25.8.2011   

EN

Official Journal of the European Union

C 248/138

1.1

The EESC has consistently supported Commission proposals for legislation aimed at reducing both the frequency and potential impact of major-accident hazards. Insofar as the scope of these Directives depends critically on other EU legislation in particular in respect of the classification and labelling of dangerous substances, the EESC agrees that a new Directive is now required following changes resulting from the recent adoption of a globally harmonised system (GHS) for classification and labelling developed and proposed by the UN. The difficulties of so doing, with few anticipated benefits other than in facilitating world trade, are recognised and have been discussed at length in a previous Opinion (1).

1.2

The EESC also agrees fully with the stated view of the Commission and of the majority of interested parties that no other significant changes are required – and indeed that changes should be kept to the minimum to avoid any loss of focus on the key objectives of this long-established, effective and well-supported legislation.

1.3

The EESC therefore believes that every effort should be made to assess critically, if necessary on a product by product basis, whether or not the changes in classification are relevant to the likelihood of a major-accident occurring. If they are not, and/or if they would significantly increase the number of smaller, lower risk, establishments and SMEs affected, then care should be taken not to dilute the impact of the proposal. This is particularly true in the detergents sector where the new classifications bear little relationship to actual experiences of household products in regular daily use. In these cases, the threshold tonnages should also be considered carefully, in particular where there is little likelihood of fire or explosion and where the goods have been packaged in smaller quantities for retail sale.

1.4

Where raw materials, intermediates and finished products are subject to more than one piece of legislation undergoing revision on differing timetables, great care must be taken with the overlapping transition periods to ensure that overall costs for operators and for Member States are minimised and confusion reduced to a minimum for all concerned.

1.5

Given that the general opinion of the competent authorities seems to be that the most important establishments are indeed already captured under this legislation, every effort should be made to improve the efficiency and effectiveness of controls and subsequent reporting on these and, where appropriate, adjacent sites. As far as possible this should not rely merely on increased demands for information to be gathered by Member States and passed to the Commission. The EESC notes that the system as currently structured is barely fit for purpose and welcomes efforts by the Commission to agree changes with the Member States upon whose open and timely inputs it depends. Changes proposed to the list of affected products and establishments should continue to receive scrutiny by the other EU institutions and advisory bodies of the EU before adoption.

1.6

The EESC strongly supports the provision of relevant, comprehensible and timely information to the general public. There will be a continuing need for hard copy, although other electronic means, including social networks, will be increasingly used, especially at the local level. All organisations representing civil society in the vicinity of a ‘Seveso’ (or any other manufacturing or storage) establishment have roles to play in both seeking to prevent and in responding to accidents of all kinds, including major-accident emergencies as defined in the Seveso Directive.

1.7

New proposals relating to ‘environmental justice’ are relevant only if ‘environmental injustice’ in respect of major-accident hazards can be demonstrated. Given the rather low frequency of accidents reported under this Directive, in particular for Lower Tier establishments, it is difficult to see this as being the case. Any information provided should be made available to all elements of organised civil society. The EESC therefore believes that this requirement should be replaced by a more modern and more widely agreed approach to safety information management, properly supported by evidence and the required impact analysis.

1.8

The EESC notes that the EU lags behind the United States in recognising and rewarding good practice, in particular with respect to process and individual safety, and believes that this would bring better returns than some of the measures proposed here.

1.9

The EESC therefore supports this proposal but suggest that a number of points are revisited to ensure that the well-established long term objectives of this legislation to reduce the frequency and impact of major-accident hazards are fully met.

2.1

The need to classify, label and package (CLP) ‘substances’ - initially at least, a finite list of elements and their compounds - defined to be ‘dangerous’ or ‘hazardous’, on a variety of scales affecting human health, safety and the environment, was recognised more than 40 years ago in the Dangerous Substances Directive 67/548/EEC. Just over 20 years later this was extended to cover ‘preparations’- a wider, and potentially infinite, list of deliberately produced mixtures of two or more substances in varying but defined proportions - in the Dangerous Preparations Directive 88/379/EEC.

2.2

These two Directives and their many amending Directives and adaptations to technical progress provide the backbone for a harmonised system of protection of workers, consumers, manufacturers, marketers, distributors and the environment. They also ensure a Single Market across the EU for the products that are affected, including raw materials, intermediate and waste streams, and finished products to be placed on the market. Furthermore, the Directives interact with and provide input to virtually all other EU legislation aimed at protecting human health, safety and the environment. Any changes to this underlying system are therefore likely to be complex and expensive for all concerned.

