Introduction
There are many possible reasons for conducting environmental monitoring. The base objective of monitoring exercises is to measure a particular substance or its environmental impact in a systematic way. Normally a single sample collected on one occasion will not provide adequate information, leading to the widespread adoption of more extensive monitoring. The outcomes of such programs may, for example, include the demonstration of regulatory compliance. Regulators themselves conduct long-term monitoring to satisfy the requirements of various Directives and other legislative drivers. Shorter-term monitoring can inform specific research and development objectives, or can help assess the effectiveness of newly installed technology for pollution abatement or management.
Whatever the reasons for monitoring might be, it is usually an expensive undertaking. Intensive sample collection can be automated in some circumstances, but simply collecting representative environmental samples may require several people to be engaged in collection over a prolonged period, absorbing many man-hours to achieve the required spatial and temporal coverage for the programme. Particularly with aquatic sampling, specialised sub-surface sampling devices may need to be hired or purchased and crewed boats may be the only viable way to gain access to essential locations. Laboratory analysis of collected samples can also be expensive, particularly if samples need extensive pre-treatment or if large or varied suites of contaminants are to be determined.
A poorly thought-through monitoring programme will often result in important information gaps only discovered at the end of the exercise, once the data are collected and processed. Despite the heavy expense incurred in implementing even a poor programme, repeating the monitoring process properly to satisfy planning or other legislative requirements will inflate the costs further.
In this article we focus on the decisions that need taking in drawing up an effective environmental monitoring programme and highlight some options available for the successful implementation of a designed programme. Although on-site effluent sampling may constitute part of such a programme, this discussion will also feature aspects of sampling the receiving aquatic environment.
Reasons for monitoring
The outcome of an environmental monitoring exercise is to gain information on the existing and possible future effects of harmful or hazardous pollutants on the environment or on the life (flora, fauna, human) existing within it. Several specific reasons may underpin the need to monitor:
- Monitoring may be needed to evaluate the need for legislative contaminant emission controls, or to demonstrate compliance with any such existing controls.
- Monitoring may be used to apportion contamination to specific emitted (or natural) sources.
- Monitoring may be used to assess human activity impacts on habitat or species health.
- Human impacts on environmental change can be identified, and information on baseline (historical) contaminant concentrations can be determined.
- Pollutant pathways can be identified, a key link in current widespread ‘source-pathway-receptor’ impact assessment approaches.
- Monitoring may provide essential background information for predictive modelling exercises. Modelling often forms part of impact assessment with regard to current emission effects and to potential improvements derived from upgrades to emission abatement measures.
Emergency remedial procedures can be initiated where monitored treatment processes lose effectiveness or fail, or when environmental conditions become dangerously contaminated.
Our particular focus here is on the aquatic environment. However, monitoring environmental waters may necessitate atmospheric, solid (e.g. suspended solids, sediments) or biological sampling to complete the information. A monitoring programme may need to cover point sources (e.g. specific identifiable outlets, chimneys, etc.) and/or regional or area diffuse contamination. Contaminants are considered to enter the environment either by planned discharge (normally consented), fugitive release (unplanned, often only partially abated), or accidental emission (often during temporary abatement failure). Clearly, a number of factors deserve consideration in implementing a monitoring programme for whichever specific function this systematic data gathering exercise is required.
Setting clear monitoring objectives
Careful consideration of the purpose of undertaking environmental monitoring must be the first stage of the process. Decisions taken here influence subsequent decisions on monitoring coverage, duration, sampling interval and the range of determinants measured (Figure 1). One or more of the objectives outlined above may need to be achieved by the proposed monitoring programme, but the organization necessary to meet different objectives should not necessarily overlap.
For example, effluent source monitoring to ensure discharge consent compliance need only consider effluent quality and quantity for a pre-determined range of contaminants, possibly with abatement failure monitoring alerts built-in. Whilst satisfying certain legislation (e.g. European Integrated Pollution Prevention and Control, IPPC, which considers emission limits at the specified boundaries of specific industrial operations) information required elsewhere might not necessarily be gathered (e.g. the combined influence of nitrogen and phosphorus present in, but not monitored under IPPC, may be of concern to the regulator with regard to the Water Framework Directive, WFD).
| A poorly thought-through monitoring programme will often result in important information gaps only discovered at the end of the exercise, once the data are collected and processed. |
The setting of sampling intervals and monitoring duration particularly can be incompatible if monitoring objectives are not clearly defined. For example, quarterly monitoring required to fulfill long-term Europe-wide marine nutrient studies (the ‘harmonised monitoring plan’) will not satisfy the intensive sampling over short periods needed to assess the impact of individual sewage releases on possible coastal eutrophication which covers tidally as well as seasonally varying water conditions.
