EUROPEAN COMMISSION

Brussels, 18.7.2025

SWD(2025) 192 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT REPORT

Accompanying the document

Proposal for a Directive of the European Parliament and of the Council

amending Directive 2004/37/EC as regards the addition of substances and setting limit values in its Annexes I, III and IIIa

{COM(2025) 418 final} - {SEC(2025) 217 final} - {SWD(2025) 191 final} - {SWD(2025) 193 final}

Table of contents

1.Introduction: Political and legal context

1.1.Political context

1.2.Legal context

1.3.Evidence-based procedure supporting the continuous revision of the CMRD

2.Problem definition

2.1.What is/are the problems?

2.2.What are the problem drivers?

2.3.How likely is the problem to persist?

3.Why should the EU act?

3.1.Legal basis

3.2.Subsidiarity: Necessity of EU action

3.3.Subsidiarity: Added value of EU action

4.Objectives: What is to be achieved?

4.1.General objectives

4.2.Specific objectives

4.3.Consistency with other EU policies

5.What are the available policy options?

5.1.What is the baseline from which options are assessed?

5.2.Description of the policy options

6.What are the impacts of the policy options?

6.1.Cobalt and inorganic cobalt compounds

6.2.PAHs

6.3.1,4-dioxane

6.4.Welding fumes

7.How do the options compare?

7.1.Cobalt and its inorganic compounds

7.2.PAHs

7.3.1,4-dioxane

7.4.Welding fumes

8.Preferred option

8.1.Selection of the preferred policy options and analysis of impacts

8.2.Subsidiarity and proportionality

8.3.REFIT (simplification and improved efficiency)

8.4.Application of the ‘one in, one out’ approach

9.How will actual impacts be monitored and evaluated?

9.1.Monitoring arrangements

9.2.Evaluation arrangements

Annex 1: Procedural information

Annex 2: Stakeholder consultation (Synopsis report)

Annex 3: Who is affected and how?

Annex 4: Analytical methods and multi-criteria analysis

Annex 5: Competitiveness check

Annex 6: Relevant sectors, uses, exposure concentrations and national regulatory frameworks

Annex 7: Route(s) of exposure and adverse health effects

Annex 8: Process for setting binding limit values and associated provisions under the CMRD

Annex 9: Data supporting the setting of policy options

Annex 10: Accompanying measures and options discarded

Annex 11: Consistency and complementarity with the REACH Regulation

Annex 12: SME test

Glossary

1.Introduction: Political and legal context

1.1.Political context

1 2 In June 2021, the Commission presented the EU strategic framework on health and safety at work 2021-2027 (‘EU OSH strategic framework’). This strategic framework aims to turn into concrete actions principle 10 enshrined in the European Pillar of Social Rights (‘EPSR’) related to workers’ right to a high level of protection of their health and safety at work. In particular, it strengthens prevention culture, both in organisations and among individual workers to reduce work-related deaths as much as possible, in line with a Vision Zero approach to work-related deaths in the EU.

One of the objectives of the strategic framework is the prevention of workplace accidents and illnesses, which can cause serious adverse health effects on workers, including fatalities, and which lead to an estimated annual loss of around 3.3% of gross domestic product in the EU.

As shown in figure 1, cancer remains the first cause of work-related death in the EU, followed by cardiovascular diseases (CVD). Every year, according to estimates, about 80,000 people in the EU lose their lives due to exposure to carcinogens at the place of work.

Figure 1: work-related deaths in the EU, ICOH 2019 3

4 Exposure to reprotoxic substances at the place of work can also lead to serious adverse health effects, by affecting sexual function and fertility. It might also cause adverse effects on the development of workers’ offspring. Up to 1,274 reproductive ill-health cases in the EU could occur every year following exposure to reprotoxic substances that are not classified as carcinogenic or mutagenic.

5 To improve the prevention of work-related diseases in the EU, the Commission is pursuing a continuous update of the Carcinogens, Mutagens and Reprotoxic substances Directive 2004/37/EC (‘CMRD’). This revision process is in line with the obligation laid down in the CMRD to set out limit values for all those carcinogens, mutagens or reprotoxic substances for which this is possible. It also responds to requests made notably by: 

·stakeholders for considering the need to adopt limit values for more substances, as highlighted in the REFIT occupational safety and health (‘OSH’) evaluation 6 (‘REFIT OSH evaluation’); and

·the European Parliament and the Council as part of the fourth revision of the CMRD 7  (‘CMRD 4’) to achieve new or revised occupational exposure limit values (‘OELs’) for at least 25 substances; and

·14 Member States 8 in a letter dated 6 March 2024 and sent to the Commission to request limit values for more substances in the future.

9 Since the initiation of this continuous process in 2016, the EU has already adopted five revisions of the CMRD, addressing more than 40 key hazardous chemicals and contributing to saving the life of more than 100,000 workers over the next 50 years. The fifth legislative proposal for a revision of the CMRD, addressing lead, a key reprotoxic substance, was formally adopted in February 2024 and entered into force on 8 April 2024. The initiative subject to this impact assessment is therefore the sixth revision of the CMRD. It aims to further improve workers’ protection and save lives in the future. 

This continuous revision process relies on an inclusive approach where the Commission involves scientists from the European Chemicals Agency’s (‘ECHA’) Risk Assessment Committee (‘RAC’), Member States and social partners represented in the Advisory Committee on Safety and Health at Work (‘ACSH’), and the social partners through the formal consultation provided by Article 154 of the Treaty on the Functioning of the European Union (‘TFEU’). The ACSH is composed of three full members (and six alternates) per Member States, representing national governments, trade unions and employers’ organisations. Its composition reflects the various economic sectors concerned (e.g. construction, industry, etc.) and the different sizes of companies (incl. SMEs), so that its opinions are based on the realities on the ground of the different sectors and companies. Furthermore, as it also include one full member (and two alternates) from each Member States, its opinions also take account of the economic and social specificities of each country.

As part of the 2nd stage consultation of social partners, BusinessEurope and the European Trade Union Confederation (‘ETUC’) supported the ACSH opinions while SMEunited called on the Commission not to go beyond what was agreed by the ACSH. This inclusive process allows for revisions of the CMRD which provide a high level of protection of workers’ health while being socio-economically viable and technically feasible for businesses. More information on the evidence-based procedure supporting the continuous revision of the CMRD is available in section 1.3 and annex 8.

10 11 The planned sixth revision of the CMRD (‘CMRD 6’), subject to this impact assessment, fits into this continuous revision process and contributes to the European Health Union. It also delivers on the European Parliament’s and the Council’s request through CMRD 4 to set an OEL for cobalt and inorganic cobalt compounds, and the substances identified for this revision are based on the Commission Staff Working Document (SWD) listing the priority substances to be scientifically assessed.     

12 Importantly, hazardous substances are not limited to carcinogens, mutagens and reprotoxins. There are other types of substances that fall under EU OSH legislation, in particular the Chemical Agents Directive (‘CAD’), that cause other types of ill-health effects (e.g. respiratory ill-health) and should be tackled.

13 14 15 By improving the prevention of work-related diseases, the CMRD 6 will also contribute to the United Nations 2030 Agenda for a Sustainable Development, in particular its goal 3 (“Ensuring healthy lives and promote well-being for all at all ages”) and goal 8 (“Promote sustained, inclusive and sustainable economic growth, full and productive employment and decent work for all”).

16 In order to collect more recent information to support this impact assessment, the Commission contracted an external study. This study was scrutinised by the relevant Commission services and members of the ACSH’s Working Party on Chemicals (WPC). 

1.2.Legal context

17 The main legislative instrument to ensure workers' protection against risks related to carcinogenic, mutagenic or reprotoxic chemicals is the CMRD, an individual directive under the OSH Framework Directive.

18 Whether a substance or mixture is under the scope of the CMRD is primarily based on its classification as a carcinogen, mutagen and/or reprotoxic substance (category 1A or 1B) according to the criteria established under the CLP (classification, labelling and packaging of substances and mixtures) Regulation.

There is also a possibility to bring a substance/mixture under the scope of the CMRD by including it in Annex I to the CMRD. This Annex covers substances, mixtures or processes (or substances/mixtures released by a process referred to in that Annex) which are not classified according to the CLP Regulation as carcinogens, mutagens or reprotoxic substances but are for example recognised by other international bodies (e.g. the International Agency for Research on Cancer - IARC) as substances, mixtures or processes of equal concern. These substances are often referred to as Process-Generated Substances (PGS). Annex I currently contains 8 entries.

To reduce the occupational exposure to these substances or mixtures, the CMRD provides for a hierarchy of preventive and protective measures. The employer has the obligation to substitute these chemicals by less or non-hazardous substances, mixtures or processes as far as technically possible. If substitution is not technically possible, other measures to prevent exposure have to be put in place by the employer, such as working in a closed system or reducing the number of workers potentially exposed.

Another obligation of the employer is to ensure that binding OELs and biological limit values (BLVs) set out in Annex III and Annex IIIb to the CMRD respectively shall not be exceeded. For binding EU OELs and BLVs adopted under the CMRD, Member States must establish a corresponding national limit value at least at the EU level or lower value (more protective).

1.3.Evidence-based procedure supporting the continuous revision of the CMRD

Setting new or revised limit values for carcinogens, mutagens or reprotoxic substances, or enlarging the list of substances, mixtures and processes in Annex I, requires a revision of the CMRD. To support this revision process, the Commission relies on a sound, comprehensive and evidence-based procedure, during which the consultation of the relevant stakeholders is key. This procedure ensures that the revisions of the CMRD are based on scientific evidence, while technically and socio-economic feasible for businesses of all sizes.

The figure below summarises this stepwise procedure starting with the prioritisation of the substances to be addressed and ending with the publication of the Directive adopted by the European Parliament and the Council, based on the legal proposal by the Commission, in the Official Journal of the EU. The figure 2 includes a brief description of each step. The procedure is explained in more detail in annex 8.

Figure 2 : evidence-based procedure supporting legislative initiatives in the area of OSH

The social partners and Member States play an important role throughout this procedure. In particular, they are involved in the following steps:

·Selection of chemicals for scientific evaluation: the members of the tripartite ACSH’s Working Party on Chemicals make recommendations to the Commission on the substances that should be prioritised in the subsequent revisions of the CMRD;

·Scientific opinions: stakeholders, including businesses and trade unions, are given the possibility to express their views and concerns in the early phases of developing a scientific report on limit values on a substance at the workplace;

·Two-stage social partners consultation: EU social partners are consulted on the possible direction of each initiative and on its content. As part of this two-stage consultation, social partners share among other their views on the prioritised substances, the suggested approaches to address them and the opinions of the ACSH;

·ACSH opinions: the ACSH is composed of three full members per Member States, representing national governments, trade unions and employers’ organisations. It adopts opinions for each substance, in which it recommends limit values or other measures (e.g. inclusion of a process-generated substance in Annex 1 to the CMRD). In its opinions, the ACSH may recommend transitional measures aiming at giving sufficient time to businesses to find technical solutions enabling them to comply with the new limit values or measures. The broad composition of the ACSH allows for the covering a large number of sectors and activities. This also ensures that its opinions reflect the current situation of each sector on the ground, including in terms of technical and socio-economic feasibility;

·External study supporting impact assessment: some members of the ACSH’s Working Party on Chemicals representing the three interest groups are actively involved in the steering committees for the external studies supporting impact assessments. Their practical expertise contributes to the quality of the studies and their estimates, including the setting of thorough and realistic assumptions.  

This extensive consultation and involvement of relevant stakeholders, in particular the ACSH and its Working Party on Chemicals, allows for effective, feasible and proportionate regulatory measures. It also enables to identify whether transitional measures are necessary for the development of technical solutions allowing for compliance with these new regulatory measures. Annex 2 provides a summary of stakeholder consultation and their feedback that informed the development of the policy options outlined in this report.

2.Problem definition

2.1.What is/are the problems?

The REFIT OSH evaluation 19  concluded that the OSH directives, including the CMRD, remain highly relevant and effective according to the available evidence. They contributed to the decrease in the incidence and number of accidents at work. 

