Different forms of cancer may be initiated or promoted by the exposure to carcinogenic and/or mutagenic chemical agents at work 10 . In the EU 32 mln workers (23% of people employed) were exposed to chemical agents classified as carcinogenic or probably carcinogenic to humans by the WHO's International Agency for Research on Cancer (IARC) in 1990-93. 11 In France, estimates from 2010 suggest that approximately 2 mln workers were exposed to at least one carcinogen in their most recent week of activity at work, approximately 10% of the working population. Workers in maintenance activities (42.6% exposed) and young workers (below 25 years of age) (15.2% exposed) were found to be particularly exposed, and many of them to several carcinogens, mutagens and reprotoxins at the same time 12 .

According to the EU Classification, Labelling and Packaging Regulation (CLP) 13 , 1017 chemical agents (and groups of chemical agents) have received mandatory 'harmonised classification' as 'category 1' carcinogens, attracting the label hazard statement 'may cause cancer'. If the assessment was based primarily on human evidence category 1A is assigned (this is the case for 336 agents), with category 1B assigned where animal evidence is primarily used (681 agents are in this category). Chemicals suspected to cause cancer in humans but where data is not sufficiently convincing for category 1 may be classified as 'category 2 carcinogens'.

IARC has identified nearly 500 agents that are carcinogenic for humans (Group 1; 118 agents), probably carcinogenic to humans (Group 2A; 75) or possibly carcinogenic to humans (Group 2B; 288). 14

Table 1 in the section 1.1.2. presents estimates of numbers of workers exposed to the carcinogens covered in this report. The numbers of exposed workers vary significantly across the agents. Some of the carcinogens, like Respirable Crystalline Silica (RCS), chromium (VI) compounds, hardwood dust or hydrazine, affect very high numbers of workers. For some others there are indications that use patterns may be lower. However, the agents are themselves important – for example the ratio between the number of exposed workers and cancer cases may be higher for some of these chemical agents. It is also important to recognise that use patterns change rapidly and for complex reasons. Market forces such as raw material and energy prices, developing technology, as well as regulatory changes can drive increases in use which are not easy to predict. This is the case for example for acrylamide, used for production of polyacrylamide, which is a growing sector.

1.1.2The Carcinogens and Mutagens Directive is outdated

The current rules

While cancer is a complex disease and some causal factors are difficult to identify, it is clear that cancers caused by work can be prevented by reducing or eliminating the exposures leading to the disease.

The EU principles of worker protection from carcinogens are laid out in the over-arching Occupational Safety Health (OSH) Framework Directive 89/391/EEC and those Directives specifically dealing with chemicals risks – notably the Chemical Agents Directive (CAD) and the Carcinogens and Mutagens Directive (CMD). The latter includes a commitment to set Occupational Exposure Limit (OELs) values for all carcinogens or mutagens for which this is possible. 15

Under the OSH framework risks to the safety and health of workers shall be eliminated or reduced to a minimum. In the case of carcinogens, CMD sets a number of concrete provisions.

Employers must identify and assess risks to workers associated with exposure to specific carcinogens (and mutagens), and must prevent exposure where risks occur. Substitution to a non- or less-hazardous process or chemical agent is required where this is technically possible. Where carcinogens cannot be substituted they must, so far as is technically possible, be manufactured and used in a closed system to prevent exposure.

Where this is not technically possible either, worker exposure must otherwise be reduced to as low a level as is technically possible. This is the so-called minimisation obligation under Article 5 of the CMD. This is a more strict standard than for other hazardous chemicals, where the duty to control risks is always qualified by an assessment of risk by the employer.

CMD provisions apply to any chemical agent which is classified as a 'Category 1' carcinogen (or mutagen) under the EU CLP Regulation, thus to the 1017 substances above mentioned. They also apply – in full – to any chemical agent which would meet the criteria for such classification if it were placed on the EU market, which means that employers have a responsibility to identify occupational carcinogens to which workers may be exposed but which have not already been classified by their suppliers.

This is the case for example of the so-called 'process-generated substances' (PGS). These are hazardous 'chemical agents' such as dust, fumes, and gases which may, for example, be generated during the combustion of fuel by diesel engines used for mining, or as by-products during production processes, etc.

PGS have the potential to be major sources of occupational exposure to chemical carcinogens – but because they are never 'placed on the market' in the EU they are never subject to the CLP classification system. This can be confusing for employers, workers, and enforcers. It is clear that CMD controls apply to workplace carcinogens – but it may in some cases be unclear whether a 'process generated substance' is hazardous in this way.

CMD therefore also includes a list of identified process generated substances in its Annex I. The aim of this list is to clarify for workers, employers, and enforcers whether a given chemical agent, if it has not otherwise been classified according to CLP, is in scope of the CMD controls. Currently, Annex I has 5 entries:

1.Manufacture of auramine.

2.Work involving exposure to polycyclic aromatic hydrocarbons present in coal soot, coal tar or coal pitch.

3.Work involving exposure to dusts, fumes and sprays produced during the roasting and electro-refining of cupro-nickel mattes.

4.Strong acid process in the manufacture of isopropyl alcohol.

5.Work involving exposure to hardwood dusts.

CMD also provides that, in any case, exposure of workers must be kept below 'binding occupational exposure limit values' (OELs).

An OEL addresses the inhalation route of exposure, describing a maximum airborne concentration level for a given chemical agent above which workers should not be exposed, on average, during a defined time period.

OELs can further be annotated with appropriate indications of additional non-inhalation hazard such as, for example, a 'skin' notation where the dermal route of exposure is scientifically considered to be relevant.

OELs are set in Annex III of the CMD for some chemical agents which fall under the scope of the Directive - including some of the PGSs identified in Annex I. 16 Annex III currently includes limit values for occupational exposure for three agents: benzene, vinyl chloride monomer (VCM) and hardwood dusts.

As explained above, employers must prevent or minimise exposure to occupational carcinogens where risks occur. The principle of minimisation of the exposure is stated in article 5.3 of the CMD: 'the employer shall ensure that the level of exposure of workers is reduced to as low a level as is technically possible'.

CMD OELs do not directly affect in theory the legal standard of control, which is in any case for minimised exposure. In practice, however, the existence of an OEL provides a clear benchmark that enables professionals to 'operationalise' the concept of minimised exposure, thereby allowing them to easily determine the level to which the exposure should at least be reduced.

From a more general perspective, OELs promote consistency by defining a 'level playing field' for all users and a common objective for employers, workers and enforcement authorities. This leads to a more efficient system of protection of workers' health.

It should be added that for most carcinogens and mutagens substances it is not possible to identify a safe threshold below which the adverse health effects of exposure can be prevented completely. The genotoxic mode of action of a substance entails that extremely low amounts of a substance reaching the appropriate target (deoxyribonucleic acid or DNA) may initiate a tumoral process 17 . Therefore, although for a genotoxic carcinogenic substance the risk cannot be eliminated by setting an OEL, the probability that the effects occur can be lowered by complying with a limit value. In terms of population, it can be said that the percentage of workers affected by a cancer would be lower if the level of exposure is kept lower. This is the reason why the minimisation of the exposure is a basic principle set up in CMD and why OELs for carcinogenic substances are useful to prevent occupational cancer although for most substances (non-threshold substances) there will always be a residual risk that could only be reduced to zero by fully eliminating the presence/use of the substance in the workplace.

The CMD includes a general obligation for the employer to reduce the use of a chemical carcinogen by, where technically possible, substituting to a less- or non-hazardous chemical agent or process and, when requested to do so by the relevant authorities, submit the findings of the associated investigations (CMD Article 4). The adoption of an OEL does not replace this obligation.

The practical implementation of effective chemical substitution policies can deliver significant benefits in terms of protecting the health and safety of workers.

Effective substitution is associated with a number of issues that are not always easy to evaluate in order to facilitate the decision making process. It requires judgment to take account of workers health and safety protection, process performance, the ease and cost of introducing substitutes, environmental considerations and other factors in making a substitution choice. Several approaches to substitution exist ranging from ad-hoc approaches to methods that are defined, structured and documented. Less sophisticated substitution approaches may be more suitable for smaller companies compared to larger better resourced organisations that have a high level of technical expertise. To make substitution happen in practice requires a raised awareness and involvement of all stakeholders.

The European Commission recognises the multi-attribute challenges that substitution presents to individual employers and has published guidance to analyse and evaluate the practical implementation of the principle of substitution of hazardous chemicals at the workplace with a view to further enhance the protection of workers health and safety while taking into account the above-mentioned factors. 18

Annex I and III of the CMD are not in line with scientific evidence.

Firstly, there is important scientific evidence that the process-generated substance, respirable crystalline silica, is carcinogenic 19 .

Secondly, two of the three current OELs defined in Annex III of the CMD, namely on 'work involving exposure to hardwood dust' and vinyl chloride monomer are estimated to be too high to adequately protect workers. Available scientific evidence points also to the need to complete Annex III with OELs for several additional agents 20 .

As has been established, over 1000 chemical substances have been given 'harmonised' classification as category 1 carcinogens under the CLP Regulation. Many of these chemical agents are, however, almost certainly no longer used in Europe, having been superseded or otherwise phased out of use – and fewer again will be present as occupational carcinogens today. There are, however, many chemical carcinogens which are known to both be in current use in Europe and which pose a risk to workers.

The Commission initiated work to amend or establish OELs for 25 priority chemical agents in 2004. 21 The selection of these agents took into account the views of stakeholders, principally at MS authority level as communicated to the Commission during exchanges of views including at the meetings of the National Experts Working Group on OELs. In addition for the two of the three existing EU OELs (hardwood dust and VCM) the Scientific Committee on Occupational Exposure Limits (SCOEL 22 ) had adopted new/revised Recommendations, in 2003 and 2004 respectively that indicated a need to consider the revision of the existing OELs.

Among hundreds of carcinogenic substances, these 25 chemical agents are considered as a priority for protection of workers and the choice is quite consistent with subsequent third party priority lists. For example, of the 25 chemical agents, 11 are listed among the 30 highest priority substances in a report published by the Netherlands National Institute for Public Health and the Environment 23 , and 21 were included in a list of 65 candidates for OELs published by European Trade Union Congress in 2015. 24 As shown by a study contracted by the Commission to evaluate impacts of introducing OELs for those 25 chemical agents 25 (later on referred to as the IOM study), at least 20 million EU workers are considered to be potentially exposed to one or several of them.

This report deals with 13 chemical agents from the original 25 for which data is available 26 .

For the second group of 12 chemical agents, further information gathering and consideration is necessary. 27

In the course of the preparatory work, the Commission considered also the possibility to extend the scope of CMD to apply also to chemical agents which are toxic for reproduction. However, as set out in more detail in annexes 1 and 2 (sections 8.4.3, 9.1, and 9.2.2-3) several policy and technical factors, including the already-adequate scope of the CAD, resulted in a conclusion that there is no need for further action at this time.

The first 13 chemical agents and their key characteristics are presented in the Table 1 below.

Table 1.    Sectors, types of cancer caused and estimated exposure levels for 13 chemical agents under consideration 28

Adoption of the proposal to review Annexes I and III of the CMD with regard to a first batch of carcinogens is envisaged for the first semester of 2016 in the Commission Work Programme. Preparatory work has been going on since 2004, and representatives of Member States authorities, employers' and workers' representative bodies within the framework of the tri-partite Advisory Committee on Safety and Health (ACSH) strongly anticipate a Commission proposal. In its Resolution of 25 November 2015 on the EU Strategic Framework on Health and Safety at Work 2014-2020, the European Parliament 'firmly reiterates its call on the Commission to present a proposal for a revision of Directive 2004/37/EC on the basis of scientific evidence adding more binding occupational exposure limit values where necessary'. 34

1.1.3Inadequate OELs at MS level have negative consequences for workers and businesses across the EU

Under the CMD, Member States may adopt a national limit value that is lower (i.e. more stringent) than the EU value. They can also adopt OELs for chemical agents for which it has not been done at EU level. This is consistent with the ultimate objective of the Directive, which is to minimise the level of exposure, as it allows for further advances to take place at Member State's level in light of country-specific and industry-specific developments.

One or more Member States have indeed set limit values for all of the agents covered in this IA. For example, most Member States have set limits for ethylene oxide and acrylamide. Less than half have done so for bromoethylene or o-toluidine.

In the absence of EU OELs, national scientific committees and/or other national bodies need to independently evaluate each carcinogen leading to a repetition of identical tasks in several Member States.

While the possibility to set national OELs is in line with the CMD provisions, in practice, where national OELs exist they vary considerably in some cases, leading to significantly different levels of protection. For example for 1,3 butadiene, the values range from 4.5 to 100 mg/m3. For ethylene oxide, values range from 0.84 to 90 mg/m3. A comprehensive overview of the national OELs for each of the chemical agents covered in this report is provided in Annex 9 and a summary overview table can be found in Annex 6.

The lack of national OELs for some chemical agents, and the high levels of others, not only leads to inadequate protection for EU workers but can also have negative consequences for the internal market, because businesses located in Member States with less stringent levels (i.e. with absent or higher OELs) would benefit from an undue competitive advantage. For example, 1,2- epoxypropane producers in FR and RO have to comply with exposure limits 10 times higher (i.e. less constraining) than in other producing countries, such as DE, ES and NL (50 mg/m3 vs 5 mg/m3). Even if the minimisation obligation applies to all employers in Europe, these undue competitive advantages would continue to exist in the absence of an EU-wide OEL, because both employers and surveillance authorities would need to rely on their own judgement on what is "technically possible".

Divergence in national OELs may be partially due to differences in the methodology used for the scientific evaluation of a carcinogen by the respective national limit setting committees. More recently established national OELs may be different (for example more stringent) than those set earlier as a result of improved scientific understanding or availability of data. Further, specific production process(es) used in one Member State may allow for lower OELs to be set compared to use of alternative production process (for example) in another Member State.

Where Member States have, in the absence of EU OELs, set national OELs that vary from each other this has the potential to create uncertainty on what is expected of employers in terms of risk management. This may be perceived as inefficiency of the system, in particular by employers that operate in more than one Member State.

EU-level OELs do not intend to completely eliminate the variation in national OELs but rather ensure that all Member States introduce a minimum level of protection which is considered appropriate in light of scientific knowledge and the state of technological development. Even if Member States retain the possibility to adopt stricter levels, EU OELs significantly reduce the scope of variation by introducing an upper limit for acceptable exposure levels.

As explained earlier, employers have the obligation to achieve exposure levels as low as technically possible to be compliant with the minimisation principle under the CMD. What is technically possible, however, may be different for the same chemical agent, depending for example on the sector of the industry and the specific type of use of the chemical agents.

Substances such as RCS, for example, can occur in very diverse workplace situations and therefore the possibilities to control exposures will differ. Without an EU-wide OEL, employers who would like to be sure to comply with the legislation ("minimisation to as low a level as technically possible") would need to embark on an extensive research into different practices existing across EU Member States and in different types of industries to arrive at an understanding of what might be the appropriate and feasible exposure limit review for their type of business. In that sense an EU-wide OEL, based on a generally accepted methodology and validated through tripartite discussions as a common denominator should be feasible for all uses of the carcinogen, which simplifies compliance for business.

Similarly, for workers, in the absence of an OEL or where national OELs are too high, it would be challenging to collect evidence that exposure levels in a specific factory are higher than what should be possible. An OEL facilitates therefore discussions between employers and workers on this matter by setting an agreed benchmark for compliance.

1.2Who is affected by the problem and how?

As mentioned above, the total number of deaths attributed to occupational cancer in the EU is reported to be 102,500 for 2011 35 .

Epidemiological studies indicate that occupational exposures cause 5.3–8.4 per cent of all cancers 36 . However, the rates are not the same for all cancers, for some, such as sinonasal cancer linked to wood dust exposure, work-related cancers are estimated in some countries to account for almost half of all cancers in some countries of this type. 37

In 2012, cancers of the lung, colorectal, breast and stomach were the most common fatal forms of cancer in Europe. The most common cancer sites were cancers of the female breast, followed by colorectal, prostate and lung, which all together represent half of the overall burden of cancer in Europe. 38

According to a UK study based on 2005 data, industries and occupations with high cancer registrations include construction, metal working, personal and household services, mining, land transport, printing/publishing, retail/hotels/restaurants, public administration/ defence, farming and several manufacturing sectors. 56% of occupational cancer registrations in men are attributable to work in the construction industry (mainly mesotheliomas, lung, stomach, bladder and non-melanoma skin cancers). 39

The chemical agents subject to this IA have a direct impact on developing the above mentioned types of cancer and are relevant to the sectors with high rates of cancer registrations. 40

For the workers and their families, cancer results not only in substantial quality of life losses, but also in direct health care costs and indirect loss of present and future earnings (net of taxes) both for the person affected and for the carers, in addition to the administration costs related to the time and expenses incurred claiming for benefits, waiting for treatment etc.

Occupational cancer also impacts the economy at large, reducing labour supply (either temporarily or permanently) not only by the person affected but also by his/her carers, decreasing labour productivity, and increasing the burden on public finances through avoidable public expenditure on health care, disability benefits, pensions for early retirement, and other benefits.

As an illustration, according to a study published in November 2013, around 8,000-8,500 deaths per year due to occupational cancer are estimated to occur in Italy alone, corresponding to 170,000 Potential Years of Life Lost and more than 16,000 Potential Years of Working Life Lost, leading to around 360 mln  euros in indirect economic loss.  41 To these add the direct health care costs of occupational cancer in Italy, estimated at around 456 mln euros. Another study conducted in Spain estimated direct costs for diagnosis and treatment of occupational bladder and lung cancer to be about 88 mln euros. 42 This did not cover indirect costs such as costs of absences, production loss, loss of income and impairment. 43

For Member States, occupational cancer leads to increased healthcare costs related to treatment and rehabilitation, as well as to higher expenditure on associated inactivity and early retirement and compensation for recognised occupational diseases. It also increases administrative and legal costs related to the handling of requests for benefits and dealing with recognized cases. Foregone earnings and income as a result of ill health also lead to tax revenue losses for social security systems.

For business, occupational cancer also implies costs in terms of productivity, as they lose skilled workers and need to spend more in job search and training of new workers. Given the often long time lag between exposure and illness and the probability of workers changing employers during their work career, the risk of future productivity losses is unlikely to be internalised by companies, and therefore not factored into present businesses decisions. Depending on the relative strength of collective bargaining, some of the affected sectors/companies may also need to pay higher wages to compensate for the higher occupational risk, which would affect their competitiveness vis-à-vis otherwise similar companies which are confronted with less organised labour. Finally, businesses located in Member States with where national OELs are relatively stringent may be at a competitive disadvantage vis-à-vis enterprises in Member States with no or higher OELs.

An indicative estimate of the direct and indirect health costs related to the agents covered in this impact assessment can be found in the following section.

1.3How would the problem evolve?

Estimations on the numbers of deaths and health costs between 2010 and 2069 have been made in case no action is taken (baseline scenario) regarding the chemical agents subject to this IA. 44 , 45 These are summarized in the Table 2 below.

Table 2.    Estimated cancer deaths, cancer cases and related health costs in case no action is taken (baseline scenario), 2010-2069

It is anticipated that there would be around 440,000 deaths from exposure to respirable crystalline silica (RCS) and a similar number of new cancer cases.  The related health costs would range between 192 bln and 493 bln euro. The social partners, representing 18 European industry sectors signed in 2006 a European Multi-Sectoral Social Dialogue Agreement on Workers' Health Protection through the Good handling and Use of Crystalline Silica and its products (NEPSi). The impact of NEPSi on the reduction of cancer cases and deaths in those sectors 50 implementing the Agreement is under evaluation – results are not available to inform this assessment. Exposure to hard wood dust, which potentially concern 3 mln workers in Europe, is expected to cause 12,000 cases of cancers and 5000 deaths over the next 60 years, with estimated health costs between 3 bln and 16 bln euro. On the other hand, relatively smaller effects are expected from exposure to o-Toluidine: 490 new cancer cases and 150 deaths before 2050.

These estimates, derived from the IOM study, are based on the assumption that for many of the concerned chemical agents, past trends of declining exposures will continue. These trends were related to technological progress, changes in work organisation and relative weight of different industrial sectors but also to past legislative developments. It is difficult to predict whether such trends would indeed continue in the absence of further EU action. The 60-year time frame of the assessment poses also the challenge of anticipating future industrial developments whereby uses of the chemical agents under consideration could either decline or grow and where potential new uses could lead to new workplace risks.

Another important assumption in the study is that for some of the chemical agents the industry has already achieved relatively low exposure levels, sometimes lower than the proposed OELs. As this varies across the chemical agents, more information will be provided in Section 5 of the report about the specific situation for each of the carcinogens as well as on the sources of the information. Generally speaking, however, even where it is estimated that current exposure levels are already very low, lack of EU OELs or too high EU OELs mean that it will still not be clear for employers and workers and enforcing authorities whether the achieved exposure level is satisfactory from the point of view of compliance with the minimisation principle of the CMD.

