Derivation of Aquatic Quality Standards for Priority List Substances not Covered by Existing UK Regulations
WQD0005
July 2003

Executive summary

A key element within the Water Framework Directive concerns the derivation of Environmental Quality Standards (EQSs) as a means of setting benchmarks for the quality of surface waters, sediments or biota. Article 2 of the Directive defines an EQS as the concentration of a particular pollutant or group of pollutants in water, sediment or biota that should not be exceeded in order to protect human health and the environment. Annex 5 then goes on to describe the procedure for deriving these standards. This report is concerned with the procedure described in Annex 5 of the Water Framework Directive, and in particular how the outcomes from using that procedure (i.e. the resulting EQSs) compare with current UK practice.

The report summarises the findings of the project to derive quality standards for seven substances/ groups of substances using the UK method and that proposed by the Fraunhofer Institute for deriving Quality Standards under the Water Framework Directive and on the assessment of the cost implications of the standards proposed (Parts 1,2 and 3 of the work programme).

Specifically, the following chemicals were considered:

• Anthracene
• Brominated diphenylethers (penta, octa and decabrominated homologues)
• C10-C13 chloroalkanes
• Chlorpyrifos
• Dichloromethane
• Polycyclic aromatic hydrocarbons (‘PAHs’), specifically: benzo-a-pyrene, benzo-b-fluoranthene, benzo-g,h,i-perylene, benzo-k-fluoranthene and indeno(1,2,3-cd)pyrene
• Fluoranthene

The UK method only derives standards for freshwater and marine waters, while the proposed EC methodology also includes thresholds for the protection of freshwater and marine sediments, secondary poisoning to higher life forms, dietary exposure in humans and abstraction to drinking water.

Based on the information available it has been possible to derive quality standards for the protection of fresh and marine water using the UK methodology for seven chemicals: anthracene, pentabromodiphenyl ether, C10-13 chloroalkanes, chlorpyrifos, dichloromethane, benzo-a-pyrene and fluoranthene. Using the proposed EC methodology, it has also been possible to derive ‘overall’ standards for the same substances and also benzo-k-fluoranthene. Insufficient data are available to derive quality standards using either the UK or proposed EC methodology for six chemicals; octabromodiphenyl ether, decabromodiphenyl ether, benzo-b-fluoranthene, benzo-g,h,i-perylene, benzo-k-fluoranthene (only for the UK methodology) and indeno-1,2,3-cd-pyrene.

In carrying out a critique of the EC and UK methodologies, some important points have arisen. In principle, the EC method (that proposed by the Fraunhofer Institute) has much in common with the current UK approach, being based on a ‘critical data/safety factor’ paradigm. However, there are important differences in detail. In the EC method, reliance is placed on the effects assessment part of the EU Technical Guidance Document (EU TGD) and similar schemes used to evaluate plant protection products. This reflects an understandable desire to promote consistency between Member States and with existing regulatory schemes. However, a number of issues have come to light:

1. The EU TGD was designed principally with small datasets in mind. It consequently places emphasis on the use of default safety factors to extrapolate from the available data to the standard. These factors can be large, especially when only small datasets are available. For many of the substances for which quality standards are required, a substantial quantity of data will be available where greater flexibility and smaller safety factors may be justified. Although some provision for peer review and flexibility in the choice of safety factors is available under the EC approach, it is given less emphasis than in the UK approach. Formal peer review of critical data and safety factors is to be encouraged and should be transparent and auditable.
2. Our investigation, based on standards for the protection of aquatic life, shows that standards using the EC approach tend to be more stringent (i.e. lower concentrations) than those derived by the UK approach. This bears out an earlier study using a different range of substances. This could lead to increased costs to industry with uncertain environmental benefits. However, this outcome does not incorporate the effects of any peer review of standards by Expert Groups at EC level, which might lead to closer agreement.
3. The EC approach has been drawn from a tiered risk assessment approach in which an adverse outcome would trigger further investigation to refine the understanding of the level of risk. This is simply not possible when a quality standard is to be used as a formal compliance measure. This is a consequence of adopting an effects assessment methodology from tiered risk assessment schemes, but not the context in which it was placed. In a tiered approach, a high degree of conservatism is entirely appropriate at the first stage but we argue that such ‘structural’ precaution may not be appropriate when defining a statutory standard.
4. Whilst the UK approach deals only with the protection of aquatic life, the EC approach also covers risks to biota from food chain transfer, drinking water abstraction and risks to human health from consuming fish/shellfish. Significantly, the most stringent of these is taken as the ‘overall’ standard that will apply to a watercourse. This means that a degree of over-protection for all but the most sensitive receptor is inevitable.
5. Standards aimed at protecting ‘food uptake by man’ and ‘protection of aquatic’ life are particularly influential in determining the ‘overall’ quality standard. Basing ‘overall’ standards on the standard for ‘food uptake by man’ is of concern because the methods used to derive this standard in particular are rather crude and are not yet validated. Such an approach is acceptable as a first step in a tiered risk assessment, but not as a means of determining a quality standard for a waterbody. In our view, the protection of human health from food taken from the aquatic food chain should not be adopted until the methodology has been further developed and validated. Alternatively, a tiered assessment should be adopted in which the human food chain effects assessment is used not as a statutory quality standard (and certainly not the ‘overall’ standard) but rather, as a trigger for a more rigorous assessment. The same concerns apply to the methodology for assessing risks to biota from secondary poisoning.

