Increasingly sophisticated laboratory tests are been developed for analysing soil, water or gas samples from contaminated sites. Methods such as gas chromatography-mass spectrometry and inductively coupled plasma atomic emission spectroscopy (G E April 1999) allow increasingly stringent regulations and requirements of modern risk assessment procedures to be met.
However, this complex (and often very expensive) equipment may as well be consigned to the dustbin if perhaps the most important step of investigation - sampling - is not carried out correctly.
The most obvious point to make is that the laboratory test results are only as good as the sample received. Anyone involved in the site investigation business should, I hope, not have to be told that a sample should be as representative of the area as possible.
Admittedly this can be difficult with a finite budget and a finite number of samples that can be taken, especially on particularly heterogeneous sites such as old gasworks or mixed industrial premises.
Learned practitioners, such as the late Professor Colin Ferguson, have written authoritatively on this subject, advocating various sampling patterns to minimise the chance of missing a hotspot.
But assuming the sample is representative of the site, the critical factor is the chemical testing laboratory's advice. The laboratory should be contacted before an intrusive site investigation or monitoring programme begins.
A lot can go wrong on the sample's transit to the laboratory and a number of issues need to be addressed as early as possible.
What are the samples being tested for?
Different determinands require different containers and preservation. For instance, looking for BTEX (benzene, toluene, ethyl benzene and xylene) compounds in groundwater following a suspected petrol spill requires at least special VOC (volatile organic compound) vials, cool boxes and ice-packs. Such vials should be filled to the brim with no headspace, minimising loss by evaporation.
Other so called labile [unstable] determinands (such as ammonia, biochemical oxygen demand or microbiological parameters) require either cooled and rapid transit to the laboratory or sample preservation - such as metals in water. A reputable laboratory will supply bottles specially preserved with acid.
If such bottles are used, waters must be filtered into them on site, otherwise the acid preservative used may leach out metals from particulate matter, thus positively biasing results.
Different determinands require different amounts of sample. More stringent limit of detection (LOD) requirements will impact on this.
Do not assume that the laboratory will know the LOD requirements: it must be told.
Lower LODs normally require more sample, which can be difficult in the case of a dry well.
The laboratory will always advise on this, but ideally will be sent sufficient sample to analyse for all the determinands requested, to the LOD required, and then some.
Additional material is needed because Further analysis for other determinands may be required following preliminary laboratory findings.
A laboratory needs extra samples for quality assurance purposes: reputable laboratories will analyse replicate samples normally at a minimum frequency of 1 sample in 10, to give an estimate of precision. If the amount of sample is insufficient, this cannot be done.
Things can go wrong in the laboratory, for example instrument failure or the quality assurance criteria for analysis of a sample not being achieved. This will necessitate re-analysis.
But waste disposal costs are now a significant outlay for testing laboratories, so they will not appreciate half a tonne of each sample when 500g would do.
Courier costs, normally based on weight, should be kept to a minimum.
How is the data going to be used?
Is a proprietary software package such as HoleBase being used that requires data in a certain format (for example AGS format). Can the laboratory supply this? It is no good finding out halfway through the project that it cannot.
Is the data going to be reformatted (rather than say, just appending the laboratory's certificate of analysis for the site report)? If so, it is best to get the data sent electronically, either on disk or via email, for two reasons.
It avoids rekeying and transcription errors, which should be minimised at the laboratory by independent transcription checks, or better still, double data entry.
Is QA/QC data required with the test data? No reputable laboratory should blink an eye at this request: my firm actively encourages it (for obvious reasons), but it saves time and money if this is requested before the project starts. Otherwise, spreadsheets may have to reformatted and raw data retrieved.
Minimum QA/QC data supplied should include blank data, spike recovery or certified reference material (CRM) data and duplicate data. Chromatograms can normally be supplied where appropriate.
Do chain of custody procedures need to be reported with the results? This is facilitated with chain of custody documentation which will be supplied by a reputable testing house, but requires active co-operation.
Is work being carried out to specific risk assessment or client specific procedures? If so, will the tests requested give the information required or follow these procedures? Petroleum hydrocarbon analysis is a particular minefield in this area, and the advice of the laboratory should be sought in such cases.
Some clients are becoming increasingly prescriptive with testing procedures.
Are there any special sample preparation requirements?
Solid samples may need particular treatment rather than the standard air drying, sieving and fine grinding. Normal preparation procedures may be inappropriate for the samples or the investigation.
Trained laboratory sample preparation staff will be on the lookout for anything out of the ordinary, but normal preparation procedures can end up rejecting a major part of a sample, which may not be what is needed.
Similarly, you may want floating product on a water sample analysed separately.
Legibility and ambiguity
All samples must be legibly and securely labelled. Testing requests must be clear and unambiguous.
Planning will facilitate this, particularly if lines of communication with the laboratory have been opened before the project starts.
However, it is common for laboratory sample reception staff to waste hours every week trying to decipher illegible or ambiguous instructions or chain of custody documents that do not tally with the samples received.
Wasted hours obviously mean delays in commencing analysis of the sample. If a specific quotation for a project has been given by the laboratory, refer to this on the sample documentation (usually the chain of custody).
This will ensure that any special technical, turnaround or other requirements are met. It will also mean that invoicing is correct.
When is the data needed?
Some tests just cannot be done in a day - or even three. All laboratories have a finite capacity, but most will be receptive to offering rapid turnarounds if it is technically feasible.
Planning will ensure that requirements are technically feasible and that the project size and logistical requirements are not beyond the laboratory's capacity.
Advance warning will allow it to prepare a slot for the work, and make overtime working arrangements if necessary.
Health and safety
The desk study or site walkover may have identified particular hazards that the laboratory must be informed of.
Potential asbestos, pathogenic or radiological hazards are examples where a phased approach with the use of specialist experts is advocated.
Otherwise both laboratories and couriers could be inadvertently exposed - for example through inhalation of asbestos fibres while drying and grinding samples.
Where is the site?
Most laboratories provide couriers as an integral part of their service. However a site in the Outer Hebrides may preclude testing for particularly labile determinands in a laboratory further south.
Soil samples from a non-European location will require a MAFF Soil Import Licence to enter the UK. Make sure the laboratory has one and supplies one to get the sample through customs.
Apologies if this advice seems to insult some readers' intelligence, but this is written with many years' experience in the laboratory testing business.
A lot of heartache and waste could have and can be avoided with a little planning and by taking the advice of laboratories.
Paul Board is a chartered chemist and business development manager of Robertson Laboratories.
Visit Robertson Laboratories' Frequently Asked Questions at www. robresint. co. uk