2.3

In recent years, two such changes have been made. In 2006 the Council adopted Regulation (EC) 1907/2006 for the ‘registration, evaluation, authorisation and restriction of chemicals’, (REACH), together with an accompanying Directive 2006/121/EC to further amend Directive 67/548/EEC and to bring these two major pieces of legislation into line. In 2008 the Council and European Parliament adopted Regulation (EC) 1272/2008 to implement a new globally harmonised system (GHS) for the classification and labelling of chemicals, developed after many years of work by the United Nations. This would lead in many cases to changes in the names, pictograms and standard phrases attributed to various hazards and to classified ‘substances’ and ‘mixtures’. The actual risks for workers, distributors, consumers and the public at large from a given product or process would of course remain the same.

2.4

It was recognised at the time that the benefits of replacing one well-established, fully functioning and effective system with another were likely to be slight, with some potential for reduced costs in international trade being more than offset by increased regulatory and compliance costs within the EU. The technical problems arising from the introduction of new classifications and end points would also be considerable, leading to increased costs of reformulation or changes to the range of products available to consumers – and with considerable potential for confusion both during and after the transition periods, for each piece of legislation affected.

2.5

These problems are now evident in, and are to some extent dealt with by, the Commission's proposal COM(2010) 781 final, otherwise known as Seveso III, for a Directive replacing existing legislation on the control of ‘major-accident hazards’, specifically involving ‘dangerous substances’, both as defined within this legislation.

2.6

This legislation was introduced in 1982 by Directive 82/501/EEC, following a major accident in Seveso (leading to widespread dioxin exposure) in 1976. It was amended following accidents in Bhopal (a major leak of methyl isocyanate) and Basel (a series of fires and toxic releases). It was replaced in 1996 by Council Directive 96/82/EC. Following major accidents in Toulouse (with ammonium nitrate), Baia Mare (a cyanide spill) and Enschede (an explosion in a fireworks factory), this Directive was subsequently amended by Directive 2003/105/EC, setting out a series of well-defined procedural and reporting obligations for manufacturers and the Member States.

2.7

This legislation is generally believed to have had a profound and positive effect on the safety and control of manufacturing plants where dangerous substances are used, manufactured or stored. Some 10 000 manufacturing establishments are now covered, of which around 4 500 are currently designated as being ‘Upper Tier’, i.e, requiring more stringent reporting and control than the 5 500‘Lower Tier’ establishments. Regular inspections are undertaken. National and EU-wide reporting systems are in place. The system is well supported and appreciated by all those involved. Accidents still occur – but hopefully fewer and with less impact on human health and the environment than otherwise would have been the case.

2.8

According to on-line Commission statistics, 745 such accidents have been reported over the 30-year lifetime of the Directive. A further 42 accidents have been reported but not yet added to the publicly available eMARS database (2). Although the statistics are neither complete nor easily available, 80 % of these are believed to be on establishments designated as being Upper Tier, the remainder on establishments designated as being Lower Tier. 35 of the above were reported on a voluntary basis by OECD countries not in the EU. The number of accidents reported per year peaked in the period 1996-2003 and has declined sharply since. It is unclear whether this represents a real improvement in plant safety or merely the rather long delays in analysing and reporting accidents by the Member States and by translation delays thereafter.

2.9

Military establishments, and hazards created by ionising radiation, off-shore mineral and hydrocarbon exploitation, transportation and waste disposal sites, together with specific substances listed in Part 3 of Annex 1 of this Directive, are all excluded from these controls.

2.10

Unfortunately, this legislation depends critically for its scope upon the CLP legislation described above for any dangerous substances involved. Whether or not, and to what extent, a specific site is required to comply with the Seveso controls is defined by the classifications and tonnages of the substances used, manufactured or stored on site. The controls are designed to avoid or minimise the effects of ‘major-accidents’ only, which are defined as those involving any one or more of the following: one or more deaths; injury and hospitalisation of six or more persons; damage to property in or outside the plant; a significant evacuation of personnel or plant neighbours; or long-term damage to the external environment. The truly ‘major-accidents’ noted in 2.6 above which led to changes in the legislation were of course on a larger scale altogether and are therefore not typical of the accidents generally reported.

2.11

The introduction of the GHS legislation now requires changes, in particular to the Annexes of the Directive where specific hazard classes and classified ‘substances’ and ‘mixtures’ thereof are named for inclusion or exclusion based on their revised hazard classifications.