20 Despite the protective EU legal framework and progress achieved over the last decades, according to estimates, more than 170,000 workers still lose their lives each year due to work-related illnesses. This shows the need to further improve prevention of work-related diseases in the EU, which has consequences for workers and their families, businesses and public authorities.

To address this problem, the Commission is continuously updating the CMRD to improve the protection of workers from the exposure to hazardous substances, in particular the 28 carcinogens, mutagens or reprotoxic substances to be scientifically assessed identified in the Commission’s  SWD 21  published in 2022. These scientific assessments to be carried out by ECHA’s RAC are key as they provide most recent scientific data on the hazards of a substance and contribute to the decision of whether further legislative actions are required.

Providing scientific assessment requires time and resources. It is therefore not possible to address these 28 substances through one single initiative. Therefore, the Commission has prioritised 5 substances, groups of substances or process-generated substances (hereafter ‘substances’) 22 , namely cobalt and its inorganic compounds, polycyclic aromatic hydrocarbons (‘PAHs’), isoprene, 1,4-dioxane and welding fumes. The selection of these 5 substances was based on a consultative approach including social partners and Member States represented in the WPC, and taking into account other considerations, such as

·Cobalt and its inorganic compounds: legal obligation 23  under CMRD 4 to set limit values for cobalt and importance to improve workers’ protection from a substance whose consumption is expected to increase by 330% by 2050 with the green transition.

·PAHs: high number of workers potentially exposed to this group of substances.

·1,4-dioxane: recent classification as carcinogen 1b in the CLP Regulation.

·Isoprene: high number of Member States with no national limit value.

·Welding fumes: high number of workers potentially exposed to these process-generated substances and absence of classification at EU-level in the CLP Regulation despite the IARC classification from 2018, see section 2.2.2.  

According to the evidence gathered, including scientific and socioeconomic data, workers, businesses and Member States are impacted by the insufficient prevention of occupational exposure to cobalt and its inorganic compounds, PAHs, 1,4-dioxane and welding fumes. Conversely, the evidence gathered indicates that workers are exposed to levels of isoprene which are lower than the health-based limit value 24 derived by RAC in its opinion 25 , suggesting that the current prevention of occupational exposure to isoprene is sufficient.

For workers, the insufficient prevention may lead to serious health problems, including cancer. The table 1 below presents estimates of the number of cancer and other ill-health cases occurring in 2023 due to past workers’ exposure (current disease burden) and over the next 40 years due to future exposure (future disease burden). For cancer cases, welding fumes, PAHs and cobalt and its inorganic compounds are the most problematic. For example, workers’ exposure to welding fumes alone would lead to close to 30,000 lung cancer cases over the next 40 years. Although no cancer case is expected for 1,4-dioxane, workers’ exposure to this substance causes other diseases as shown in table 1. More information on trends used for estimates is provided in section 5.

Table 1: estimates on number of workers exposed in the EU and estimated current and future burden of disease

For businesses, work-related diseases (cancer and other diseases) imply costs, including in terms of reduced productivity. 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. The table below includes an estimate of the costs to employers for the 5 substances subject to this impact assessment, which includes the cost due to insurance payments and absence from work, and the productivity loss due to mortality.

In addition to significant social and financial burden to those affected by work-related diseases and deaths, the diseases are also associated with significant costs to society. For public authorities, work-related diseases lead 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. According to estimates 26 , direct costs of work-related cancer in terms of healthcare and productivity losses at the EU level are between €4 – 7 billion per year; and the indirect costs may reach between €242 and €440 billion each year. The table below includes an estimate of the costs to public authorities for the 5 substances subject to this impact assessment, which includes healthcare costs and loss of tax revenues.

Table 2: costs of ill-health for businesses and public administration (in € million, over 40 years)

The problem tree below presents the identified problem, as well as its drivers and consequences. The problem and its drivers are further described above. More detailed information on the four identified drivers is provided in the next section.

Figure 3: problem tree

2.2.What are the problem drivers?

2.2.1.Driver 1: high exposure levels to cancer-causing substances 

As shown in section 1, cancer is the main cause of work-related diseases and deaths in the EU. Substances classified as carcinogenic can also cause other diseases. For instance, exposure to PAHs can lead to male infertility and developmental toxicity.

According to the evidence gathered 27 , more than 2.6 million workers in the EU are exposed today to the 5 substances considered in this impact assessment.  In order to adequately protect their workers, employers have to perform a risk assessment and implement the necessary risk management measures (RMMs). The type of RMMs required depends on several factors such as the nature of the work, the levels of exposure or the duration of exposure.

Today, some workers are still exposed to levels of substances that are too high and might lead to hazardous health problems, including cancer, as summarised in table 1. For instance, the evidence gathered indicates that workers operating in the manufacture of tools sector are exposed on average to 4.8 µg cobalt/m³ (inhalable fraction), while in 1% of these workers are exposed to levels higher than 27.9 µg cobalt/m³ (inhalable fraction). These levels are higher than the health-based limit value of 1 µg/m³ (inhalable fraction) derived by RAC in its opinion 28 , which corresponds to the value below which the health risk for workers is nonexistent.

Conversely, the evidence gathered 29 indicates that the current levels of workers’ exposure to isoprene are lower than the threshold limit value derived by RAC in its opinion 30 . Therefore, workers exposed to isoprene in the EU are not expected to experience adverse health effects. More information on the current exposure levels of workers to cobalt, PAHs, 1,4-dioxane and isoprene are available in annex 6.

These estimates of cancer and non-cancer cases reflect the insufficient RMMs currently in place in some sectors or companies resulting in some workers being exposed to hazardous levels. Following a survey 31 conducted for the purpose of this impact assessment, businesses indicated if additional RMMs would be necessary to comply with certain limit values.

To continue with the example of manufacturing of tools sectors, 30% of the respondents indicated that 30% of processes are fully enclosed, 20% partially enclosed, 20% have open hood systems, 10% have high efficiency particulate air (HEPA) filters and for the remaining 20%, workers wear simple masks. According to these respondents, complying with the health-based limit value for cobalt would require additional RMMs, in particular fully enclosed systems (33%). Only 10% of the processes would not require any additional RMMs. More information on the current and required RMMs for complying with certain limit values is provided in Annex 6.

In conclusion, the insufficient take-up of the relevant RMMs results in workers exposed to hazardous levels of carcinogenic substances.

2.2.2.Driver 2: misalignment of the EU Regulatory framework with the latest scientific data

Article 16 of the CMRD provides that EU-wide limit values should be established for all those carcinogens, mutagens or reprotoxic substances for which the available information makes this possible. Indeed, EU-wide limit values must be regarded as an important component of the general arrangements for the protection of workers. They provide clear benchmark for employers when implementing RMMs so workers’ exposure cannot exceed limit values.

Despite the existence of EU-wide limit values for 39 hazardous substances within the CMRD, the following three substances subject to this assessment do not have yet a binding EU limit value although scientific data are available: cobalt and inorganic cobalt compounds, polycyclic aromatic hydrocarbons, 1,4-dioxane.

The necessary scientific data was provided by RAC in substance-related opinions in 2022. A summary of this scientific information is provided in annex 2. These opinions provide all the necessary information to establish limit values in the CMRD, as required by Article 16 of the CMRD.

The misalignment of the CMRD with the latest available scientific data may affect its effectiveness with insufficient RMMs in place in some sectors. As a result, workers are more likely to be exposed to too high levels of these 4 substances. More details on exposure levels are provided in section 5.

2.2.3.Driver 3: disparate or inexistant limit value at national level 

While employers must comply with their minimisation obligations stemming from Articles 4 and 5 of the CMRD in all cases, limit values provide very relevant benchmarks enabling them to know exactly the levels above which exposure cannot occur. Limit values also allow employers to determine the level below which their RMMs should ensure compliance with the obligation to reduce the exposure to as low a level as it is technically possible. Likewise, limit values support enforcement authorities in controlling that employers are putting in place the relevant RMMs, including those that could contribute to lower the exposure below the limit values.

The absence of EU-wide binding limit values for the three substances does not prevent Member States to establish their own national limit values. According to the evidence gathered 32 , several Member States have already set binding limit values for one or several of these substances. The table below includes the lowest and highest existing binding OELs in Member States for each substance, as well as the number of Member States with no binding OEL. More information on the current national OELs is available in Annex 6.

Table 3: national binding limit values in EU Member States

A high number of Member States have currently no binding OEL for PAHs. With regard to cobalt and inorganic cobalt compounds and 1,4-dioxane, a significant number of Member States already have an OEL. However, the OELs currently in place in most of the Member States are much higher than the values recommended by RAC and hence likely not sufficiently effective for protecting workers’ health. Furthermore, diverging limit values across the EU result in an unlevel playing field within the single market. Indeed, businesses operating in Member States with more stringent limit values will have to bear higher costs than their competitors subject to less strict rules. Aside from improving workers’ protection to exposure to hazardous substances at the workplace, employers’ representatives are of the opinion that setting binding limit values at EU level helps to provide a level playing field for industry 35 .    

An inadequate OEL creates no incentive for the implementation of the most modern production technologies and RMMs allowing for a reduction of the occupational exposure to hazardous substances.

The evidence gathered 36 suggests that some companies would invest in more effective RMMs if they would have to operate under stricter OELs.

2.2.4.Driver 4: insufficient awarness of the possible dangerousness of welding fumes

In 2018, IARC classified welding fumes and UV radiation from welding as “carcinogenic to humans” (Group 1). Exposure to welding fumes may lead to lung damage and various types of cancer, including of the lung, larynx and urinary tract. As process-generated substances and despite the IARC classification, welding fumes are not classified according to the CLP Regulation.

As mentioned by ECHA in its scoping study 37 , welding fumes are complex and have variable compositions. The content of welding fumes depends largely upon the materials being welded. The majority (95%) of the components of welding fumes are emitted from the fillet or consumables used and only 5% from the base material 38 .

The dangerousness of welding fumes depends also on the process used, as shown in the ECHA scoping study. For instance, beam welding 39 is unlikely to lead to workers’ exposure as it is almost completely automated while arc welding 40 workers are likely to be exposed to carcinogens, particularly in the craft sector where automated process is less prevalent. A summary of the welding processes, the generated substances, the indication of the presence of carcinogens, mutagens or reprotoxic substances, and the workers’ exposure is available in Annex 6.

Due to the complexity and heterogeneity of the composition of welding fumes, as well as the absence of harmonised classification in the CLP Regulation, employers might ignore that the fumes generated by their welding activities might include carcinogens, mutagens or reprotoxic substances. Consequently, those employers might not carry out the mandatory risk assessment provided in the CMRD when workers are or are likely to be exposed to carcinogens, mutagens or reprotoxic substances as a result of their work. In the absence of such a risk assessment, employers might not implement all necessary risk management measures to comply with the CMRD and adequately protect their workers.

The evidence gathered 41  confirms that some employers are not aware that welding fumes may contain carcinogens, mutagens or reprotoxic substances (such as chromium (VI) compounds, nickel compounds, cadmium and its inorganic compounds, and beryllium and its inorganic compounds) and lead to health risks. Further, according to the evidence gathered, this lack of awareness is considered as an important driver by stakeholders 42 .  

These employers may therefore ignore to what extent exposure to their welding processes and materials used can be carcinogenic, mutagenic or reprotoxicant for their workers and do not apply the measures prescribed by the CMRD to protect workers. According to the evidence gathered 43 , the improvement of risk management measures following the introduction of welding fumes in Annex I to the CMRD will concern about 1% of workers involved in welding activities 44 .

As a consequence of the legal unclarity, and despite the fact that welding can be a relatively safe activity if all the safety and exposure control measures are in place, it remains a prominent concern that welders are at high risk from various diseases, including cancers 45 (see Table 1). In the absence of sufficient data, the geographical distribution of the level of compliance with the CMRD in individual Member States cannot be provided.

