I look forward to reading the new AGS guide to selection of geotechnical soil laboratory testing (Ground Engineering April) as this seems to complement research which we first reported in 1992 (and about which we were consulted).
In general, our advice to young geotechnical engineers faced with deciding the quantity and type of geotechnical testing to be ordered is four-fold:
Make trial calculations to establish the possible line and level and also what data will be needed for the final calculations.
Ask supplementary questions, such as: will the soil rot the concrete? Might the invest-igators get killed? etc.
Check against a proforma schedule of tests, which lists all the standard tests which should be considered, and which gives provisional recommendations for how many of each test to use for particular combinations of type of soil and type of works.
Obtain approval from the boss before wasting the client's money.
The new AGS guide will strengthen the priori recommendations which we give in step 3.
The University of Glasgow
Limits of analysis
Further to your article 'Finding the best test' (Ground Engineering May), while we agree with the thrust of Dr Cairney's argument, we feel that it does not take us much further in the search for appropriate analysis. We certainly agree however that pages of expensive analytical results often do not reveal the levels of risk on a site.
Leachability testing, in accordance with the Environment Agency method described in Research and Development Note 301, goes some way to addressing the problem of appropriate metals analysis. The results indicate likely risks to groundwater and surface water quality. In addition, they also indicate, in a more realistic way, the level of risk to humans, as water soluble forms of metals are generally more available as toxins to humans.
Despite the recommendations in virtually all documentation providing advice on investigating potentially contaminated sites, phased investigation and therefore analysis is only undertaken occasionally. This is a major source of waste of available resources, perhaps most particularly those for analysis. Inadequate desk studies, or interpretation of information from them frequently results in inappropriate scheduling of analysis.
Analysis for contamination by organic compounds is particularly difficult to schedule if ICRCL documents are relied upon for advice. The toluene extractable matter test is, in our experience, not a bad starting point on sites where potential contamination by a relatively limited range of organic compounds with reasonably high trigger values is suspected. The method allows some inexpensive screening of samples. The more expensive analytical techniques can then be scheduled for the usually small number of samples with a significant TEM concentration.
It is important to know the limitations of TEM analysis however, as sulfur, in addition to the organic substances listed by Dr Cairney, is also dissolved. However it must be remembered that toxic effects from many organic compounds occur at very low concentrations and at sites where, for example, chlorinated solvents may be present, TEM would be too blunt an instrument to use as a starting point.
Analysis by GC-MS (gas chromatography with mass spectroscopy) is a powerful tool for the analysis of organic compounds. However, many commercial laboratories screen for a limited number of volatile and/or semi-volatile compounds usually depending on the extent of the library of standards available to them. As a result, analysis can result in very little speciation, most of the extract being left as an 'unidentified mixture of hydrocarbons'.
This practice may result in a great deal of the story about contamination of a site, and the associated risks, remaining untold. For example, speciation of the otherwise 'unidentified mixture of hydrocarbons' by GC-MS results in the identification of alteration/ degradation products, giving a clear indication of the extent of natural remediation and attenuation of contamination (or lack of it).
Finally, rather than refining the scheduling of analyses, the use of Microtox assessment may offer an alternative to the initial assessment of soils which may be contaminated. This method exposes organisms to test samples and measures the toxic effects. It is interesting to note that, so far, results for 'total' metals have had little practical significance, whereas 'available' metals have proved far more important for toxicity.
More precise scheduling of chemical testing is then possible for those samples shown to contain toxic substances.
Dr John Pitts
Director, GEM Services, Nottingham