Contaminated soil analysis is a relatively recent phenomenon compared with water and air pollution testing. Reflecting this, the National Measurement Accreditation Service (NAMAS), which accredits the quality control procedures of laboratories for various tests, has over 30 laboratories accredited for water analyses and five for some air pollution tests.
However, only one laboratory is accredited for analysis of soil contaminants. But applications for accreditation have recently surged, and NAMAS expects another dozen or so to gain accreditation within the next year.
NAMAS accreditation is gradually becoming the entrance qualification for environmental testing laboratories under pressure from customers. Stuart Newstead, head of HM Inspectorate of Pollution's monitoring unit, told an LGC conference in October that he is "very frustrated by the quality of work that is performed for us." From next year, HMIP will be stipulating that for large contracts the laboratory must gain NAMAS accreditation during the course of the contract if it does not already possess it.
While NAMAS accreditation certifies that a laboratory has adequate quality control procedures and therefore that the results of chemical analysis have greater precision, it provides no guarantee that results are accurate. A laboratory's results may be consistently biased.
The results from two laboratories even with good quality control procedures may vary by 20-30%, according to Ian Rix of the LGC - and divergences may be much greater for organic pollutants, says Tony Ellis of Clayton Environmental. An interlaboratory comparison project in the Netherlands gave results between 50-145mg/kg around a mean of 79mg/kg for chromium in soils, although different testing methods were used. The "real" value, according to more precise testing, was between 107-132mg/kg.
Given that expensive contaminated land clean-up decisions are based on whether certain "trigger values" of contamination are exceeded, these variations can prove costly to customers, especially if the results are close to the trigger values.
In the water and indoor air pollution testing fields the accuracy of results has been improved by interlaboratory proficiency testing, in which laboratories check their results for the analysis of a particular sample against those of other laboratories. Laboratories with results at variance from the norm are therefore warned that they may be doing something wrong. WRc's Aquacheck has been in operation for over three years for water pollution analysis, while for toxic substances in workplace atmospheres the Health and Safety Executive operates WASP.
The LGC initiative, named CONTEST, is the first in the UK for soil analysis. LGC sends a sample of homogenised naturally contaminated soil to participating laboratories which perform a standard test on the sample. The laboratories receive a report after each testing round on their relative performance against other laboratories, an annual report, invitations to meetings to discuss any implications, and advice over a telephone "hotline" if they cannot recognise where they are going wrong. This service costs £1,000 in membership fees and further costs are incurred in performing the tests. To date, 44 laboratories representing over 50% of the contaminated soil testing market have joined up.
So far, according to Ian Rix, the results have been "better than expected" for the first testing round on toxic metals. Subsequent rounds will probably deal with other easily measured contaminants such as cyanide, but greater divergence is expected when traces of contaminants such as dioxins have to be detected.
The testing methods stipulated are generally based on draft British Standards - there are no agreed procedures. The feedback from CONTEST will, it is hoped, assist in the development of final British Standards by highlighting inadequacies in the test methods.
At present there are no incentives, besides professional pride, for laboratories to act on poor results from CONTEST. However, the LGC hopes that NAMAS will encourage laboratories to participate in the scheme, and that NAMAS assessors will check whether laboratories have acted on its results.
NAMAS already requires accredited laboratories to participate in WASP and an asbestos analysis proficiency testing scheme. And within the next few weeks it will require accredited laboratories within WASP to maintain a satisfactory level of performance. For contaminated soil testing, Brian Kent from NAMAS envisages that participation in CONTEST will become a requirement within 6-12 months, and that satisfactory performance will be mandatory in two years.
One drawback of proficiency testing is that every laboratory with results near the average may nevertheless have a "false" result. But because naturally contaminated soils rather than spiked samples have to be used to reflect real-life situations, a round-robin approach appears to be the only solution. Accuracy can be improved by comparing the results from these off-the-shelf tests with the results of expensive precision analysis.