2.12

Given that these are changes to definitions and not to actual risks, and the Commission does not intend to significantly amend or extend the scope of the current legislation, the actual benefits for process and worker or consumer safety or the environment are expected to be minimal. The need to control costs and other impacts for operators and the Member States is therefore obvious, as is the need to avoid any weakening of the current focus on major-accident hazards.

3.1

The Commission's proposal for a new Directive is based on Article 191 TFEU. The Directive is addressed to Member States and will come into effect 20 days after publication in the Official Journal. Directive 96/82/EC will be repealed effective 1 June 2015. Stakeholders have been consulted. Overall there was agreement that no major changes were required, other than to align Annex 1 to Regulation (EC) 1272/2008.

3.2

The Commission does however seek to clarify and update some of the procedures and definitions, as well as to introduce new measures, in particular with respect to the frequency of inspections, the content of the operator's major-accident prevention policy (MAPP), requirements for a safety management system (SMS), the provision of information to the public, the rights of access to environmental justice, the provision of reports from Member States to the Commission, and to the process of amending the Annexes via delegated acts.

3.3

The Commission recognises that the main difficulties lie in aligning the existing categories of ‘Very Toxic’ and ‘Toxic’ with the new categories ‘Acute Toxic 1, 2 and 3’, now split by exposure routes (oral, dermal and inhalation). There will be new and more specific categories for oxidising, explosive and flammable hazards, including ‘flammable aerosols’. A number of other products, including ammonium nitrate and heavy fuel oil, in general use despite their occasional use as explosives precursors, receive special mention.

3.4

The proposal is accompanied by a staff working paper and impact assessment, two external impact assessments prepared by COWI A/S (Denmark, international consulting group) on possible options for the overall proposal and for adapting Annex 1, and by a report from a JRC Technical Working Group (TWG) on classification criteria for the identification of Seveso establishments. Additional information on proposals for the reform of the eMARS database was provided on request.

3.5

Despite some accrual of powers and responsibilities to the Commission, there is said to be no impact on the Community budget. The impact assessment does not fully quantify the costs and benefits for Member States or operators – but suggests that both should be significantly lower than when the legislation was first introduced. It also notes that costs are generally minor compared to those incurred in an actual incident. The 2005 fire at the Buncefield terminal in the UK is quoted as an example. Some new proposals for communicating to the general public or for providing data to the Commission were not evaluated for cost or actual effectiveness. The application and presumed outcomes of the current legislation were examined in related documents but these were not included in the impact assessment.

4.1

The EESC has commented favourably via its Opinions on all the above Directives and has strongly supported the series of proposals seeking to minimise the frequency and impact of major-accident hazards under the general names of Seveso Directives I, II and now, III. It therefore supports the new proposal, its chosen legal basis and choice of instrument. Concerns however exist with respect to the proportionality and likely effect of the proposal, where some of the provisions clearly do go beyond what is strictly necessary to achieve the desired objectives.

4.2

The EESC also strongly supported the aim of a ‘globally harmonised system’(GHS) for the ‘classification, packaging and labelling of chemicals’, as set out by the United Nations, to support world trade and to assist less developed economies in their efforts to protect the safety and health of workers and consumers.

4.3

The EESC however set out a number of caveats on the above in a previous Opinion published in OJ C 204 p. 47 dated 9 August 2008. Many of these would apply in any process of trans-national or, in this case, global, harmonisation, however well-intentioned, where one well-functioning system is replaced by another in the name of some greater good, i.e. facilitating world trade. Bureaucracy and costs can escalate. Long established procedures and definitions can be weakened. Essential aims can be diluted. Manufacturing and marketing practices may require revision at considerable cost and with no benefit to workers or consumers. Confusion at all levels is likely both during and after the inevitable transition periods for each piece of legislation affected. Benefits, if measured, are likely to be low or non-existent and additional costs are hard to justify.

4.4

Many of the above points were recognised in the preparation of this proposal, especially with the almost universal agreement that no major revisions were required to the focus, scope and general implementation of the existing legislation, other than to revise Annex 1 in line with the new definitions for the CLP of ‘dangerous substances’ upon which this legislation depends.

4.5

Sadly a number of problems remain. Some of these were raised during consultation, but not addressed in the current text. Other concerns of a general nature have been lost altogether.