2.2.5.Driver 5: lack of inspections for non-compliance with the CMRD

The lack of inspections was also identified as a driver of the insufficient prevention in the EU to reduce prevalence of work-related diseases due to welding fumes-related activities 46 . However, as labour inspection is a prerogative of Member States and cannot be addressed by the CMRD, this driver was not reflected in the intervention logic.  

2.3.How likely is the problem to persist?

In the absence of action at EU level, workers exposed to cobalt and its inorganic compounds, PAHs, 1,4-dioxane or welding fumes are expected to continue to be at risk of contracting work-related cancer and other adverse health effects. Estimates on the number of future ill-health cases are provided in table 4. Additional information, including the monetised health costs in case of no EU action (baseline scenario), are provided in section 5.

The general obligations set by the CMRD, employers’ actions and the corresponding implementing measures adopted by Member States contribute overall to lowering exposures. According to the evidence gathered 47 , exposure levels to the five substances have generally been decreasing in the past years and this positive trend is expected to continue in the future for PAHs and welding fumes 48 . Substitution may be possible for some chemicals and uses in the future and technological developments could lower exposure concentrations. Furthermore, the number of workers employed in the industries using these carcinogens may change, as further described in section 5. Despite these trends, the insufficient prevention of work-related diseases occurring from the exposure to these 4 substances is expected to persist.

Future forecasts in this area are however far from certain due to scarcity of relevant data and the fact that market forces such as raw material and energy prices, developing technology, as well as regulatory changes can drive decreases or increases in use which are not easy to predict. Even if trends were overall positive, the existing employers’ practices as well as protective measures at Member State level do not always reflect the most recent scientific and technological knowledge. For that reason, the retained assumptions are based on the legal provisions contained in the CMRD, including the minimisation requirements, on information gathered from the stakeholders (via targeted consultations), as well as on the expertise of the employers, workers and government representatives that steered the supporting study (RPA 2024) analysis. The objective has been to define a baseline scenario as close as possible to the future situation. More information about the past and future trends that would influence the future burden of disease is available in section 5.1.

Member States usually do not inform the Commission on their intentions to revise existing or determine new OELs in their national legislation. However, national administrations represented in the ACSH are aware of the preparatory work at EU level and therefore it is likely they will await its results in order not to duplicate efforts.

Table 4: estimated future burden of disease over 40 years

3.Why should the EU act?

3.1.Legal basis

Article 153 of TFEU 49  empowers the EU to support and complement the activities of the Member States as regards improvements, in particular of the working environment to protect workers' health and safety and to adopt, by means of directives, minimum requirements for gradual implementation, having regard to the conditions and technical rules obtaining in each of the Member States.

On the basis of this provision, Article 16 of the CMRD sets out that limit values are to be adopted in respect of those carcinogens, mutagens or reprotoxic substances for which this is possible, having regard to the available information, including scientific and technical data.

3.2.Subsidiarity: Necessity of EU action

Cancer can affect every citizen in one way or another, regardless of their country of origin or residence. Faced with this issue, the Commission adopted in 2021 its Europe’s Beating Cancer Plan 50 . In this plan, the Commission recognises the need to mobilise the collective power of the EU to drive change to the benefit of the EU citizens.

As part of this plan, the Commission highlighted the need to reduce the exposure to carcinogens at the workplace, through the revision of the CMRD. It is in line with the REFIT OSH evaluation 51  which considers the CMRD of high relevance to protect workers at the EU level. According to contributions received from national and EU stakeholders during that evaluation process, further occupational limit values should be established at the EU level.

A consistent level of protection from the risks related to carcinogens, mutagens or reprotoxic substances must be established for the EU as a whole by a framework of general principles to enable Member States to apply the minimum requirements consistently. Action at EU level ensures that all workers within the EU will benefit from a similar high level of protection, in line with the latest scientific evidence.

EU-wide limit values must be regarded as an important component of the general arrangements for the protection of workers. Article 16 of the CMRD provides that EU-wide limit values should be established for all those carcinogens, mutagens of reprotoxic substances for which the available information makes this possible. Following the request from the European Parliament and the Council made as part of the fourth revision of the CMRD, the Commission published in 2022 a SWD 52 listing 28 priority chemicals for scientific assessment in view of the subsequent revisions of the CMRD. Among them are the 5 substances subject to this impact assessment. 

As shown in table 4 above, future occupational exposure to cobalt and its inorganic cobalt compounds, PAHs and welding fumes is expected to result in more than 30,000 cancer cases over 40 years. Therefore, it is urgent to improver workers’ protection from the exposure to these 4 substances, in line with the Europe’s Beating Cancer Plan. With regard to 1,4-dioxane, no cancer cases are expected from occupational exposure over the next 40 years. However, workers’ exposure to this substance can lead to other diseases which represent a burden for workers, employers and Member States across the EU, justifying an action at EU level.

As for the fifth substance, isoprene, the evidence gathered 53 indicates the current and future levels of workers’ exposure are already lower than the health-based limit value derived by RAC in its opinion 54 . Despite the unanimous support of representatives from businesses, workers and Member States within the ACSH for setting a limit value for isoprene, an action at EU level does not seem necessary for improving workers’ protection.

3.3.Subsidiarity: Added value of EU action

Improved clarity and enforcement

Establishing limit values for cobalt and inorganic cobalt compounds, polycyclic aromatic hydrocarbons and 1,4-dioxane will provide common reference points that are used as a practical tool by employers, workers and enforcers to assess compliance with the general requirements, in particular in those Member States with no existing limit values. Limit values can also be used by process plant and machinery designers when planning new or considering alterations to existing process plants.

The complexity and heterogeneity of the composition of welding fumes, together with the absence of harmonised classification for welding fumes in the CLP Regulation, contribute to a lack of clarity on the possible dangerousness of welding fumes for workers, and therefore a lack of appropriate RMMs. Addressing this absence of classification for welding fumes at EU level would ensure more legal clarity, which would result in better implementation of the existing rules.

Governments’, employers’ and workers’ representatives have expressed clear support for establishing limit values for the substances subject to this initiative, and including welding fumes in Annex I to the CMRD, as it clearly results from the two-stage consultation of social partners and the opinions of the tripartite ACSH. Detailed information on the outcome of the social partners consultation and the ACSH opinions is available in annex 2.

Ensuring a similar minimum level of protection across the EU

The CMRD aims to improve workers’ health by protecting against risks arising or likely to arise from exposure to carcinogens, mutagens or reprotoxic substances at work. This objective is in line with the and Europe’s Beating Cancer Plan 55 .

Limit values under the CMRD are an important component of the general arrangements for the protection of workers 56 and should be established for all those carcinogens and mutagens for which the available information makes this possible.

The CMRD does not prevent Member States to adopt national limit values in the absence of EU values. However, the national limit values for the 3 substances subject to the setting of a limit value  may vary considerably between Member States, where they exist. For instance, and as set out in table 3, the national limit values for cobalt range from 10 to 500 µg/m3. Some Member States have therefore set limit values fifteen times lower than others for the same substance, while five Member States have no limit values for cobalt 57 . In the absence of any action at the EU level, very different levels of workers’ protection are likely to persist.

Contribution to a more levelled playing field

Some employers' organisations stressed in their response to the social partner consultation that setting EU limit values helps to provide a level playing field for industry. The costs of complying with lower national levels are generally higher and entail, therefore, a competitive advantage for enterprises operating in markets with no or less stringent national limit values.

Setting EU limit values will not completely eliminate the differences between Member States since they can adopt more protective (lower) limits. However, it will reduce the scope for divergences and enhance certainty that there is a core definition and/or enforceable exposure limit for all concerned carcinogens in all Member States. It will also reduce regulatory complexity resulting from highly diverging rules between Member States, contributing to reduce the administrative burden of compliance for businesses operating across the single market.

Reducing burdens related to derivation of limit values

The process of establishing limit values is very complex and requires a high level of scientific expertise, as well as an impact assessment and discussions with the stakeholders (depending on national practices). An important advantage of setting OELs at EU level is that it eliminates the need for Member States to conduct their own analysis, including scientific assessment, with likely substantial savings on administrative costs.

4.Objectives: What is to be achieved?

4.1.General objectives

This initiative contributes to the improvement of health and safety of workers pursuant to Article 153 of the TFEU. It aims at complying with the legal obligations laid down in Article 16 of the CMRD related to the setting of limit values, and to prevent work-related deaths and diseases in line with the second key objective of the EU OSH strategic framework 58 .

This report evaluates the impact of setting limit values for cobalt and inorganic cobalt compounds, PAHs and 1,4-dioxane, as well as to include welding fumes in the Annex I to the CMRD for the purpose of legal clarity. As no problem with workers’ exposure to isoprene was identified, the main report does not include any assessment of options for limit values. However, as the ACSH recommended a limit value for isoprene, a thorough assessment of this option is provided in Annex 10.

4.2.Specific objectives

The specific objectives are:

·To further improve workers’ protection from exposure to cobalt and its inorganic compounds, PAHs and 1,4-dioxane in the EU through the adoption by employers of appropriate risk management measures;

·To increase the clarity and effectiveness of the CMRD by keeping it up to date with the latest scientific data allowing the establishment of limit values;

·To facilitate implementation, and contribute towards a better level playing field for economic operators by adopting minimum requirements at the EU level that apply to all businesses, irrespective of their location; 

·To bring more clarity on the scope of the CMRD with regard to welding fumes so that businesses perform the mandatory risk assessment and apply, if workers are likely to be exposed to carcinogens, mutagens or reprotoxicants according to the risk assessment, all requirements when these fumes contain carcinogens, mutagens or reprotoxic substances.

The specific objectives of the initiative are relevant for the sustainable development goals 59 on good health and well-being (3rd goal) and decent work and economic growth (8th goal).

4.3.Consistency with other EU policies

4.3.1.Charter of Fundamental Rights of the EU

The objectives of the initiative are consistent with Article 2 (Right to life) and Article 31 (Right to fair and just working conditions) of the EU Charter of Fundamental Rights 60 .

4.3.2.REACH Regulation

The Regulation concerning the Registration, Evaluation, Authorisation and Restriction of Chemicals 61 (‘REACH Regulation’), entered into force in 2007, establishes among others two distinct EU regulatory approaches that are restrictions and authorisations.

The applicable provisions of REACH authorisation or restriction of cobalt and its inorganic compounds, PAHs and 1,4-dioxane are summarised in the table below. As process-generated substances, restrictions and authorisations within the meaning of REACH are not applicable to welding fumes.

Table 5: applicable provisions of REACH authorisation or restriction

None of the existing REACH restrictions set work-related limit values for cobalt and inorganic cobalt compounds, PAHs or 1,4-dioxane. Therefore, there is no overlap between the initiative subject to this impact assessment and the existing restrictions under REACH. Furthermore, the relevant services within the Commission will ensure a successful complementarity between REACH and OSH legislations in case of future action under REACH for any of the substances subject to this impact assessment. More information related to the REACH requirements for substances subject to this impact assessment is available in annex 10.

4.3.3.Europe’s Beating Cancer Plan 

The aim of Europe’s Beating Cancer Plan 64 is to tackle the entire disease pathway. It is structured around four key action areas where the EU can add the most value: (1) prevention; (2) early detection; (3) diagnosis and treatment; and (4) quality of life of cancer patients and survivors.

The occupational exposure to the five substances subject to this initiative can cause cancer, as outlined in section 2.1. Therefore, addressing them through the setting of limit values or an inclusion in the scope of the CMRD will contribute to preventing cancer.

5.What are the available policy options?

The policy options include a no-policy-change scenario, which serve as a baseline for assessing and comparing the policy options described under section 5.2. A timeframe of 40 years is assumed for the baseline due to the long latency periods 65  for carcinogenic substances. Cancer might indeed take a long time to develop after the exposure to a carcinogenic substance. This means that several future cancer cases would occur independently of this initiative. Therefore, the baseline does not include cancer cases resulting from past exposure 66 .

5.1.What is the baseline from which options are assessed?

