risk assessment rests on measured concentrations. The National Planning Policy Framework puts responsibility for securing safe development on the developer, with the authority checking that the site can be made suitable for its proposed use. Officers are not equipped to audit a laboratory, so they use accreditation as the proxy for reliability.

Accreditation is not a badge the laboratory holds in general. It is granted for named methods applied to named matrices, and it appears on a schedule of tests held by the accreditation body. A laboratory can be accredited for metals in soil and not for the same metals in water, or accredited for total petroleum hydrocarbons by one method and not by another. The schedule is public, and reading it takes five minutes.

Where the site may fall within Part IIA of the Environmental Protection Act 1990, the stakes rise further, because the data may end up supporting or resisting a determination of contaminated land. The Contaminated Land Assessment report types for Phase 1 and Phase 2 are among those where a suitably qualified and experienced professional is legally expected to take responsibility for the work. An AI-assisted draft can structure the reporting and flag what is missing. It cannot sign off analytical data, and it is a starting point for that professional rather than a substitute.

The schemes recognised in practice

Three names come up on English planning sites.

Accreditation to BS EN ISO/IEC 17025 by the United Kingdom Accreditation Service is the baseline for any testing laboratory. It covers technical competence, method validation, equipment calibration, staff training and quality control.

The Environment Agency's Monitoring Certification Scheme, usually shortened to MCERTS, sits on top of that for environmental measurement. For chemical testing of soil it adds performance requirements specific to the matrix, including precision and bias criteria and participation in proficiency testing. Where a report is going to an Environment Agency officer as well as a planning officer, expect the absence of it to be raised.

For potable water, the drinking water testing specification applies instead, and it is a separate scheme with its own requirements. Asking a soil laboratory to run a private water supply sample and quoting the result against drinking water standards is a common error on rural sites.

Matching the test schedule to the conceptual site model

The analytical suite should follow from the conceptual site model, not from a standard package. The model identifies sources, pathways and receptors, and the tests exist to establish whether the sources are present at concentrations that matter to those receptors.

A former petrol filling station needs a hydrocarbon suite with speciation into aliphatic and aromatic fractions, plus the volatile compounds that drive vapour risk. A former orchard needs organochlorine pesticides and arsenic. A made ground site with demolition rubble needs asbestos screening and identification. Ordering a generic metals and hydrocarbon package on all three wastes money on two of them and misses the answer on the third.

The Environment Agency's land contamination risk management guidance sets out the tiered process the whole exercise follows, and it is the document to cite when a planning officer asks why a particular suite was chosen. Write the justification into the report rather than leaving the reader to infer it.

Sampling, containers and holding times

Data quality is decided in the trial pit, not the laboratory. Four things routinely invalidate results.

• The wrong container. Volatile organic compounds need a sealed vessel with minimal headspace, and samples decanted into a tub on site have already lost part of what you are measuring.
• Temperature. Samples that sit in a van in July have a different chemistry by the time they arrive.
• Holding times. Each determinand has a maximum period between sampling and analysis, and results outside it should be reported as deviating samples.
• Chain of custody. Every transfer of a sample from hand to hand needs a record with date, time and signature, so the sample analysed can be tied to the location logged.

Laboratories will flag deviating samples on the certificate if they are told when the sample was taken. They cannot flag what they are not told, which is why sampling sheets and the laboratory instruction need to carry the same information.

Detection limits against screening values

The limit of detection has to sit below the assessment criterion, or the result is meaningless. A soil result reported as "less than 50mg/kg" against a screening value of 20mg/kg tells you nothing.

This is a live issue for a handful of determinands where the generic assessment criteria are low relative to routine laboratory practice. Check the laboratory's reporting limits against the criteria you intend to use before the samples go in, and ask for a lower limit method where the standard one will not resolve the question. Asbestos in soil is the clearest case: a screening result of "not detected" by a method with a high reporting limit will not satisfy an officer who has seen fragments in the site photographs.

State in the report which generic assessment criteria were used, who published them and when, and whether they were derived for the land use proposed. Residential with plant uptake, residential without, allotments and commercial give materially different numbers for the same substance.

Reading a certificate of analysis

A certificate is a technical document and it repays a careful read. Look for the accreditation mark against each determinand rather than at the top of the page, because certificates routinely mix accredited and unaccredited tests, with a footnote explaining which is which. Check that results for soil are reported on a dry weight basis with the moisture content given. Check the date of sampling against the date of receipt and the date of analysis. Check the sample references against your own logs.

Then check what is missing. A suite that was ordered but not run, a sample marked as insufficient, a result suppressed because of interference: all of these should be visible, and all of them leave a hole in the risk assessment that someone will find later.

Where unaccredited data undermines a report

Field screening with a photoionisation detector, portable X-ray fluorescence readings and on-site colorimetric kits all have a proper place. They guide where to sample and where to dig deeper. They are not a substitute for laboratory analysis in a risk assessment, and presenting them as if they were is one of the fastest ways to lose credibility with a regulator.

The same applies to historic data. Results from an investigation carried out 15 years ago by a laboratory whose accreditation status cannot now be established may still be useful context for the conceptual model, but they should be labelled as such rather than tabulated alongside current data as though the two carry equal weight.

Where a council queries the author or the data behind a report, the position is worth understanding before you respond, and the note on when a council can reject a report because of its author sets out the ground. Budget for the analysis properly at the outset too, because laboratory costs are one of the items most often excluded from a fee proposal, as the guide to reading a consultant fee proposal explains.

For the review stage, the professional directory is the route to a reviewer who can check the interpretation and take responsibility for it.

This is general guidance, not legal advice. Check the position with your local planning authority or a qualified professional before relying on it. For specialist technical reports, qualified professional input may be required. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.