Table 1 in Annex 6 presents the current limit values in the EU Member States. The information regarding existing national OELs 51 was gathered through an extensive 2014 survey. 52 Lack of EU action will most likely mean that there will remain Member States where no limit values exist for certain carcinogens or where those values are too high to ensure adequate worker protection. A minimum standard across the EU will not be ensured, to the detriment of both worker protection and the internal market.

2Why should the EU act?

2.1Does the EU have the right to act?

The Treaty on the Functioning of the EU (TFEU) in Article 153 empowers the EU to support and complement the activities of the Member States as regards the protection of workers' health and safety, and to adopt by means of Directives, minimum requirements for gradual implementation. On this basis the CMD provides a specific basis for action. For more details regarding the full legislative framework see Annex 7.

2.2Why is EU action needed and what is its added value?

Limit values under the CMD, which according to the Directive are regarded as an important component of the general arrangements for the protection of workers 53 , should be established for all those carcinogens and mutagens for which the available information, including scientific and technical data, make this possible 54 , and must also be revised whenever this becomes necessary in the light of more recent scientific data.

One of the most important advantages of setting OELs is that they provide robust and objective benchmarks to help demonstrate compliance with the current legal framework. They establish ceiling exposure values above which exposure should not occur (in a weighted average over one working day). This enables the employer, based on the risk minimisation approach, to implement the OSH risk management measures that reflect the needs of their sector. In some cases it may be possible to achieve exposure reductions well below the level at which the OEL is set, for other sectors this may be less achievable. However, in all cases the resultant exposure should not exceed the OEL and over time, with developments in technology, there should be a trend for overall exposures to reduce to a minimum.

The concept and use of OELs is well understood by all the main stakeholder groups. As such they are a common reference point that can be used as a practical tool by employers, workers and enforcers to assess compliance with the general CMD requirements. They can also be used by process plant and machinery designers when planning new, or considering alterations, to existing process plant.

OELs therefore make an effective contribution to the practical implementation of the requirements in the legislation for prevention and reduction of exposure to chemical carcinogens.

The continued support from stakeholders for OELs was restated during a meeting on 18th November 2015 of members of the tri-partite Working Party on Chemicals to discuss those aspects of the overall evaluation of the EU OSH acquis that are relevant to chemicals risk management. During this meeting the Members States', employers' and workers' representatives agreed that 'OELs are an important tool for chemical risk management at the workplace and there is a need to adopt values for more substances based on duly justified reasoning. For example an early focus should be on carcinogenic, mutagenic and reprotoxic substances'.

In addition, during 2015, Commission services asked the Senior Labour Inspectors Committee, via its working group on chemicals (CHEMEX), to assess the utility of OELs from an enforcer's perspective, including the question 'Are OELs considered (by labour inspectors) to be critical/not important when carrying out inspections'? 55 All respondents considered OELs to be essential and particularly valuable when carrying out inspections and enforcement.

It can be concluded that OELs provide an added value as a common, and well understood, objective measure to better facilitate the practical implementation of the general requirements for prevention and exposure reduction. There is a clear support for their continued use from key stakeholders.

For some sectors it may be that a given OEL is the best that can be achieved but in others sectors, or for specific uses, it is possible that much better exposure values can and should be achieved – this is taken into account in the minimum standard nature of OSH in general and OEL setting in particular.

Data presented in this report indicate wide differences in the Member States regarding the setting of OELs for the identified carcinogenic chemical agents 56 . Some Member States have already established legally enforceable OELs which are at the same value or lower than the value recommended by the Advisory Committee on Safety and Health at Work (ACSH). This demonstrates that unilateral national action is possible as regards setting an OEL for these chemical agents. However, as shown in Table 2 in Annex 6 there are also many cases where Member States have no OELs or ones which are less protective of worker health than the value recommended by ACSH.

Under such circumstances a minimum basis of protection against the risks arising from workers' exposure to these carcinogens cannot be ensured for all EU workers in all Member States by an action taken by Member States alone. Table 4 in Annex 6 provides an overview of the proportion of potentially exposed workers who lack such legal protection and this factor is taken into account in the analysis of impacts of introducing an OEL for each of the considered carcinogens. It follows that an action taken at the European Union level to achieve this objective appears to be necessary and in line with Article 5(3) of the Treaty on European Union.

Absent or too high OELs also provide a potential incentive for companies to locate their production facilities in Member States with the lower standards. This distortion of the internal market may be reduced by establishing clear minimum standard OELs at EU level.

Amending the CMD can only be done by action at EU level and after a two-stage consultation of the social partners (management and labour) in accordance with Article 154 TFEU.

2.3What are the EU instruments relevant to dealing with carcinogens at the workplace?

From the mid-1960s the European Union developed a legal framework addressing control of chemical risks under both social policy and internal market areas of EU policy. This framework places duties on authorities, businesses, and individuals to manage risks associated with the sale and use of chemicals. By the late 1990s a comprehensive and sophisticated system of interdependent Directives and Regulations had been established to protect humans and the environment from hazardous chemicals.

To protect workers from chemical carcinogens both employers and authorities must apply the same three stages necessary for all risk control: i) identify what harm might be caused ('hazard identification'); ii) identify the likelihood and severity of the harm actually being caused ('risk assessment'); and iii) where necessary put in place measures to mitigate harm ('risk management').

Directives setting out an agreed system for chemical hazard identification (i) were developed – augmented where necessary by additional provisions in topic specific Directives. The outcomes of this hazard identification are then applied in Directives and Regulations focussed on the key areas of chemical risk – in the case of worker protection from chemical carcinogens these were laid down in the CMD.

A subsequent process of chemicals risk assessment (ii) by European authorities built on this to identify measures needed to protect the environment or humans as consumers, members of the public, or workers. This risk assessment resulted in recommendations for regulatory risk management (iii) from a suite of available options.

The key EU risk management outcomes for the protection of workers from chemical carcinogens were CMD OELs (under social policy) and/or 'restrictions' on the placing on the market and/or use of certain chemicals (principally under internal market policy).

The REACH Regulation 57 , which was adopted in 2006, consolidated and evolved several parts of the EU chemical risk control system – principally those relating to risk assessment and internal market risk management measures. REACH further augmented these changes by establishing a new 'authorisation' risk management option.

Both CMD and the REACH Regulation therefore protect workers from risks associated with exposure to carcinogens. The two legal acts are complementary. CMD is a more targeted measure intended for protection of those employed by others from occupational carcinogens, while REACH applies more broadly to protect the health of humans and also the environment and also offers mechanisms which can be used to target risk controls to protect workers.

A chemical carcinogen may appear on both CMD Annex III and REACH Annex XIV without systematic conflict 58 . The OSH ʻFramework Directiveʼ – under which CMD is made – applies without prejudice to existing or future national and EU provisions which are more favourable to protection of the safety and health of workers at workʼ. REACH in turn applies without prejudice to worker protection legislation, including the CMD.

REACH is designed to integrate with the rest of the EU legal framework with which it co-exists – including CMD. As a key example of this, REACH does not include a mechanism to set EU OELs for carcinogens – these are properly established under social policy Directives where the social partners and co-legislators play a more prominent role.

REACH generates information on chemical agents which are 'placed on the market' (above one tonne per year) and used in the EU. It also includes the 'authorisation' and 'restriction' mechanisms – both of which may, in some cases, be available as risk management options for occupational carcinogens.

Carcinogens are one of several types of hazardous substance which can be added to Annex XIV to REACH, meaning they become subject to 'authorisation'. Once a given 'sunset' date is passed, these substances may only be placed on the EU market and/or used when an authorisation has been applied for and granted by the European Commission, taking into account the advice of the European Chemicals Agency, that either the risks associated with use are adequately controlled or that, if this is not possible, ongoing use of the substance is socioeconomically justified and there are no suitable alternatives.

The Commission may make exemptions from REACH authorisation under certain conditions, including where specific Union legislation already exists which imposes minimum requirements relating to the protection of human health for the use of a substance. So far no exemption has been made for substances used at the workplace.

Restrictions set conditions under which placing on the market and/or use of a chemical substance (or mixture or article containing it) may be permitted. These range from a near-total ban through to substance concentration specifications or established risk management measures to be observed by 'downstream users'.

Restrictions are a flexible, targeted chemicals risk management option carried forward from the pre-REACH EU chemicals framework. Restrictions and OELs have long co-existed as different possible outcomes from the EU chemicals risk assessment process – indeed four of the five existing binding OEL values established at EU level under the OSH Directives (asbestos, benzene, lead, and vinyl chloride) are already partnered with REACH restrictions.

As a result of their flexibility it is, conceptually, possible to establish an 'OEL-like' exposure limit via a REACH restriction – in 2013 the Netherlands REACH authorities proposed such a measure. In response, in November 2013, the lead Commission services for the REACH Regulation wrote to the European Chemicals Agency (ECHA), with the common Commission services' view that 'any proposal for adoption of an exposure limit value at the occupational premises should not be implemented under REACH and should only be set under the appropriate workers' protection legislation, which is specifically designed to establish and implement OELs'. In the same month the ACSH adopted an Opinion formally expressing the views of the Member States, employers, and workers' representatives that the OSH acquis is the more appropriate way to establish exposure limits for worker protection to chemicals. 59

Clear synergies between REACH and worker protection legislation can be seen – including in particular that REACH 'registration' should result in more information being available to inform chemicals risks assessment. REACH 'authorisation' also both establishes, for a given chemical agent, a clear and renewed pressure to substitute for safer alternatives, and can drive applicants to improve their worker protection risk assessments and controls. At the same time, adoption of EU OELs can be useful inputs for REACH risk characterisation.

The REACH status (authorisation and/or restriction) of the 13 chemical agents under consideration in this report is as follows:

Out of scope of REACH: hardwood dust, process-generated RCS.

On the Candidate List of 'substances of very high concern' potentially to be made subject to authorisation: acrylamide, hydrazine, RCF, o-toluidine.

Currently subject to authorisation: some chromium VI compounds.

Currently subject to restriction: acrylamide, restricted above a certain concentration for the placing on the market or use in grouting (REACH Annex XVII Entry 60); chromium (VI) compounds, restricted above a certain concentration for the placing on the market or use in cement, with certain exemption (REACH Annex XVII Entry 47); vinyl chloride, restricted (REACH Annex XVII Entry 2) for use as an aerosol propellant. All of the chemical agents in question, where placed on the market, are restricted for supply to the general public (REACH Annex XVII Entries 28 and 29).

A more detailed list of REACH status of the concerned chemical agents can be found in Annex 7 (14.2.1).

2.4How does this initiative relate to the envisaged review of the OSH acquis?

The CMD is part of an overall OSH acquis, comprising the Framework Directive 89/391/EEC and 23 individual directives. The Commission is currently finalising an ex-post evaluation of the acquis, covering the period 2007-2012 and 27 Member States. This evaluation consists of a wide assessment of the legislation, including in terms of benefits, research and new scientific knowledge. It also falls under the remit of the Commission's Regulatory Fitness and Performance Programme (REFIT), which aims to ensure that the EU regulatory framework is relevant, coherent, effective, efficient, provides EU added value and targets ensuring the avoidance of any unnecessary regulatory burden.

The results of the ex-post evaluation may lead to the further assessment of specific issues and potentially impact assessments covering possible initiatives to improve the operation of the regulatory framework. The Commission will be working throughout 2016 on the follow-up of the evaluation as regards collection of any necessary further evidence and justifications for the modernisation of OSH framework with the aim of increasing effectiveness and efficiency, reducing regulatory burden whenever possible without undermining the public interest objectives of the legislation.

Within the framework of the overall ex-post evaluation the CMD was, along with the other individual OSH Directives, subject to a specific assessment through an external evaluation study. The conclusions of this study refer mainly to broader issues such as the interface between the CMD and REACH, or the possibility to merge the CMD with the Chemical Agents Directive (CAD) 60 . The former is discussed at section 2.3 in this report. As for the latter, no clear conclusions can be drawn as views of stakeholders at national and EU level are quite polarised. There is however consensus on maintaining the material provisions of the two acts.

Secondly, concerning specifically OELs, experts discussions conducted in the framework of the evaluation confirmed that these are an important tool for chemical risk management at the workplace. It was indicated that the terminology used to describe different types of OELs (and corresponding procedures) should be aligned and guidance on the practical use of OELs should be developed.

The fact that the evaluation did not question the existence of OELs and pointed to the need to revise them and adopt new ones, together with the further evidence and analysis gathered in the framework of the impact assessment, support the conclusion that the two technical annexes of the Directive should be updated.

3What should be achieved?

3.1What are the general policy objectives? What are the more specific objectives?

The main general policy objective of this initiative is to ensure and maintain a high level of protection of worker's health and safety in the European Union.

The specific objectives are:

To ensure up-to-date protection from occupational exposure to chemical carcinogens in the European Union;

To increase the effectiveness of the EU framework by updating it on the basis of scientific expertise;

To ensure more clarity and create a better level playing field for economic operators.

3.2Are these objectives consistent with other EU policies and with the Charter for fundamental rights?

The objectives of the initiative are consistent with the fundamental rights as set out in the EU Charter of Fundamental Rights, in particular article 2 (Right to life) and article 31 (Right to fair and just working conditions which respect his/her health, safety and dignity).

Ensuring a safe and healthy work environment for over 217 mln workers in the EU is a strategic goal for the European Commission according to the recent Communication from the Commission on the EU Strategic Framework on Health and Safety at Work 2014 – 2020. 61 One of the main challenges identified in the EU OSH Strategy is to improve the prevention of work-related diseases by tackling existing, new and emerging risks.

Improving working conditions and preventing workers from suffering serious accidents and/or occupational diseases and promoting workers’ health throughout their working life, is a key principle in line with the ambition for a Triple A Social Europe rating set by the Juncker Commission. It also has a positive impact on productivity and competitiveness and is essential to promote longer working lives in line with the Europe 2020 strategy’s objectives for smart, sustainable and inclusive growth. 62

The objectives of this initiative are also in line with and complementary to the ILO Convention no. 170, concerning safety in the use of chemicals at work. The Convention, adopted so far by six EU Member States 63 , establishes minimum universal standards on OSH to be observed worldwide. CMD and CAD reflect these standards and go further, establishing more stringent and protective conditions that EU Member States are in a position to implement. They also provide a stronger motive for MS to enact national measures as they are legally bound to do so. ILO 170 is intended to complement and explicitly envisages, inter alia, (Article 13), that 'Employers shall: (a) ensure that workers are not exposed to chemicals to an extent which exceeds exposure limits or other exposure criteria for the evaluation and control of the working environment established by the competent authority, or by a body approved or recognised by the competent authority, in accordance with national or international standards (…)'.

4What are the various options to achieve the objectives?

4.1Complementary measures

During discussions in the ACSH 64 the need for further guidance was called for, in particular on:

the minimisation obligation under Article 5 of the CMD Directive;

methodologies for the development of OELs for carcinogens;

measurement methodologies regarding exposure to hardwood dust,  taking account of mixed exposure to both hard- and softwood dust;

how to take into account in the cost-benefit analysis the EU Charter of Fundamental Rights, in particular Article 1 (Human dignity), Article 2 (Right to life) and Article 3 (Right to integrity of the person).

In the same vein, one of the conclusions of the Social Partners consultation was that there was a need for effective implementation of training and information requirements, which are a key aspect of the prevention policy. Workers called the Commission to set up a strategy to improve coordination and sharing of information at EU level. Employers indicated that there was an added value in the preparation of guidance documents with recommendations on workers protection against carcinogens and mutagens exposure.

Most of these actions are ongoing or will be undertaken as complementary measures to this initiative. These are not alternative options to updating the CMD but are rather part of the baseline and will further reinforce potential positive effects of the considered options. Non-binding guidance cannot address the issues identified in the problem definition; only clear listing of chemical agents under the CMD and setting OELs will introduce the legal certainty needed for employers, workers and enforcers with regard to managing the exposure to carcinogenic chemical agents in the workplace.

4.2Discarded options

Several other options which could be considered as alternative mechanisms to control and limit exposures to carcinogens in the workplace have been discarded as they were considered disproportionate or less effective in reaching the objectives of this initiative.

A. Banning the use of the carcinogenic chemical agents

As explained above, for most carcinogens even a very low OEL does not completely eliminate the risk of triggering a cancer. In that sense the risk could only be reduced to zero by eliminating the presence/use of the substance in the workplace.

Indeed, substitution is the first option in the hierarchy of risk management measures under the CMD that an employer needs to consider. This means that if it were technically feasible, employers should have already replaced use of the concerned chemical agents with safer alternatives.

Wherever substitution is a suitable alternative for use of the 13 chemical agents in question then the CMD already requires this, regardless of the existence or otherwise of an OEL. As this legal standard already establishes that these carcinogens should not be used in the workplace where alternatives are available, establishing a more strict prohibition in the form of a ban would constitute a disproportionate measure with a strong negative impact on businesses. Moreover, in the case of process-generated substances such as hardwood dust or RCS, banning is not only a counterintuitive option but it would also result in operational challenges which could result in closure of otherwise viable and compliant businesses – potentially across whole industry sectors.

B. Self-regulation (industry voluntary agreements)

The NEPSi Agreement, including the practical guidelines to risk management, contributes to the overall protection of workers health in the affected employment sectors.

NEPSi was established in part as a self-regulatory alternative to measures reflected in the current proposal. Stakeholders implementing NEPSi have therefore indicated concern that the current proposal may pose a challenge to industry participation in self-regulation initiatives in the future. However, in addition to any recommendations of the NEPSi evaluation report, which shall be issued by the European Commission in Q2 2016, it should be noted that the effectiveness of this self-regulation initiative is hindered by the following:

it does not apply to all sectors where exposure to RCS occurs, in particular not to the construction sector, which, has the largest burden of occupational cancer 65 ;

it is not binding and cannot therefore be enforced by national authorities.

Consequently, the NEPSi agreement and any similar future initiatives would be an important complementary measure but could not be as effective as OELs in setting and enforcing the same level of minimum protection across the EU and in all sectors of industry.

C. Sector or industry specific OELs

It could be envisaged that, where justified, sector, industry or use specific OELs could be set or derogations could be allowed for specific sectors, industries or uses for a defined time-period or subject to specific conditions.

As a result of such an approach we would have different minimum standards of worker protection across the EU for the same type of work and the same chemical agent, which is not permissible in the context of binding OELs as established under the CMD.

Further the process by which the tripartite ACSH agree their Opinion on appropriate EU OELs under CMD includes extensive consideration of socioeconomic and technical feasibility factors – including sector, industry and use-specific concerns. As a result these concerns are already reflected in the value supported by the ACSH and taken forward in the current proposal.

This option, therefore, had to be discarded as the current legal framework does not provide a legal basis for such provisions.

Where an EU OEL is established, Member States and employers retain the right to implement more protective values than the proposed minimum standards, also taking into account the general CMD duty to substitute, eliminate or otherwise minimise exposures.

D. Providing industry-specific scientific information without setting OELs

Another option could be for the Commission to collect and provide industry-specific scientific information to support employers in complying with the CMD obligations.

Apart from the practical difficulties related to collection of relevant data for the multitude of sectors concerned, it is considered that this option would not be effective in achieving the objectives of the initiative for the following reasons:

the way the information is used by employers would not be enforceable by surveillance authorities;

such an option would not fit with the overarching legal framework of the CMD, which provides for general exposure management requirements to be specifically supplemented by EU-wide minimum standard OELs:

in some cases, extensive industry- and chemical agent- specific information and guidance already exists and should be taken into account by employers during risk assessments – but this has not demonstrably addressed harmful exposures at EU level as identified in the IOM study.

E. Market-based instruments

Market-based instruments such as subsidies, tax breaks or reductions of social insurance contributions, are sometimes used by Member States to incentivise business to comply with health and safety rules. Such instruments can effectively support compliance with exposure limits. However, to be applied effectively in this context, such mechanisms would need to be linked (directly or indirectly) with the actual levels of exposure at firm level. These are generally not observed (as it also emerges from the analysis presented in Section 5 below), and improved data collection would likely result in being extremely costly and cumbersome. It should also be noted that these instruments remain in the hands of Member States and the extent to which they are used vary significantly 66 . This option alone would therefore not be effective in ensuring the same level of minimum protection across the EU.

F. Regulation under other EU instruments (REACH)

As detailed in section 2.3, other EU regulatory options exist for managing risks to workers associated with exposure to some chemical carcinogens – most notably REACH 'authorisation' or 'restrictions'.

As previously indicated, REACH is a relevant regulatory instrument for protection of workers from hazardous chemicals, including in particular chemical carcinogens. However, in the case of the present proposal CMD is the more appropriate regulatory instrument for the following reasons:

Hardwood dust and respirable crystalline silica, which are process generated in the workplace, are outside scope of REACH.