This section outlines the baseline scenarios for the four substances subject to the initiative. It describes how the problem is expected to evolve over 40 years in case no further action is taken at EU level. These baseline scenarios are based on the evidence gathered 67 to support this impact assessment and take account of the impact of the implementation of limit values for other substances, the effects of forthcoming changes in national legislations, the future trends (e.g. in use and recycling of the substances, in exposure concentrations due to technical improvements or in exposed workforce), and any other relevant insights for the future (e.g. impact of the green transition). More detailed information on the relevant sectors, number of enterprises with exposed workers, number of workers, exposure levels and some methodologic aspects (including trends, quantification and monetisation of costs and benefits) is provided in annexes 4 and 6.

5.1.1.Cobalt and inorganic cobalt compounds

113,000 workers in the EU are estimated to be currently exposed to cobalt and its inorganic compounds. The main sectors with exposed workers are manufacture of tools (26.5% of the total exposed workforce), machining (22.1%), ceramics (6.6%), biogas (4.8%), surface treatment of metals (4.6%), vehicles (4.6%) and medical and dental devices (4.4%).

The data and evidence gathered 68  indicates that the use of cobalt over the next 40 years will be influenced by the green transition, in particular the growing manufacturing of batteries, the transition from fossil fuels to alternatives and the growing recycling and recovery of critical raw materials. The following sectors will particularly be impacted: batteries (+5% increase in the use of cobalt every year), petrochemicals (-6%), catalysts (-6%) and metal recovery (+5%). Based on this, the exposed workforce for these sectors is assumed to follow the same trends than for the uses over 40 years, as described in table 6. For the other sectors, although the use and number of exposed workers may slightly increase, this could be counterbalanced by further automatization in some sectors. In the absence of evidence supporting another scenario, it was assumed that the number of workers would remain stable for the other sectors over 40 years.  

The current levels of exposure of these workers differ depending on their sectors of activity and the RMMs currently in place (see section 2 and annex 6).  The sectors with the highest average levels of exposure 69 are welding (12.9 µg/m³), medical and dental devices (11.5), production of electronic components (7.6) and manufacture of other fabricated metal products (7.4). Furthermore, the evidence gathered from stakeholders 70 indicates that the ongoing implementation of the EU-wide limit values for nickel compounds, chromium VI compounds and cadmium and its inorganic compounds is expected to also decrease the exposure to cobalt and its inorganic compounds in some sectors 71 , due to co-exposure. For instance, three respondents have indicated that the limit value for nickel compounds may reduce the exposure concentrations for cobalt and inorganic cobalt compounds by less than 50%. Indeed, although nickel and cobalt are typically not used together in catalysts, the same production lines are used for production. Therefore, it is important to take into account the effect of the limit value for nickel compounds on workers’ exposure to cobalt when assessing costs and benefits. A table summarizing the adjusted concentration for all relevant sectors is available in Annex 6.

The levels of exposure might theoretically decrease over the next 40 years due to possible actions at company level in favour of workers’ protection (e.g. further use of closed systems), an increased use of automatization in some sectors and forthcoming regulatory changes within Member States. However, in the absence of any evidence to support these assumptions and as agreed by the employers, workers and government experts steering the work of the supporting study’s contractor, a trend of 0% was used for the baseline.

With these assumptions, about 19,000 workers exposed to cobalt and its inorganic compounds in the EU will become ill over the next 40 years in case of no EU action (76 lung cancer cases, 4,365 restrictive lung disease cases and 14,152 upper airway irritation cases).

Table 6: baseline scenario for cobalt and its inorganic compounds over 40 years

5.1.2.Polycyclic aromatic hydrocarbons - PAHs 

About 1.3 million workers in the EU are currently estimated to be exposed to PAHs across a wide range of sectors, as detailed in annex 6. More than 90% of exposed workers are active in one of the following five sectors: firefighters 74 (both professionals and volunteers – 34.2%), motor vehicle maintenance and repairs (28.9%), road paving (16.6%), steel and iron foundries, ferroalloys (7.5%), and casting of metals (5.3%).

According to the evidence gathered 75 , some sectors should be impacted by a number of long-term driving forces, in particular, climate change and other environmental factors, aggravating resource scarcity and the acceleration of technical change. As a result, it is assumed that the exposed workforce will vary annually over time for the following four sectors: 

·refineries (-6%/year): refinery capacity is expected to decrease by 6% per year, due to the transition to electric vehicles in the automative sector as well as alternative fuels in the maritime and aviation sectors resulting in less demand for petroleum products; 

·aluminium manufacturing (+1%): demand for aluminium is expected to increase with the green transition (e.g. aluminium is used for electric vehicles for reducing weight and in batteries, electricity networks and solar panels). According to a study conducted by Eurometaux 76 , the demand for aluminium will increase between 0.6-1.4% per year; 

·silicon metal manufacture (+1.95%): the shift towards renewable energy is expected to result in an increase in demand for silicon metal, which is key among others for photovoltaic panels, batteries and semi-conductors. Eurometaux 77 estimates that silicon metal demand will increase by between 1.6-2.3%; and

·coal fired power plants (-2.5%): the climate-neutrality objective of the EU is the main driver behind coal power plant closures. According to the European Commission’s Joint Research Center (‘JRC’), up to 100% of existing coal power plants will be decommissioned by 2030. However, coal power plants have experienced a recent resurgence due to the current geopolitical climate. Based on this, an annual decrease of 2.5% in the number of workers is expected.

However, and despite the above estimated impacts on exposed workers, these sectors represent less than 4% of the total exposed workforce and it is therefore assumed that their evolution should not influence future trends in terms of the number of workers exposed.

Another factor that could impact the number of exposed workers over time is the increased automatization in some sectors. Nevertheless, no robust data to support the impact of automatization was identified. It is therefore assumed that the overall number of exposed workers will keep stable.

On the other hand, the exposure levels to PAHs are expected to decrease annually by 1.2% over the next 40 years. The current average levels of exposure to PAHs differ depending on their sectors of activity and the current RMMs currently in place (see section 2 and annex 6). The sectors with the highest average levels of exposure 78 are welding of train tracks (300-400 ng/m³), coking plants (329), manufacture of refractory products (207), coal tar distillation (152) and other non-ferrous metallurgical processes (152). According to the evidence gathered 79 , the implementation of other limit values, including from previous revisions of the CMRD, is not expected to lead to significant impact on PAHs exposure.

A trend of -1.2% per year for the benzo[a]pyrene 80 (BaP) exposure concentration between 1975 and 2009 was reported 81 , due to several factors, including changes in the production methods or application (e.g. banning of the use of coal tar for paving, automation of oven operations in coking plants) or development in the engines releasing less PAHs. Furthermore, the evidence gathered has not identified any forthcoming changes in national limit values or protective regulation and self-regulatory initiatives. Although some possible upcoming PAHs-related REACH restrictions were identified (more details are available in annex 11), it is not possible to anticipate how they may affect the baseline. In this context, and in the absence of any evidence questioning the continuous past trend for concentration exposures, it was decided to set the same trend for the future (-1.2%).  

With regard to the FDB, 3,355 workers are expected to get ill over the next 40 years due to the exposure to PAHs (184 lung cancer cases, 38 developmental toxicity case and 3,157 infertility cases). Among them, 128 workers are expected to die from cancer.

Table 7: baseline scenario for PAHs over 40 years

5.1.3.1,4-dioxane

Around 30,000 workers are currently exposed to 1,4-dioxane in the EU. Almost 90% of these workers are active in three sectors: pharmaceutical production (48%), laboratories (24%) and industrial use as solvent and generation as by-product in the chemical sector (17.5%).

As a result of the recent reclassification of 1,4-dioxane as carcinogenic 1b 82  in the CLP Regulation and its inclusion in the list of SVHC 83 , which are subject to potential regulatory attention under REACH, we could expect that the past trend of -2% 84  related to the number of exposed workers continues. However, the evidence gathered 85  also shows that certain sectors manufacturing or using 1,4-dioxane have seen an increase in their number of companies between 2011 and 2020 of approximatively 2% per year in total. Therefore, it is assumed that both trends will cancel each other out, resulting in no change in the number of workers exposed during the next 40 years. No forthcoming changes in national limit values or other relevant national legislation have been identified. 

According to the evidence gathered, including the RMMs currently in place (see section 2 and annex 6), the highest average exposure is observed in the chemical sector (10.5 mg/m3) where 1,4-dioxane is industrially used as a solvent and generated as by-product. Workers operating in the other sectors are exposed on average to a level lower than 3.9 mg/m3. In the absence of data, including on technical improvements that could result in decreasing workers’ exposure, no past and future trends could be identified. It is also not expected that any of the EU-wide limit values for other substances that have already been adopted but are not yet in force are likely to have a significant impact on occupational exposure to 1,4-dioxane. It was therefore assumed that the exposure concentrations would remain stable over 40 years (conservative approach).

In case of no EU action, and based on the assumptions described above, 5,512 workers would suffer from non-carcinogenic diseases over the reference period (497 with kidney effects, 633 with liver effects and 4.382 from local irritation).

Table 8: baseline scenario for 1,4-dioxane over 40 years

5.1.4.Welding fumes

Around 1.2 million workers are estimated to be currently exposed to welding fumes 86 . They are active in a wide range of sectors as described in annex 6. Based on Eurostat data, three quarters are operating in the following four sectors: specialised construction activities (36%), construction of buildings (19%), manufacture of fabricated metal products (12%), and wholesale and retail trade and repair of motor vehicles and motorcycles (10%).

A limit value for most of the metals and metal compounds classified as carcinogenic have been already set in the Annex III of the CMRD. Some of them (for chromium (VI) compounds and nickel compounds) have a transitional period until end of January 2025. These two groups of compounds are of particular concern as carcinogens in welding fumes from welding stainless steel (food industry, medical industry, petrochemical and power/heating plants). They can also be contained in welding fumes released when welding mild steel, although in less quantity. Theoretically, the introduction of limit values for cobalt and PAHs can also be beneficial, although to a much lesser extent. However, due to the paucity of data for welding fumes exposure in general, it was not possible to take account of the potential impact of implementation of other limit values in the calculation of the baseline.

The evidence gathered indicates 87 that the annual EU welding market growth forecast is 4.5%. This increased demand is partly attributed to the EU green transition, which requires investments in energy infrastructure (e.g. power networks, electric vehicle charging infrastructure and wind turbines). Other megatrends such as the continuing urbanisation, growing consumerism are also expected to support this growing demand. With the ongoing investment in robotics 88 and the current shortage of welders, it is expected that most of this growth is to be carried out by new and automated technologies. In the absence of robust data, it was assumed that this welding market growth forecast will lead to a growth of 0.45% in the number of welders per year (10% of the growth forecast). This assumption is supported by the current skills shortage of trained welders in Europe which encourages businesses to invest in automated technologies.

The evidence gathered 89  indicates that the past exposure levels to welding fumes declined between 2% and 4% per year. The increasing automation of welding processes will lead to further reduction in exposure to welding fumes, in particular, in the automotive sector. However, this decrease depends on the current RMMs in place. According to the evidence gathered 90 , it is assumed that many businesses, in particular larger enterprises, are already applying best practices measures.

The evidence gathered suggests that EU Western Member States are more advanced in the implementation of RMMs and adoption of innovative processes than other Member States. According to some stakeholders 91 , the best practice is already applied and additional RMMs will not reduce worker exposure any further. Therefore, countries (or enterprises) with lower current workers protection from exposure to welding fumes have more to gain in RMMs. Furthermore, innovative practices to optimise welding process could also lead to reductions in exposure to welding fumes. In the absence of further data and given the above information, it was assumed that the exposure levels will continue declining but at a lower pace (1% per year) as best practices may already be implemented in some countries such as Germany and the Netherlands.

In the absence of robust data on non-cancer endpoints, the FDB is limited to lung cancer cases. Considering the future trends summarised in table 9, around 28,000 workers would get lung cancer over the next 40 years. This FDB are likely to be underestimated since the exposure of workers being in the vicinity of welding processes were not considered, due to insufficient data. As a result of their exposure to welding fumes, about 23,000 workers would lose their lives.