4.6

The EESC particularly regrets that the adoption of a globally negotiated, essentially monolingual GHS has led to the loss of meaning of long established key words, such as ‘substance’, which can now include both ‘preparations’ and ‘mixtures’; and that these last two words are taken as meaning the same, which was not the case in the General Preparations Directive; and that still no effort has been made to confirm that the three terms - in English and in some other EU languages - ‘chemical’ (as a noun), ‘chemical substance’ and ‘substance’ are synonymous in EU legislation, despite different usages and attributed meanings. For some, it may be necessary to explain that there are no ‘non-chemical’ substances. References to ‘M-factors’ or ‘R & S Phrases’ again only make sense in one language and can present problems when translated into other languages.

4.7

This is therefore a missed opportunity to establish a glossary of key terms, in all the languages of the EU, as previously suggested, which is essential as legislation moves into new areas affecting the same group of products – for instance for limiting the availability for use by terrorists as explosives precursors – as well as dealing with overlapping and interacting horizontal and vertical legislation, for instance REACH, Industrial Emissions, Water Quality and WEEE with product-specific legislation on solvents, detergents, cosmetics, aerosols, fertilizers and pesticides.

4.8

A similar point follows regarding the proposed process for amending the Annexes, which essentially serve to increase or decrease the number of products and therefore sites subject to this legislation, by the Commission acting alone by ‘delegated acts’. These will require clearly written guidelines, acceptable to all the affected parties. These have yet to be developed. The scientific bases for these decisions must be set out in full and all pre-agreed procedures closely followed. In the case of a challenge by the European Parliament or Council, a full review by the other EU institutions and advisory bodies should be mandatory. Provision should also be made for the lodging of objections by individual Member States or by other affected parties.

4.9

This is also relevant to the scope of application of the proposal. The Seveso II Directive applies to around 10 000 named establishments across the EU. Of these, around half are also covered by the newly adopted Industrial Emissions Directive, replacing IPPC, which will cover more than 50 000 sites in all. The so-called ‘Seveso establishments’ include chemicals manufacturing, petroleum refining, consumer products and other downstream manufacturing, and some waste processing sites. There seems to be agreement by the competent authorities in the Member States that the existing definitions capture reasonably well the sites at which a major-accident is at all possible. Certainly all major sites are listed. Any increases in the level of the product classifications to meet the requirements of the GHS without any changes to the actual underlying hazards will merely add smaller sites, with ever decreasing actual risks, or increase, with little justification, the number of sites classified as being of higher risk. There are specific concerns in the case of detergent raw materials where, due to changes in definition, a significant number of Lower Tier sites could be added. Given that, according to Commission statistics, as above, the 5 500 Lower Tier establishments account for no more than five to ten reported accidents per year, this does not seem to be a priority area for further regulatory attention. Indeed, for any one Lower Tier Seveso establishment, a reportable accident every 500-1 000 years (or even for a Higher Tier Seveso establishment, one accident every 100-200 years) suggests that greater personal hazards exist at home or are encountered on the way to work – although these of course rarely have a major consequential impact on others or are treated as being as serious by regulators or the public at large. Ensuring that the SMEs responsible know and comply with the law and that the sites are regularly inspected by the competent authorities brings rapidly diminishing returns. In times of reduced budgets and manpower restrictions, this could increase the likelihood of serious accidents occurring elsewhere.

4.10

The EESC therefore strongly requests that the original focus of the Directives on avoiding or minimising the effects of major-accident hazards, as defined, should be maintained. Any decision to water down this effect, merely through the introduction of the new GHS for CLP, or through changes to the local and EU-wide reporting systems, should be strongly resisted. This requires careful consideration of not only the new classification limits but also of the threshold tonnages for products stored. Where this includes goods already packaged in smaller quantities for retail sale, and where there is little likelihood of fire or explosion, the risk of a major-accident as defined is greatly decreased.

4.11

The EESC also notes that this proposal specifically and rightly excludes accidents such as the recent off-shore explosion on a drilling rig in the Gulf of Mexico, where new legislation may be required, and, more locally, the ‘red mud’ leak in Hungary which was covered, at least in theory, by the 2006 Mining Waste Directive (MWD). Proper implementation and inspection at national level is of course critical, whatever legislation is in place at EU level.

5.1

The EESC notes that there are reporting requirements for manufacturers and Member States on differing timescales under many of the above-mentioned Directives. Adding to the frequency and depth of reporting under this one heading, without clear evidence of positive effect, adds to the burden of all those concerned. Accumulating data centrally, in Brussels or elsewhere, adds to the problems of maintaining both data quality and of confidentiality where appropriate.