When contaminant pathway studies or current or predicted impact assessments are monitoring objectives, multiple environmental media my need to be collected and analyzed to acquire the necessary information. Some contaminants, for example metals, can partition strongly from the dissolved phase in water onto suspended particles which may eventually become part of the underlying sediment. Partitioning is itself influenced by the acidity of the water environment and its suspended solids loading, amongst other factors. This in turn determines the differing environmental impacts of a dissolved metal-rich effluent on algae and fish (which tend to absorb metals from the dissolved phase) and on filter-feeding biota including shellfish. Hence, to assess the impact of such an effluent upon the receiving environment dissolved metal analysis in water samples alone will provide only part of the information sought and a wider suite of sample types including suspended solids, sediments and biological specimens, will need to be collected.
Assessing process and environmental variability
Clarity of monitoring objectives helps determine the range of sample types that should be collected in during monitoring. Similarly, the range of analytes (contaminants) to be measured in the collected samples can be established. This will normally include any prescribed harmful substances that could potentially be present in effluents, but might also include substances of wider concern in terms of regional diffuse pollution (e.g. ammonia, nitrite, nitrate) in areas prone to eutrophication. For industrial operators, the analytes that require measurement would normally be identified by the environmental regulator.
Effluent monitoring is normally straightforward. Operators will have good knowledge of the quantity and quality of effluents their processes produce, and monitoring decisions usually include determining appropriate sampling intervals and sampling locations on-site to collect representative effluent samples. More care is needed to evaluate fugitive releases to surface, to sub-surface drainage or to groundwater, although such sampling can be incorporated into plant design with forethought. Sampling intervals can be problematic for fugitive releases affected by storm-water, since the times during which releases are likely to occur are often when weather conditions are unpleasant or even dangerous to work in.
Another common monitoring objective for plant operators is impact assessment of their process to the wider environment. Recently, implementation of IPPC across the European Union has required operators to demonstrate their emissions are environmentally acceptable. Many operators employ the services of consultants to produce impact assessment reports together including pollutant measurements and modelling. Impact assessments may be required in future to fulfil IPPC improvement programs or where processes or plant is modified to predict the effects. It behoves operators to have an appreciation of whether a commissioned monitoring exercise will provide the requisite information.
When sampling environmental waters the selection of sampling interval and programme duration must allow for a thorough assessment of environmental temporal variability. Water residence times in flowing water (streams, rivers) may range from minutes to hours to days, whilst in standing water (lakes, reservoirs) much longer residence times may restrict the dispersal of discharges over periods of weeks. Thus sampling intervals in the former case will typically be far more frequent than the latter, whilst effective monitoring of standing freshwater will also often need be of sufficient duration to incorporate seasonal variations. In some cases standing freshwater will stratify, particularly in the summer, meaning multiple-depth sampling may be appropriate where simple surface water sampling of rivers will usually suffice. If the receiving water is tidal twice-daily changes of water flow may require upstream sampling in addition to downstream sampling, and many estuaries and coastal waters are also vertically stratified, stratification here being a function of water mixing energy, salinity and temperature. To further complicate assessments of effluent dispersion and fate in coastal waters, stratification here can be variable even within a single day.
In dynamic surface waters it is often necessary to collect large numbers of samples over an extensive spatial range to assess environmental impact. Where the affected surface water is uni-directional (non-tidal rivers and streams) spatial impact assessment is easier although flow variations can be rapid, so short-interval sampling over a restricted spatial range is more usual. In contrast, groundwater residence times are much longer and impact assessment can often be made with just a few samples collected from a limited number of sites. Where water quality is very dynamic, it may prove beneficial to monitor a pollution proxy such as sediment or indigenous or placed sedentary biological indicator species. Such proxies can be viewed as integrating ‘average’ contamination loads at a given location even under highly variable water quality conditions, although care must be taken with data interpretation, particularly with biota where uptake can be influenced strongly by species, age, sex, the specific organ analyzed, and season.
Increasingly, impact assessment studies involve some degree of predictive computer modeling. For models to be accurate a range of ancillary parameters might require measurement to evaluate dispersive and other physical processes. Thus, monitoring programs often include various water temperature, level, velocity and direction data, conductivity, pH, oxygenation and turbidity; these parameters also need vertical resolution in stratified waters. Robust, reliable continuous monitoring devices are available for purchase or hire, many of which include telemetry. Other, less obvious background information may also be essential to incorporate for accurate impact assessment. For example, for biologically active contaminants (nutrients, some metals) photosynthetic activity will sometimes need to be measured (e.g. chlorophyll), or the organic content of suspended solids, a key control for the partitioning of organic pollutants, might need to be determined.