Table 9: baseline scenario for welding fumes over 40 years

5.2.Description of the policy options

5.2.1.Structure and logic of the policy options

To meet the first three objectives outlined in the table below, limit values should be established for cobalt and inorganic cobalt compounds, PAHs and 1,4-dioxane. For each substance subject to the setting of limit values, several policy options for limit values are identified to further improve the existing regulatory framework and further levelling the playing field at EU level. These policy options always include the limit values derived by RAC (scientific experts), where any, and those recommended by the tripartite ACSH (stakeholders). These limit values are key as they represent either the limit values below which workers’ exposure is not dangerous (health-based limit value derived by RAC) or the most proportionate limit values according to social partners and Member States authorities represented in the WPC/ACSH as they ensure values which are protective for workers and technically and socio-economically feasible for businesses. There are no existing health-based limit values for some substances (PAHs). In this case, RAC provides an exposure-risk relationship, which provides the correspondence between the substance concentration and the statistical probability of developing cancer (additional risk beyond the background incidence).

In addition, other relevant reference points are chosen as policy options to ensure a wide range of levels for assessment. These additional policy options were mainly based on the new methodology establishing risk-based limit values for non-threshold carcinogens 92 and the existing national limit values. According to this new methodology, limit values for non-threshold carcinogens as set out in the ACSH opinion should be set between the upper risk level of 4:1,000 (corresponding to 4 predicted cancer cases in 1,000 employees) and the lower risk level of 4:100,000. The next section provides more detailed information regarding the reasons for choosing each option. All these policy options are summarized in table 10 and further discussed in annex 9.

For each policy option, the employers would need to put in place the necessary additional RMMs to keep workers’ exposure below the limit value(s). As a result, workers would be less exposed to the below hazardous substances, which should further improve OSH at EU level.

In addition to the policy options for OELs and BLVs, RAC also recommended the setting of short-term exposure limit (STEL) for 1,4-dioxane, and notations for cobalt and its inorganic compounds, PAHs and 1,4-dioxane. These accompanying measures are not expected to result in significant costs and are therefore not assessed individually. More information on these measures is provided in annex 10.

An inclusion of welding fumes in Annex I of the CMRD has been identified by relevant stakeholders represented in the ACSH as a means to meet the fourth specific objective outlined in the table below. Furthermore, the existing Senior Labour Inspectors Committee’s 93 (SLIC) Guidance for National Labour Inspectors on addressing health risks from Welding Fume 94  could be updated in the coming years, which would contribute to improving awareness of the possible dangerousness of welding fumes. However, in the absence of specific information on if and how this guidance document will be updated, the effects of this option could not be assessed. More information on this possible revision is available in annex 10, under the accompanying measures.

Figure 4: Intervention logic

5.2.2.Substances prioritised for the setting of limit values

This section outlines the reasons for choosing the different options for each substance, except option 1 which corresponds to the baseline scenario for all substances.

(a)Cobalt and its inorganic compounds

Option 2 are the limit values (both inhalable and respirable fraction) recommended by RAC, below which workers are assumed to be safe. Options 4 and 5 correspond to the final limit values and the transitional limit values recommended by the ACSH, respectively. Option 3 is an intermediate limit values between option 2 and option 4 to avoid a significant gap between two values. Finally, option 6 corresponds to the combination of limit values (final and transitional), as recommended by the ACSH in its opinion 95 . This transitional limit value of 6 years at the level of option 5 would apply to all sectors, giving more time to businesses to plan their investments.

(b)PAHs

Options 2 and 4 correspond to the limit value with an associated risk of 4:100,000 and 4:1,000 according to RAC opinion 96 . Option 3 is the final limit value recommended by the ACSH. Option 5 corresponds to the combination of limit values (final and transitional), as recommended by the ACSH in its opinion 97 . This recommended transitional value of 140 ng/m³ (double the preferred option) would apply during 6 years and would be limited to 9 sectors: (1) steel and iron foundries, which includes ferroalloy manufacturers, (2) aluminium manufacturers, (3) carbon and graphite electrode manufacturers, (4) coking plants, (5) coal tar distillation, (6) refractory products manufacturers, (7) welding of train tracks, (8) other non-ferrous metallurgical processes, and (9) casting of metals. According to the ACSH, these 9 specific sectors are expected to face high costs and/or are at risk of discontinuation.

(c)1,4-dioxane

Option 2 corresponds to the health-based limit value derived by RAC in its opinion 98  and is also the limit value recommended by the ACSH 99 . As both RAC and the ACSH recommended the same limit value, only one additional option (option 3) was considered and corresponds to the lowest existing national limit value (LV and NL).

According to RAC, occupational exposure is expected to occur via inhalation and dermal route. As an OEL only protects workers from oral exposure, it is also important in this case to set a BLV. In its opinion, RAC identified 2-hydroxyethoxyactic acid (HEAA) as the main metabolite. It also established a relationship between the mean urinary HEAA levels after the end of exposure and the air concentration of 1,4-dioxane. This function is urinary HEAA level (in mg/g creatinine) at end of exposure = 17.82 x (air concentration of 1,4-dioxane in ppm) + 9.52. This function is used in this impact assessment to calculate BLVs corresponding to each OEL option.

Table 10: policy options for each substance subject to the setting of limit values

5.2.3.Welding fumes - substance subject to inclusion in the scope of the CMRD

The evidence gathered indicates that some employers are not aware that their welding processes may contain carcinogens, mutagens or reprotoxic substances. This driver was also supported by the tripartite ACSH in its opinion 107 . These employers may therefore ignore to what extent exposure to welding fumes is dangerous for their workers and include substances that are carcinogenic, mutagenic or reprotoxicant. Consequently, and despite the fact that welding can be a relatively safe activity if all the safety and exposure control measures are in place, it remains a prominent concern that welders are at high risk from various diseases, including cancers 108 . 

According to ECHA 109 in its scoping study on welding fumes, an entry of welding fumes in Annex I of the CMRD would bring clarity about employers’ duties and which measures have to be taken. The introduction of welding fumes in Annex I to the CMRD would ensure a better implementation of the existing obligations laid down in the CMRD. This introduction aims to draw employers’ attention to the possible dangerousness of the fumes generated by their welding process. As provided by the CMRD, employers have to perform a risk assessment as soon as an activity is likely to involve a risk of exposure to carcinogens, mutagens or reprotoxic substances. The ECHA scoping study will support this assessment as it provides information on welding processes likely to generate carcinogens, mutagens or reprotoxic substances. If this risk assessment reveals that their welding fumes contain such hazardous substances, employers will need to take all the necessary RMMs to comply with the obligations laid down in the CMRD, including the hierarchy of controls in Articles 4 and 5. The introduction of welding fumes does not result in new rules but aims to ensure a correct implementation of the existing ones, which contributes to the proportionality of the proposal.

The policy option of an entry of welding fumes in Annex I was supported in its opinion 110 by the ACSH. The three interest groups consider that an inclusion into Annex I might indicate to stakeholders that more needs to be done to ensure that the RMMs are in place and will bring clarity about employer’s duties on measures to be taken. 

Given the broad support from the three interest groups within the ACSH, the inclusion of welding fumes in Annex I was kept as policy option.

No additional policy option other than the inclusion in Annex I and the complementary measure related the possible revision of the SLIC guidance was retained for welding fumes. Indeed, in the absence of scientific opinion from RAC on the possible setting of a limit value for welding fumes, the latter could not be considered as an option for the current initiative. Furthermore, the heterogeneous character of welding fumes makes it very difficult to establish a common OEL for all welding processes.

Furthermore, several guidance documents already exist at EU-level, including the Senior Labour Inspectorate Committee’s guidance on addressing health risks from welding fumes 111 and the European Welding Association’s (EWA) guidance 112 . Despite the existence of guidance at EU level, the evidence gathered shows that the welding fumes-related problem persists. Therefore, the development of new guidance was not considered relevant to address the problem. More information on both guidances is available in annex 10.

5.2.4.Isoprene-related discarded options 

The evidence gathered 113 indicates that the exposure of workers to isoprene is lower than the threshold limit value derived by RAC. Therefore, no problem was identified for isoprene, as described in section 2. Consequently, no policy option was assessed in the main report of this impact assessment. However, as the ACSH recommended a limit value for isoprene, and for the sake of transparency, a fully-fledged analysis of this option was carried out and is available in annex 10.

5.2.5.Options discarded at an early stage

A number of policy options that have been considered at the early stages of this initiative have been discarded without a detailed assessment of their impacts, either because they are clearly disproportionate (banning the use of substances, directly adopting the most stringent national limit values, less stringent rules for SMEs resulting in less ability to attract workers as the latter would favour better working conditions), or unsuitable to achieve the policy objectives (providing industry-specific scientific information without amending the CMRD, guidance documents (including to address the lack of awareness related to welding fumes)), or there is insufficient scientific information (setting a limit value for welding fumes). Annex 10 contains a description of those policy options and the reasons for discarding them.

6.What are the impacts of the policy options?

Section 6 outlines for each substance the identified relevant positive and negative impacts associated with the selected policy options against the baseline. Overview tables per substance summarising all impacts are available in Annex 4. The impacts are classified in three main categories: economic, social and environmental. They mainly concern three groups of stakeholders: workers (and their families), businesses and public authorities. These impacts are mostly monetised or quantified on a 40-year reference period, to calculate the total risk over the working life. When the available data does not allow for this, impacts are discussed qualitatively. Data collection for the purpose of this impact assessment includes literature review (including scientific literature) and stakeholder consultation (targeted questionnaires, interviews, site visits and conversations). More information on stakeholder consultation is available in annex 2.

The introduction of limit values for cobalt and its inorganic compounds, PAHs and 1,4-dioxane, and inclusion of welding fumes in Annex 1 of the CMRD, is expected to reduce the occupational exposure to these hazardous substances as businesses with exposure levels higher than the options will have to invest in additional RMMs. The extent of such reduction depends on the current levels of exposure and RMMs, as well as the projected future levels of exposure in the absence of any action (‘baseline scenario’). Annex 6 outlines the current RMMs in place by sector, as well as the required additional RMMs by sector for each option.

For a given reduction in exposure levels of all substances except welding fumes, the expected decrease in the incidence of ill-health over 40 years was calculated. This involved estimating the risks of adverse health effects, derived from the existing toxicological and epidemiological literature, and information about the current level of worker exposure (number of workers exposed, level, duration and frequency of exposure). The health benefits of the new limit values are calculated in terms of the costs of ill-health avoided. They are expressed in monetary terms by applying standard valuation methods 114 .

The estimate of the costs for all substances except welding fumes was based on a literature research and data obtained from stakeholders and considers the following factors: the RMMs needed to comply with the proposed limit values, the costs of these RMMs for each company, the life span of the RMMs, and the number and size of companies.

While acknowledging the importance of occupational health and safety, the employers organisations’ feedback during social partners consultation focused on the feasibility and economic implications of new regulations, stressing the need for transitional periods, where necessary, and careful socio-economic impact analysis to avoid harming competitiveness. Additional details on social partners consultation are available in Annex 2

Some policy options for cobalt and its inorganic compounds and PAHs include a transitional value (combination of two different values). Transitional values are expected to reduce the adjustment costs and to make the financing of adjustment costs cheaper. Indeed, the longer time companies have to implement compliance measures, the lower are the likely costs. They can for instance better plan investments. As further described in annex 4, the total discounted value for adjustment costs would be 16% lower. Furthermore, transitional values are also expected to spread the substantial one-off investments over a number of years. Therefore, companies could already start planning their investments.

The methodology to calculate the costs and benefits of placing welding fumes in Annex 1 of the CMRD is different from the methodology applied for the substances subject to the setting of limit values. The costs and benefits presented below relate to those additional companies which would apply the already required RMMs following the increased awareness of the dangerousness of welding fumes induced by their inclusion in Annex 1. In particular, it was assumed that for 1% of welders will improve their current RMMs following the introduction of welding fumes in Annex I to the CMRD.To validate these assumptions, interviews were held with key stakeholders representing a large number of welders in the EU, who agreed that the estimates were reasonable.

