Archive:Chemicals - monitoring REACH with indicators

Data extracted in 2012

Article updated: March 2015

This Statistics Explained article has been archived - for recent articles on Environment see here.

Since June 2007, REACH, the new European Union (EU) Regulation on the Registration, Evaluation, Authorisation and restriction of CHemicals', is in force.

Eurostat, in collaboration with the services responsible for environment and for industry of the European Commission, has developed a baseline study and a set indicators to monitor the implementation of REACH.

The baseline study can measure progress towards two major objectives of REACH: to ensure a high level of protection of human health and the environment as well as enhancing innovation in the development of safer chemicals.

In 2009, the first assessment was made. It presented a baseline of the (nominal-) 'risk' caused by chemicals and the 'quality' of the underlying data which was available when REACH came into force in June 2007. In 2012 a second 'risk and quality snapshot' was taken.

This article provides an overview of the methodology and the assessment of the main changes from the first study in 2009 to its update in 2012.

Full article

Background

Work carried out by Eurostat in the field of 'Chemicals management statistics' dates back to the period from the mid-1990s to 2000 when some Environmental Pressure Indicators (EPI) related to chemicals were developed. Recent priorities of Eurostat have been the development of indicators on toxic- and environmentally harmful chemicals. They measure progress towards the headline objective for ‘public health’ established in the EU - Sustainable Development Strategy and towards the long term vision of a non-toxic environment as set out in the 7^th Environmental Action Programme for the European Union.

REACH introduces a new era of chemicals policy in Europe. It will increase our knowledge of the hazardous properties of chemicals. It is expected to enhance the communication and implementation of conditions of safe use in supply chains and the substitution of dangerous substances by less dangerous ones. Through different types of measures, REACH is expected to cause a decrease in risks to human health and the environment associated with the use of chemicals (see Figure 1).

Figure 1: Possible evolution of the risk caused by chemicals

Did activities related to REACH already reduce the risk caused by chemicals? How did the quality of the underlying data evolve? This article, based on the two comprehensive assessment reports aims to make the methodology better known and to trigger the discussion on the results.

Policy monitoring by indicators

It is difficult to foresee which reductions in the real risks caused by chemicals will be a result of REACH; different risk evolutions are also possible. The European Commission has therefore developed the REACH Baseline Study as an instrument to monitor which evolutions really occur. Has our limited knowledge about the properties of substances and their safe uses increased due to REACH? The REACH Baseline Study aims to provide an indicator set to monitor whether such changes take place or not — indicating the success of REACH. In doing this, the focus of the Baseline Study is not restricted to indications for the risk itself. The indicator system also enables changes in the quality of the public data on substances and their safe use to be detected. A first ‘snapshot’ has been taken in 2007 for a picture of the situation before REACH (“Pre-REACH-Baseline”). A second snapshot has been made in 2012. Further assessments can follow. Comparing the pictures allows for monitoring and assessing the success of the REACH legislation.

The 'risk and quality' indicator system

The central element of the REACH Baseline Study is the risk and quality indicator system. It tracks two major goals of REACH:

reduction in the risks of chemicals for humans and the environment as well as
improvement in the quality of publicly available data.

Risks to three impact areas (see Figure 2) are addressed: the environment, workers and the general population. Impacts on the general population are divided into direct impact on consumers (resulting from the use of chemicals e.g. paints or glues) and impact on humans via the environment (e.g. drinking water).

Figure 2: The impact areas addressed by the REACH baseline study

This indicator system directly assesses the nominal risk caused by exposure to chemicals and characterises the quality of the data on which this risk assessment is based. These characterisations can be followed over time.

Since the calculation of the risk and the assessment of the quality is not manageable for all of the (approximately) 30 000 substances within the focus of REACH, a subset of 237 substances has been selected from the known high (HPV), medium (MPV) and low (LPV) production volume chemicals (approx. 10 000 existing substances in volumes >10 tonnes/year as reported to the European Commission). It includes 27 substances of very high concern (SVHC).

This set is considered large enough to detect with sufficient sensitivity changes taking place in the risks related to these chemicals and the quality of the databases for their assessment. Two figures are calculated for each reference substance: the 'Risk Score' and the 'Quality Score'.