CMD covers worker exposure to carcinogenic agents released by any work activity, whether produced intentionally or not, and whether available on the market or not.

REACH sets directly-acting harmonised standards from which Member States cannot deviate except in exceptional (and possibly time limited) circumstances. CMD sets 'minimum standards' which allow Member States to maintain or implement more protective measures – which is appropriate in the interest of worker protection. OELs are an example of this.

REACH places the onus of risk assessment on the supply chain, and is 'chemical agent specific'. CMD risk assessment is more likely to be workplace- and process-specific and should take into account aggregated exposure of workers to all occupational carcinogens. From the point of view of preventing exposure to carcinogens CMD offers therefore a more holistic approach to workplace risks.

REACH authorisation is a risk management measure covering all risks arising from given intrinsic properties of a substance, including the risks for workers. In the absence of explicit derogations it applies to all placing on the market for a use and use of subject chemical substances and so is a less targeted measure than a CMD OEL, which applies in the long-established OSH regulatory context and relates specifically to the workplace. REACH authorisation can complement CMD, in particular by strengthening the substitution principle and its full implementation, as well as driving toward additional risk management.

It should also be noted that, as it is based on social policy provisions of the TFEU 67 , CMD reflects the role foreseen there for the social partners in establishing standards for worker protection.

OELs are an important part of CMD and of the wider occupational safety and health approach to managing chemical risks. REACH, on the other hand, is not intended to set OELs. The concerned Commission services, Member States, and the social partners have all expressed their view that OSH Directives are the appropriate EU legislative framework to establish limit values for the protection of workers.

A comparison table summarising key characteristics of the CMD and REACH approaches to risk assessment and control of occupational carcinogens can be found in Annex 7 (14.2.3).

From the point of view of enterprises, it is important to note that compliance costs associated with workplace legislation (such as protective measures and equipment) directly contribute to better risk management, and that administrative costs triggered by workplace legislation are relatively small.

In 2015 an ongoing collaboration of 40 European sectoral and cross-sectoral associations, nine national associations, and several corporations, all with interests in manufacturing, importing, and using chemical substances began actively engagement with the Commission services in order to ensure more efficient risk management measures that can combine OSH OELs in addition to REACH risk management options as the more proportionate measure where chemical risks identified during risk assessment relate principally to worker protection. 68 The Commission services concerned are working with this 'Cross Industry Initiative' 69 and to discuss the relevance of their proposals in the operation of EU chemicals policies regarding worker protection.

Targeted CMD OELs are appropriate for protecting workers from occupational exposure to the chemical carcinogens subject to the present proposal. The synergies in the relationship between CMD and REACH provide further opportunities for regulatory measures as appropriate.

It should be noted that this impact assessment does not aim to assess whether REACH risk management measures are proportionate and effective. The present proposal does not preclude relevant REACH measures from being proposed where this is justified, and in some cases a combination of CMD OELs and other regulatory (or non-regulatory) risk management measures, including under REACH, may be justified.

4.3Options retained for consideration

Identification process

Reviewing or setting new OELs under CMD follows a specific procedure involving seeking scientific advice and consulting the ACSH. Article 16 of the CMD, which states that scientific/technical data should be included in the basis on which OELs are set, does not determine which scientific body should be the source of such data. In practice, the Commission and the ACSH principally seek the advice of SCOEL, but can also refer to scientific information sourced elsewhere as long as the data is adequately robust and is in the public domain (e.g. IARC monographs or conclusions of national OEL-setting science committees). 70

Figure 3.    Simple representation of EU OEL setting procedure

The SCOEL is an independent scientific committee, established by a Commission Decision and composed of 21 experts appointed in their personal capacity as leading experts in fields relevant for protection of workers from risks associated with workplace exposure to hazardous chemicals. 71

SCOEL carry out scientific evaluation at EU level and as a result publish a single evaluation document (previously a "SCOEL SUM", more recently a Recommendation or Opinion) for hazardous chemicals where there is priority concern for worker protection. SCOEL procedures for the adoption of a Recommendation by SCOEL include an external consultation with identified contact points in all of the Member States; this ensures scrutiny of the scientific evidence and methodological approach used by SCOEL and ensures transparency of the process.

The Advisory Committee on safety and health at work (ACSH) is a tripartite body set up in 2003 by a Council Decision (2003/C 218/01) to streamline the consultation process in the field of occupational safety and health and rationalise the bodies created in this area by previous Council Decisions. The Committee's remit is to assist the European Commission in the preparation, implementation and evaluation of activities in the fields of safety and health at work. The Committee is composed of three full members per Member State, representing national governments, trade unions and employers' organisations, also organised in three separate interest groups within the Committee. The ACSH is supported by working parties of experts on given topics of interest – also tripartite but with smaller selected expert membership. The Working Party on Chemicals (WPC) serves the ACSH according to a mandate agreed by the plenary Committee, and in particular undertakes detailed technical and policy negotiation of EU limit values as well as broader chemicals policy support for the ACSH and Commission.

The ACSH discusses adopted SCOEL Recommendations (and/or other appropriate scientific evidence) and adopts a formal Opinion on what, in practice, is considered to be achievable by employers whilst ensuring that workers' health is adequately protected.

In the case of the carcinogens considered in this report, SCOEL concluded recommendations on all but two. 72 , 73 Table 1 in Annex 5 sets out the status of relevant adopted SCOEL recommendations.

While the aim of ensuring the protection of the health of workers is maintained, binding OELs set under CMD are usually based on factors beyond the independent scientific advice, because they must also reflect other factors such as 'feasibility' and taking into account the views of the social partners.

Amending the CMD can therefore only be proposed after a two-stage consultation of the social partners (management and labour) in accordance with Article 154 TFEU. This consultation took place in 2004 and 2007 and addressed the following elements: the possibility of extending the scope of the Directive to include reprotoxic substances, the revision of existing OELs and the establishment of new ones for more substances, need to develop a EU-wide methodology for carcinogens and mutagens OELs setting, and the need to improve the training and information requirements for workers.

Between 2009 and 2011 an external contractor evaluated, on behalf of the Commission, health, socio-economic and environmental aspects of the proposed amendments to CMD in order to inform impact assessment according to the regulatory procedures in place at that time. Between 2010 and 2013 the Working Party on Chemicals of the ACSH undertook detailed discussion on these issues in an increased work schedule, aiming to secure stakeholder engagement and agreement on values to propose for ACSH adoption.

The consultation process resulted, amongst others, in the support of the following:

to bring a limited number of so-called process generated substances (PGSs) under the scope of the Directive by including them in Annex I,

to revise existing OELs in Annex III in the light of most recent scientific data, and

to add additional OELs for a limited number of substances in Annex III where available information, including scientific and technical data supports this.

The ACSH, in an Opinion adopted in December 2012, and in supplementary Opinions adopted in May 2013 and November 2013, confirmed (with a few dissenting opinions) the OELs initially developed by the SCOEL (where relevant) and approved inclusion of these into CMD.

The three adopted ACSH Opinions include, where necessary, specific comments from the interest groups (the social partners and Member States) which broadly reflect the principal points maintained by each interest group throughout discussions of the Working Party on Chemicals (WPC). In many cases there are no specific comments as there was a consensus view of the three interest groups. As such, the final ACSH Opinions should be taken as representative of the views of stakeholder groups represented.

In addition, the amendment of the CMD is routinely discussed at meetings of the tri-partite WPC. Following adoption of the three ACSH Opinions this discussion has primarily focussed on the state of play and timing of the next steps of the planned amendment of the Directive. They have not expressed additional views on the content of the adopted opinions.

In principle, revision of the Annexes should be conducted on regular basis, as soon as new scientific data is available and the technical feasibility of introducing new or revised OELs has been established. The duration of the current preparatory work reflects that it is a first preparation of a proposal for a relatively large number of new/revised limit values. Appropriate modes of cooperation between all the concerned actors as well as working methods within the advisory bodies had to be developed.

It is expected that the procedures and working methods described above and established in the course of the current exercise will give a good basis to complete preparatory steps leading to eventual future revisions of the CMD in a much shorter time. The fact that a second proposal for the introduction of OELs for additional chemical agents may be presented in the near future should be seen precisely as a step of a continuous, regular process of updating of the CMD, where necessary.

Identified options

The Table 3 below summarises options for different OELs for each of the 13 chemical agents.

A baseline scenario of no further EU action is Option 1 for each chemical agent represented in this initiative.

As explained above, for each of the chemical agents scientific and technical data has been considered and discussions at the ACSH have taken place, resulting in values to be proposed into co-decision as an OEL. Directly adopting the values agreed by the ACSH forms Option 2 for each chemical agent.

Where appropriate and depending on specific characteristics of the agents, flanking options to either propose a OEL which, compared with the ACSH value, is lower (theoretically more protective of worker health) or higher (theoretically less protective of worker health) are also presented as Option 3 and/or 4 respectively for each chemical agent.

These flanking values are drawn from the IOM Study, for which they were established by preference: i) from a SCOEL Recommendation if available; ii) as values reflecting available data (for example taking account of existing national OELs); or iii) on the basis of recommendations from the contractor (for example taking into account non-EU OELs). Where available data do not support setting a lower or higher OEL than the ACSH value, these options are discounted. Further detailed information regarding the source of these additional options may be found in the relevant IOM Study reports for each of the chemical agents in question.

In the case of RCS, options 2, 3 and 4 include the possibility of inclusion in the Annex I of CMD along with OEL proposals under Annex III.

In addition, for each chemical agent where SCOEL has identified significant risk of adverse systemic affects resulting from dermal uptake a 'skin notation' ('sk.') is indicated alongside the numerical OEL.

Table 3.    Options matrix

The next section presents an analysis of impacts of the different policy options for each chemical agent. For reasons of space the analysis for each chemical agent is contained in a table which is the basis of the comparison ratings presented 74 .

Unless otherwise specified all data in the agent-specific analysis comes from the IOM study, with reference periods as specified in that study.

5What are the impacts of the different policy options and how do they compare?

Before looking at the comparison between different policy options, it should firstly be noted that the standards of legal control established in the CMD are strict for all in-scope carcinogens, whether or not entries in Annex I (process generated substances) or Annex III (OELs) have been established or amended.

The proposed amendments are thus intended to enhance worker protection by improving clarity for employers and enforcers. The inclusion of chemical agents in Annex I (the list of 'process-generated substances') confirms that they fall in scope of the Directive control provisions. OELs set out in Annex III establish clear compliance benchmarks for exposure control. These considerations will be inherent in any investment decision, profit projection etc. and should be based on as much legal clarity as possible.

Concretely, comparing the different policy options requires estimating the effects of introducing new or more stringent OELs for the chemical agents under consideration. The methodology followed can be summarised as follows:

In general, the introduction of an OEL is expected to determine a reduction in the occupational exposure to the carcinogen concerned (as full compliance with the proposed new OEL is assumed). The extent of such reduction depends on the current levels of exposure, as well as on the projected future levels of exposure in the absence of the proposed OEL ("Baseline scenario", corresponding to Option 1).

For a given reduction in exposure levels, it is then necessary to estimate the expected decrease in the incidence of cancer cases over a given timeframe attributable to the carcinogen in question. This requires estimates of the risk of carcinogenicity, which can be derived from the existing toxicological and epidemiological literature, as well as information about the actual level of worker exposure (numbers, level, duration and frequency of exposure).

The health benefits of avoided cancer registrations and deaths can then be expressed in monetary terms by applying standard evaluation methods (value of life years lost, cost of illness, willingness to pay to avoid cancer). These monetised health benefits can in turn be compared to the expected monetary costs that would have to be incurred in order to comply with the proposed OEL. This methodology is further explained in Annex 4.

A number of issues need to be taken into consideration in order to better understand the figures and conclusions presented in the following sections, which are mostly derived from the results of the IOM study. As set out above, the CMD requires employers to take measures to protect workers from occupational use of carcinogens. All stakeholders as consulted in the ACSH are in favour of the proposed legislative initiative albeit with some differences regarding appropriate OEL values.

In terms of cost and benefit projection, this renders the updating of the CMD particularly complex; independent of further clarity under the CMD, employers will include more or less stringent protection of workers in their facilities, e.g. costly industrial ventilation combined with personal protective gear. Furthermore, Member States with substantial production involving a certain carcinogenic chemical agent are likely to introduce national OELs irrespective of EU action.

In order to take proper account of the cancer latency period (which the IOM study assumes to be of 10-50 years for solid tumours and of 0-20 years for haematopoietic neoplasms), the future cancer burden is estimated over a 60 years period.

Looking at the methodological challenges in more detail:

First of all, for most chemical agents under consideration, data on the number of workers exposed is scarce and unreliable (especially for some sectors and/or for some Member States), and data on the current exposure levels across EU Member States is generally not available. Member States record statistics relating to cancer in different ways which cannot be readily aggregated. 75 Where exposure data is available, its use as an evidence base for regulatory decision-making is often confounded by the sensitive and sometimes confidential nature of the information, and the potential for source bias.

For many of the carcinogens, the baseline scenario taken from the IOM study foresees a constant reduction in average exposure levels (e.g. of 7% annually). This projection of future exposure levels is obtained by extrapolating past declining trends in average exposure levels. However, for some substances this (large) declining trend assumption is contested by other studies. 76 In addition, even when declining trends in average exposure levels are observed, it may be misleading to regard these as exogenous. Recent reductions in exposure may have been precisely the result of OELs having been introduced or as an anticipation of those changes. With respect to the cost and benefit analysis, therefore, the projected decline in average exposure levels under the baseline scenario may bias the estimated health impacts downward.

The available epidemiologic evidence is scarce and not always sufficiently robust, inevitably affecting the reliability of the derived estimates for the number of cancer registrations and deaths. Among the factors contributing to the scarcity of reliable data are the complexity of cancer development and also of workplace exposures. Different carcinogens can, for example, result in the same type of cancer (e.g. lung cancer), and occupational exposure to hazardous agents is characterised by simultaneous exposure to multiple chemical agents. It can therefore be difficult to establish a causal relationship between cancer cases and exposure to a specific carcinogen.

The cost-benefit analysis underestimates benefits as only the cancer-related health impact is considered. Exposure to the chemical agents under consideration is also associated with additional non-cancer health effects which can induce further health costs (such as for example neurotoxicity, severe skin damage, respiratory diseases or renal toxicity).

When a declining trend exposure is considered under the baseline scenario, it would be incorrect to factor in among the costs of compliance with OELs based on the proposed OEL the full value of the investment required to reduce exposure: such investment would have occurred in any case also under the baseline scenario (in order to justify the decline in exposure), but possibly only later or more gradually over time. As a result the cost estimates of introducing an OEL reported in the IOM study would be overestimated.

Finally, to allow for a comparison between the monetised health benefits and compliance costs, the net present values of the streams of costs and benefits over the 60-year period under consideration are computed. The values originally reported in the IOM study, based on a constant discount rate of 4%, have been recalculated applying a declining discount rate (4% for the first 20 years, 3% thereafter) in line with the most recent better regulation guidelines. Still, benefits estimates are disadvantaged as discounting reduces much more the present value of impacts taking place in the longer term (typically health benefits) than those happening at the beginning of the period (typically compliance costs).

To allow for a better comparison between net health benefits and net compliance costs, as it was not possible to obtain new estimates of health benefits assuming a constant level of exposure under the baseline scenario for all chemical agents, the costs presented in the IA report are indicative estimates of the actual additional costs of compliance assuming some delay (e.g. of 10-20 years) in the realisation of the investment needed to achieve a certain level of compliance.

The lack of reliable exposure data on both the numbers of workers exposed and on the levels of exposure is recognized. To address this data gap the Commission initiated a study in 2013 77 . The outcome of this work is expected to contribute to a better definition of the baseline situation for possible future initiatives on developing OELs for other priority occupational carcinogens.

Analysis of impacted Member States and proportionality

For each of the 13 carcinogens, an analysis of the need for and benefit of action at the EU level is presented by chemical agent in the following sub-sections (and in Annex 9).

The IOM study identifies the number of EU workers exposed to each chemical agent and, in most cases, also identifies exposed working populations by Member States. An illustration of the OELs currently in place at national level is provided in Annex 9, as well as the estimated number of workers potentially exposed by Member State. Similar information is also summarised in Table 4 in Annex 6, which identifies the population of workers occupationally exposed to each chemical agent and compares this to the overall EU population of exposed workers, resulting in the percentage of workers in the EU for whom legal protection would be improved by adoption at EU level of an OEL under CMD.

This analysis bears the condition that reliable exposure data is scarce. In particular it should be noted that it is not possible to identify the proportion of workers in each Member State who may be employed in tasks where closed system or effective exposure controls are already either eliminating or reducing exposures below the level represented by the OELs recommended by ACSH. The probable over-estimation of exposed populations resulting from this data scarcity is likely to be mitigated by the converse inability to identify specific populations of workers who may experience exceptionally high exposures.

5.11,2 Epoxypropane

This chemical agent is associated with lymphopoietic cancer, haematopoietic cancer and increased leukaemia risk.

The major use of this chemical agent is to make 1,2 epoxypropane polymers that are used in the manufacture of polyurethane foams. Its second most important use is in the production of propylene glycol. About 5% of all 1,2 Epoxypropane production is used in a diverse range of applications such as the manufacture of surfactants and as a stabiliser for dichloromethane.

The productive capacity within the EU is 2.75 mln tonnes per year and it is produced in 14 facilities in eight Member States (BE, FR, DE, NL, PL, SK, ES and RO). There are estimated to be 150-300 user facilities across the EU.

Between 35-70 EU workers are estimated to be exposed to 1,2 epoxypropane during its manufacture and between 450-1,500 EU workers in the chemical industry.

Table 4.    1,2 Epoxypropane – Types of impacts

Estimates of exposure levels were based on samples obtained during manufacturing and downstream use in 7 facilities. According to these estimates most companies already comply with a level of 1 ppm. For those that do not yet comply, the additional costs are expected to be very low. It is therefore assumed that the introduction of an OEL of 1 ppm would not imply any important costs associated with compliance, neither major social, macro-economic or significant environmental impact.

During the ACSH discussions employers' representatives noted that in routine and manufacturing practices workers' exposures are already well below an OEL of 2 ppm and that a level of 1 ppm during maintenance and loading operations could be ensured through breathing personal protective equipment.

Introduction of an OEL of 1 ppm can be expected to be more effective in reducing exposure to 1,2 epoxypropane compared to the baseline, however the monetised health benefits are minor.

Table 5.    1,2 Epoxypropane – Comparison of options

The scientific advice and the opinion of the social partners and Member States in the ACSH support option 2, and the CMD further establishes an expectation that OELs be set where it is possible to do so.

Impact on MS and proportionality

In the case of 1,2 epoxypropane, 26 Member States have so far opted either to not to set a limit or to set one which is less protective of worker health than the value recommended by ACSH. Figure 4 in Annex 9 illustrates the ranges of existing national OELs compared to Option 2.

Member States where the production of this chemical agent is concentrated (such as BE, DE, ES, FR, NL, PL, RO and SK)) have no OEL or OELs above the proposed limit value. Introducing an OEL would bring a greater clarity for economic operators across the EU and ensure that lack of an OEL or a less stringent OEL does not act as an incentive for business in decisions concerning the plant location.

The introduction of an OEL of 1 ppm would require changes to the national frameworks of most MS. Due to the generally low number of exposed workers across the EU, no estimates were made about the numbers of exposed workers in those 26 MS. Even if current exposure levels in the EU are estimated to be well below 2 ppm, a minimum basis of protection against the risks arising from workers' exposure to these carcinogens cannot be ensured under the baseline for all EU workers – it follows that an action taken at the European Union level to achieve this objective is proportionate.

5.21,3 Butadiene

This chemical agent is associated with an increased risk of lymphohaematopoietic cancer, mainly lymphosarcoma.

Most 1,3 butadiene is polymerized at a relatively small number of sites in Europe to form synthetic rubber. It is also used as a chemical intermediate in the production of neoprene for automotive and industrial rubber goods, in the production of methylmethacrylate-butadiene-styrene polymer, which is used as a PVC reinforcing agent, and in the production of adiponitrile (a nylon precursor). The production capacity in the EU is estimated to be 2.9 mln tonnes.

There were nine plant sites producing emulsion of SBR in July 2010 (IOM report), four producing solution of SBR, seven producing polybutadiene or butadiene rubber and six producing nitrile butadiene rubber. Some of the sites produce two or more of these elastomers at the same location. All companies affected by the proposed OEL value of 1ppm would be SMEs, and the great majority of them (about 90%) would be microenterprises (less than 10 employees).

About 27,600 workers in the EU are estimated to be potentially exposed to this chemical agent.