Regarding benefits, in the absence of Exposure-Risk Relationship and exposure data for welding fumes, a tentative extra cancer risk of 2.7% was derived from existing toxicological and epidemiological literature. This excess risk is subject to an average trend of reducing by 1% per year 115 , and this was assumed to continue for the next 40 years for the baseline. The impact of policy option 2 is assumed to reduce the excess risk by a further 1% for the first five years after the policy option takes effect 116 , due to those companies which would carry out risk assessments and apply adequate RMMs following the increase of awareness of the possible presence of carcinogens, mutagens or reprotoxicants in welding fumes.

With regard to costs, two approaches 117 are applied:

·Bottom-up: this is based on the number of welders that are estimated to move from having poor or no RMMs to adequate RMMs as a result of policy option 2 and multiplying this by the estimated additional average cost of these RMMs 118 ; and

·Top down: based upon the current market value of RMMs being used annually, an estimate of an assumed 1% increase in the sale of RMMs as a result of policy option 2 was applied.

More information on the overall methodology, including the key assumptions, is available in annex 4.

Due to the lack of data, the analysis of the impact on workers’ health is limited to the most relevant diseases. For instance, exposure to PAHs can also result in skin cancer. However, the paucity of data does not allow for a quantification or monetisation of the impact on health. Due to this limitation, the benefits of the options are likely to be underestimated.

Furthermore, the discontinuation costs presented below are subject to uncertainty 119 . First, the number of companies that are expected to discontinue might be lower than estimated because they could find technical solutions or alternatives to ensure compliance. Secondly, if some companies close down effectively, it is likely that a part of their activities would be absorbed by other companies in the same sector to meet the overall demand. The same uncertainty also applies to the estimated job losses. First, an overestimate of discontinuations would also mean an overestimate of job losses. Secondly, the possible transfer of activities mentioned in this paragraph would also lead to the creation of jobs that would partly offset the quantified impact below. However, in the absence of relevant data, these likely effects couldn’t be quantified. Nevertheless, it is likely that the estimated discontinuations and job losses below are overestimated.

Other limitations were identified and are summarised in annex 4, in particular table 65 summarising the key limitations or uncertainties. This table indicates if these limitations might lead to under- or overestimates.

With regard to the impact on Fundamental Rights, the Article 31 of the EU Charter of Fundamental Rights states that workers have the right to fair and just working conditions that respect their health, safety and dignity. This is further developed in the EPSR which aims to build a fairer and more inclusive European Union. Each policy option (‘PO’) that would further improve the protection of workers from the health risks posed by exposure (resulting in avoided ill-health), would also positively impact this right. For those POs with no impact on workers’ health, there would be no impact on Fundamental Rights. Given that the impact on Fundamental Rights would be equivalent for each substance, it is not further discussed below.

6.1.Cobalt and inorganic cobalt compounds

The estimates related to cobalt and inorganic cobalt compounds outlined below take account of the impact assessment commissioned by Cobalt Institute and data on exposure levels and chemical safety report also provided by Cobalt Institute. The data have been assessed along with data obtained from stakeholders’ survey and other available data sources. With regard to exposure levels, exposed workforce, and number and companies with exposed workers – for those sectors covered by the Cobalt Insitute’s impact assessment – the estimates used for the study 120 supporting this impact assessment are well in accordance with the data provided by Cobalt Institute.

The Cobalt Institute assessment, however, estimates the total costs of compliance significantly higher than estimated below.

The differences between the Cobalt Institute's assessment of impact and the external study primarily arise from variations in methodology, assumptions about company structures, and compliance strategies.

Detailed information on why Cobalt Institute’s estimates differ from this impact assessment is provided in annex 2.

6.1.1.Social impacts 

(a) Impacts on workers’ health

Setting limit values at the level of policy option 2 (‘PO 2’) would allow for eliminating most of the health burden over the reference period. Under these levels of exposure, no worker would suffer of restrictive lung disease and upper airway irritation (representing 4,365 restrictive lung disease cases and 14,152 upper airway irritation cases avoided) and the number of lung cancer cases would be reduced by more than 90% vis-à-vis the baseline scenario (71 out of 76 lung cancer cases).

PO 3 would allow for fully protecting workers from restrictive lung disease and for avoiding 51 lung cancer cases and 12,266 upper airway irritation cases. Overall, about 90% of all cancer and non-cancer cases would be avoided.

The limit values provided under PO 4 would reduce by 55% the total number of cancer and non-cancer cases over the reference period. As shown in table 11, 23 lung cancer cases, 2,842 restrictive lung disease and 7,363 upper airway irritation cases would be avoided.

PO 5, which provides for the less stringent combination of limit values, would have a limited impact on the total health burden. 12 (out of 76 under the baseline scenario) lung cancer cases would be avoided and 3,135 non-cancer cases would be avoided (1,000 restrictive lung disease cases and 2,135 upper airway irritation cases). The corresponding health savings are presented in the table below.

The benefits in terms of the reduced number of cases of ill-health under PO 6 would be approximatively the same as under PO 4. However, since a part of these benefits will be postponed on average by 3 years (see annex 4) due to the transitional period, the monetary value will decrease by 8%, accounting for between €68 million and €69 million.

Table 11: number of avoided cases of ill-health and monetised health benefits over 40 years (compared to the baseline)

(b)Impacts on employment

The impact on employment varies significantly depending on the PO. The more stringent the PO, the higher the need for additional risk management measures to ensure compliance and the higher the impact on employment. The evidence gathered 122 suggests that 1,086 companies will discontinue, resulting in job loss for almost 26,000 workers under PO 2, representing a social cost of about €1.95 billion. More than a half of job losses would occur in the manufacture of tools sector (17,977 job losses).  

Under PO 3, the estimates indicate that more than 2,700 workers would lose their jobs. As for PO 2, the manufacture of tools would be the most impacted sector with a loss of about 1,750 jobs, representing two thirds of total job losses. POs 4 and 5 would have less impacts on employment with 1,124 and 77 job losses, respectively.  

The transitional period provided under PO 6 is expected to result in less job losses as companies should use this period to avoid discontinuation (see section 6.1.2). However, the evidence gathered does not allow for any quantification.

While some of the workers made redundant might easily find another job, notably in the faster growing sectors (e.g. batteries, metal recovery), the lack of data didn’t allow to estimate the number of redundant workers. The above job losses numbers can therefore be considered as the worst-case scenario. Furthermore, as mentioned in introduction of section 6, estimates of job losses are subject to high uncertainties.

6.1.2.Economic impacts 

(a)Impacts on businesses, including SMEs

Under all POs, some companies will need to invest in additional RMMs to comply with the new limit values, entailing direct adjustment costs. The magnitude of these costs varies significantly between the different POs.

For a limited number of companies, these additional costs or the absence of appropriate RMMs available on the market that could enable companies to ensure compliance with some POs might endanger their continuation 123 . Under PO 2, 1,086 businesses (7.1% of the total number of companies with exposed workers to cobalt) are expected to discontinue at least a part of their activities, representing a discontinuation cost of almost €6 billion. Among these businesses, 947 are small- and 109 medium-sized enterprises. The less stringent the POs, the lower the discontinuation costs as shown in table 12. Under PO 5, 10 businesses, all being small enterprises, are expected to discontinue, representing a discontinuation cost of €20 million. 

With PO 6, businesses would have more time to implement major and expensive changes to RMMs; developing, finding and testing substitutes or avoiding discontinuations. The transitional value provided in PO 6 was highly recommended by both the ACSH and social partners, in particular businesses, as part of their formal consultation. It is expected therefore that the transitional period will result in less discontinuations than those estimated under PO 4. However, the evidence gathered does not allow for a quantification of such a decrease.

The manufacture of tools would be the most impacted sector in terms of discontinuation under the first three POs.

Table 12: number of discontinuations per company size and total discontinuation costs

For the businesses remaining in the market, PO 2 would be particularly costly as it would require a total investment in additional RMMs of over €3.65 billion over 40 years, of which €1.5 billion would need to be invested during the first year.

Under POs 3 and 4, businesses would need to dedicate over the next 40 years about €1.15 billion (of which €708 million during the first year) and €350 million (€237 million), respectively. PO 5, which provides for less stringent limit values, would naturally require less investments. In total, businesses would need to spend about €110 million in additional RMMs, of which €88 million during the first year.

If we bring these figures down to the scale of a company, each of those would on average need to make investments in RMMs over the next 40 years for about €250,000 under PO 2, €75,000 under PO 3, €22,800 under PO 4 and €7,200 under PO 5.

The machining sector would bear a significant share of the investments in additional RMMs, in particular under POs 2 and 3 for which these investments would represent more than half of the total investments in RMMs for all sectors (about €1.95 out of €3.65 billion for PO 2 and €615 million out of €1.15 billion for PO 3). The medical and dental devices sector would be the second most impacted sector with investments representing 10% and 15% of the total investments to be carried out under POs 2 and 3, respectively.

With less discontinuations expected under PO 6 than under PO 4, the compliance costs for PO 6 are also expected to increase when companies are able to finance the investments in RMMs instead of closing down. On the other side, it might also be that the costs of RMMs will be reduced if companies can plan and do investments as part of investment in new production facilities and equipment. The possibility to plan and finance investments in RMMs over the transitional period will in particular be important for SMEs. The evidence gathered does not allow for a monetisation or quantification of the impact of a transitional period on the compliance costs.

The introduction of new limit values may entail air monitoring costs for those businesses which would need to carry out additional monitoring campaigns. The number of additional campaigns varies from 0 to 2, depending on several factors, including the existing national limit values, the size of companies 124 .

According to estimates, the total air monitoring costs under PO 2 would account for €112 million, of which €23.6 million and €52.9 million would be supported by small- and medium-sized enterprises, respectively. The less stringent the limit values under the POs, the lower the air monitoring costs, as shown in table 13.

In percentage of the total compliance costs, the air monitoring costs is about 1%, 4%, 8% and 22.5% for the first four POs, respectively, with the highest percentage for PO 5.

Under PO 6, it is expected that the air monitoring costs would amount to €46.8 million. Although the final limit values under PO 6 are equivalent to PO 4, businesses will have more time to put in place the necessary RMMs and demonstrate compliance. Therefore, the total air monitoring costs should also be lower, mainly due to the discount factor 125 .

Table 13: total estimated costs of up to two additional air monitoring campaigns in € million over 40 years by policy option and size of company

The total costs related to the administrative burden of undertaking the additional air monitoring campaigns for companies are shown in table 14, broken down by PO and size of company. The lower the limit values, the higher the RMMs and the higher the number of companies that need to carry out additional monitoring campaigns to demonstrate compliance. Therefore, administrative costs are proportional to the degree of severity of the limit values considered for each PO. Overall, the administrative costs represent a negligible share of the total compliance costs under all POs, ranging from 0.1% (PO 2) to 2.5% (PO 5). Similarly to air monitoring costs, the costs related to the administrative burden are expected to amount to €6.1 million, as costs will be delayed by 3 years compared to PO 4.

Table 14: total estimated costs of administrative burden of additional air monitoring by policy option and size of company over 40 years

If compared with turnover, small- and medium-sized enterprises would be more impacted than larger companies under all POs. The impact would be particularly significant under POs 2 and 3 as small-sized enterprises would have to dedicate on average 1.16% and 0.32% of their turnover for compliance (including adjustment, monitoring and administrative costs) over 40 years, and medium-sized enterprises 0.65% and 0.11%, respectively. In some sectors, the share of the adjustment costs compared to the turnover under PO 2 is significantly higher than the average share for small-sized companies, in particular in the manufacture of other porcelain and ceramic products (3.08%), manufacture of tools (2.17%), manufacture of other fabricated metal products (2.10%) and manufacture of medical and dental instruments and supplies (2.04%). The impact of POs 4, 5 and 6 is more limited, with small enterprises having to invest between 0.1% and 0.02% of their turnover, and medium enterprises 0.04% and 0.009%.