Risk score

The methodology to calculate the nominal risk of the reference substances has the same structure as the chemical safety assessment under REACH. Exposure estimates and toxicity estimates are the key parameters to calculate the risk characterisation ratios (RCRs) and Risk Scores for the reference substances. The risk score is a nominal value that indicates to what extent a risk could be associated with the use of the substance. This score can range from far below 1 to 1.000 and more. High risk scores are indications of high risk. However, no attempt has been made to define a value discriminating risk from no-risk.

Quality score

Data on use patterns, volumes and physico-chemical properties are needed to assess the exposure to humans or/and the environment. Toxicological data are required for the toxicity assessment. These data sets can be of very different quality: If input data are highly uncertain (default assumptions) the quality of the assessment is regarded as poor. If representative monitoring data and well defined threshold values are available, the quality of the assessment is regarded as good. Data of high quality have a Quality Score of 1, data of low quality have a Quality Score of up to 10. The Quality Score for the exposure data (QS_exp) and the Quality Score for the toxicity data (QS_tox) are multiplied to give the (total) Quality Score QS_total, the latter ranging from 1 (best quality) to 100 (lowest quality). This quality assessment is a key element of the risk and quality indicator system.

Figure 3 illustrates how the quality score depends on the kind of data.

Figure 3: Types of data used for the risk assessment, and related quality of the estimated risk

REACH will lead to more complete testing of toxicological properties, to better data provided by alternative testing methods such as modelling, improved reporting, and better information on exposure. By this means, the quality of the data (i.e. the completeness of the databases and to a lesser extent the quality of the individual data) is expected to improve and the uncertainty will, consequently, be reduced.

Results: The 2009 ‘snapshot’ and the 5 years update in 2012

In the assessment based on 2007 data (“Pre-REACH Baseline”) the aggregated geometric mean of the Risk Scores for the 237 reference substances has values of:

16 for workers,

34 for consumers,

0.06 for the environment and

30 for humans exposed via the environment.

The 5 years update found clear indications that registration under REACH leads to a significant improvement of our knowledge on substance properties. For the first time, for many substances, existing data have been used to derive toxicity estimates such as DNELs (Derived No Effect Levels), DMELs (Derived Minimal Effect Levels) and PNECs (Predicted No Effect Concentrations), and to perform exposure estimations and risk characterisations.

In addition, for a relevant part of the substances analysed the nominal risks show a clear decrease. The fraction of reference substances with risk characterisation ratios at or below 1 increased. This can be seen as an indication for a better control of risks due to the chemical safety assessments required by REACH.

The key findings of the 5 years update are described in the related comprehensive report. They can be summarised as follows:

Key question 1: Does REACH lead to an improvement in the quality of publicly available data for the assessment of chemicals?

Development of the Quality Scores:

1. The 5 years update found a considerable improvement of the quality of the underlying data for the assessment of the 62 substances (46 HPV chemicals and 16 SVHC) . This is expressed in an improvement of the total Quality Score from 2007 to 2011 (with lower Quality Scores indicating higher quality).

Figure 4 shows the improvement of the Quality Score for the 62 reference substances in the impact area ‘workers’. The figure shows a decline of the geometric mean of the Quality Score for this impact from 21 based on 2007 data to 11 with 2011 data.

Figure 4: Aggregated Quality Score. Comparison 2007 – 2011. Impact area workers (n= 62)

2. The improvement in quality is evident in all four impact areas.
3. For the majority of HPV chemicals and SVHC, the quality of the data underlying the exposure estimate (Quality Score Exposure) and the toxicity estimate (Quality Scores Toxicity) improve.
4. For the first time, some of the reference substances reach the best quality possible (total Quality Score equal to 1) in some impact areas.
5. Due to the registration, DNELs, PNECs and more detailed information on uses and exposures become available for a large number of substances.

Conclusion on key question 1: The results of the 5 Years Update show a marked increase in the quality of the data available for the assessment of the registered substances included in this evaluation.

Key question 2: Does REACH lead to a reduction of the risks which are posed by chemicals to humans and the environment?

