"Risk assessment is like a political prisoner: if you torture it long enough, it will give you any answer you want" - reputedly uttered by a United States Environmental Protection Agency administrator. Paul Board, Robertson Laboratories, doesn't agree.
But what remains unchanged, however robust and rigorous a risk assessment model may be, is this: if the analytical data is wrong, or just inappropriate, no amount of "rocket science" wizardry will fix it. As the old saying goes, "Garbage In, Garbage Out".
Whereas laboratory users can take comfort in UKAS accreditation and subscription to Proficiency Testing (PT) schemes (such as CONTEST and Aquacheck for soils and waters analysis respectively), nowadays, this is not enough. Formal accreditation of a method does not make it "fit for purpose" for all scenarios, particularly with the new risk assessment packages. Laboratories have risen to the challenge, and now provide a number of analysis packages specifically geared to feed appropriate and meaningful data into these new models.
Nowhere has a more intelligent approach to analysis been required than in the area of organic analysis. Carbon's ability to join to four other atoms, including itself, has enabled both nature and man's chemical ingenuity to fabricate thousands upon thousands of organic compounds. The following quotation was made by a French chemist in 1854:
"When we consider the great number of organic substances that have been discovered during the last dozen years, and the increasing rapidity with which chemistry daily discovers fresh ones; when we see that, from a simple hydrocarbon and chlorine, we may produce a hundred compounds, and that from them we may obtain a great number of others; lastly, when we reflect upon the absence of all system, all nomenclature, for the classification and denomination of this multitude of bodies, we demand with some anxiety, whether in a few years' time, it will be possible to direct ourselves on the labyrinth of organic chemistry."
At least now we have a logical (if less romantic) system for naming chemicals, but the list has grown. This has certainly led to the betterment of our lot, but has also placed with analytical chemistry a tall order: to elucidate and quantify the downside effects of unwanted pollution. The measurement of the now ubiquitous and long-lasting residues of the anthropogenic PCBs (Polychlorinated Biphenyls) or the sourcing and remediation of petroleum hydrocarbon spills both require sophisticated analysis. We will look at the latter, by way of example.
Total petroleum hydrocarbons, or "TPH" for short. A misnomer if ever there was one. For as yet, no single analytical procedure will provide a result for Total Petroleum Hydrocarbons, however persuasive the salesperson or sales literature may be.
Why? Petroleum hydrocarbons span a broad range of compounds from the very volatile (such as benzene, C6H6) to the very non-volatile or "sticky" end of the petroleum hydrocarbon series (C30 and beyond). To encompass this spectrum of compounds with very different physical properties requires ingenuity and an arsenal of modern analytical equipment.
Petrol/gasoline range organics (PRO/GRO) analysis, covering the range C6-C10, can be carried out using Headspace GC-FID (Gas Chromatography-Flame Ionisation Detection) or GCMS (Gas Chromatography Mass Spectrometry) and can be used to quantify the lighter components of petroleum hydrocarbons. Petrol or specific compounds that comprise this lighter boiling range, such as the BTEX compounds (Benzene, Toluene, Ethyl Benzene and Xylenes) or additives such as the hydrophilic MTBE (Methyl Tertiary Butyl Ether), are amenable to this procedure.
Diesel Range Organics (DRO - C9-C34) are extracted with a solvent (e.g. dichloromethane, or DCM) and the extract is analysed by GC-FID. This covers a number of different types of petroleum product, including white spirit (C9-C12), kerosene (C9-C16), diesel (C9-C26), transformer oil (C13-C30), and lube/motor oil (C20-C30+).
Providing chromatograms of nice fresh standards is all very well, but real life is just a little bit different. Evaporation of more volatile components and biodegradation can paint quite a different picture. For example, even if the contaminant in your sample is diesel, the chromatogram may not always look exactly the same as a sample taken from your local pump, depending on the degree of degradation.
Some indication of biodegradation can be acquired by measuring the abundance of phytane and pristane - important biomarkers derived from the degradation of the chlorophyll molecule in the original plant material from which the crude oil was formed (many moons ago), and consequently from which the diesel was refined.
The above techniques will estimate the amount of pollution from petroleum hydrocarbons, normally by calibrating against the closest match of standard (e.g. if the sample chromatogram looks like diesel, a set of various diesel standards will be run to quantify against), but this approach has its limitations. Toxicological data is normally most readily available for single compounds rather than mixtures (as would make up diesel). For this reason, modern risk assessment models often require TPH analysis to be split into groups depending on the carbon chain length.
Petroleum hydrocarbon analysis is only one example of the various ways there are of analysing for what can be perceived by the uninitiated as a single entity (and indeed very often appearing as such on a Bill of Quantities). Get the advice of your analytical laboratory prior to commissioning a project (however large or small) - talking to analytical chemists will help to ensure that the maximum amount of meaningful analytical data will be squeezed out of a finite budget.