All POs would also positively impact businesses as they would allow for cost savings from reduced ill-health. PO 2 would entail the highest benefits for employers as it would also have the highest positive health impacts. Under this PO, businesses would save €1.6 million over 40 years against 0.4 million under PO 5 over the same period. The costs savings under POs 3, 4 and 6 would be in between with €1.5, €1 million and €0.9 million, respectively. These benefits are negligible compared to the costs to businesses and are therefore not expected to mitigate the impacts described above.

(b)Impacts on competitiveness

Although of different extent, all POs are likely to have impacts on cost and price competitiveness as some businesses are expected to discontinue, partly or totally, due to the high compliance costs or the current absence of appropriate RMMs enabling some companies to comply with some POs. Such an impact would particularly occur under PO 2, especially for the manufacture of tools sector where more than 4% of the companies would discontinue (30% if we only consider companies with exposed workers to cobalt), which could affect product availability and prices. Under all other options, estimated discontinuations are not expected to result in significant changes in any sector.

Under all POs except PO 2, the total adjustment costs are below 1% of turnover for all sectors (including SMEs), and in most cases even below 0.1%. With regard to PO 2, adjustment costs can represent up to 3% (small-sized companies in the manufacture of other porcelain and ceramic products sector). Only SMEs in the following sectors would face adjustment costs representing more than 1% 126  of the turnover: manufacture of other porcelain and ceramic products, manufacture of tools, manufacture of other fabricated metal products, manufacture of medical and dental instruments and supplies, machining, manufacture of electronic components and boards, manufacture of motor vehicles, manufacture of air aid spacecraft and related machinery, and forging, pressing, stamping and roll-forming of metal; powder metallurgy.

Therefore, overall impacts of POs 3, 4, 5 and 6 on businesses should have no or limited impacts on consumer choices and prices as companies are expected to absorb costs which represent a small share of their turnover and no sector should be significantly impacted by discontinuations. Regarding PO 2, consumers are expected to be impacted by higher prices or lower diversity of supply in sectors where SMEs would face high adjustment costs compared to their turnover 127 or sectors would be characterised by a significant number of estimated discontinuations compared to the total number of companies with exposed workers 128 . 

With regard to the EU’s main competitors, only Canada and Japan have binding limit values for the inhalable fraction or total dust comparable to PO 5. The other competitors have limit values higher than PO 5, or no binding limit values for cobalt. Therefore, under all POs, the international competitiveness of EU-based companies could be impacted when compared to their competitors who will operate without incurring the capital and operating costs of compliance with a lower limit value. Non-EU countries often have different compliance rules and methods to define exposure, which makes the comparison uncertain.

Consequently, some companies operating within the most impacted sectors might consider the possibility to relocate their activities outside the EU 129 . However, except under PO 2, the average compliance costs per company compared to the turnover are not expected to be particularly significant and should therefore not exceed the costs of relocation of activities. Therefore, the impact on international competitiveness from POs 3 to 6 should be limited and is not expected to impede the objective of increasing the overall competitiveness of the EU. 

The impact on capacity to innovate is expected to be significant in some sectors under PO 2. According to the evidence gathered, the total compliance costs in the following 5 sectors would represent more than 20% of their expenditures in research and development (R&D) over 40 years: manufacture of tools (53%) manufacture of other fabricated metal products (36.5%), forging, pressing, stamping and roll-forming of metal; powder metallurgy (25%), manufacture of electric components and boards (23.5%) and recovery of sorted materials (20.5%). The resources used to comply with PO 2 would therefore not be available for R&D activities. Under the other four POs, the share of the total compliance costs compared to the R&D expenditures is lower than 10%, and even 1% in most sectors and POs.

(c)Impacts on the single market

No Member State has an equivalent combination of limit values to those assessed in this report. For instance, despite that several Member States already have a limit value of 20 µg/m3 (inhalable fraction), it does not apply to the same scope 130  or is not combined with a limit value for the respirable fraction. Furthermore, 5 Member States currently have no limit value for cobalt. Therefore, introducing a combination of limit value covering both the respirable and inhalable fraction, whatever their levels, would lead to a more level playing field within the single market, improving its functioning. In the absence of further information, this impact could not be quantified.

(d)Impacts on public authorities/Member States 

No Member State has currently an existing combination of limit values for cobalt and its inorganic compounds in its national legislation. Therefore, they will all need to revise their legislations, which will entail transposition costs. Consequently, the total costs for transposition will be the same under each PO: €910,000 131 .

All POs would improve the protection of workers, resulting in less ill-health cases over the next 40 years. POs 2 and 3 in particular would allow for a significant reduction of the health burden, resulting in avoided healthcare treatment costs for public authorities of about €8.1 and €7.2 million, respectively. Although the positive impact on health is less significant, public authorities would still benefit from a reduction of costs under POs 4, 5 and 6 (€4.4, €1.5, and €4 million, respectively).

All POs might also be beneficial for those Member States which currently have no limit value and would like to set one. Under all POs, the 5 Member States would not have to carry out an assessment in view of setting limit values for cobalt. This could represent a benefit of €500,000 in total for the 5 Member States, irrespective of the PO.

Under all options, the benefits to public authorities would outweigh the costs.

The Member States with the highest share of companies comprise Czechia, France, Germany, Italy and Poland. It can therefore be expected that industries in these Member States will face the largest compliance costs. These five Member States accounts for about 60% of the companies and therefore, they could also face around 60% of impacts (depending on the current RMMs in place). The mentioned Member States are also for the most the larger Member States. If compared to the total size of the industry, there is overrepresentation of companies in Czechia, Italy, Poland and Slovakia. This is a reflection of large metal industries in these Member States. Therefore, also considering that the larges costs by sector include Machining and Manufacture of metal tools, these Member States are likely to be most affected. Among these 4 Member States, Poland, which has an existing binding limit value of 20 µg/m³ (inhalable) should be the less impacted. While Italy has no limit value for cobalt, Czechia and Slovakia have a binding limit value of 50 µg/m³ (inhalable).

It also means that the expected reduction in discontinuation from the transitional value provided by PO 6 will also benefit all Member States identified in the previous paragraph.

(e)Impacts on digitalisation

No PO is expected to have an impact on digitalisation.

(f)Impacts on EU Open Strategic Autonomy

Cobalt is a critical raw material and is used in key technologies ensuring a successful green transition (e.g. batteries for electric vehicles and storage and magnets in wind turbines). Depending on their levels, limit values for cobalt and inorganic cobalt compounds may result in higher costs of recycling cobalt and thereby have a negative impact in this area. As a possible consequence, the EU would remain dependent on other countries for its supply of cobalt, which goes against its objective of open strategic autonomy.

Under PO 2, 10% of businesses with workers exposed to cobalt operating in metal recovery sector are estimated to discontinue their activities. The compliance costs would represent up to 0.7% of the average turnover in the sector (7.83% of the gross operating surplus). Therefore, PO 2 could negatively impact the EU objective of open strategic autonomy. PO 3 is expected to result in the discontinuation of 0.5% of businesses operating in this sector. Furthermore, the compliance costs would not exceed 0.04% of the turnover (0.49% of the gross operating surplus). Therefore, the impact under PO 3 would be negligible. Under PO 4 and PO 5, no discontinuation is estimated for the metal recovery sector and the compliance costs would represent less than 0.003% of the turnover (0.04% of the gross operating surplus). Therefore, POs 4 and 5 would have no impact on the EU objective of open strategic autonomy. Finally, the transitional period provided under PO 6 would give more time to plan investments, which would further reduce the already limited impacts arising from a limit value at the level of PO 3.

6.1.3.Environmental impacts 

(a)Direct impacts on environment

To meet the new limit values, businesses might need to invest in alternative technical RMMs such as closed systems, partially closed systems, self-contained breathing apparatus, half and full facemasks and open hoods. The choice of these RMM will depend on the RMMs currently in place (see annex 6) and the PO. However, it is unlikely that the alternative RMMs contribute to higher releases of cobalt in the air. Therefore, the direct impact of all POs on the environment should be negligible or even zero.

(b)Indirect impacts on environment

The table below provides theoretical impacts on some of the objectives of the EU Green Deal.

Table 15: theoretical impacts of the introduction of limit values for cobalt and its inorganic compounds on the objectives of the EU Green Deal

The key sectors for the green transition are: manufacture of batteries and accumulators, manufacture of motor vehicles, manufacture of air and spacecraft and related machinery and recovery of sorted materials. PO 2 would hamper the Green Deal’s objectives as some companies operating in the above-mentioned sectors would discontinue (batteries and accumulators: 1% of businesses; motor vehicles: 10%; air and spacecraft: 11%; and recovery: 10%). Furthermore, the total adjustments costs would represent a high share of the gross operating surplus, in particular for SMEs operating in the manufacture of motor vehicles (up to 33%) and in the manufacture of air and spacecraft and related machinery (15%).

Although to a lesser extent, PO 3 would also negatively impact the capacity of the Green Deal to deliver on some of its objectives. While the number of discontinuations for the sectors relevant to the Green Deal should remain limited (motor vehicles: 1%; air and spacecraft: 2.5%; and recovery: 0.5%), the adjustments costs compared to the gross operating surplus would approach the 5% for SMEs operating in the manufacture of motor vehicles and of air and spacecraft and related machinery.

POs 4 and 5 are not expected to have a negative impact on the Green Deal. There would indeed be no discontinuation except under PO 4 for businesses operating in the manufacture of air and spacecraft and related machinery (0.5% of businesses discontinuing). Furthermore, the total adjustment costs should remain very limited compared to the operating surplus, representing up to 0.5% of the gross operating surplus, except for SMEs manufacturing air and spacecraft and related machinery which would have to dedicate about 1% of their gross operating surplus under PO 4. Therefore, the impact of POs 4 and 5 on the Green Deal should be negligible or even zero.

The transitional period provided by in PO 6 should be beneficial to SMEs operating in the manufacture of air and spacecraft and related machinery, which would have more time to plan their investments. Therefore, this could result in less discontinuations, which would further reduce the already negligible impact of PO 4 on the Green Deal.

In the longer term, the introduction of limit values may be a driver for development of alternatives and reduce the demand for cobalt, and thereby have a positive impact on the supply of clean, affordable and secure energy. However, the evidence gathered did not allow for a more detailed assessment of this theoretical positive impact.

6.2.PAHs

6.2.1.Social impacts 

(a)Impacts on workers’ health

All POs would entail benefits for workers by reducing the number of ill-health cases compared to the baseline over 40 years. However, the magnitude of this impact varies greatly between the POs. The lowest limit value, PO 2, would have the most significant impact with 3,336 (out of 3,355 estimated under the baseline) ill-health cases avoided over 40 years, including nearly 90% of the cases of lung cancer and 100% of the cases for the other estimated endpoints.

PO 3 would also be effective in protecting workers from non-cancer ill-health since it would eliminate all cases of developmental toxicity and male infertility. However, it would only prevent less than one fourth of the lung cancer cases compared to the baseline (35 out of 160). PO 4 would not bring significant improvement for preventing work-related diseases related to exposure to PAHs. At 700 ng/m³, less than 20% of the total ill-health cases would be prevented, including 2 lung cancer cases only. Despite the transitional value, it is expected that the number of avoided ill-health cases under PO 5 should be similar to PO 3 132 . The table below summarizes the avoided cases over 40 years for each PO and also provides a monetisation of these direct positive impacts.

Table 16: n° of avoided cases of ill-health and monetised health benefits over 40 years (against the baseline)

(b)Impacts on employment

Impacts on employment are expected under POs 2 and 3, as a number of companies might be forced to discontinue (see economic impacts related sections). As a result of these discontinuations, about 47,500 (PO2) and 2,700 (PO3) workers are expected to lose their jobs out of the 1,285,000 workers currently exposed to PAHs in the EU. The social costs related to those job losses would be of about €4 billion and €0.25 billion for those two options, respectively.

Even if most of the sectors would be impacted under PO 2, the impact on employment would represent more than 60% of workers active in the following sectors: welding of railroad tracks (17,735 workers, representing 37% of the total number of job losses), petroleum refineries (5,739, 12%) and steel and iron foundries, ferroalloys (5,499, 11.5%).