Development of the Risk Scores:

1. A marked decrease has been found of the Risk Scores in the aggregated evaluation of 62 substances.

The change in the Risk Scores from 2007 to 2011 is summarised for the four impact areas in Figure 5. Median values are included in addition to the geometric means and confirm the trend of decreasing Risk Scores. This indicates a clear reduction in the risks which are associated with the 62 reference substances (High Production Volume chemicals (HPVCs) and substances of very high concern (SVHC).

Figure 5: Summary of aggregated Risk Scores. Comparison 2007 – 2011 (n= 62)

2. The decline in Risk Scores is almost entirely due to decreases in Risk Characterisation Ratios.
3. The analysis shows a pronounced reduction of the fraction of substances with RCRs above 1 and/or RCRs above 10 in all four impact areas.
4. For almost all substances, changes in at least one of the key input parameters for the RCR (toxicity estimate, exposure estimate) took place indicating changes in the knowledge about the substances.

Figure 6 shows the movement of the risk scores and the total quality scores (QS_total) for the high production volume chemicals. The figure indicates a general movement of the data points towards the lower left corner. This movement means: lower Risk Score, better quality in 2011. The decline of QS_total is somewhat more eye-catching in the scatter plot than the decline in the Risk Score. This observation is in line with the statistical data. ´

Figure 6: Scatter plot Risk Scores versus Total Quality Scores for HPV chemicals. Comparison 2007 – 2011 (n= 46)

Conclusion on key question 2: The results of the 5 years update show a marked decrease in the Risk Characterisation Ratios and the Risk Scores from 2007 to 2011. This indicates a better control of risk, which is largely believed to be due to REACH.

Additional key findings

In 2011, a remarkable number of reference substances still show RCRs above 1. This is mainly due to four reasons:

1. REACH does not require a chemical safety assessment (e.g. for intermediates);
2. REACH does not require an exposure assessment and risk characterisation (e.g. for non-classified substances);
3. limited scope of exposure assessment by some registrants;
4. lack of reliable DMELs for SHVC.

These findings highlight the fact that appreciable risks can be associated with substances, which are not classified. In most of the CSRs analysed, no detailed quantitative risk assessments have been made for the impact areas consumers and humans via the environment.

Many additional findings are specific for individual impact areas. They are described for each impact area in the subchapters “Summary and conclusion” of the comprehensive report.

Causal link between the findings and REACH registrations:

The main difference in the data sources between the analysis 2007 and 2011 has been the availability of REACH registration dossiers in 2011. The changes in the Risk Scores and Quality Scores are calculated based on the data in these documents (to a minor extent additional information came from REACH documents for SVHC due to authorisation and restriction).

Therefore, it is reasonable to state that the registration obligation under REACH leads to the improvements of data availability and the reduction of the RCRs, which have been found in the REACH baseline 5 years update for registered reference substances.

Conclusions on the risk and quality indicator system

The risk and quality indicator system shows reasonable results, differentiation and sensitivity to changes. The calculations permit relative comparisons between substances from the different production bands (HPV, MPV and LPV) and allow comparisons to future points in time. Both changes in risk and changes in the quality of information can be observed and analysed in comparison to the baseline.

This indicator system apparently provides sufficient sensitivity to demonstrate REACH-related changes. It indicator system does not provide results on the absolute and 'real' risk at baseline or in the future. However, the calculated figures correlate with a plausible risk profile, established by scientific approximations and widely agreed conventions in handling uncertainty. The depth of assessment is balanced against transparency for non-experts (e.g. only a limited number of sources were used) and the handling of a sufficiently large number of substances to create a meaningful index.

However, because of the relatively simple assessment procedures used and the 'nominal' character of the calculated risk, no absolute interpretation of the risk score or the risk characterisation ratio (RCR) should be made. For example, the interpretation of an RCR >1 as a 'dangerous' situation at the workplace or an RCR of <1 as a 'safe' situation is clearly an over-interpretation of this parameter. Changes in risk characterisation ratios and in risk scores as seen in the 5 years update are not trivial. Due to the better quality of information available in the REACH process some toxicity values will have to be corrected downwards, other upwards. For some substances poor modelling results for exposure can be substituted by better calculations or measurements. Again, this leads to upward or downward corrections of the exposure estimate and thus unpredictable changes in the risk characterisation ratio.