5.2

This is also relevant to a new requirement for ‘establishments’ to provide details of their ‘neighbours’ to avoid ‘domino effects’ in adjoining sites which may or may not be covered by this legislation. It is unclear how this can be handled under EU competition law. It is however certainly relevant to the development of plans to respond to local emergencies and in this respect has the full support of the EESC.

5.3

The requirement for operators to draw up reports including evidence for the presence of a ‘safety culture’ follows accidents in the US with respect to the Space Shuttle disaster, and more recently to major-accidents in Texas and the Gulf of Mexico where such a culture was said, in retrospective reports, to be missing. These are however subjective comments which are hard to evaluate or quantify. Providing regular and meaningful assessments in advance would present problems for the competent authorities, as currently staffed, in most Member States. This proposal was therefore rejected at a meeting of safety experts in Ispra in 2010 and it is unclear why it has been reintroduced in the current proposal.

5.4

Overall the EESC would prefer the reporting requirements to be kept at a realistic, meaningful, comparable and enforceable level across all Member States, with every effort to exchange best practice across borders. The EESC in particular regrets that the ‘lessons learned’ sections of the 745 on-line reports currently available in the eMARS database are typically blank and that the remainder of the ‘tick-boxes’ provide little useful information, despite being offered to the general public, as well as to safety experts, as a key source of relevant data. Some data, for instance on the distribution of accidents between Upper and Lower Tier establishments have apparently not been systematically collected, making it difficult to evaluate the effectiveness of the various components of the legislation and of the actions required at national level. The EESC therefore welcomes actions by the Commission to agree new reporting standards with the Member States and trusts that sufficient resources will be allocated to keep the system fit for its original purpose.

5.5

The EESC also questions why the reporting requirements for the two Tiers of operators, defined by the tonnages and hazard classifications of the substances produced, used or stored on site, differ with respect to the need for a Major-Accident Prevention Policy (MAPP) and subsequent Safety Management System (SMS) and safety report. Given that the first (MAPP) has little value without the second (SMS), the EESC believes that this requirement should apply equally to all establishments listed under this Directive. However, the specific requirements for Lower Tier establishments should be more closely adapted to the much reduced risk of a major-accident actually occurring.

5.6

The EESC notes that proposals to supply information to the public have been considerably extended, although it is not always clear why this has been done. ‘Schools and hospitals’ are specifically identified in one paragraph – but it is unclear whether they have been defined for educational purposes, as sites employing or containing large numbers of individuals requiring specific evacuation plans and training, or as key resources in the case of an emergency. This should be clarified so that the appropriate actions can be taken by those concerned.

5.7

In all such cases the requirement should be that the information supplied is relevant, comprehensible and timely for a particular purpose. Electronic delivery will serve some sectors of the community but not others. Hard copy information will be required for many years to come. Novel forms of communication such as the use of targeted emails, social networks and even Twitter® could be explored at local level within plans to deal with specific emergencies.

5.8

Finally the Commission includes a new article to ensure ‘environmental justice’, a concept that emerged in the United States in the 1980s, based on the civil rights movement of 20 years earlier, where colour, poverty and lack of civil rights, and therefore an obvious lack of ‘justice’, were closely correlated. The same principles were incorporated in the Aarhus Convention of 1998. A Regulation setting out the obligations of the EU institutions was adopted in 2006. The accompanying Opinion of the EESC published in OJ C 117 p. 52, dated 30 April 2004, supported the proposal but expressed concern over the narrow definition of ‘organisations promoting environmental protection’ where ‘other non-profit organisations, such as trade unions, social economy and socio-occupational organisations, consumer associations, etc., also play an important role in protecting the environment at local, regional, national and European level.’ This point remains valid today, with all the elements of organised civil society quite properly involved in making informed judgements on the issues relevant to this legislation and in ensuring that the health and safety of workers and of the surrounding general public are properly protected in the event of a major-accident actually occurring. According to on-line Commission reports on the implementation of the 2006 Regulation, the rather few recorded information requests have been in support of on-going pan-European campaigns rather than on specific site issues. It is therefore unclear why this particular requirement has been added at this point, rather than, for instance, proposals to exchange, recognise and reward best practice. This is an area where the EU does lag significantly behind the United States and where real gains in process and personal safety could be achieved, in line with the declared objectives of this proposal.