This will be my last article for AWE, as I will be retiring at the end of March, so it seemed fitting to look back on the many changes in laboratory testing over the last 40 years.
Thoughts about this were triggered last year at a CIRIA conference, where one of the speakers stated that not very much had changed in site investigation – it still came down to a bucket and trowel for soil sampling. However, when I started to think about all the changes in the labs, it was really very different.
Size of instruments
First of all, size – the common saying is that size doesn’t matter, but it does… When I joined Geochem in 1980 the mass spectrometer (there was only one) took up the whole of a (very) large room. It was also water cooled, using gallons of water every day, and was temperamental, so only two highly trained analysts could operate it. The situation is very different today, with over 20 benchtop mass spectrometers in the current laboratories at DETS, with the same (or even better) resolving power on a much smaller platform. It is much easier to train staff, and most of our analysts are multi-tasking, so can switch easily from instrument to instrument.
Automation
Probably the biggest series of changes came with increasing automation, as the IT revolution gathered momentum from 1990 onwards. The instrument companies began to add auto samplers and this gave twofold benefits – once samples were loaded, the analysts could work on other tasks such as processing and interpreting data, and secondly, the instruments could run overnight, thus greatly increasing the productivity of each instrument. Once laboratories started to use LIMS (Laboratory Information Management Systems) to hold all the data, then software facilitated results from instruments to be downloaded automatically, allowing much faster transfer of data and reducing the risk of transcription errors by manual inputting. This has spread to include balances and simple meters, so human inputting is now reduced to a minimum. It even spread to Admin tasks – I think it was around 1993 when I was given a desktop PC, and from then on, I had to type my own letters and reports – luckily, I had learnt to touch type at evening classes in my late teens – my mother thought it would be useful, especially if I couldn’t get a job as a scientist, then I could always be a secretary…
“working in any lab can never be 100% risk free, but it is the responsibility of any manager to ensure risks are controlled and reduced wherever possible”
Health and safety
Health and safety has proceeded in leaps and bounds – I can remember at Geochem, we would often clean oily residues off the benches with dichloromethane, happily inhaling the vapours. Nowadays DCM is only handled inside a fume cupboard. And the more mature readers may even remember pipetting by mouth – not a good experience, if you misjudged it. Laboratories now have dedicated HS managers, everyone is given extensive training, all requisite PPE is issued to staff and checks are performed regularly to ensure they are wearing it, most laboratory areas are monitored for dust/asbestos, safety notices are everywhere, and there is a much higher awareness of responsibility for personal safety and also checking for the welfare of your colleagues. This can sometimes go a little too far – I remember one young technician stating that she did not want to work in the cyanide lab due to the possible risks of cyanide exposure, but it was pointed out that she couldn’t pick and choose lab areas, and also that if soils did contain cyanide, then the risk of exposure was the same in any of the lab areas. However, we did review the safety procedures to ensure all samples are treated as potentially toxic, and that adequate precautions were in place, with additional fume cupboards to ensure minimal possibilities of exposure. Working in any lab can never be 100% risk free, but it is the responsibility of any manager to ensure risks are controlled and reduced wherever possible.
Quality
Quality systems have also metamorphosed over the years – again, more mature readers will probably remember NAMAS, the forerunner of UKAS. This name change brought with it a more modern approach and auditors from UKAS would visit the labs regularly (at least annually) and audit a selection of methods in great detail, against the standard ISO 17025. Increasing IT changes also made an impact here – the saying used to be “install a quality system and kill a forest” as it generated so much paper, but the benefits of working electronically soon became apparent – even the auditors now have to type their reports whilst still on site. One significant change was the recognition of ‘deviating samples’ – many auditors had documented the problems with samples not being fit for purpose (even if the analysis was good), and the emphasis was placed on the importance of correct procedures for sampling. Although this is usually not performed by the lab staff, UKAS made it the responsibility of the labs to flag up samples which are taken incorrectly and report these back to the clients. If the lab was instructed to continue with the testing, then a comment had to be inserted into reports, stating that these samples were deviating, and the integrity of the data may be compromised. It took several years to educate all the clients to these requirements, and I have spent many, many hours providing technical support and giving seminars to clients explaining the reasons for this and how they should sample correctly. In fact, we still receive a percentage of samples which are incorrectly sampled, so there’s still work to do here.
MCERTS
In the late 1990s, the Environment Agency brought out the MCERTS standards (Monitoring Certification Scheme). One criticism of the labs was that validation of methods was only performed on prepared ‘clean’
standards, and that this was not reflective of the often difficult matrices encountered in the type of environmental samples received by the laboratories. For labs to achieve MCERTS certification, they had to validate their methods on a range of matrices to prove they were robust. For soils, these are commonly clay, sandy soil and loamy soil, but for waters it could be far more extensive – landfill leachate samples, saline waters, industrial effluents, etc. This placed a heavy burden on the labs in terms of instrument and analyst time, plus the cost of different standards in different matrices, but it improved the certainty of the data and MCERTS also stated that Limits of Detection should be calculated based on bias and precision of the validation data, which also had an impact on the LoDs quoted by the labs in their reports.