Table 6.    1,3 Butadiene – Types of impacts

The exposure levels were estimated across the industries of manufacture of refined petroleum products and manufacture of rubber. The estimates of exposure levels in the manufacture of refined petroleum products (NACE code 23) were based on different studies covering Finland, 13 EU countries, EU and UK. The estimates of exposure levels in the rubber industry (NACE code 251) were based on studies in 27 EU plants, in the Netherlands, Finland and Czech Republic.

Given the estimated decreasing exposure levels to 1,3 Butadiene in the baseline scenario and the fact that it is not possible to identify an exposure level at which there is no risk of lymphohaematopoietic cancer, none of the assessed options would allow a reduction in the numbers of attributable deaths or cancer registrations in the 2010-2069 period. However, introducing an OEL would be effective in ensuring a greater clarity for economic operators across the EU. Option 4 is the least effective as it is estimated to slightly increase the number of years of life lost (YLL) over the 2010-2069 period as compared to the baseline scenario, while YLL are expected to slightly decrease under options 2 and 3. Of those two, option 2 brings about the least additional costs for enterprises.

Employers' representatives at the ACSH accepted the option of an OEL of 1 ppm as feasible while expressing concerns about potential economic impacts of lowering an OEL below that level.

Table 7.    1,3 Butadiene – Comparison of options

The opinion of the social partners and Member States in the ACSH supports option 2, and the CMD further establishes an expectation that OELs be set where it is possible to do so. Introducing an OEL would also be effective in introducing a greater clarity for economic operators across the EU. An OEL of 1 ppm is also set in the US.

Impact on Member States and proportionality

In the case of 1,3 Butadiene 23 Member States have no OEL or one that is less stringent than the ACSH recommended 1 ppm level. Figure 5 in Annex 9 illustrates the ranges of existing national OELs compared to Option 2. The following Figure 6 shows distribution of exposed workers across the Member States.

The Member States where the production or use of this substance is concentrated (such as DE, FR, UK, ES, PL, RO) have no OEL or OELs above the proposed limit value. Introducing an OEL would bring a greater clarity for economic operators across the EU and ensure that lack of an OEL or a less stringent OEL does not act as an incentive for business in decisions concerning the plant location.

The introduction of an OEL of 1 ppm would require changes for a substantial number of Member States. It is estimated that approximately 93% of exposed workers are located in those 23 Member States. Even if current exposure levels in the EU are estimated to be well below 2 ppm, a minimum basis of protection against the risks arising from workers' exposure to these carcinogens cannot be ensured for all EU workers under the baseline scenario – it follows that an action taken at the European Union level to achieve this objective could be justified.

5.32 Nitropropane

According to animal toxicity studies 2 nitropropane may cause liver tumours in humans.

Occupational exposures to this chemical agent occur primarily in its production and use as a solvent in inks, adhesives, paints and coatings. It is produced in relatively low volumes. According to the IOM report there was only one plant in Germany which produced 2-nitropropane. In downstream uses exposures were considered only likely to occur in the manufacture of aircraft and spacecraft (NACE code 35.3) and possibly at very low levels in the recycling of non-metal waste and scrap (NACE code 37.2).

No information on EU total production is available.

About 51,400 EU workers are estimated to be exposed to 2 nitropropane.

Table 8.    2 Nitropropane – Types of impacts

Estimates of exposure were assumed to decrease by 7% per annum (reference: Creely et al. 2007). The estimates for 2010 exposures were based on worst-case measurements obtained/reported in 1984 from the industries of manufacture of 2-nitropropane and of automotive manufacturing.

Given the estimated low exposure levels (it is likely that no worker in the EU is exposed in excess of the proposed limit value of 5 ppm), option 2 is expected to have no health impact. For the same reason, establishing this OEL at EU level would not impose additional costs on firms, while remaining a measure in terms of promoting a levelled playing field for companies across the EU.

Table 9.    2 Nitropropane – Comparison of options

The opinion of the social partners and Member States in the ACSH supports option 2, and the CMD further establishes an expectation that OELs be set where it is possible to do so. Introducing an OEL would also be effective in introducing a greater clarity for economic operators across the EU.

Impact on Member States and proportionality

In the case of 2 Nitropropane 14 Member States have so far opted either to not to set a limit or to set one which is less protective of worker health than the value recommended by ACSH. The Figure 7 in Annex 9 illustrates the ranges of existing national OELs compared to Option 2. The following Figure 8 shows distribution of exposed workers across the Member States.

Some of the Member States where production or use of this chemical agent is concentrated and therefore where the numbers of exposed workers are the highest (such as DE, FR, PL or IT) have no OEL or OELs above the proposed limit value. UK and ES have OELs, which are slightly higher than Option 2. Introducing an OEL would bring a greater clarity for economic operators across the EU and ensure that lack of an OEL or a less stringent OEL does not act as an incentive for business in decisions concerning the plant location. 81

Introduction of an OEL of 5 ppm would require changes for a substantial number of MS. It is estimated that approximately 62% of exposed workers are located in those 14 Member States. Even if current exposure levels in the EU are estimated to be below 5 ppm, a minimum basis of protection against the risks arising from workers' exposure to these carcinogens cannot be ensured for all EU workers under the baseline scenario. It follows that an action taken at the European Union level could be justified.

5.4Acrylamide

Acrylamide may cause pancreatic cancer, and is also a skin irritant and may be a tumour initiator in the skin, potentially increasing risk for skin cancer.

99% of acrylamide in the EU is used in the production of polyacrylamide 82 . The main uses of polyacrylamide are in wastewater treatment, paper and pulp processing and mineral processing. Three companies are reported as producing acrylamide within the EU (in UK, Germany and the Netherlands). There were also thought to be firms in Spain, Finland and Italy who either supply (from imports) or produce acrylamide. The total plant capacity within the EU is estimated at between 80,000-150,000 tonnes per annum. The IOM report found seven producers of polyacrylamide within the EU (two of which also produced acrylamide), as well as a number of smaller producers throughout the EU.

54,000 EU workers are estimated to be exposed to acrylamide.

It should be noted that the global market for polyacrylamide has been growing and is expected to continue expanding. 83

Table 10.    Acrylamide – Types of impacts

Estimates of exposure levels are based on a study by Bull et al. (2005), who monitored personal inhalation and dermal exposure at a UK acrylamide and polyacrylamide manufacturing facility, based on measurements between 1992 and 1995. The results of this study were extrapolated to the whole acrylamide and polyacrylamide manufacturing industry in the baseline scenario even if the 10.5% annual decline rate appears optimistic.

On this basis it is assumed that industry is already complying with an OEL of 0.03 mg/m3, and thus no major health impact is to be expected from any of the two options. No economic cost is expected for the implementation of any of both options, although, if any, this would be even lower for an OEL of 0.1 mg/m3 compared to a more stringent OEL of 0.03 mg/m3.

Where there is the possibility of a significant uptake via dermal exposure SCOEL recommend that any OEL be accompanied by a 'skin notation' (in the case of acrylamide ACSH did not comment on this aspect of the SCOEL Recommendation). If adopted and accordingly transposed by Member States, employers are required under the national transposing legislation to take this into account in selecting appropriate risk management measures to protect workers.

Once the need for managing exposure has been established the only risk management measure available in practical terms is to avoid skin contact. Acrylamide has a harmonised classification as a skin irritant (category 2) and skin sensitiser (category 1). These hazards would normally result in employers taking steps to avoid skin contact as a part of routine OSH risk control. Adoption of a 'skin notation' should therefore result in no additional cost for employers.

Table 11.    Acrylamide – Comparison of options

The opinion of the social partners and Member States in the ACSH supports option 2, and the CMD further establishes an expectation that OELs be set where it is possible to do so. Introducing an OEL would also be effective in introducing a greater clarity for economic operators across the EU.

Impact on Member States and proportionality

In the case of acrylamide 13 Member States have so far opted either to not to set a limit or to set one which is less protective of worker health than the value recommended by ACSH. Figure 10 in Annex 9 illustrates the ranges of existing national OELs compared to Option 2. The following Figure 11 shows distribution of exposed workers across the Member States.

The three Member States where the production and/or use of this chemical agent is concentrated and the numbers of exposed workers are the highest (DE, UK or FR) have no OEL or OELs above the proposed limit value of 0.1 mg/m3. Introducing an OEL would bring a greater clarity for economic operators across the EU and ensure that lack of an OEL or a less stringent OEL does not act as an incentive for business in decisions concerning the plant location.

Introduction of an OEL of 0.1 mg/m3 would require changes for a substantial number of Member States. It is estimated that approximately 64% of exposed workers are located in those 13 Member States. Even if current exposure levels in the EU are estimated to be below 0.03 mg/m3, a minimum basis of protection against the risks arising from workers' exposure to these carcinogens cannot be ensured for all EU workers under the baseline. It follows that an action taken at the European Union level could be justified.

5.5Hardwood dust

Hardwood dust is listed in the CMD but not classified according to the EU CLP regulation (because it is a PGS); however it is classified as a Group 1 carcinogen ("carcinogenic to humans") by the International Agency for research on cancer. Hardwood dust may cause sinonasal and nasopharyngeal cancers. In addition, hardwood dust may cause non-malignant respiratory health problems, including occupational asthma.

Exposure to hardwood dust occurs mainly in the wood working industry, furniture manufacturing and construction sectors. Over three million EU workers are potentially exposed in over 340 000 companies, mostly SMEs, with a production value of around €230 bln/year.

The estimates of the prevalence and level of exposure to hardwood dust were based on the results of the 2001-2006 WOODEX project and on the timber statistics UNECE. The WOODEX project aimed at estimating occupational exposure to inhalable wood dust by country, industry, level of exposure and type of wood dust in 25 Member States. National labour force statistics, a country questionnaire (in 15 Member States, EU-15), a company survey (in Finland, France, Germany and Spain), exposure measurements (from Denmark, Finland, France, Germany, the Netherlands and the United Kingdom), and expert judgments were used to generate preliminary estimates of exposure to different types of wood dust. These estimates were reviewed and finalised by national experts from 15 Member States.

The study showed that construction employed 33% of exposed workers, mostly construction carpenters. 20% of exposed workers worked in the furniture industry, 9% in the manufacture of builders’ carpentry, 5% in sawmilling, 4% in forestry. In addition, 20% were employed in miscellaneous industries employing carpenters, joiners, and other woodworkers. The numbers of exposed workers varied by country ranging from <3,000 in Luxembourg and Malta to 700,000 in Germany. According to the IOM report, the countries with the highest numbers of exposed workers are Germany (19.75% of all EU workers), Spain (12.18%), the UK (19.91%), Italy (9.88%) and Poland (9.29%).

The WOODEX study concluded that the high wood dust concentrations were measured in the furniture industry, in the manufacture of other wooden products, and especially in construction. The mean wood dust concentrations were in fact the highest in construction. Regarding the manufacture of wood products industry, the IOM report estimates that more than 88% of enterprises have between 1 and 9 employees, while only 1.5% have between 50 and 250 employees. The furniture manufacturing sector is similarly mostly composed of very small enterprises (86% with 1 to 9 employees) and few large companies (2.3% with 50 to 250 employees).

Table 12.    Hardwood dust – Types of impacts

Assumptions concerning declining exposure trends in the IOM study were based results of two studies – one from UK, another from the US, both showing steady decrease of exposures over approximately 20 years' period. It should be noted however, that other studies do not support the resulting estimate of current compliance with an exposure level of 3 mg/m3. The WOODEX project provides data on numbers of exposed workers in the EU as well as on levels of exposure 85 . The WOODEX data on levels of exposure for 2000-2003, shows that 25% of all affected workers are exposed to a level of 2-5 mg/m3 (897 000 workers) and as much as 16% are even exposed to levels higher than the legally binding 5 mg/m3 (563 000 workers). In construction, which has the highest number of exposed workers (1,190,000), the proportion of those exposed to levels higher than 5 mg/m3 reaches 21% or 254 000 workers.

Option 3 would be the most effective option in reducing occupational exposure to carcinogens. It also leads to the highest reduction of estimated health costs. However, it does so at the cost of imposing a disproportionate burden on firms and in particular SMEs. It may force SMEs to use other type of material or to close down business. On the other hand, option 2 should lead as well to a substantial reduction in the health costs, while the negative impact on firms remains very low or non-existent. Thus, option 2 is the best scoring option in terms of efficiency.

The impact on mobile machinery might be higher than on stationary machinery as it is considered, based on the results of a report by the French authority quoted in the IOM study, that more companies have already put in place ventilation systems for stationary machinery than they have for mobile machinery. Moreover, concentration exposures vary according to sectors and occupation. It is estimated that the highest concentrations (2-3 mg/m3) are in the construction sectors, while for the other sectors concentrations are estimated to be generally below 1 mg/m3 for those working far from machinery in control rooms or maintenance tasks.

Exposure concentrations for installers and carpenters across all industries were generally estimated to be between 2-3 mg/m3. Therefore it is assumed that there will not be any significant cost to most enterprises to meet Option 2. However, data is missing to assess the impact of the OEL on non-substitutable hand-held machines and whether there are alternative solutions to ensure that exposure levels will be properly implemented. Employers emphasised that some hand-held machinery does not allow going below 5 mg/m3. On the other hand the workers favoured a 2 mg/m3 limit in view of current limit values of 2 mg/m³ or below in the majority of Member States. For some MS, 3 mg/m3 is at the limit of feasibility particularly for SMEs while other MS favour a 1 mg/m3 limit to further reduce occupational cancer cases.

Hardwood exposure is also associated with increased symptoms of the upper respiratory tract and to allergic (asthma) and non-allergic effects in the lower respiratory tract. As explained above, the direct and indirect costs of illnesses other than cancer were not taken into account in the IOM evaluation of the health costs and benefits, leading to a clear under-estimation of the health costs benefits of introducing a lower OEL. A 2014 British study 86 showed that occupational asthma had a poor prognosis; once a worker develops occupational asthma after a latency period the chances of recovery are small. The average direct costs per annum per case were found to range from £530 to £715 (670 to 900€), whereas the indirect costs range from £1525 to £1685 (1930 to 2140 €).

Table 13.    Hardwood dust – Comparison of options

The preferred option is 2 (3 mg/m3) because it is the most cost-effective and at the same time is the level which came out of the discussions between employers, employees and MS representatives at ACSH discussions.

Impact on Member States and proportionality

In the case of hardwood dust 18 Member States have so far opted to set a limit which is less protective of worker health than the value recommended by ACSH. Figure 12 in Annex 9 illustrates the ranges of existing national OELs compared to Option 2. The following Figure 13 shows distribution of exposed workers across the Member States.

Among the six Member States with the highest numbers of exposed workers, three have an OEL which is less protective than the value proposed by the ACSH (ES, UK, IT) and thus there is a case of competitive advantage compared to Member States which have adopted more stringent measures. On the other hand, the fact that an OEL at 3 mg/m3 or below is already in place in the other three Member States with high numbers of exposed workers (DE, PL and FR) as well as in six other Member States (AT, BE, CZ, DK EE, NL, SE) indicates that solutions exist to practically implement the OEL. An even lower OEL exists for example in Switzerland (2 mg/m3) as well as in Canada and Australia (1 mg/m3).

It should also be noted that as more than 75% of workers exposed to wood dust are exposed to both hardwood and softwood dusts, the prevention of the risks posed by hardwood dust will also include softwood dust. So in practice, the OEL on hardwood dust would also be beneficial to protection against exposure to softwood dust.

Introduction of an OEL of 3 mg/m3 would require changes for a substantial number of MS. It is estimated that approximately 48% of exposed workers are located in those 18 Member States. Even if it is estimated that EU industry is already compliant with an exposure limit of 3 mg/m3, which appears to be an optimistic assumption when the WOODEX data is taken into account, an action at EU level could provide a minimum basis of protection against the risks arising from workers' exposure to these carcinogens for a significant proportion of workers at risk.

5.6Chromium (VI) compounds

Occupational exposure to chromium (VI) compounds – also known as hexavalent chromium compounds - has been associated with an increased risk of lung cancer and sinonasal cancer.

Hexavalent chromium compounds are no longer manufactured in Europe and they are imported less than in the past. The main use of hexavalent chromium is found in wood preservatives, metal coatings, chromium production and catalyst manufacture. In 2006 about 917,000 workers in the EU were exposed. Since 2006, hexavalent chromium has been banned for certain uses (for instance in new vehicles or electronic equipment); the number of workers with high level of exposure is likely to have declined further since then.

The number of workers' exposed to hexavalent chromium was estimated on the basis of the Finnish, Spanish and Italian CAREX estimates. The proportion of exposed workers in each industry was taken from each of these three CAREX estimates and the average proportion exposed across all three countries was found for each industry. The average proportion of exposed workers was applied to information on the number of employees in each industry obtained from Eurostat. To find the number of exposed workers in each industry in each country the average proportion of exposed workers in those three countries was used and multiplied by the number of workers employed in each industry in each country in 2006.

There are no current hexavalent chromium exposure level data available. Exposure levels presented are the result of the analysis done on a sample taken between 1990 and 1999, with geometric mean exposure representative of 1995. The 2010 geometric mean exposures was extrapolated from the 1995 geometric mean exposure based on an estimated decrease in exposure of 7% per year. Due to the limited availability of exposure data it is not possible to present a systematic description of differences in exposure across the EU. The assumption used to determine exposure of different countries consider the extrapolated 2010 exposures level in high exposure industry groups (Table 17) as typical of exposures throughout the EU.

The exposure data available for high exposure NACE industries 27, 28 and 35 were collected from the UK National Exposure Database (NEDB) and data for NACE 24 were compiled for the EU Risk Assessment Report for Chromates using data from industry and the HSE NEDB. No data were found for NACE 28 but exposures are likely to be very similar to NACE 28 87 .

A subset of chromium (VI) compounds (out of over 100), as shown in table 14, are listed in Annex XIV of the REACH Regulation and so are, or will be, subject to 'authorisation' for continued use.

Table 14.    Chromium (VI) compounds currently in REACH Annex XIV

Available evidence (e.g. from the CAREX database) indicates that sodium, potassium, calcium and ammonium chromates and dichromates have been identified as the most important in terms of workers' exposures, a number of which are already regulated under REACH.

It is not possible based on available data to identify exposed populations or impacts associated only with chromium (VI) compounds which are not listed on REACH Annex XIV, nor to account for the impact REACH authorisation may have on exposures for those important compounds to which it will apply.

The IOM study assessed impacts of introducing an OEL for all chromium (VI) compounds, including those subject to REACH.

While this analysis is based on best available data, it should be noted that this results in likely overestimates of both the costs and benefits associated with setting an OEL under CMD for chromium (VI) compounds.

Table 15.    Chromium (VI) compounds – Types of impacts

The baseline scenario for all industries assumes a 7% annual decline in exposure levels and standards change in employed numbers up to the 2021-30 estimation interval. Due to cancer latency, cancer cases - derived as a consequence of past exposure - will continue to affect workers' population until 2030. This situation urges a timely intervention in regulating exposure levels since the introduction of an EU-wide OEL could have a significant positive long-term impact on workers' health. The number of future cancer cases can be most substantially decreased through full compliance with an OEL of 0.025 mg/m3. In addition to the importance of combating cancer it should be noted that the chemical agents may cause other adverse health effects at comparable threshold levels including nasal irritation, severe skin damage or renal toxicity.

Figure 14 in Annex 9 synthetises the changes in the valence state of chromium at all stages of industrial use from mining production, to downstream industry activity and to end use. Hexavalent chromium and trivalent chromium have been coloured coded to emphasise the different states of chromium at the different stages of use. 88

In 2005 the main industrial uses of the chromium (VI) compounds are summarised in Table 16. In 2007 the last EU producer of hexavalent chromium compounds, located in the UK, closed down following declines in the market.

Table 16.    Main industrial uses of the chromium (VI) compounds

Source: HSE (2007) Human Health Risk Reduction Strategy for Chromates in IOM (2011) Hexavalent Chromium, pp. 5

In the EU certain industries have been classified as high in terms of workers exposure to hexavalent chromium. These industries, grouped by NACE code, were identified from CAREX data and are summarised in the table below.

Table 17.    Classification of Industries with High Level Exposure

Compliance cost analysis shows that the majority of firms within affected industries would meet the most stringent proposed OEL (0,025 mg/m3) given that the estimated geometric mean (2010) is between 0.002-0.005 mg/m3. However, risks for workers' health are still consistently high as shown in Table 18. There are substantial percentages of workers within each industry (NACE) who are exposed above OELs proposed under options 2 and 4. This situation calls for timely and effective actions at the EU level to ensure that workers' health and safety at work is safeguarded.

