Ascertaining the level of remediation required without compromising data
The quality of soils is paramount, primarily for human health risk reasons, but also for the potential growth of plants. Prior to development or redevelopment of a site, a developer and their consultant must consider the potential of contamination of the soil from previous usage of the site.
Which industries were present here? What buildings existed? Were chemicals or fuel stored here? Was there leakage into the groundwater or runoff into local watercourses? What is the hydrogeology? How much contamination is still present, is it likely to pose a risk and how can it be remediated? These are some of the questions which must be answered by constructing a Conceptual Site Model (CSM) and possibly conducting an intrusive survey using trial pits and boreholes, followed by analysis of both soils and groundwater. The consultant will then assess the risk of harm using all the available data. The future use of the site will also determine the level of remediation required.
For test results to be meaningful, the samples must be taken following well documented methodologies. A recommended document is BS 10175 – 2011: Investigation of potentially contaminated sites – Code of Practice. The Association of Geotechnical Specialists (AGS) also produces useful guidelines for site sampling.
Suitable equipment and containers should be available for taking and storing samples. The following are typical examples of suitable containers:
Certain types of sample, such as ballast or aggregate, may require more than this, possibly as much as 10 kg.
Samples should be stored under appropriate conditions throughout the whole process, including storage prior to and during transportation to the laboratory, and while in the laboratory. A cool environment, with a typical storage temperature of 5 +/- 3°C is recommended, and samples should be despatched to the laboratory as soon as practicable.
If samples are not sampled and stored correctly, they may be reported as ‘deviating samples’.
It is a requirement of UKAS and the European Accreditation service for a laboratory to report samples as deviating if they do not conform in some way. Examples include:
• Plastic containers used for organic parameters • Headspace in volatile jars or vials • No preservatives used when required • No sampling date provided • Sample exceeds holding time • Sample not stored in controlled temperature • Sample matrix unsuitable for test method • Sample volume not adequate for test method
Any of the above may compromise the integrity of the data, depending upon the analyte in question, and will be listed in the laboratory report. This is now a crucial area audited by UKAS, aimed at improving the quality of samples submitted to laboratories.
Some parameters can be tested on site, for example, hydrocarbons in soil vapour can be analysed using a hand held Potentiometric Ionisation Detector (PID). Groundwater can be tested for a range of parameters such as pH, EC, oxygen content and ammoniacal nitrogen, using a range of meters and probes. This real time data may be very useful in delineating the extent of contamination or monitoring remediation, but it is usually not accredited, as there are too many variables on site. It is important that all meters and probes are calibrated prior to use, so that the data is as reliable as possible.
Consultants may notice differences between site data and laboratory data, and this may not always be due to instrumental differences. The pH of borehole water, for example, will become more acidic upon exposure to air due to carbon dioxide dissolving in the water, thus lowering the pH by the time the sample is tested in the laboratory.
The laboratory results must be reliable, accurate, fit for purpose, cost effective, and provided within an agreed timeframe. So how can the buyer of analytical services ensure the end data meet all these criteria? The most important factor is to ensure the laboratory is accredited. This is usually provided by UKAS – the United Kingdom Accreditation Service.
Laboratories apply to UKAS for accreditation, and UKAS will then audit the laboratory to ensure stated methods and the infrastructure meet the requirements of the requisite standard. For testing and calibration laboratories, this is usually ISO 17025, and then there may be additional and more specific standards relating to certain industries. For example, UKAS will also audit against the Environment Agency’s Monitoring Certification Scheme (MCERTS), which provides the standard for the analysis of air, water, soil, and equipment used in environmental monitoring. MCERTS sits over and above ISO 17025 and has more stringent requirements for method validation on at least three different matrices. For soils, these are usually sand, clay and loamy soil.
ISO 17025 is a comprehensive standard covering all aspects of laboratory testing from contract review, training and monitoring of staff, equipment maintenance, laboratory conditions, controlled methods, analytical quality control, use of reference standards, recording of data, and final reports. An important aspect of 17025 is proof of continuous improvement – laboratories are expected to continuously improve methods by the use of new technology or more stringent quality control. Innovation plays a major part in this, but any new changes must be robust and validated to satisfy UKAS requirements.