Legislation
This brings us to the next major factor for change: the ever increasing number of EU Directives, combined with the ever decreasing EQS levels (Environmental Quality Standards). In 1986, we only had the ICRCL tables for soil (Interdepartmental Committee for the Redevelopment of Contaminated Land). This was one of the great benefits of being in Europe – successive UK governments had to comply with Waste, Drinking Water, Surface Water, Air Quality, and Bathing Water Directives, to name just a few, and compliance with these has greatly improved our environment. In addition to these, the EA produced a small number of SGV’s (Soil Guideline Values), and when they did not publish any more, other organisations produced GAC’s (Generic Assessment Criteria) for a much wider range of compounds. We are now working with C4SLs (Category 4 Screening Levels). For the labs, these all meant improving the detection limits of methods to meet these criteria, and instrument manufacturers were also involved in producing models with increased sensitivity and/or high volume injectors.
“in addition to the Directives, there is also a list of Priority Pollutants, which is added to on a regular basis, this has increased now to forty eight listed substances”
In addition to the Directives, there is also a list of Priority Pollutants, which is added to on a regular basis, and this has increased now to forty eight listed substances. There is a Watch Committee, which regularly reviews monitoring data produced by most EU countries and if contaminants are deemed to be widespread and toxic enough to provide a risk, then these are put forward to be added to the list. Laboratories must then constantly add to their range of analytes and provide testing for these new contaminants, although the target levels are often very challenging, to say the least.
EAGLE
In the mid 1990s, the government set up EAGLE (Environmental Analysis Group for Laboratory Excellence) which was one of a series of focus groups to review the quality of data – MERLIN was for medical labs and FALCON for food testing (collectively known as the Birdie groups). The objective of the groups was to review methods and proficiency testing data and then to improve on the consistency of data by harmonisation of methods. This was achieved by EAGLE through the Standing Committee of Analysts (SCA), who formed several working groups to publish the series of Blue Books, also known as MEWAM (Methods for the Examination of Waters and Associated Materials). These methods are widely used by laboratories in the UK, and there are over 200 of them, although the majority are for testing water. The soils working group was set up in 2001, and has slowly produced about a dozen methods now, although it can take many years. A subgroup for asbestos quantification in soils was set up in 2010 and only succeeded in publishing the method on the SCA website in 2018 (although it still has ‘Draft’ written on the first page). This is not as bad as the method for total petroleum hydrocarbons, which has been going on for over 15 years, and is still not completed…
Costs
Despite all this improved instrumentation, automation and extended range of methods, there is unfortunately a downside for the labs (but not for the clients), and this is the very much reduced cost of testing. Back in 1990, an ICRCL suite consisting of pH, moisture, a suite of 10 metals, chloride, sulphate, cyanide, phenol, TPH, PAHs and asbestos would cost about £150 per sample. Current day prices can be as low as £30 – £35 in a very competitive market, and clients also expect a much more rapid turnaround – five days is now standard (it used to be 15) and three days or less is often requested. Laboratories are now run on a production based model, with ever faster throughput, which is fine when the samples are straightforward matrices and nothing breaks down, but there is no such thing as a standard soil. Analysis often does not always lend itself to cheap and cheerful, and like many things in life, you get what you pay for.
Summary
Overall, the quality, range and detection limits of the analyses are all very much better than forty years ago, and the training, opportunities and health and safety of the staff have also much improved.
My own career has involved training both staff and clients, serving on committees, writing articles, speaking at conferences and providing technical support and advice. I have worked on some demanding projects, including assisting laboratories in Bulgaria and Romania to monitor the Black Sea (driving in Romania in 1999 was an experience in itself), integrating laboratory acquisitions into Alcontrol, project managing the relocation of the labs from Chester to Hawarden (that took a year of my life), helping to set up a lab in Dublin, and giving seminars for the UN to environmental scientists from Iraq (in Jordan – I didn’t have to go to Iraq).
It has been interesting and worthwhile to work in an area of analysis which has constantly raised the bar and worked to improve the quality of data and the training of staff. It is also a privilege to work in an industry which is improving the environment and protecting people and wildlife. I have met some extremely interesting and knowledgeable people, and I don’t think there has ever been a dull day – I would recommend a career in environmental science to anyone.
My thanks to the staff at AWE for publishing all my articles and being very understanding over deadlines.