Exposure risks are particularly high in SMEs. Small companies have been found 89 to be at particular risk of lacking adequate control especially in the manufacture of pigments and dyes, the formulation of metal treatment products electroplating and wood dyeing. An inspection of 29 chrome planting facilities found that those that used hexavalent chromium were considered not to be adequately controlling the risks and most were not complying with the relevant legislation. An EU-wide OEL would then have the positive effect of raising public awareness of these important occupational health hazards and contribute to better monitoring of exposure levels and increased compliance of industries.

Table 18.    Estimated percentages of workers with exposures exceeding the proposed OEL (Option 2) in high exposure industries

Some firms within affected industries would require further control measures to meet the proposed OEL given that estimated geometric standard deviations range from 3.8-14. For enterprises that would need ventilation systems to comply with an OEL of 0,025 mg/m3 there would be a significant up-front capital cost. 90

It is possible that some firms might be able to pass through additional costs in the form of higher prices for their final products since the OEL would be applied consistently across the EU. This should create a 'level playing field' for firms across the EU and reduce competitiveness distortions created by differences on OELs across the EU.

On the macroeconomic side, short term spending on risk management measures may also be good for the economy as equipment manufacturers (ventilation systems), installers and others will benefit with money flowing through the economy. With fewer life year lost, there should be also a benefit to the economy through avoided loss of output and consumption in the future (post 2040). However, since compliance with an OEL would not change the current manufacturing process there is unlikely to be any significant change to macro-economic impacts.

It should further be noted that the impact of REACH authorisation on relevant chromium (VI) compounds cannot at this stage be reliably assessed, but should be expected to reduce exposures and result in reduced costs and benefits associated with setting an OEL for these chemical agents. When applying for REACH authorisations, enterprises need to also invest in risk management measures and demonstrate that risks are controlled (either 'adequately' or with 'appropriate and effective risk management measures' according to the case) – potentially driving investment in additional risk management measures.

A targeted REACH restriction (entry 47 of REACH Annex XVII) applies to placing on the market or use of (soluble) chromium (VI) compounds as a component of cement, which is prohibited above a certain concentration. While this restriction is relevant for worker protection, it should not be expected to significantly affect the overall exposure patterns and associated cost benefit assessments made in this report. In any case, this restriction was established prior to REACH (in 2005) and hence its effects have been taken into account in the IOM study.

Table 19.    Chromium (VI) compounds – Comparison of options

The preferred option is 2 which is more effective reducing exposure and therefore the number of deaths, while slightly more costly than option 4. In addition, it might trigger more investment in research and innovation as well as the adoption of products that do not contain chromium. Option 2 would then be expected to increase the dissemination of improved technologies and production methods. The potential volume of ventilation systems being required across the EU may also stimulate investment in R&D to produce more cost-effective systems. Furthermore, with fewer life years lost and cancer registrations, there should be a benefit to the economy through avoided loss of output and consumption in the future, for example due to greater productivity as well as greater consumption and greater taxes raised.

It should be noted that the workers group in the ACSH argued that even a binding limit of 0.025 mg/m3 would correspond to a high cancer risk. Employers' representatives did not make any specific comments in the ACSH opinion but the metal industry expressed their positive view on the establishment of an EU-wide binding OEL as a way to demonstrate appropriate control of the exposure at the workplace in the course of meetings between Commission services and industry and workers representatives (see point 9.2.6 in Annex 2).

Impact on Member States and proportionality

In the case of chromium (VI) compounds 21 Member States have no OEL for this agent or have one that is less protective of worker health than the 0.025 mg/m3 value recommended by ACSH. This includes 5 Member States where the national OEL is provided as a range, given that different OELs apply for different specific Cr(VI) compounds. Figure 15 in Annex 9 illustrates the ranges of existing national OELs compared to Option 2 the upper limit has been considered for those countries providing a range of values). The following Figure 16 shows distribution of exposed workers across the Member States.

Among the six Member States with the highest numbers of exposed, two have no OEL (DE, IT) Introduction of an EU OEL could ensure that lack of a national OEL or a less stringent OEL does not act as an incentive for business in decisions concerning the plant location. On the other hand, the fact that an OEL equal or close to the value proposed by the ACSH is already in place in one of the Member States with high numbers of exposed workers (FR) as well as in the US indicates that solutions exist to practically implement the OEL such that businesses are able to make appropriate investment to protect workers and demonstrate compliance.

It is estimated that approximately 83% of exposed workers are located in Member States with no OEL for this agent or one that is less protective of worker health. An action at EU level could provide a minimum basis of protection against the risks arising from workers' exposure to these carcinogens for a significant proportion of workers at risk.

5.7Ethylene oxide (EO)

EO is classified according to the criteria in the CLP Regulation as a category 1B human carcinogen based on limited human epidemiological evidence and other data that it may cause leukaemia.

It is used in the manufacture of production of consumer goods such as anti-freeze solvents and cosmetics. About 3.8 mln tonnes of EO are produced in Europe each year. The majority is used in the manufacturing of ethylene oxide derivates such as ethylene glycols, which are used in the production of consumer goods.

It is estimated that approximately 15,600 workers in the EU are potentially exposed and that in the period 2000-2010 in the EU there were about 5 deaths or less per year from leukaemia that were attributable to exposure to ethylene oxide before the early 1980s, which corresponded to about 0.01% of all deaths from leukaemia and a loss of 82 Disability-Adjusted Life Years (DALYs) each year.

Global market for EO is expected to continue expanding. 91

Table 20.    Ethylene oxide (EO) – Types of impacts

No ethylene oxide exposure data were available from industry. The peer-reviewed scientific literature for occupational exposure data for the three main uses of EO (i.e. chemical manufacturing, industrial sterilization, hospital sterilization) shows a baseline scenario that suggest that EO exposures is generally already below 1 ppm. As a consequence there are not health benefit are expected since exposure is already estimated to be controlled to below 1ppm under the baseline scenario.

The economic impact of the baseline scenario highlights that industries affected have already incurred costs of installing control measures to reduce EO. On the other hand, the economic benefits could affect mostly research and development due to the introduction of non-EO alternatives in hospitals.

Other costs and benefits related to 'social', 'macro-economic', and 'environmental' impact of the baseline scenario are negligible.

Administrative costs of setting the limit of 1 ppm for employers (i.e. minimising exposure of workers; redesign work processes; hygiene measures; information of workers; training staff; make information available; consultation with employees; etc.) are estimated to be low since firms have already been required to carry out these measures following national legislation.

Where there is the possibility of a significant uptake via dermal exposure SCOEL recommend that any OEL be accompanied by a 'skin notation' (in the case of ethylene oxide ACSH endorsed this aspect of the SCOEL Recommendation). If adopted and accordingly transposed by Member States, employers are required under the national transposing legislation to take this into account in selecting appropriate risk management measures to protect workers.

Once the need for managing exposure has been established the only risk management measure available in practical terms is to avoid skin contact. Ethylene oxide has a harmonised classification as a skin irritant (category 2). This hazard would normally result in employers taking steps to avoid skin contact as a part of routine OSH risk control. Adoption of a 'skin notation' should therefore result in no additional cost for employers.

Table 21.    Ethylene oxide (EO) – Comparison of options

The opinion of the social partners and Member States in the ACSH supports option 2, and the CMD further establishes an expectation that OELs be set where it is possible to do so.

Impact on Member States and proportionality

In the case of ethylene oxide nine Member States have no OEL for this agent or have one that is less protective of worker health than the value recommended by ACSH, so introduction of an OEL of 1 ppm would represent a significant change in this sense.

The Figure 18 in Annex 9 illustrates the ranges of existing national OELs compared to Option 2. The following Figure 19 shows distribution of exposed workers across the Member States.

The two Member States where the production and/or use of this chemical agent is concentrated and the numbers of exposed workers are the highest (DE, UK) have no OEL or OELs above the limit proposed by the ACSH. Introducing an OEL would bring a greater clarity for economic operators across the EU and ensure that lack of an OEL or a less stringent OEL does not act as an incentive for business in decisions concerning the plant location. An OEL of 1 ppm is already in place e.g. in Australia, Canada, US, Switzerland, New Zealand and Japan.

It is estimated that approximately 43% of exposed workers are located in the nine Member States with no OEL or an OEL less protective than proposed. Even if current exposure levels in the EU are estimated to be below 1 ppm, a minimum basis of protection against the risks arising from workers' exposure to these carcinogens cannot be ensured for all EU workers – it follows that an action taken at the European Union level to achieve this objective could be justified.

5.8o-Toluidine

This chemical agent may cause bladder cancer. Occupational exposure is most likely to occur through inhalation and dermal contact.

o-Toluidine is utilised in the production of dyes and pigments. The industries in Europe that report the highest workers' exposure levels are: i) Manufacture of chemicals, chemical products and man-made fibres; ii) Manufacture of rubber products. In the chemical industry exposure to o-toluidine can occur during its production and during its use in the production of herbicides, dyes and pigments, rubber chemicals, epoxy resin hardeners, fungicide intermediates, and pharmaceutical intermediates. o-Toluidine appears to have been relatively well controlled in most chemical manufacturing facilities for the past several decades.

The available data indicates that the EU o-toluidine production is centred in France, Germany, Italy, the Netherlands and the United Kingdom 93 . The available information suggests that there is no o-toluidine production in the rest of Europe.

In the EU there are approximately 5,500 workers who are potentially exposed to o-toluidine. The prevalence of exposure to o-toluidine was estimated from the Finnish CAREX estimates. The proportion of exposed workers from the Finnish estimate was applied to information on the number of employees in each industry obtained from the structural business statistics and the Labour Force Survey available on the Eurostat database. The Finnish proportion of exposed workers was multiplied by the number of workers in each industry in each other Member State.

Since Finland does not manufacture o-toluidine, the estimate of workers' exposure in manufacturing countries (France, Germany, Italy, the Netherlands and the United Kingdom) might be underestimated.

It was judged that 98% of workers are exposed to less than 0.1 ppm. In recent years exposure levels have been decreasing by about 8.8% per annum.

Table 22.    o-Toluidine – Types of impacts

On the basis of estimates of the number of workers exposed and the Eurostat data on the distribution of firms by size, broad estimates of the number of enterprises requiring further action to comply with each proposed EU-wide OEL were produced (Table 23-24).

Tables 23-24 show that a total of nine firms are estimated to be affected by an EU-wide OEL at 0.1 ppm while no firms would be affected if an OEL of 1 ppm were to be introduced. It is not expected that there would be any significant and additional potential closure of companies as a result of introducing an EU-wide OEL of 0.01 ppm because compliance costs are likely to be minimal.

Table 23.    Numbers of Enterprises affected in NACE group 25.1

Table 24.    Numbers of Enterprises affected in NACE division 23

The industrial sector estimated to benefit most from the introduction of an EU-wide OEL is the manufacture of chemicals. The administrative costs for employers are estimated to be low in both baseline scenario and in the intervention scenario of OEL 1 ppm, since most of the investments needed to decrease the exposure have already been made. Low to medium costs would involve firms that do not currently comply with baseline scenario in case an OEL of 0.1 ppm was going to be implemented.

Table 25.    o-Toluidine – Comparison of options

Neither option is predicted to give rise to any important reduction in bladder cancer death or registrations over the baseline assumptions, primarily because exposures are already very low. However, the preferred option is 2 as it brings some health benefits while generating only slightly higher costs than option 4. Industry is almost fully compliant with this OEL already and 0.1 ppm has been accepted by all parties in the ACSH discussions. Switzerland and New Zealand have set lower OELs for this chemical agent.

Impact on Member States and proportionality

In the case of o-Toluidine 25 Member States have no OEL for this agent or have one that is less protective of worker health than the value recommended by ACSH. The Figure 20 in Annex 9 illustrates the ranges of existing national OELs compared to Option 2. The following Figure 21 shows distribution of exposed workers across the Member States.

Thus, introduction of an OEL of 0.1 ppm would represent a significant change in this sense. It is estimated that approximately 98% of exposed workers are located in those 25 Member States. Even if current exposure levels in the EU are estimated to be largely below 0.1 ppm, a minimum basis of protection against the risks arising from workers' exposure to these carcinogens cannot be ensured for all EU workers under the baseline scenario – it follows that an action taken at the European Union level to achieve this objective could be justified.

5.9Refractory Ceramic Fibres (RCF)

Occupational exposure to RCFs is associated with adverse respiratory effects as well as skin and eye irritation and may pose a carcinogenic risk based on the results of chronic animal inhalation studies. However, epidemiologic studies have found no association between occupational exposure to airborne RCFs and an excess rate of pulmonary fibrosis or lung cancer. Based on the animal test results, certain RCFs are classified according to the CLP Regulation as Carc. 1B (may cause cancer by inhalation).

RCFs are synthetic vitreous fibres or man-made mineral fibres used in industry for their properties of heat resistance, tensile strength and durability. The total tonnage of RCF used in the EU is about 25,000 tonnes per year, 90% of which is used for industrial insulation. Ceramic fibres are typically used only in industrial settings, implying no consumer exposure.

For the estimations, due to unavailability of epidemiological data, exposure to RCFs has been assumed to be no worse than exposure to chrysotile (white) asbestos in terms of cancer risk.

On the basis on information provided by the European Ceramic Fibre Industry Association, it is estimated that about 10,000 workers are exposed to RCF across the EU. Of these, about 730 are employed in RCF production plants in Germany, France and the UK, and about 9,270 in the downstream user industry (a breakdown by Member State not available. Geometric mean exposures in the industry are less than 0.2 fibres/ml and it is estimated that about 7% of workers in manufacturing facilities and 12% of workers at downstream user facilities have been exposed above 1 fibres/ml. More than half of workers are exposed above 0.1 fibres/ml. A majority of the companies in question (downstream use) are SMEs.

It should be noted that, in December 2011, certain RCFs were identified as 'substances of very high concern' under REACH owing to classification under CLP as a category 1B carcinogen. This constitutes the initial step toward possible inclusion into REACH Annex XIV and hence REACH 'authorisation'. In February 2014 the European Chemicals Agency further recommended to the European Commission that RCF, among other chemical substances, be included into REACH Annex XIV.

Two possible substitutes for RCF are Alkaline-Earth Silicate Glass Wool (AES), which however is an imperfect substitute (having a lower applicable temperature range for thermal insulation), and Polycrystalline Wool (PCW), which however is 20 times more expensive than RCF.

Table 26.    Refractory Ceramic Fibres (RCF) – Types of impacts

The values of health benefits are relatively low for all the options because of the low level of assumed cancer incidence under the baseline and the existing controls in place.

Option 3 would be the most effective option in reducing occupational exposure to carcinogens and levelling the playing field across the Union. However, it raises significant problems of technical feasibility (the OEL may de facto not be achievable in some instances) and imposes therefore excessive burden on firms even if some of the costs associated with substitution of RCF with more expensive alternatives (such as PCW) would be passed through to consumers via higher prices. Option 2 appears to be more efficient. It leads to a similar reduction of Years of Life Lost and Disability Adjusted Life Years as Option 3, but at a significantly lower cost.

Table 27.    Refractory Ceramic Fibres (RCF) – Comparison of options

Option 2 because is the most effective and corresponds to the limit recommended by SCOEL. Some MS have already imposed nationally a similar OEL, and there are no concerns regarding technical feasibility. In addition, this is the minimum common denominator of ACSH discussions. Employers advise to follow the recommendation of SCOEL. They also stress that industries have been working on a 0.5 f/ml exposure limit for many years, and further reductions of exposure are technically difficult. Views of Member States range from 0.3 f/ml to 0.1 f/ml. The workers argue for an OEL 0.1f/ml on the basis of considerations detailed in the scientific explanations given for the exposure risk relationship on aluminosilicate fibres derived in Germany, according to which aluminosilicate fibres exhibit a carcinogenic potency comparable to asbestos.  94

Impact on Member States and proportionality

In the case of RCF, 24 Member States have no OEL for this agent or have one that is less stringent than 0.3 f/ml. Figure 22 in Annex 9 illustrates the ranges of existing national OELs compared to Option 2. There is insufficient data to provide a breakdown by Member State.

RCF is only manufactured in three MS (FR, UK, DE) but exposure from use and disposal (e.g. from insulation in furnaces) may be located across Europe. Thus, introduction of an OEL of 0.3 f/ml would require changes for a substantial number of MS. The analysis shows that a minimum basis of protection against the risks arising from workers' exposure to these carcinogens cannot be ensured for all EU workers under the baseline scenario – it follows that an action taken at the European Union level to achieve this objective could be justified.

5.10Respirable Crystalline Silica

RCS is taken here to mean 'the respirable fraction of crystalline silica dust generated by a work process'. RCS, so-defined, is a process generated substance which is not placed on the market and, as such, is not classified under the CLP Regulation. This is the reason why it is proposed to be included in Annex I of the CMD. In addition, a few options for an OEL value to be included in the Annex III are considered.

IARC has stated its opinion that 'crystalline silica inhaled in the form of quartz or cristobalite from occupational sources is carcinogenic to humans (Group 1)'. 95

Crystalline silica is abundant in rocks, sands and soils. Exposure to RCS therefore occurs in many industries. Common exposure scenarios include earth moving (eg. mining, quarrying, tunnelling), crushing or grinding of silica-containing material such as concrete, aggregate or mortar, the manufacture of glass and other non-metallic mineral products and the use of sand as moulding media in foundries.

It is estimated that 5,300,000 EU workers are potentially exposed, more than 70% of them in the construction sector. Beside construction, the following industries are estimated to have the highest exposure levels compared to the other industry sectors as well as the highest numbers of exposed workers: manufacture of other non-metallic mineral products, other mining and quarrying, manufacture of basic metals, manufacture of fabricated metal products, electricity, gas, steam and hot water supply.

Table 28.    Respirable Crystalline Silica – Types of impacts

As in the case of all other considered chemical agents, the exposure trends for RCS used to establish the baseline are quoted from the IOM study. The validity of these assumptions is discussed in the introduction to section 5. Specifically for silica the correctness of the assumed 7% decline can be further questioned based on the fact that among the three studies that IOM used to build this assumption, two are only looking at quarries. The only one that analyses all industries comes from the US. Lack of data for the two principal exposure sectors ("construction" and the "electricity, gas, steam and hot water supply") weakens the assumption of exposure decline.

A different exposure trend can be derived for example from a Finnish study (Kauppinen et al., 2013) 96 , which looked at data from FINJEM database, starting in 1950 and projected trends till 2020. This study found that an overall exposure decline was 1% rather than 8%. The data presented in this study shows also that exposures were declining significantly faster between 1950 and 1990, compared to more recent years. 97

Even with the IOM studies estimates, a high proportion of workers are considered to still be at risk of being exposed above the considered OELs (14% above 0.2 mg/m3, 26% above 0.1 mg/m3, 41% above 0.05 mg/m3), and the estimated mean exposure is 0.07 mg/m3.

The prevalence of exposure to RCS in the IOM report was estimated from a Finnish, a Spanish and an Italian study. The proportion of exposed workers in each industry was taken from each of these three studies and the average proportion exposed across all three countries was found for each industry. The average proportion of exposed workers was applied to information on the number of employees in each industry obtained from the structural business statistics and the Labour Force Survey available on the Eurostat database. The average proportion of exposed workers was multiplied by the number of workers employed in each industry in each country in 2006 to estimate the number of exposed workers in each industry and country.

However there is a limited availability of exposure data, and it is not possible to determine exposure differences across the EU. Moreover, exposure levels and therefore cancer risk vary greatly according to the sectors.

According to the IOM report, the countries with the highest numbers of exposed workers are Spain (26.8% of all EU workers), Germany (11.8%), France (11%) and the UK (9.4%).

In all industries with high prevalence of RCS exposure, the majority (between 58 and 94% depending on the industry) of companies are very small, with 1 to 9 employees. The industrial minerals industry alone consists mainly of SMEs with some large multinational companies. Members of the Industrial Minerals Association operate over 810 sites throughout Europe.

The IOM report recognises that the number of workers and enterprises affected by the proposed reduction in the OEL are likely to be an overestimate since the NACE codes include activities in which workers may not necessarily be exposed to RCS.

In all the considered options, introducing an OEL for RCS will contribute to reducing attributable deaths compared to no action, and the benefits of introducing an OEL may outweigh compliance costs. Moreover, any reduction of exposure will also lower the risk of silicosis. Indeed exposure to respirable crystalline silica in workplace air is associated with the development of silicosis, an irreversible scarring disease of the lung. Silicosis also appears to be a significant risk factor for the development of lung cancer. As the health costs of silicosis were not taken into account in the IOM report, the total health benefits of introducing an OEL are clearly higher than was has been indicated for the purposes of this proposal.

While the compliance costs are relatively high for all three options, they should be compared to the total value of goods and services in the affected sectors, which was €5 trillion in 2006. Moreover, the bulk of compliance costs will be already incurred under the baseline scenario (i.e. in the absence of EU action). Imposing an EU OEL would however oblige businesses to anticipate the corresponding investments.