The 5 years update of the REACH baseline study found clear indications for significant improvement of our knowledge on substance properties. An overall trend can be seen towards RCRs close to 1 or below 1. It will be interesting to observe in future updates the speed of this development, the differentiation for substances from the different production bands and the simultaneous changes in information quality.

The risk and quality indicator system shows detailed substance-specific risk and quality scores. By indicating different tonnage bands, plausible results appear for workers, consumers, the environment and humans via the environment alike, giving the reader a deeper understanding of the existing knowledge of risk.

Outlook

The REACH baseline study has set up a consequent and coherent indicator system which allows to monitor the effects of REACH. At the core of the indicator set is the innovative risk and quality indicator system. It combines a risk-based approach with a clear procedure to deal with different data sources, modelling, and even data gaps. As a result, every risk score comes with a quality tag providing the reader with transparent results.

Eurostat, supported by a Steering Committee and in close collaboration with the other involved Commission Services and the European Chemicals Agency, will continue the work on chemical indicators. Progress regarding the administrative- and supplemental indicators is periodically presented by the evaluation reports of the European Chemicals Agency. Also progress in the development of alternative testing methods to replace animal testing is part of the reporting obligations laid down in REACH.

Every five years the effectiveness of REACH will be monitored due to the REACH Article 117. The 5 Years Update was part of the first effectiveness review. The detailed analysis of the development of the risk and quality scores had to be restricted to HPV chemicals and SVHC. In the second and third registration phase, registrations of substances with medium and low production volumes take place. In the next REACH reviews, the methodology developed in the REACH Baseline Study allows to analyze whether the findings of the first update can be confirmed for medium and low production volume chemicals. The 10 Years Update and the 15 Years Update will show whether REACH is successful to reduce the risk caused by chemicals and to increase the quality of the data, available for the safe use of chemicals.

Data sources and availability

The data used for the calculation of the baseline scenario and the indicators have been obtained from different data sources or by modelling according to the Guidance Documents on Information Requirements and Chemical Safety Assessment. The chemical and toxicological properties and the information on use patterns have been obtained from IUCLID (International Uniform Chemical Information Database, ECHA and Eurostat). For the 5 Years Update, the main data sources have been registration dossiers for of the substances.

Context

The Sixth Environment Action Programme 2002-2012 (sixth EAP) requires a complete overhaul of EU policies on chemicals management, of which REACH is the central element. The major objective of REACH is to ensure a high level of protection for human health and the environment. This includes promoting alternative methods (to animal testing) to assess the hazards of substances, the free circulation of substances on the internal market,and the enhancement of competitiveness and innovation in the EU chemical industry.

By increasing knowledge about the hazardous properties of chemicals, REACH is expected to enhance the communication and implementation of conditions of safe use in supply chains and the substitution of dangerous substances by less dangerous ones. Through different types of measures, REACH should lead to a reduction of the risks to human health and to ecosystems.

The General Union Environment Action Programme to 2020 — also referred to as the 7th Environment Action Programme (7th EAP) — guides the EU’s environment policy up to 2020. It contains proposals for further actions that promote the implementation of REACH and aims to develop — by 2018 — a ’Union strategy for a non-toxic environment’ that is conducive to innovation and the development of sustainable substitutes including non-chemical solutions, building on cross-cutting measures to be undertaken by 2015 to ensure:

the safety of manufactured nanomaterials and materials with similar properties;

the minimisation of exposure to endocrine disruptors;

appropriate regulatory approaches to address combination effects of chemicals; and

the minimisation of exposure to chemicals in products, including inter alia imported products, with a view

to promoting non-toxic material cycles and reducing indoor exposure to harmful substances.

After the United Nations Rio+20 summit there was agreement on a set of Sustainable Development Goals. The agreement called on its signatories to take action on two specific areas related to the production of chemicals, namely, that ’by 2020 the environmentally sound management of chemicals and all wastes throughout their life cycle shall be achieved’ (Goal 3.9: ensure healthy lives and promote well-being across all ages) and that ’by 2030 the number of deaths and illnesses from hazardous chemicals will be substantially reduced’ (Goal 12.1: ensure sustainable consumption and production patterns).