The proposal to derive sediment quality standards and to allow compliance assessment on the basis of contaminant levels associated with suspended particulate matter (SPM) gives rise to some particular concerns:

1. The understanding of the science of sediment chemistry and toxicology is not yet at a level where robust sediment quality standards can be set. This is exacerbated by a lack of sediment toxicity data for many substances and this would mean that reliance on equilibrium partitioning approaches would, in our view, be the norm.
2. Reviews of this subject have consistently advocated the use of sediment quality thresholds as guidelines within a tiered risk assessment process rather than statutory compliance thresholds. We understand this is also the view of several regulatory agencies throughout the world.
3. Sediment quality standards could be expressed either as the concentration of a substance in bed sediment or that associated with suspended particulate matter (SPM) in the overlying water. The latter option is fundamentally flawed from a scientific point of view and has practical difficulties:

• The assumption that recently deposited superficial sediment deposits are of most ecological importance is without foundation. Deeper bed sediments can be equally important to benthos.
• It is illogical to monitor compliance in an environmental compartment (suspended sediment) that does not support organisms we are seeking to protect. The aim is to protect benthic organisms and so compliance should be based on bed sediments.
• Decisions based on contaminant levels in SPM are subject to much uncertainty. This is because sedimentation of suspended particles is unpredictable so the assumption that contaminant levels in SPM will determine those in bed sediments cannot be sustained. Furthermore, the chemical and physical environment in the suspended phase may differ greatly from that in bed sediments with important consequences for chemical speciation and bioavailability.
• Costs of monitoring SPM would impose an additional cost with dubious benefit. Where low levels of SPM are present, it will be impractical to collect sufficient volumes of water. The option to monitor compliance either on the basis of bed or suspended sediment would simply lead to confusion and an inability to compare results.

In our view, a more effective strategy would be to develop non-statutory sediment guidelines within a tiered risk assessment framework.

Finally, we draw attention to concerns raised by the Royal Commission on Environmental Pollution about the expression of standards as an absolute threshold (such as a Maximum Allowable Concentration, MAC) without proper consideration of the sampling protocol and measures to assess the level of confidence in determining compliance or non-compliance. The UK government have accepted these principles and so we recommend that further consideration of the compliance assessment measures is needed if a MAC is to be derived.

With respect to the cost assessment this can be summarised by substance as follows:

Benzo-b-fluoranthene, benzo-g,h,i-perylene, indeno(1,2,3-cd)pyrene

No standards are proposed and therefore no cost assessment has been undertaken.

Decabromodiphenyl ether

No standards are proposed and therefore no cost assessment has been undertaken. However, available data on solubility and effects suggest no additional measures will be required and hence no cost implications are expected.

Octabromodiphenyl ether

No standards are proposed. However, the uses of the substance that contribute to emissions to the water environment are to be banned under Directive 76/769/EEC. Therefore no additional measures are expected to be required and no cost implications are expected.

Pentabromodiphenyl ether

Standards have been proposed. However, the uses of the substance that contribute to emissions to the water environment are to be banned under Directive 76/769/EEC. Therefore no additional measures are expected to be required and no cost implications are expected.

C10-C13 chloroalkanes

Standards have been proposed. However, the uses of the substances that contribute most of the environmental emissions, are expected to be banned. Therefore, no additional measures are expected to be required and no cost implications are expected.

Chlorpyrifos

Standards have been proposed. Exceedances could be considerable but monitoring data are incomplete. Remediation measures are likely to focus on changes to farm practices, with costs expected to be moderate overall, although potentially significant for individual farmers. The cost estimates are insensitive to the methodology used for the derivation of the EQSs.

Dichloromethane

Standards have been proposed. Monitoring data are patchy, but point towards limited exceedances in well-defined locations, possibly linked in certain serious cases with historic pollution, and hence outside of the scope of Article 16. This suggests that costs of compliance measures are likely to be low overall, but could potentially be significant in some local cases. EQSs derived using the UK methodology are much less stringent than the ones derived using the EC methodology, although based on the data the implications for exceedance appear limited.

Polycyclic aromatic hydrocarbons

The assessment is limited by a lack of monitoring data, and refers only to a comparison with the situation observed over the period 1993-1995. Since then we would have expected possibly significant improvements in water quality status.

• Exceedances of EQSs for fluoranthene in coastal and estuarine waters are limited, and tend to occur in heavily industrialised and/or urbanised locations. No substantial change in this pattern is observed when switching between methodologies for deriving EQSs.
• Exceedances of EQS for anthracene appear to be relatively rare in coastal and estuarine waters. No significant change in this pattern results from different EQS methodologies.
• Exceedances of EC EQSs for benzo-k-fluoranthene cannot be assessed due to a lack of suitable monitoring data. Therefore no cost assessment is possible. No EQS was derived using the UK methodology and a comparison could therefore not being made.
• Exceedances of the EQSs for benzo-a-pyrene are widespread in comparison with the more stringent MAC EQS derived using the EC methodology, with order of magnitude failures in some locations. This result is sensitive to the choice of EQS methodology, with non-compliance falling when using the MAC derived by the UK method.

The cost implications for meeting the PAH EQSs could be considerable. If the strictest EQSs are adopted, these could amount to billions of pounds. Adoption of less stringent standards is likely to significantly reduce non-compliance, but remaining exceedances could still require considerable expenditure, possibly hundreds of millions of pounds, or more. Cost implications can only be described as high, therefore. However, this is subject to information on the impacts of recent and planned investments in wastewater treatment by industry and water companies, as well as expenditure to address combined sewer overflows, which has not been available in structured form.

Copies of this report may be available as an Acrobat pdf download under the 'Completed Research' heading of the Research Page on the DWI website.