It should also be added that the options have variable impacts on different sectors. For example, for quarrying, costs are significantly lower under option 4 compared to both 2 and 3. For some other sectors, such as foundry, brick manufacture or silica sand production options 2 and 4 result in the same costs. As for the construction sector, which represents a majority of the affected firms, average cost per enterprise is the same under all three options.

Option 3 would be the most effective in reducing occupational exposure to carcinogens but will have a negative impact on some sectors of the industry 98 , and in particular on SMEs. Option 2 has similar characteristics, however its impact should be lower on the industry. 99 At least 17 countries are reported to have already introduced an OEL of 0,1 mg/m3 or below. 100 A majority of EU countries are therefore already compliant with Option 2, which indicates that the impact of introducing Option 2 would be lower than Option 3. Option 4 would imply the lowest compliance costs and the least negative impact on SMEs, however it would be less effective in reducing cancer. Moreover, the OEL of 0,2 mg/m3 would be higher than the majority of national OELs already in place. No significant environmental impact is foreseen in none of the options.

Some stakeholders argue that the Carcinogens Directive is not the appropriate legal framework to introduce OELs for RCS – instead, they argue for introduction of an EU OEL via the Chemical Agents Directive as the most efficient risk management tool.

A mechanism exists to establish EU OELs via the Chemical Agents Directive – these may be either 'binding' (mandatory maximum) or 'indicative' (indicated maximum) in nature, with Member States committed, as a minimum requirement, to in any case ensure that national OELs are established for the chemical agents in question, taking into account the values set at EU level.

However, RCS is a long- and widely-recognised occupational carcinogen – including long-standing classification by IARC. A clear threshold for silicosis development cannot be identified, and there is sufficient information to conclude that preventing the onset of silicosis will also reduce the cancer risk. 101 It is therefore most appropriate to establish an EU OEL under CMD for this chemical agent where it is either generated as a result of a work process or otherwise classified as a category 1 carcinogen according to the CLP criteria.

Stakeholders also put forward existing guidelines and good practices. Most notably, as reported above, NEPSi requires signatories to follow good practices and monitoring protocol, and to provide quantitative data.

Table 29.    Respirable Crystalline Silica (RCS) – Comparison of options

Option 2 (0,1 mg/m3) is a feasible option for EU industries, while still guaranteeing a substantial reduction of estimated attributable deaths due to lung cancer. Introducing an OEL of 0.05 mg/m3 would increase the estimated health benefits by around 10%, but would also increase the estimated compliance costs by close to 80%. It is therefore considered that the cost/benefit ratio of introducing an OEL 0,1 mg/m3 is better than introducing an OEL of 0.05 mg/m3.

It is also the level agreed by the ACSH even if there were some questions on the part of employers and some MS whether it would not be more appropriate to legislate under the CAD. Workers stated that further exposure reduction below the proposed binding OEL of 0.1 mg/m³ is paramount to reduce the risks of lung cancer, and silicosis.

According to the IOM report there is a risk that production would be affected as a result of the need to comply with more stringent OELs and with high compliance costs. Indeed, the majority of the companies that would be affected by the imposition of an OEL are SMEs and the costs may be very expensive for a large proportion of the affected sectors. If the costs cannot be passed through to consumer prices and/or some companies will not be able to get access to loans to pay for the necessary investment, there is a possibility of some company closures and, for industries for which it is possible, some relocation of activities in non-EU countries. Such impacts may be differentiated at the sector and sub-sector level, by geographical location, and/or at operator level.

Among the six Member States with the highest numbers of exposed workers (see Figure 25 in Annex 9), two have no OEL (DE, IT) and one has an OEL significantly less protective than 0.1 mg/m3 (PL) thus there is a case of competitive advantage compared to Member States which have adopted more stringent measures. On the other hand, the fact that an OEL equal or below the value proposed by the ACSH is already in place in the other three Member States with high numbers of exposed workers (ES, FR, UK) and in 14 other Member States indicates that solutions exist to practically implement the OEL.

It should also be noted that an OEL at 0,1 mg/m3 is comparable or higher to the levels applied in most non-EU countries (except notably China). For the USA, a new value of 0,05 mg/m3 is being proposed for RCS 102 . The proposed revision of the existing value is based on evidence that indicates employees exposed to RCS well below the current value of 0,1 mg/m3 are at increased risk of lung cancer mortality and silicosis mortality and morbidity. At the same time, it is assumed that an OEL of 0.05 mg/m3 would be technologically feasible for most affected industries, except two out of 12 construction activities, and would not have any major micro- and macro-economic impact.

Finally, the reluctance on the part of employers to accept that RCS is regulated under the CMD appears to be related to concern regarding the stricter standard for substitution and exposure control (process enclosure) resulting from application of the CMD over CAD. There is also some concern regarding stigmatisation of silica as a carcinogen and associated negative impacts on industry, for instance in case of 'social license' 103 to operate mines and quarries. It should however be noted that some forms of RCS have been classified as carcinogenic by the IARC since 1987. Potentially, compliance with an OEL provides an opportunity for industry to demonstrate to workers and others that the necessary measures to address the risks of handling this chemical agent are in place.

Impact on Member States and proportionality

In the case of RCS, 11 Member States have no OEL or have one that is less stringent than the ACSH recommended 0.1 mg/m3 level. Figure 24 in the annex 9 illustrates the ranges of existing national OELs compared to Option 2. Figure 25 shows distribution of exposed workers across the Member States.

33% of exposed workers are estimated to work in those 11 MS. Under such circumstances a minimum basis of protection against the risks arising from workers' exposure to these carcinogens cannot be ensured for all EU workers under the baseline scenario – it follows that an action taken at the European Union level to achieve this objective would be justified.

5.11Vinyl Chloride Monomer (VCM)

Exposure to VCM is associated with increased risks of the usually rare form of liver cancer, angiosarcoma, and possible increased risks of hepatocellular carcinomas. VCM is classified as a Group 1 carcinogen by IARC, and as Cat 1 carcinogen in the EU under the classification and labelling legislation. An OEL of 3 ppm is currently in place at EU level.

95% of VCM 104 produced worldwide is used in the manufacture of Polyvinyl Chloride (PVC) and its associated polymers. The latter is then used to manufacture automotive parts and accessories, furniture, packaging materials, pipes, wall coverings, and wire coatings. VCM production is often located on the same site as the production of PVC. In 2007, 7.2 mln tons of VCM were produced in the EU and Norway in 30-40 plants located in 13 EU Member States 105 and Norway. In total, there are approximately 100 VCM manufacturing or polymerisation sites. There are no substitutes to VCM for VCM production.

In the EU, about 15,000 workers from VCM and PVC plants are estimated to be potentially subject to high exposure levels. No breakdown by Member State is available. Exposure to VCM outside VCM and PVC manufacturing industry is limited, with about 5 000 more workers (in manufacture of rubber and plastic products, water transport, supporting and auxiliary transport activity, R&D) potentially subject to (negligible) background exposure.

The risk estimates used in the IOM study for the assessment of health impacts are based on evidence provided by the existing epidemiologic literature. The estimates on current exposure levels used are derived from the existing scientific literature 106 , and from exposure data collected in 2008 from 36 VCM and PVC plants (a non-random sample of the around 100 plants present in the EU). No detailed information is available on differences in exposure levels across Member States.

Table 30.    Vinyl Chloride Monomer (VCM) – Types of impacts

Option 2 would be the most effective option in reducing occupational exposure to carcinogens and levelling the playing field across the EU. It also leads to the highest reduction of Years of Life Lost and Disability Adjusted Life Years. Already under the baseline the majority of companies is able to control exposure under the 1ppm exposure level, but option 2 would imply that those enterprises who are not currently below that level would no longer be able to delay investment to introduce preventative measures and effectively minimise exposure. There is a ready market for VCM and no plant closures are expected to result from the implementation of a more stringent OEL.

Table 31.    Vinyl Chloride Monomer (VCM) – Comparison of options

Option 2 is the most effective and the closest to the limit recommended by SCOEL. In addition, this is the level that has been agreed at ACSH discussions. Some MS have already imposed nationally a similar OEL, and there are no concerns regarding technical feasibility. The same OEL is already in place in the US and Canada.

Impact on Member States and proportionality

In the case of VCM, 25 Member States have so far opted to set a limit which is less protective of worker health than the ACSH recommended 1 ppm. Figure 26 in the annex 9 illustrates the ranges of existing national OELs compared to Option 2.

There is insufficient data to provide a breakdown by Member State. However, given the number of Member States with higher OELs, a minimum basis of protection against the risks arising from workers' exposure to these carcinogens cannot be ensured for all EU workers under the baseline scenario – it follows that an action taken at the European Union level to achieve this objective could be justified.

5.12Bromoethylene (vinyl bromide)

Bromoethylene may cause liver cancer.

Bromoethylene is used as a flame retardant in the production of acrylic fibres carpet backing materials. Other uses include children’s sleepwear and home furnishings.

Number of people exposed in the EU likely to be small, i.e. less than a few hundred, but there is not enough information to assess the actual extent of exposure. Given the uncertainty about the number of exposed workers it is not possible to provide a health impact assessment. Similarly, there are no estimates of health costs of inaction for this chemical agent.

Table 32.    Bromoethylene (vinyl bromide) – Types of impacts

Based on the available measurements and the annual reduction in exposure the consultant judged that occupational exposure levels to bromoethylene are currently low, with the highest exposures probably about 3 mg/m3 (less than 1 ppm).

There are no predicted health benefits from setting an OEL at 5 ppm, although it can be assumed that the impact would be relatively small because current exposures are estimated to be much lower than 5 ppm. There are no additional costs associated with compliance with an OEL of 5 ppm. There are also no social or macro-economic costs associated with introducing such an OEL. There are no significant environmental impacts foreseen. Lack of data for option 2 does not allow assessing its impacts.

Table 33.    Bromoethylene (vinyl bromide) – Comparison of options

The opinion of the social partners and Member States in the ACSH supports option 2, and the CMD further establishes an expectation that OELs be set where it is possible to do so. Introducing an OEL would also be effective in introducing a greater clarity for economic operators across the EU. At the same time, lack of data does not allow to assess costs/benefits options 2 (ACSH value).

Impact on Member States and proportionality

In the case of bromoethylene 22 Member States have no OEL for this agent or have one that is less stringent than the value recommended by ACSH, so introduction of an OEL of 1 ppm would require changes for a substantial number of MS. Figure 27 in Annex 9 illustrates the ranges of existing national OELs compared to Option 2.

There is no data on numbers of exposed workers per Member State but it is possible that the overall number of exposed workers in the EU is close to zero. An action taken at the European Union level to achieve this objective could be justified as a way to ensure minimum basis of protection against the risks arising from workers' exposure to this carcinogen.

5.13Hydrazine

Exposure to hydrazine may increase the risk of lung and colorectal cancer. It is classified by IARC as a Group 2B carcinogen (Possibly carcinogenic to humans), and as a Cat 1B carcinogen in the EU under the classification and labelling legislation.

The principal applications of hydrazine solutions include chemical blowing agents, agricultural pesticides and water treatment. Production levels in Europe are estimated to be around 20-25 thousand tonnes per year, and in the EU the largest producers of hydrazine are Germany and France.

The risk estimates used in the IOM study for the assessment of health impacts are based on existing epidemiologic evidence. Estimates of exposure prevalence in the manufacture of basic chemicals for the EU were based on 2009 exposure prevalence data available for Finland (although the Finnish exposure prevalence data may not be applicable to all EU countries). As data on exposure prevalence in agriculture is not available, it is assumed that exposure prevalence is similar to that of fungicide captafol (based on Italian 2005 data).

Data on possible current exposure levels is extremely limited (the most recent data available is from 1998 for Japan). Assuming a declining trend in exposure of 7% per year, the assumed (upper) estimate of current exposure for high, medium and low group industries are of 0.7, 0.1 and 0.06 mg/m3, respectively.

In the EU, about 2,1 mln workers are estimated to be exposed to low levels of hydrazine, about 14,600 to medium levels (in agriculture) and 833 to high levels (in manufacturing of basic chemicals). Overall, it is considered that about 8% of workers are exposed above 0.13 mg/m3, and about 75% above 0.013 mg/m3.

Cost estimates include exposure control measures undertaken by firms in the manufacturing of basic chemicals and in agriculture. As the number of firms potentially affected is not known, an estimate is derived on the basis of the (estimated) number of workers potentially exposed, combined with the available information on firm distribution by size. For agriculture, it is expected that the costs related to use of hydrazine as a herbicide can be controlled through good practice and use of appropriate personal protective equipment (PPE). No significant additional costs are therefore expected, as the latter is already considered to be good practice. In the manufacturing of basic chemicals, the installation of new local exhaust systems (LEVs) may be required by some firms to ensure compliance. Estimates on the number of firms potentially affected are subject to high uncertainty.

Table 34.    Hydrazine – Types of impacts

Option 2 would be the most effective option in reducing occupational exposure to carcinogens and levelling the playing field across the EU (as some MS already have set OELs at 0.013 mg/m3). In terms of compliance costs, in agriculture the costs related to use of hydrazine as an herbicide can be controlled the use of appropriate personal protective equipment, which is considered already as good practice. In manufacturing, no significant additional costs (vis-à-vis option 1) are expected since the costs associated with purchase, maintenance and use of local exhaust systems would have been incurred in any case under the baseline scenario, only possibly more gradually over time.

Where there is the possibility of a significant uptake via dermal exposure SCOEL recommend that any OEL be accompanied by a 'skin notation' (in the case of hydrazine ACSH did not comment on this aspect of the SCOEL Recommendation). If adopted and accordingly transposed by Member States, employers are required under the national transposing legislation to take this into account in selecting appropriate risk management measures to protect workers.

Once the need for managing exposure has been established the only risk management measure available in practical terms is to avoid skin contact. Hydrazine has a harmonised classification as corrosive to the skin (category 1B) and skin sensitiser (category 1). These hazards would normally result in employers taking steps to avoid skin contact as a part of routine OSH risk control. Adoption of a 'skin notation' should therefore result in no additional cost for employers.

Table 35.    Hydrazine – Comparison of options

Some Member States have already imposed nationally an OEL of 0.013 mg/m3, and there are few concerns regarding technical feasibility. Compliance costs are small, not hampering production processes nor business viability.

The opinion of the social partners and Member States in the ACSH supports option 2, and the CMD further establishes an expectation that OELs be set where it is possible to do so. Introducing an OEL would also be effective in introducing a greater clarity for economic operators across the EU. At the same time, option 2 brings only a minor benefit, accompanied by a limited cost for enterprises.

Impact on Member States and proportionality

In the case of hydrazine 24 Member States have no OEL for this agent or have one that is less stringent than the value recommended by ACSH, so introduction of an OEL of 0.013 mg/m3 would require changes for a substantial number of MS. The Figure 28 in the annex 9 illustrates the ranges of existing national OELs compared to Option 2. Figure 29 in Annex 9 shows distribution of exposed workers across the Member States.

It is estimated that approximately 91% of exposed workers are located in those Member States. An action taken at the European Union level could improve legal protection of exposed workers.

5.14Summary of the retained options

It has been shown in the previous sections that the considered chemical agents vary significantly. The table below summarises the retained options on the basis of several criteria:

i) Stakeholders' acceptance

For all the considered carcinogens the stakeholders represented in the ACSH support the retained options. However, for a few there have been some dissenting opinions in the course of the discussions. The following rating is applied:

- XX - full support in the ACSH;

- X - partial or conditional support in the ACSH.

It should be noted that in case of partial or conditional support by the stakeholders represented in the ACSH, diverging views concerned e.g. ranges of values or feasibility considerations rather than the principle of setting an OEL at EU level.

ii) Legal clarity

Introduction of OELs for all the considered chemical agents will improve legal clarity for employers and workers. For some, however, the effect will be more significant as currently fewer MS have introduced OELs corresponding to the advised level.

- XX - legal clarity will be improved in half or more of the MS

- X - legal clarity will be improved in less than half of the MS

iii) Size of the problem

The numbers of workers potentially exposed to the carcinogens vary substantially. While an introduction of an OEL will be useful even if currently few workers are exposed (in the future, due to new uses of the chemical agents this might change), an immediate impact will be greater when exposed populations are bigger.

- XXX - - over 500,000 exposed workers

- XX - between 50,000 and 499,999 exposed workers

- X- less than 50,000 exposed workers

iv) Health benefit

There is also a divergence in the size of monetised health benefits of introducing OELs.

- XXX - benefits over 100 mln EUR

- XX - benefits between 10 mln EUR and 100 mln EUR

- X - benefits of less than 10 mln EUR

v) Limited costs for business

While all the retained options are expected to be feasible for business, there are different levels of associated costs for business.

- XXX - costs below 10 mln EUR

- XX - costs between 10 and 100 mln EUR

- X - costs over 100 mln EUR

Table 36

6Overall impact of the package of retained options

6.1Impact on workers

The retained options package (henceforth 'the retained option') should result in benefits in terms of avoided work-related cancer cases and related monetised health benefits as follows:

Respirable Crystalline Silica: an OEL at 0.1 mg/m3 will provide for 99,000 avoided cancer cases by 2069 for a total monetized health benefit quantified between 34 billion and 89 billion EUR;

Hardwood dust: an OEL of 3 mg/m3will provide for a total monetized health benefit between 12 million and 54 million EUR;

Benefits are also expected in relation to introducing a OEL at 0.025 mg/m3 for all chromium (VI) compounds, although it is likely these will have been overestimated as a result of reductions in exposure which may be expected for some important compounds when subject to REACH authorisation.

Remaining substances: The avoided cancer cases and the monetized health benefits are less significant.

The introduction of retained option would decrease the burden of economic costs derived by workers' exposure to hazardous substances. The main economic costs caused by disability and premature death at work are:

the worker’s lost wages during the period of absence from work;

the subsequent severe consequences for household's well-being as well as a reduction in tax collection.

By acting as a clear and homogeneous playing field for workers' rights enforcement, the introduction of the retained option will reduce health and safety risks for disadvantaged workers such as precarious and informal workers. The scientific literature (Aronsson, 1999; Quinlan and Mayhew, 2000; Letourneux and Thebaud-Money, 2002) agrees in pointing out that precarious workers have less control over fundamental determinants of their health and safety such as: the ability to change temperature, lighting, ventilation, and work location, and the freedom to choose when to take personal leave; and they are also more likely to be employed in dangerous working environment.

As a consequence, the retained option have the advantage of shielding workers and families from suffering financial and social costs which would otherwise occur in a baseline scenario.

It is important to note that the study underlying the Impact Assessment report was limited to assessing health benefits resulting solely from avoided cancer cases. However, the chemical agents under consideration cause a range of other occupational diseases. These include primarily different forms of respiratory diseases (caused e.g. by hardwood dust, RCS, RCF) and dermatological diseases (e.g. acrylamide, RCF). Enhanced workplace control of exposure to the considered chemical agents will also contribute to decreasing the risk of those occupational diseases.

The available data is not sufficient to estimate the magnitude of related health benefits. However, taking into account general estimates of costs related to these diseases, it could be expected that limiting the risks leading to their onset could be considerable.

Concerning occupational asthma, according to a review of existing research 111 , costs are related to the poor prognosis of the disease and small chances for recovery. 112   Estimated 70% of workers diagnosed with occupational asthma show symptoms even several years after complete cessation of exposure 113 . According to estimates of the direct and indirect costs of occupational asthma in the United Kingdom 114 , the average annual direct costs per case are £530-£715. The annual indirect costs range from £1525 to £1685. The total costs of an average case to society is some £120,000-£130,000. The total lifetime costs of new cases to society could lie between £95 and £135 million. Further research showed that some 49% of the present value total costs are borne by the individual, 48% by the state and only 3% by the employer. 115

On the subject of occupational skin diseases the same research review quotes a study on costs of occupational hand eczema, conducted in 2013 in Germany. 116 The annual direct and indirect costs per worker diagnosed and treated were on average 2646 and 6152 EUR, respectively. An Italian study estimated societal costs of severe chronic hand eczema to be on average 5016 EUR per person-year (min. 411 EUR, max. 27648 EUR). 117 The need to conduct occupational retraining, job change, or adverse psychosocial effects result in further costs. 118 As for severe occupational skin diseases more than a half of all cases may become persistent and continue even after exposure to a chemical agents discontinued, further direct and indirect costs might occur depending on the degree of disability. 119 , 120 Also research conducted in Australia confirmed that occupational skin diseases had a significant socioeconomic impact, with an estimated annual cost of over $33 million. 121

6.2Impact on businesses

From an economic standpoint, the total cost to an economy of occupational morbidity and mortality is the sum of all private economic costs that are also social costs, plus the social costs that are external to all private parties. Thus, the illness of a worker results in lost output for the employer. If the worker is paid during the period of non-production, this mitigates the private cost to the worker but increases the cost to the employer. A loss of production may lead to a loss of profits, which would then be a social as well as private cost, but the firm might have the ability to raise prices, maintain profits, and shift the cost to consumers.