Soils present several difficulties in analysis due to their variability and complexity. Chemically they can consist of a single constituent such as silica in sand or calcium carbonate in limestone, or much more complex clay minerals or loamy soils containing vegetative matter and animal residues. On top of this they may contain a wide range of contaminants – everything from building materials, hydrocarbons, toxic industrial chemicals, waste, explosives, plastics, and paper, to the often overlooked asbestos.
From an analytical perspective, the first step is to try and obtain a representative subsample. This can be done by a variety of methods, including stirring, cone and quartering, shredding, kneading or other mixing protocols, and using appropriate equipment, depending on the sample.
It is important for some tests to be performed on the as-received soil, with no drying or crushing, as these processes would cause changes to some parameters, such as volatile organic compounds, ammoniacal nitrogen, cyanide, phenols and hexavalent chromium. For other more stable compounds such as metals, sulphate and chloride, the soil samples can be dried and crushed to provide a much more homogenous subsample.
This describes the routine monitoring of every analytical method to ensure validity and consistency of data. With every batch of samples (a batch can be anything between two and 20 samples) an analyst will always run, as a minimum, a reagent blank and an analytical quality control (AQC) standard. Blanks may give slight positive values, which can be subtracted from the sample results. An AQC standard has a known value, and the result achieved is plotted onto an AQC chart to ensure it falls within calculated acceptance values (Figure 1).
These values show both warning and action values, and if even one AQC result is outside the action values then the whole batch is discarded and has to be repeated. UKAS will expect to see an investigation into any failed AQC standard when they audit the laboratory. Laboratories should willingly provide details of monthly AQC performance to their clients upon request.
In a well-run laboratory between 25% and 30% of all samples analysed each day will be part of the quality system, which includes AQC standards, calibration standards, blanks, instrument system checks, repeats, PT samples and validation samples.
Another requirement before accreditation can be awarded for a particular method is for the laboratory to demonstrate a proven track record for proficiency testing (PT) samples. PT schemes allow a laboratory to monitor the robustness of its methods and the competency of its staff using blind samples. The laboratory pays to join a PT scheme, blind samples are then sent out every month or bimonthly, and the laboratory sends back the results obtained. The PT organising body then sends a report to every participating laboratory to provide an indicator, known as a z score, of how well the laboratory has performed (lab IDs are kept confidential). If the z score is <2 it’s good, between 2–3 is questionable, and >3 is a fail.
Again, UKAS will expect to see an investigation into any failed PT score. Any laboratory should be willing to provide copies of recent PT data to its clients, upon request, and one aspect to check for is that the laboratory participates in a wide range of PT testing, not just for a few analytes.
Organisations running PT schemes include the Lab of the Government Chemist (LGC), which runs both Aquacheck (waters) and Contest (soils), and also the RT Corporation (now owned by Sigma). Both provide a range of analytes in different matrices to assist laboratories in developing, validating and monitoring their methods. They also provide Certified Reference Materials (CRMs), which are used both in quality control and validation.
In 2013, the Health & Safety Laboratory (HSL) commenced a PT scheme for asbestos in soil, as this was needed by the industry and was not available from any other supplier.
If a new instrument or a change to a method is considered to be an improvement, then the laboratory must validate the method to prove its robustness before it can be offered for accreditation. This validation involves testing numerous standards in different matrices (at least three – sand, clay and loamy soil) at high and low concentrations over several days or weeks. The data sets produced must meet certain criteria for precision and bias (defined in the MCERTS standard) before accreditation can be awarded.
These targets will vary depending on the analyte in question. In the environmental sector most inorganic tests work to limits of +/- 10% or less, whereas some organic parameters can be acceptable with up to +/- 30% variation, particularly at ppb concentrations. Certified Reference Materials (CRMs) are used extensively in validation testing as a method must be able to prove itself on known standards.
Accredited laboratories are highly regulated and are required to invest in significant levels of staff, time and effort to maintain a robust quality system. This may not always be appreciated by end users of the data, who are often constrained by costs and meeting timescales – it is crucial to factor in adequate provision for both these aspects of laboratory testing. With ever increasing regulation from European standards and directives, the emphasis on lower limits of detection and additional quality checks is likely to escalate. It may be tempting to use a cheaper laboratory with limited or no accreditation but this may possibly lead to indefensible data, dissatisfied clients/regulators and possible litigation. The quality system should be the bedrock of any laboratory.
Published: 21st Aug 2014 in AWE International