Although some parameters can adequately be measured continuously using probes or similar devices, the detection limits or resolution of available devices may be inappropriate for at least some of the contaminants of interest. Here, sample collection and laboratory analysis is normally required. Obtaining a representative sample is essential. Volume-weighted composite water sample collection may be beneficial where water quality changes rapidly. In many cases simple discrete water sample collection is most appropriate. This can be done utilizing a number of pumped, depth-interval, messenger-activated vertical/horizontal water bottle, or other collection means.
Sample handling and data processing
Where samples are collected, due consideration of materials used and sample stabilization and storage must be given, particularly where contaminants are likely to be at low concentrations, redox reactive, or biochemically active.
In general where samples require metal analysis, polyethylene or Teflon sampling equipment and storage bottles should be used, preferably pre-cleaned by acid washing. Metal sampling equipment is best avoided in such cases. In contrast solvent-rinsed glassware is better for holding water samples for trace organic substances (e.g. pesticide, POP). Use of correct materials will minimise losses of dissolved constituents by adsorption to sample collection and bottle surfaces.
Where contaminants are likely to be strongly partitioned, rapid filtration of the sample (preferably upon collection) will help preserve sample integrity. Similarly, filtration will reduce biological modification of samples, and poisoning, acidification, refrigeration or freezing will further reduce this impact. Staff will need to be supplied with the appropriate equipment, chemicals and power supplies where such immediate post-collection stabilization is necessary.
Effective sample tracking from the point of collection, through sample storage to delivery to the laboratory must be robust. Unique sample identification forms a key part of attaining and maintaining quality assurance accreditation; many commercial laboratories now attain such accreditation from recognised quality assurance or regulatory bodies (e.g. the UK Environment Agency Monitoring Certification Scheme, MCERTS, system), and it is wise to investigate accreditation status for specific tests and procedures where analysis is to be outsourced. If in-house testing is to be used, effective tracking of samples (together with any derived sub-samples) using fit-for-purpose methods should be verified.
The monitoring data acquired must be collated and securely archived; accredited quality systems are normally adopted to minimise the chance of catastrophic data loss. Data must next be interpreted and transformed into information to answer the objectives set at the outset of the exercise. Particularly when continuous monitoring devices are employed, large quantities of data can be generated which may require reduction (e.g. running means or statistical representations such as ‘box-and whisker’ style plotting). Software is readily available to depict data with a number of pictorial and graphical representations, even linking data directly onto mapping tools to show spatial distributions. A full understanding of any use of interpolation between discrete sample data is advisable such as when using the contaminant ‘contour’ plotting on maps so commonly found in impact assessment reports.
Compromises in monitoring programme design
A carefully designed idealised monitoring programme is normally put together as a desk study. Costs are estimated and contingency measures built in. Practitioners quickly realise that many factors can prevent full implementation of the proposed plan. Firstly, projected costs can be prohibitively high, including wages, equipment purchase/hire and laboratory analysis. Reduced sampling frequency or sample coverage may prove financially necessary, but care must be taken that too much information will not be lost from fiscal compromises.
Environmental sampling can be hazardous, particularly under adverse weather conditions. Bank-side sampling during river-spate or sediment sampling on soft mudflats may prove too unsafe on occasion for samplers to complete collection. Forethought in site selection and sampling protocol will ameliorate this problem, but missing samples inevitably result in data gaps. The monitoring programme should be robust enough to cope with a degree of missing data.
Equipment power failure, loss or breakage can severely disrupt an entire sampling session. Where possible spare batteries, rope, tubing, etc. should be accessible in the field, and where possible low-technology back-up sampling systems should be considered to minimise the loss of sampling opportunities. These ‘common-sense’ contingencies are often be overlooked with desk study programme design, but can rescue an otherwise serious disruption to the monitoring programme.
Conclusions
In this article we have focused on some of the stages involved in drawing up an effective environmental monitoring programme. A clear exposition of the reasons for monitoring is of great importance. This informs decisions regarding appropriate sampling duration, intervals, location and methodology having due regard to process and environmental variability, both temporal and spatial.
After gathering samples and data with any necessary quality control and sample tracking in place, thought to the most effective data handling and presentation techniques to use will help greatly in transforming data into the information sought to meet the original objectives. Monitoring programme designs should be robust enough to survive partial incompleteness or lost samples where safety considerations or other practical compromises interpose.