As regards the economic impacts on costs incurred by enterprises for some carcinogens, the retained option will affect operating costs for companies which will have to put in place additional protective and preventive measures. This will be in particular the case for Chromium (VI) compounds and RCS. As regards RCS, the total costs to industry of introducing an OEL at 0.1 mg/m3 are estimated to be at 3.5 bln EUR until 2069 122 .

For the remaining carcinogens, the impact on operating costs and conduct of business (including small and medium enterprises) will be minimal as only small adjustments will need to be done in specific cases to ensure full compliance.

The retained option will not impose any additional information obligations and will not lead to an increase in administrative burdens on enterprises.

It is important, from an economic point of view, to distinguish between costs that do or do not create incentives for improvements in health and safety. The advantages for businesses of introducing EU wide OELs is that the new regulation will help firms addressing costs which would, otherwise, negatively affect their business prospect in the long-term in the case of non-compliance. According to the economic literature these costs are:

Economic - in this case these costs include unequal firms' costs related to differences in the rate of depreciation of capital equipment or loss of raw material due to the existence of different national OEL; in addition there are 'opportunity costs' – for instance if a firm loses market share due to worker absences from ill-health; finally, company's loss of 'goodwill', which may result from well-publicized cases of industrial accidents or disease, this is an opportunity cost to the firms which can have serious consequences;

Internal – if retained options are not adopted the enterprise run the risk to be considered accountable for workers' ill-health;

Variable – the cost derived by workers' ill-health might not simply resolve in a fixed premium compensation but might involve longer-term costs for the firm;

Routinely visible – cost derived by non-adoption of EU-wide OELs would increase firms' exposure to expensive process of legal and economic information acquisition to deal with the consequences of workers' ill-health.

6.2.1Impact on SMEs

None of the considered options contain lighter regimes for SMEs. SMEs are not exonerated from the obligation to eliminate or reduce to a minimum the risks arising from occupational exposure to carcinogens or mutagens.

For many of the agents covered in this impact assessment, OELs exist already at national level, even if the level as such differs between the EU Member States.

Establishing the OELs foreseen in this initiative should have no impact on those SMEs situated / located in those EU Member States were the national OELs are either equal or lower than the proposed values.

However, due to differences in OELs at national level there will in some cases depending on industry practice be an economic impact in those Member States (and economic operators established therein) which currently have higher OELs established for the chemical agents subject to the initiative.

The major economic impact for SMEs can be summarised in three points. First, some of the OELs have overhead costs, and the smaller the firm, the smaller the revenue base over which these costs can be distributed. Second, the level of expertise is frequently lower for SMEs. Third, the SMEs environment is generally more competitive and finance is more difficult to obtain, leading to shorter time horizons and fewer expenditures on what may be perceived as nonessential items different spending priorities – potentially reducing voluntary investment in equipment or training which may be necessary to implement effective risk management.

The majority of companies that would be affected by the imposition of a RSC limit at 0.1 mg/m3 are small companies. If the costs cannot be passed through to consumer prices and/or some companies will not be able to get access to loans to pay for the necessary investment, there is a possibility of some company closures and, for industries for which it is possible, some relocation of activities in non-EU countries. The extent of this risk is difficult to estimate. 17 EU Member States have already introduced an OEL of 0,1 mg/m3 or below and also from this point of view it appears to be feasible for the industry to adjust to such an exposure limit. Most of the costs incurred in relation to it will be incurred by the construction industry, in which case relocation is not a viable option, so companies are likely to pass the additional costs on to consumers to some extent. The fact that a majority of MS already have introduced the OEL under consideration and that the same level exists in non-EU countries such as the US, Canada or Australia, indicate that it should be feasible for the sectors concerned to absorb the additional costs.

For chromium VI the IOM study assessed that the costs of compliance will notably affect small firms employing less than 20 people, particularly in the manufacture of fabricated metal products. It is possible that some could either close or cease to use chromium (VI) containing components. However, those estimates are for all Chromium (VI) compounds and, as a result of anticipated exposure reductions resulting from the impact of REACH authorisation on several important compounds, both the costs and benefits can be expected to be overestimated. These figures nevertheless represent the best and most appropriate available data to inform this analysis.

SMEs indicated that OSH legislation in general creates some burdens, particularly in relation to documenting risk management and reporting obligations (even if there is evidence that some burdens might be related to national transposition rather than EU level legislation). 123  One of the goals of a future revision of the OSH acquis may be to simplify and/or reduce administrative costs, including for micro and small enterprises CMD limit values do not, however, result in additional reporting or documenting obligations and the clarifications provided in Annexes I and III should in fact decrease the burden on employers by providing clear guidance as for the scope of application and the level of expected compliance.

Moreover, the European Association of Craft, Small and Medium-Sized Enterprises (UEAPME) was among signatory parties of the Cross Industry initiative, which recognised 'EU-wide Occupational Exposure Limits (OELs) as the most effective risk management option for substances where there is a need to address a risk limited to the workplace' and called for the Commission and Member States 'to proceed to set EU-wide OELs for substances, where a risk is identified at the workplace'. 124

6.2.2Impact on competition and competitiveness

Risk prevention and the promotion of safer and healthier conditions in the workplace are key not just to improve job quality and working conditions, but also to promoting competitiveness. Keeping workers healthy has a direct and measurable positive impact on productivity, and contributes to improving the sustainability of social security systems. There is a correlation between low levels of a country's rating as regards work-related deaths and diseases and its competitiveness rank. 125

Implementing the retained option would have a positive impact on competition within the internal market. Having EU-wide OEL for those agents will decrease competitive distortion between firms located in Member States with different national OELs. Setting a minimum standard at EU level will not (and should not) prevent Member States from setting even more protective OELs. However, it will provide certainty that there is an enforceable exposure limit for all concerned carcinogens in all Member States. It will also significantly minimise the scope for variation in OELs across the EU. It may be noted that the examples of the current hardwood dust and VCM OELs show that a majority of Member States in practice adopt the EU OEL directly.

On the other hand, the retained option should not have a significant impact on the external competitiveness of EU firms as many of the values proposed are similar to those in other countries (see Table 3 in Annex 6), notably EU largest trading partners, such as US, China or Switzerland. For example, the proposed value for exposure to hardwood dust is 3 mg/m3 while the value in Canada and Australia is 1 mg/m3. The proposed value for vinyl chloride monomer is 1 ppm, the value in the USA and Canada is also 1 ppm. And the value of 0.1 mg/m3 proposed for RCS is also established in the USA, Australia and Canada.

Furthermore, it is to be expected that in some of those third countries, limit values will also continue to be adjusted to new scientific data. The US Occupational Safety & Health Administration (OSHA), for example, recognizes that many of its permissible exposure limits (PELs) are outdated and inadequate for ensuring protection of worker health. Most of OSHA’s PELs were issued shortly after adoption of the Occupational Safety and Health (OSH) Act in 1970, and have not been updated since that time. 126

6.3Impact on Member States/national authorities

Where no EU OEL exists many Members States conduct their own scientific analysis to determine the acceptable exposure level. Exact costs of this type of exercise are difficult to establish but would be significant given the level of scientific and technical expertise required. The establishment of OELs at EU level eliminates the need for national public authorities to independently evaluate each carcinogen thereby removing an inefficiency of repetition of identical tasks.

Given the substantial economic costs imposed on workers due to their exposure to hazardous substances, the retained option also contributes to mitigate financial loss of the Member State's social security system. From an economic point of view, the coverage and adequacy of EU-wide OELs is the single most important determinant of who bears the cost burden of occupational ill-health.

Administrative and enforcement costs will differ according to present status of each chemical agent in each MS, but should not be significant.

Based on the experience gathered from the work of the Senior Labour Inspectors Committee (SLIC) and having regard to the way the enforcement activities are organised in different MS it is not likely that the introduction of new OELs in the CMD would have any impact on the overall costs of the inspection visits. Those are mostly planned independently of the revised legislation, mainly based on complaints filed during a given year and according to the inspection strategies defined by a given authority. On the other hand, the existence of an OEL, by bringing clarity regarding the acceptable levels of exposure, facilitates the work of inspectors by providing a helpful tool for compliance checks.

Additional administrative costs might be incurred by authorities as regards the necessity to provide information and training on the revision to staff, as well as to revise the compliance checklists. However, these costs are minor in comparison to the overall costs of functioning incurred by the enforcement authorities.

6.4Impact on fundamental rights

The impact on fundamental rights is considered positive - in particular with regard article 2 (Right to life) and article 31 (Right to fair and just working conditions which respect his/her health, safety and dignity).

6.5How does the retained option conform to the principles of subsidiarity and proportionality given the size and nature of the identified problem?

The protection of workers health against risks arising from exposure to carcinogens is already covered by EU legislation, in particular by Directive 2004/37/EC (CMD) and the REACH Regulation. Amending Directive 2004/37/EC can only be done by action at EU level and after a two-stage consultation of the social partners (management and labour) in accordance with Article 154 TFEU.

With regard to the values proposed as retained option it has to be stressed that socio-economic feasibility factors have been taken into account after long and intensive discussions with all stakeholders (representatives from employees' associations, representatives from employers' associations, and representatives from governments).

While monetised benefits of the initiative may be modest, it should be recalled that the initiative would contribute to saving 100,000 human lives in the forthcoming 50 years. These figures are based on the estimations made the IOM study contracted by the Commission 127 . There is a range of additional benefits which have not been quantified (e.g. benefits related to avoided cases of other occupational diseases caused by the chemical agents, positive effects on level playing field, facilitation of legal compliance and enforcement etc.).

The subsidiarity and proportionality check done for each specific agent, indicated that, where relevant data was available, introduction of proposed OELs would improve legal protection for an estimated 33% to 98% of exposed workers (see Table 4 in Annex 6).

In addition, the proposal does not set levels to be directly translated into national legislation but maximum limits. Member States can decide to introduce lower levels.

The planned action therefore complies with the principles of proportionality and subsidiarity.

7How would actual impacts be monitored and evaluated?

This section presents the monitoring and evaluation arrangements that seem most appropriate at this stage in order to monitor and evaluate the planned legislative initiative. It should however be noted that the overall OSH legislative framework has just undergone a comprehensive evaluation. As a result of this, the Commission might decide to modify the monitoring and evaluation mechanisms currently foreseen in this framework. This would clearly have an impact on the monitoring and evaluation arrangements of the individual Directives within the OSH framework, including the CMD.

7.1Monitoring arrangements

The operational objectives for the retained option, in relation to the chemical agents covered are:

The reduction of occupational diseases and occupational related cancer cases in the European Union;

The reduction of costs related to occupational cancer for economic operators and for social security systems in the European Union.

The table below presents the core indicators for each operational objective and the data sources for the monitoring of the core indicators.

Table 37

The monitoring of implementation (transposition and application) is also an indicator of the effectiveness of the initiative to reach its operational objectives. A compliance assessment will be carried out for the transposition of the limit values, by the Commission in two stages: a transposition check and conformity check. The monitoring of application and enforcement will be undertaken by the national authorities in charge of the protection of workers' health and safety and in particular the national labour inspectorates. Information on the practical implementation of the Directive 2004/37/EC at national level is provided every five years by the Member States in accordance with Article 17a of Directive 89/391/EEC. At EU level, the Committee of Senior Labour Inspectors ('SLIC') established by Commission Decision 95/319/EC, informs the Commission on all problems relating to the enforcement of EU OSH Directives, including Directive 2004/37/EC. 129

7.2Evaluation arrangements

In accordance with the ex-post evaluation clause in Art. 17a of Directive 89/391/EEC, every five years, Member States are required to submit a single report to the Commission on the practical implementation of the EU OSH Directives. This report includes a chapter dealing with the implementation of particular aspects of Directive 2004/37/EC, including specific indicators, where available. On this basis, the Commission will evaluate the implementation of Directive 2004/37/EC in terms of its relevance, of research and of new scientific knowledge, in accordance with Art. 17a (4) of Directive 89/391/EEC.

In accordance with Art. 17a of Directive 89/391/EEC, the Commission is required, within 36 months of the end of the five-year period, to inform the European Parliament, the Council, the European Economic and Social Committee and the Advisory Committee on Safety and Health at Work of the results of this evaluation and, if necessary, of any initiatives to improve the operation of the regulatory framework.

Given the data challenges explained earlier in this report it is suggested to made use of the next ex-post evaluation exercise to define the baseline values (benchmark) that will allow assessing the effectiveness of the planned CMD revision. This seems reasonable considering that due to the long latency periods to develop cancer (10 to 50 years), it will not be possible to measure the real impact of the revision before 15-20 years.

Availability of exposure data is presently the subject of a contract study ('HazChem@Work'). 130  It will in due course also contribute to the definition of the benchmark and may enable some further conclusions to be drawn about the effectiveness of the proposed measure. For the foreseeable future, however, the latency of some occupational cancers and other well-explained difficulties in collecting and aggregating exposure data between member states will continue to present significant challenges to their use as indicators.

In that context aggregating exposure data collected from Member State labour inspectorates at EU level could have several important benefits for operation of chemicals policy – but would be a very significant additional burden on both national labour inspectorates and possibly employers. It is also unlikely that such data could in any case be effectively aggregated as reporting structures within companies and between sectors and Member States are unlikely to align.

Annexes

8Annex 1 – Procedural Information

Concerning the process to prepare the impact assessment report and the related initiative.

8.1Lead DG

Lead DG: Directorate-General Employment, Social Affairs and Inclusion, Unit B/3 Health, Safety and Hygiene at Work

8.2Consultation of the Regulatory Scrutiny Board (RSB)

The Impact Assessment report was reviewed by the RSB in a meeting on 17 February 2016. The RSB required a number of improvements.

The revisions introduced in response to the RSB comments are summarised in the table below:

The RSB issued a positive opinion on 1 April 2016 and several further adjustments were made in the text of the report following RSB's further suggestions, as summarised in the table below.

8.3Evidence used in the impact assessment

8.3.1IARC Monographs

Various IARC Monograph have been used as evidence in preparation of the impact assessment. The exact source of which monograph has been used for each individual chemical agents is provided for in Annex 5 of this document.

Through the Monographs programme, IARC seeks to identify the causes of human cancer. The criteria established in 1971 to evaluate carcinogenic risks to humans were adopted by the Working Groups whose deliberations resulted in the first 16 volumes of the Monographs series. Those criteria were subsequently updated by further Ad-hoc Advisory Groups (IARC 1977, 1978, 1979, 1982, 1983, 1987, 1988, 1991, 1992, 2005, 2006).

As stated in the preamble of the Monographs, "the objective of the programme is to prepare, with the help of International Working Groups of experts, and to publish in the form of Monographs, critical reviews and evaluations of evidence on the carcinogenicity of a wide range of human exposures. The Monographs represent the first step in carcinogen risk assessment, which involves examination of all relevant information in order to assess the strength of the available evidence that an agent could alter the age-specific incidence of cancer in humans. The Monographs may also indicate where additional research efforts are needed, specifically when data immediately relevant to an evaluation are not available". 131

The scope of the programme nowadays now include specific chemicals, groups of related chemicals, complex mixtures, occupational or environmental exposures, cultural or behavioural practices, biological organisms and physical agents.

For further information on for example the selection of agents, the data used for the Monographs, the selection of experts, the working procedures etc. can all be found in detail in the preamble of the monographs (see above reference).

8.4External expertise

8.4.1Use of scientific expertise / Commission expert groups / SCOEL

The Scientific Committee on Occupational Exposure Limits for Chemical Agents was set up by Commission Decision 95/320/EC 132 to evaluate the health effects of chemical agents on workers at work. The work of the Committee directly supports Union regulatory activity in the field of occupational safety and health. It develops high quality comparative analytical knowledge and it ensures that Commission proposals, decisions and policy relating to the protection of workers’ health and safety are based on sound scientific evidence.

The Committee assists the Commission, in particular, in evaluating the latest available scientific data and in proposing occupational exposure limits for the protection of workers from chemical risks, to be set at Union level pursuant to Council Directive 98/24/EC and Directive 2004/37/EC of the European Parliament and of the Council.

Members of SCOEL are highly qualified, specialized, independent experts selected on the basis of objective criteria. They are appointed in their personal capacity and provide the Commission with Recommendations and Opinions that are necessary for the development of EU policy on workers protection.

For the purpose of this initiative, the Commission services have used the relevant chemical agent-related SCOEL recommendation. The exact reference for the recommendation used for each individual chemical agent is provided for in Annex 5 of this document.

8.4.2Studies performed by external consultants

Study on health, socio-economic and environmental aspects of possible amendments to the EU Directive on the protection of workers from the risks related to exposure to carcinogens and mutagens at work

Following the two stage consultation of the European social partners (see section 9.1 of this document), DG EMPL/F/4 published on 25 July 2008 an open call for tender in order to carry out an assessment of the social, economic and environmental impacts of a number of policy options concerning the protection of workers health from risks arising from possible exposure to carcinogenic chemical agents at the workplace.

The main outputs expected were a study report containing full reports on 25 carcinogenic chemical agents and two other policy issues relating to the effectiveness of risk management measures and risk based criteria for the setting of occupational exposure limit values.

The contract started on 24 April 2009 and run until 27 April 2011.

The outcome of this study (summary report and individual chemical agents' reports) provides the main basis for this Staff Working Document and are summarised in the relevant sections of this document. The executive summary report, the summary report as well as the reports for the individual chemical agents are available on the internet. 133 .

8.4.3Study on chemical agents toxic to reproduction

DG EMPL/F/4 published on 22 May 2010 an open call for tender in order analyse at EU-level the socioeconomic and environmental impact in connection with possible amendment to Directive 2004/37/EC to extend the scope to chemical agents toxic to reproduction, category 1A or 1B according to the CLP Regulation.

The underlying consideration was, that under the REACH Regulation these chemical agents are considered as chemical agents of very high concern (SVHC), meaning that they might become subject to the so-called authorisation procedure, which in a nutshell foresees that they cannot be placed on the market of the EU unless the supplier demonstrates in a dossier that the use of the chemical agent is safe.

However, based on the current scientific knowledge, the majority of these chemical agents have a threshold below which a safe use of the chemical agent is possible. Therefore, it can be argued that the protection of workers is already covered under the CAD, and that Indicative OELVs or OELs for these chemical agents should be established under the CAD, and the more stringent protective and preventive measures under the CMD, in particular with regard to the substitution provision, are not proportionate.

Nevertheless, and following in particular the request from the workers interest group of the ACSH (see also information provided in Annex 2 of this document – Social Partner Consultation), the study was launched in order to complement the available data to enable the Commission to take an informed decision.

The contract started on 30 November 2010 and the final report was submitted to the Commission in February 2013.

However, the results of the study did not provide sufficient evidence that including these chemical agents under the scope of the CMD would lead to a higher protection of workers.

9Annex 2 - Stakeholder consultation

9.1Social partner Consultation

The TFEU foresees a two stage consultation of the European social partners for legislative initiatives in the field of social policy (article 154, ex Article 138 of the EC Treaty).

The Commission launched the first stage of consultation of the social partners on the protection of workers from risks related to exposure to carcinogens, mutagens and chemical agents toxic for reproduction at work on 6 April 2004. In accordance with Article 154(2) of the TFEU (former Article 138(2) of the EC Treaty), the social partners were asked to give their opinions on the possible direction of EU action in this field.

The first phase of the consultation confirmed that action needs to be taken at Community level to introduce better and standardised methods across the EU, and to tackle situations involving workers' exposure.

All the European social partners who replied by the end of the six-week consultation period to the consultation 134 underlined the importance they attached to protecting workers from the health risks associated with exposure to these chemical agents.

However, while all respondents acknowledged the relevance of the existing legislation, their views differed as to the strategy and direction of future action and which factors should be taken into consideration. 135

For example, whereas five organisations representing trade union umbrella organisations or the British Occupational Hygiene Society considered to be appropriate to amend or update the CMD, three other organisations representing employers felt that priority should be given to practical guidance documents and enhanced sectorial prevention.

Regarding the extension of the scope of the CMD to cover chemical agents toxic for reproduction and the inclusion of more limit values in the Directive most replies were in favour of an EU initiative. On the other hand, social partners' organisations suggested that national and sectorial approaches were more appropriate to tackle the specific issue of workers exposure to environmental tobacco smoke.

Following the first phase of consultation and due to the classification of respiratory crystalline silica as carcinogenic category 1 (proven carcinogen to humans) by IARC, the social partners of the sectors producing (quarries) and using silica (construction, glass, metal industry, pharmaceutical, etc.) embarked on negotiations in view of a European cross-sectorial agreement for the prevention of exposure to silica respirable dust. Worker's organisations agreed to negotiate on the condition that any future agreement would be without prejudice of any EU initiative setting adequate levels of protection at EU level.

Once the Silica agreement was signed in 2006, the Commission launched on 16 April 2007 the second stage of consultation of the European social partners, in accordance with Article 154(3) of the TFEU on the content of the envisaged proposal.

The specific points for consultation were: 1) Inclusion of chemical agents toxic for reproduction (categories 1A and 1B) in the scope of CMD; 2) Updating OELs for chemical agents in Annex III of CMD; 3) Including OELs for more chemical agents in Annex III of CMD; 4) Introducing criteria for setting OELs for CMR chemical agents; and 5) Focus on training and information requirements.

The Commission received replies from seven European social partner organisations: four from employers' organisations (Business Europe, Eurocommerce, European Association of Craft Small and Medium-sized Enterprises (UEAPME) and European Cement Industry), two from workers organisations (European Trade Union Confederation (ETUC), and European Federation of Building and Woodworkers (EFBWW)) and one from an independent organization (British Occupational Hygiene Society (BOHS)).

In their replies these organizations reaffirmed their approach to the prevention of occupational risks derived from carcinogens and mutagens at work, as outlined in their responses to the 1st stage consultation document. The opinions gathered are summarized below:

Inclusion of chemical agents toxic for reproduction (categories 1A and 1B) in the scope of CMD

There was no agreement on the need to initiate a EU level action, neither in the extension of the scope of the Directive to include reprotoxic chemical agents of categories 1A and 1B according to the CLP Regulation. Employers thought that the effective application of the existing legal framework is enough to attain a suitable level of protection, whilst workers called on the Commission to make legislative changes and to commit to eliminate exposure to occupational carcinogens by 2025. Workers took a positive view in order to extend the scope of the Directive to cover reprotoxic chemical agents. The possibility of launching the negotiation procedure under Article 154 (4) and 155 of the Treaty was not agreed.

Updating OELs for chemical agents in Annex III of CMD and including OELs for more chemical agents in Annex III of CMD

There was a partial agreement on the revision of existing binding OELs and on the establishment of new OELs for chemical agents not yet listed in the Directive Annex III of the Directive. While workers indicated a positive attitude based on the fact that it shall ensure equivalent protection of workers at EU level, the employers have expressed their scepticism reasoning that this action could only be justified on the grounds of an evaluation of the Directive 98/24/EC on chemical agents, on the grounds of robust scientific evidence and under the condition that socio-economic and feasibility factors must be taken into account. Furthermore, the revision of binding OELs should be examined in the light of the implementation of the REACH Regulation and of the relationship and interaction between OELs and DNELs (Derived Non Effect Levels) which will be derived under REACH for hazardous chemicals.

Introducing criteria for setting OELs for CMR chemical agents

There were no significant divergences between the replies of both employers and workers on the methodologies to be used and the criteria to be set up for the derivation of OELs. The introduction of criteria for OELs setting was seen as generally positive. However, socio-economic impact assessments and the consideration of feasibility factors should be part of the criteria. Social partners expressed the view that the ACSH should be involved.

Focus on training and information requirements

There was an overall agreement on the need for effective implementation of training and information requirements. This issue is considered to be a key aspect of the prevention policy. Workers call the Commission to set up a strategy to improve coordination and sharing of information at EU level. Employers see an added value on the preparation of guidance documents with recommendations on workers protection against carcinogens and mutagens exposure.

Following the results of the Social Partners consultation, the Commission tendered a study to assess socio-economic aspects of revising OELs and introducing new ones (see point 8.4.2). The results of the study as well as SCOEL recommendations, where available, were subsequently discussed by the tripartite ACSH (see point 9.2.3). The discussions resulted in agreement on limit values, which have been taken forward to this initiative.

9.2Other consultation of stakeholders

9.2.125 October 2006 - Workshop of setting OELs for Carcinogens

In 2006, DG EMPL organised in collaboration with the ACSH a workshop on “Setting OELs for Carcinogens”. The key questions addressed during the workshop were the following:

What is the acceptable/unacceptable level of risk?

What is the maximum level of risk?

Is it possible to quantify it in terms of incidence rate versus the number of exposed workers?

In accepting risk levels should a distinction be made for general public and workers?

What criteria are used in some Member States and what political decisions have been taken in respect to the OEL setting process for carcinogens?

What criteria should be used to define the border between the acceptable and unacceptable risk?

Should the approach to address the risk levels be systematic (quantitative/semiquantitative) or stochastic (case by case)?

Should criteria on the acceptability of risks be regulated at EU level?

Should the workability of the existing EU legal framework be safeguarded versus subsidiarity, in terms of establishment of OELs for carcinogens?

One of the main conclusions of the workshop was that the existing EU OSH legal framework and its supportive administrative, technical and scientific structure should remain in place and be used for the derivation and adoption of OELs at the EU level. However, it was also acknowledged that the derivation of OELs for Carcinogens, Mutagens and Reprotoxic chemical agents (CMRs) - both genotoxic and non-genotoxic - is a demanding task. The availability of sound and sufficient evidence, and in particular the availability of criteria and methodologies for their derivation, is a critical prerequisite for setting OELs for carcinogens.

More than 80 scientists, technicians and academics contributed to the discussions.

9.2.2EU-OSHA - Exploratory survey of Occupational Exposure Limits (OELs) for Carcinogens, Mutagens and Reprotoxic chemical agents (CMRs) at EU Member States level (published in September 2009) 136

Between late 2007 and early 2008, EU-OSHA, at the request of the European Commission, carried out a survey among its network partners aiming at increasing the Commission's knowledge on the existing situation at national levels concerning OELs for CMRs.

Part of the survey was to collect data on existing OELs values for CMRs from the 27 Member States and from selected countries outside of the EU (Australia, Canada, Japan and US). In addition, information was required on the methodology and criteria (scientific, technical and socioeconomic) used when setting an OEL for a carcinogen or a mutagen.

Based on the feedback received, the final survey covered 21 Member States 137 .

With regard to the current initiative it is worth noting that the chemical agents covered by this initiative are in most cases also included in national OEL lists for carcinogens and mutagens. The majority of the MS which reported back have between 30 to 50 OELs established for carcinogenic and / or mutagenic chemical agents (Belgium, Czech Republic, Denmark, Estonia, Latvia, Lithuania, the Netherlands, Portugal, Slovakia, Slovenia, and UK), and only 4 EU MS (Austria, Finland, Poland and Spain) have listed a higher number than 50.

With regard to the selection and prioritization of carcinogenic and mutagenic chemical agents for OEL setting, it is also important with regard to the current initiative that criteria used in EU Member States are very similar to those used in the EU.

Based on the answers of 11 countries, the most important criteria for the selection of chemical agents for setting of OEL appear to be (in order of priority): (1) epidemiological evidence, including reported cases of ill-health in the workplace, (2) availability of toxicological data, (3) severity of effects, (4) number of persons exposed, (5) availability of data on exposure, and (6) availability of measurement methods.

Results of the survey have been used to put together the lists of existing OELs in Annex 6.

9.2.3Consultation of the tripartite Working Party "Chemicals at the Workplace" (WPCs) of the ACSH

Following the Social Partner Consultation, the Commission informed the members of the WPC at its meeting in April 2008 on its intention to propose a revision of the CMD. Information were provided on a possible launch of a call for tender during 2008 with a view to appointing a contractor to carry out an impact assessment of possible amendments of the Directive, covering amongst other the inclusion of certain PGSs in Annex I to the Directive and the revision of existing and the introduction of new OELs for a number of chemical agents in Annex III to the Directive (the so-called IOM study).

At that point in time, the Commission confirmed that the option of covering chemical agents toxic for reproduction under the scope of the revised Directive was now excluded by the Commission. However, in 2010, the Commission launched another call for tender for a study to explore whether or not these chemical agents should be under the scope of the CMD (the so-called RPA study).

At various meetings of the WPC, the progress on the studies was discussed 138 , followed by a first more in-depths discussion on the results of the IOM study based on draft reports for individual chemical agents in March 2011. The discussions on the individual chemical agents took place at various meetings of the WPC in 2011 139 , 2012 140 and 2013 141 , resulting in one opinion and two supplementary opinions adopted by the plenary of the ACSH in 2012 142 and 2013 143 , 144 .

The OEL values agreed upon by the ACSH were taken forward to this initiative.

9.2.4September 2012 - Workshop in Berlin

A workshop ‘Carcinogens and Work-Related Cancer’ was organised by the European Agency for Safety and Health at Work (EU-OSHA) and hosted by the German Ministry of Labour and Social Affairs at their offices in Berlin on 3 and 4 September 2012. About 60 representatives from various European countries, the European Commission, the Advisory Committee on Safety and Health’s Working Party on Chemicals, the Chemex group of the Senior Labour Inspectors Committee (SLIC), the Scientific Committee on Occupational Exposure Limits (SCOEL), the European Chemicals Agency and the International Agency for Research on Cancer (IARC) of the World Health Organisation (WHO) attended.

The aim of the workshop was to summarize the current understanding regarding exposures to carcinogens and the causes and circumstances of work-related cancer, and to discuss how this knowledge can be used across the European Union (EU) to reduce the future burden of these cancers.

The workshop 145 highlighted the need to enhance research efforts to estimate the burden of occupational disease and build on links between occupations and exposures to set priorities for prevention, disease recognition and compensation. In this regard the on-going study HazChem@Work will collect the available occupational exposure data on chemicals across the EU countries. Interim results of this study have shown difficulty in finding data on occupational exposure as it is not routinely collected and centralised at the national level. It is been also identified that the measurements can be performed under different conditions and for different purposes – this can hamper the comparability among different data sets. The final results are expected by the second semester of 2016.

It was generally agreed that the current legislative framework in Europe and its implementation and enforcement is essential for the effective prevention of cancer in the workplace. The need to provide the same level of protection to all workers was also stressed.

9.2.5Consultation of the members of the ACSH on existing national OELs for chemical agents subject to the amendments

In order to establish a base-line scenario for the establishment of OELs subject to the initiative, the Commission services requested from the members of the ACSH at its plenary meeting on 28 November 2013 to submit updated information on the national OELs for the chemical agents covered by the IOM study.

9.2.6Meetings with Industry and Workers representatives

Between the beginning of 2013 and end of 2015, a number of meetings between Commission services and industry and workers representatives concerned about specific chemical agents subject to the initiative took place, as well.

The following organisations, among others, discussed bilaterally with the Commission services on specific chemical agents subject to the initiative:

NEPSi (European Network for Silica formed by the Employee and Employer European sectoral associations),

Euromines and IMA (Industrial Minerals Association) for Silica;

ECFIA (European Ceramic Fibre Industry Association) and Unifrax for Refractory Ceramic Fibers (RCF);

CEEMET (Council of European Employers of the Metal, Engineering and Technology-Based Industries) and Eurometaux for metals as Chromium and Beryllium

BeST (Beryllium Science & Technology Association) for Beryllium.

The main purpose of the meetings asked for by industry was to achieve information on the process for amending the legislation in general, and on the intention of the Commission with regard to the proposed value for a particular chemical agent.

In particular with regard to RCS, the Commission discussed intensively possibilities to recognise on the one hand the results of the NEPSi Agreement, and on the other hand to protect also workers currently not covered by the agreement.

The Commission also participated in meetings organised annually by DG GROW with the European Glass and Ceramic Industry, where similar information were presented.

Some conclusions can be drawn from these meetings regarding the position of the industry representative organisations on specific substances.

For silica, the main concern for industry was the inclusion of the agent in Annex I of the CMD but not the limit value itself, which is attainable in view of the current available technologies and working methods. It was argued that the inclusion of the agent under the scope of the Directive would put a more stringent obligation on substitution and process enclosure, leading to non-affordable costs. It should however be noted that, where no workers were present (e.g. long distance material conveyers), closed processes is not a CMD requirement – and that the substitution principle already applies by means of the Chemical Agents Directive.

On Refractory Ceramic Fibres, the industry expressed their positive view on the establishment of an EU-wide binding OEL as a way to demonstrate appropriate control of the exposure at the workplace – potentially of value in discussions regarding REACH authorisation. The same argument has been presented by the European metals industry as regards Hexavalent Chromium compounds.

10Annex 3 – Who is affected by the initiative and how?

11Annex 4 – Analytical model used in preparing the Impact Assessment 146

The impacts of the different policy options proposed in this impact assessment were quantified, to the extent possible, based on a methodology as described below 147 .

11.1Exposure estimation

The following occupational exposure information was required for each substance for estimating the health impact of any changes in exposure:

-Prevalence of exposure by industry (current);

-Classification of industries into high, medium, low and background exposure, or a subset of these categories;

-Distribution of exposure (the geometric mean (GM) and geometric standard deviation (GSD), ideally by country, across industries, and

-Temporal change in exposure (% change per year) arising from general improvements in European workplaces and work processes, not taking into account the impact of changes to the Carcinogens Directive.

The graphic below provides an overview of the general procedure used for estimating the prevalence of exposure:

Figure 1

Exposure prevalence data were available from the Carcinogens Exposure database (CAREX), for almost all agents analysed in this impact assessment, except 1,2-Epoxypropane, 2-Nitropropane and Hardwood Dust. For wood dust exposure, information on the prevalence (and level) of exposure was available for 25 countries following an EU project estimating the risk of exposure to wood dust (Kauppinen et al, 2005). For the remaining agents information collected from trade associations and other stakeholders was used.

The information from CAREX and other sources, were combined with data from Eurostat (number of workers exposed by relevant sector of activity) to obtain estimates of exposure prevalence.

The level of intensity was assessed using:

-Published scientific literature;

-Information from European Risk Assessment Reports compiled in relation to the Existing Substances Regulations;

-The Woodex database (hardwood dust);

-Information provided by industry stakeholders.

The overall weighted geometric mean (GM) and geometric standard deviation (GSD) exposure level for each agent was estimated across all "medium" and "high" exposure industries across the EU using @Risk (Palisade Corporation, New York).

Where possible, exposures were simulated using GM and GSD for each country. The number of values each industry contributed was weighted according to the number of workers exposed in that industry.

Temporal changes in exposure were determined from information from the literature, which was ideally specific to the substance being considered but in situations where this was not available, the study relied on the results of a systematic review of the literature (Creely et al, 2007).

11.2Health impact – methodology for estimation of the current cancer burden (baseline) as compared with the policy intervention scenarios

In order to assess the current burden of occupational cancer related to the exposure to substances subject to this impact assessment, the analysis made was built on work to quantify the burden of cancer due to occupation in Great Britain (Rushton at al, 2010).

The primary measure of the burden of cancer used in this project was the attributable fraction (AF) i.e. the proportion of cases that would not occur in the absence of exposure; this was then used to estimate the attributable numbers.

The estimates were made considering the risk exposure period (REP) for specific types of cancer: for solid tumours a latency of 10-50 years was assumed and for haematopoietic neoplasms 0-20 latency was assumed. The proportion of the population ever exposed to each carcinogenic agent or occupation in the REP was obtained from the ratio of the numbers ever exposed to the carcinogens of interest in each relevant industry/occupation over the total number of people ever employed. Estimates of employment turnover for grouped main industry sectors and of life expectancy were used to estimate the exposed population, and adjustment factors were applied to the exposure prevalence data to take account of the change in numbers in the industry sector groups.

The attributable fraction (AF) for each cancer/occupational carcinogen was estimated using Levin's method (Levin, 1953).

The relative risk (RR) for cancer(s) in question for the relevant agent or work environments, were derived from a review of the published epidemiological literature. Risk estimates were obtained from key studies, meta-analyses or pooled studies, taking into account quality, relevance to the EU, sample-size, effective control for confounders, adequate exposure assessment, and clear case definition.

For predicting the future burden, the risk exposure windows were projected forward in time, and estimation was carried out for a series of forecast target years (FTYs) that stretch far enough into the future to account for the latency of cancers initiated at the time when the study was performed (i.e. the decade starting 2060). Estimates were made for alternative scenarios of changes in exposure levels and proportions exposed, for example assuming the introduction of new or reduced exposure limits, which were assumed introduced in 2010. The projections were made each time under the assumption of full compliance with the legislation (i.e. 99% of exposures less than the limit value).

To predict future cancer numbers based on the pattern of past and current exposure either a "static" baseline, where no change in exposed numbers or exposure levels is expected beyond 2010, or a "dynamic" baseline was used, where current trends are forecast to continue until 2030.

The socioeconomic health impacts of the different policy options were then assessed in terms of the cost of the disability and death based on the estimated cancer burden under each policy option.

In this respect, several approaches exist to assessing potential health impacts ranging from non-monetary approaches such as Quality Adjusted Life Years (QALY), Disability Adjusted Life Years (DALY) and Health Life Years (HLY) to monetary methods such as the Cost of Illness (COI), Value of Statistical Life (VSL) and Value of Life Year Lost (VLYL).

As part of this study, Imperial College have developed a model to estimate DALYs based on exposure data from Institute of Occupational Medicine (IOM) and the Finish Institute for Occupational Health. The DALYs uses time as a common metric taking into consideration both premature death and ‘healthy’ years lost due to problems associated with living with the disease or health condition (i.e. cancer for this study). DALYs are calculated as the sum of the years of life lost due to premature mortality (YLL) and the years lost due to disability (YLD):

DALY = YLL + YLD

Each DALY represents one lost year of ‘healthy’ state. The DALYs can be used as one approach to compare different options (e.g. cost effectiveness analysis) especially when different Occupational Exposure Limit (OEL) values have been proposed. Entec have also used the underlying data developed by Imperial College to attempt to monetise health impacts associated with introducing EU-wide OELs. This will allow for a more formal cost benefit analysis to be used to compare different policy options. This paper sets out this approach to monetising human health impacts.

The approach used to estimate a monetary value on changes in health impacts is dependent on the data available such as the population at risk (i.e. data on the exposed population) and any evidence of dose-response relationships. Since it is not possible to develop dose-response functions for each chemical agent, the approach to valuing changes in human health is based on estimating the monetary loss (damages or costs) that might occur if no changes were made (Business-As-Usual scenario) in comparison to the avoided health related costs under the introduction of an EU-wide OEL level(s). The difference in health impacts between the BAU scenario and the scenario(s) with an OEL is the main health benefits valued in this project.

The valuation of health impacts are divided into two main aspects:

Life years lost – This is calculated by using the year’s life lost (YLL) estimated by Imperial College and multiplying this with a valuation of the Value of Life Year Lost (VLYL). This values the time (years) lost due to premature death.

Cost of Illness (COI) – This is often the main market-based approach in relation to health impact (ECHA 2008) 148 . Depending on the valuations available, it can include the direct, indirect and intangible costs of cancer. This is a monetary cost of the time spent with cancer. In this study, a unit COI estimate is multiplied by the number of cancer registrations.

Each of these two impacts is explained in more detail below as well as using willingness-to-pay (WTP) estimates as an alternative approach.

Value of life years lost (VLYL)

The years of life lost (YLL) are estimated by multiplying the number of disease specific deaths times average life expectancy after average age at death from the specific disease 149 . EU and Member State specific average life expectancies were used for this project. Essentially years of life lost are the difference between death and average life expectancy (‘premature death’). This is illustrated in Figure 2.

Figure 2    DALY component: Years of life lost (VLL)

Monetary estimates of the value of life years lost (YLYL or sometimes known as VOLY 150 ) allows us to put a value on VLL. The latest EC Impact Assessment guidance applicable at the time of the study (EC 2009) 151 suggested using estimates of €50,000-100,000 in Europe for the purpose of an Impact Assessment, if no more specific estimates are available. Markandya (2003) uses an estimate of €50,000 and is also used more widely in other assessing health policies such as CAFE. Therefore for the purposes of this study the €50,000 is used as a lower estimate and €100,000 as an upper estimate. This should therefore help encompass the uncertainties associated with the VLYL.

These valuations are increased by 2% each year in the future in part to present costs in real terms (i.e. adjusting for inflation in prices) and to reflect societies increasing value attached to their health (as economic growth typically increases over a long period of time).

The values originally reported in the IOM study, based on a constant discount rate of 4%, have been recalculated applying a declining discount rate (4% for the first 20 years, 3% thereafter) in line with the most recent EC Better Regulation guidelines 152 .    

Cost of illness (COI)

Introduction

The cost of illness (COI) is one of the most common market based approaches to valuing health impacts. It involves multiplying the number of cancer registrations occurring under each scenario (i.e. with and without proposed changes) with the valuation for COI.

The COI might include health sector costs (direct costs), the value of lost productivity by the patient (indirect cost), and the cost of pain and suffering (intangible costs) 153 . This will however depend on data availability as in most cases intangible costs are unlikely to be included in valuations of COI. These three components are described in Table 2.

Table 2 Components making up a valuation of COI

Estimating COI

Outlined below is an approach to valuing COI for cancer (excluding intangible costs):