Although its overall impact would be less significant, PO 3 would mainly impact workers operating in the following sectors: steel and iron foundries, ferroalloys (550, 20.5%), soil remediation (511, 19%), coking plants (479, 18%) and coal tar distillation (437, 16%).

Although the evidence gathered does not allow for a quantification, PO 5 is expected to result in less discontinuations than PO 3 (see section 6.2.2). Consequently, it is also expected that the number of job losses should be lower under PO 5 than under PO 3.

The evidence gathered does not allow for estimating the number of workers which would find a new job following the discontinuation of their companies. No impact on employment is expected under PO 4.

6.2.2.Economic impacts 

(a)Impacts on businesses, including SMEs

Under POs 2 and 3, some companies are expected to discontinue because of the significant investments that they would need to make for complying with the new limit values. The impact would be particularly significant under PO 2, with 3,462 companies estimated to discontinue at least a part of their activities, all sectors combined. With a higher limit value, PO 3 is estimated to entail 146 discontinuations. In total, it would represent a discontinuation cost (considered as adjustment costs) of €32 billion and €2.25 billion, respectively. No discontinuation is expected under PO 4.

Under PO 2, 3 sectors would represent together more than four fifths of the total discontinuations: welding of railroad tracks (1,109 discontinuations, representing 32% of the total discontinuations), chimney sweeps (969, 28%) and soil remediation (726, 21%). The latter would continue to face difficulties under PO 3, with 68 discontinuations (representing 46.5% of the total discontinuations), followed by steel and iron foundries, ferroalloys (28, 19%) and coking plants (25, 17%).

Whether under PO 2 or 3, most of those companies which would discontinue are small-sized enterprises (93% and 88%, respectively). Although less impacted, 103 medium-sized enterprises are expected to discontinue under PO 2, against 6 under PO 3. The share of SMEs in the total discontinuation costs is lower. They would represent just under 40% for both POs. This difference is due to the lower turnover of SMEs compared to large companies 134 .   

With PO 5, businesses operating in the sector subject to the transitional value 135 would have more time to implement major and expensive changes to RMMs; developing, finding and testing substitutes or avoiding discontinuations. The transitional value provided in PO 5 was highly recommended by both the ACSH and social partners, in particular businesses, as part of their formal consultation. It is therefore expected that the transitional period will result in less discontinuations than those estimated under PO 3. However, the evidence gathered does not allow for a quantification of such a decrease. As most of the companies estimated to discontinue under PO 3 are SMEs, the introduction of the transitional period provided in PO 5 should mainly benefit them.

Table 17: number of discontinuations, including per company size and total discontinuation costs

Some of the companies that would not discontinue would need to invest in additional RMMs to comply with the new limit value, depending on their current levels of exposure and RMMs. Under PO 2, businesses would need to invest about €4.6 billion over 40 years, of which €1.7 billion during the first year. The impact of POs 3 and 4 on the need for investments in RMMs would be less significant as businesses would need to dedicate €417 million and €3.3 million over 40 years (of which €220.90 million and 4.5 million 136  during the first year).

The most impacted sectors in terms of investments in additional RMMs, irrespective of the company size, would be casting of metals (€1.2 billion over 40 years), road paving (€1.15 billion) and chimney sweeps (€765 million). While most of the investments required for the road paving sector would be recurrent (about 93%), the two other sectors would need to make about 50% of their investments during the first year.

Under POs 3, 2 sectors would concentrate almost two thirds of the required investments: road paving (€161.5 million) and welding of railroad tracks (€ 102 million). Under PO 3, investments supported by the road paving sector would be mainly recurrent (about 90%), while one third of the investments made by companies operating in the welding of railroad tracks sector would need to be done during the first year. Under PO 4, the total investments are less significant and would be mainly supported by the welding of railroad tracks sector (€3 million out of €3.3 million).

More than 95% of the companies operating in the two sectors mentioned in the previous paragraph are SMEs, which are usually characterised by less financial resources to make investments, in particular one-off investments (during the first year).

With less discontinuations expected under PO 5 than under PO 3, the costs of additional RMMs for PO 5 are also expected to increase when companies operating in the 9 targeted sectors are able to finance the investments in RMMs instead of closing down. On the other side, it might also be that the costs of RMMs will be reduced if companies have the opportunity to plan and do investments as part of investments in new production facilities and equipment. The possibility to plan and finance investments in RMMs over the transitional period will in particular be important for SMEs. The evidence gathered does not allow for a monetisation or quantification of the impact of a transitional period on the compliance costs.

While no Member State currently has a limit value at the level of PO 2, one (DE) and four (DE, LV, BG and NL) Member States have currently a limit value below PO 3 (+ final OEL under PO 5) and PO 4, representing 11% and 18% of businesses, respectively. Consequently, all enterprises would need to undertake additional air monitoring at PO 2 and over 80% at POs 3, 4 and 5.

PO 2 would represent a total monitoring cost of €550 million, against €292, €181 and €291 million under POs 3, 4 and 5, respectively. Under the three POs, most of these costs would be supported by SMEs, since they represent the largest part of the companies with exposed workers to PAHs.

Table 18: total estimated costs of up to two additional air monitoring campaigns in € million over 40 years by policy option and size of company

The total administration costs related to the running of air monitoring are shown in table 19 below. As for the monitoring costs, most of the administrative costs are attributed to SMEs, in particular small companies. The lower the limit value associated with the PO, the higher the administrative costs.

Table 19: total estimated costs of administrative burden of additional air monitoring by policy option and size of company over 40 years

In total, compliance costs for companies under PO 2 would be of about €37.5 billion, against €3 billion, €225 million and €2.9 billion under POs 3, 4 and 5, respectively. Under POs 2 and 3, more than three fourths of the compliance costs are due to the companies discontinuing. For PO 4, monitoring and administrative costs account for more than 95% of the total compliance costs.

All POs would positively impact businesses as they would allow for cost savings from reduced ill-health. The most significant benefit would be under PO 2, with cost savings estimated at €540,000, against €130,000, €10,000 and €120,000 under POs 3, 4 and 5, respectively.

(b)Impacts on competitiveness

POs 2 and 3 would have impacts on cost and price competitiveness with some businesses estimated to possibly discontinue, partly or totally, due to the high compliance costs. Certain sectors would be particularly affected by PO 2 which would result in the above-mentioned possible discontinuations. Among these sectors are coking plants (100% of the companies are expected to discontinue), welding of railroad tracks (44%), soil remediation (17%) and steel and iron foundries, ferroalloys (10%). Furthermore, PO 2 is expected to entail first year compliance costs (minus discontinuation costs) representing more than 10% of the gross operating surplus for 13 sectors 137 . Except for graphite and electrode manufacturers, these sectors are mainly composed of SMEs, which have less capacity to absorb additional costs. Therefore, it is likely that a share of the adjustment costs generated by PO 2 will be passed on to consumers.  

Under PO 3, one sector is estimated to be particularly impacted with discontinuations, i.e. the coking plants sector with 45% of companies estimated to discontinue, which would have negative impacts on consumers both in terms of diversity of supply and prices. Furthermore, 6 sectors 138 should face first year compliance costs (minus discontinuation costs) representing more than 10% of the gross operating surplus. These sectors are mainly composed of SMEs, which are unlikely to absorb additional costs. Therefore, consumers could be impacted with a share of the costs being passed on to them. However, the shares of first year compliance costs compared to gross operating surplus are less significant than under PO 2. Therefore, the impact of PO 3 on consumers should also be less. No discontinuation is expected under option 4 and the compliance costs represent less than 10% of gross operating surplus. Therefore, the impact on consumers should be more limited, although an increase in prices cannot be excluded given the high share of SMEs. 

With its transitional period, PO 5 should have less impact on cost and price competitiveness than PO 3 (which does not have transitional period). Indeed, businesses targeted by PO 5 will have the opportunity to better plan and finance the necessary investments, which might reduce the share of costs that would be passed on to consumers. Furthermore, the possible positive impact on the number of discontinuations (expected to decrease) for the coking plants sector would reduce pressure on cost and price competitiveness, and on the diversity of supply. Nevertheless, the impact on consumers cannot be excluded as compliance costs compared to gross operating surplus will remain high for some sectors, in particular graphite and carbon electrode manufacture, coking plants and other non-ferrous metallurgy. Furthermore, while the transitional measure is expected to reduce discontinuations, it cannot be excluded that some companies would close down anyway, impacting negatively the diversity of supply and prices.

With regard to the impact on international competitiveness, certain non-EU competitors such as Canada, Japan, Norway, South Korea and Switzerland have binding limit values comparable to PO 4, or even below in some of them. Other competitors such as China, India, Australia, Turkey or Russia have no limit value for PAHs. Non-EU countries often have different compliance rules and methods to define exposure, which makes the comparison uncertain.

Since 99% of the companies with workers exposed to PAHs are small companies, they are more likely to operate at national scale and the possibility of relocation in or import from non-EU countries is limited. However, sectors such as steel and iron foundries, ferroalloys, aluminium manufacturers, other metallurgy, casting of metal and graphite and carbon electrode manufacturers already experience pressure from cheaper imports, particularly Asian countries 139 . An increase in operational costs resulting from compliance costs could lead to discussions about the possibility to relocate their activities outside the EU. It is particularly the case under PO 2, and to a lesser extent under PO 3. Under PO 5, businesses will have more time to plan their investments or find alternatives, which should reduce the impact on international competitiveness. Given the limited costs compared to gross operating surplus under PO 4, the impact on international competitiveness is considered as negligible.

Furthermore, the impact on coking plants resulting in possible discontinuations, in particular under PO 2 and to a lesser extent under PO 3, could also negatively impact downstream sectors (including coal tar distillation plants, and carbon and graphite electrode manufacturers) requiring coking plants’ by-product coal tar for their production. Among the products requiring coal tar are batteries storing energy produced by wind turbines, semiconductors and electric vehicles. A more limited availability of these products would lead to greater dependence on countries outside the EU. With its 6-years transitional period applying among others to coking plants as provided by PO 5, businesses would have more time to plan their investments or find alternatives, which is likely to result in less discontinuations, and therefore lower impact on downstream users in terms of imports from outside Europe. Under PO 4, no discontinuation is expected, and there should therefore be no impact on downstream sectors.

The impact on capacity to innovate is closely linked to the level of compliance costs, which divert resources from other purposes, potentially including R&D. Under PO 2, compliance costs significantly outweigh the available R&D funding for most companies, with a few exceptions 140 . Although compliance costs would represent a small percentage of the R&D under PO 3, the impact on the capacity to innovate cannot be excluded, in particular for SMEs operating in the following sectors: coking plants (compliance costs represent about 160% of the R&D expenditures), soil remediation (155%), welding of railroad tracks (120-135%) and other non-ferrous metallurgy (50%). The impact of PO 4 on the capacity to innovate should be negligible given the low compliance costs. By having more time to plan and finance the necessary investments in additional RMMs, the impact on capacity to innovate under PO 5 could be lower that under PO 3. However, the compliance costs for the sectors identified above would remain high compared to turnover. Therefore, the impact on capacity to innovate for coking plants, soil remediations, welding of railroad tracks and other non-ferrous metallurgy sectors cannot be excluded.

(c)Impacts on the single market

11 Member States currently have no limit values for PAHs. Among those with an existing limit value, the levels vary greatly, ranging from 70 ng/m³ to 5000 ng/m³. Most of Member States with a limit value operate at levels around PO 4, except BG, DE, LV and NL. Therefore, POs 2, 3 and 5 are likely to have a positive impact on the functioning of the single market, by creating a more level playing field.

(d)Impacts on public authorities/Member States

About 438,000 firefighters are assumed to be exposed to airborne PAHs in the EU 141 . Firefighters are employed by public authorities, not private businesses. Public authorities may therefore have to bear compliance costs. The current exposure data indicates that all firefighters are exposed to levels below 70 ng/m3 . Therefore, it is assumed that no additional RMMs (adjustment costs) would be required under PO 4. As firefighters are not subject to the transitional period provided for by PO 5, the impact of the latter on public authorities is equivalent to PO 3 and will not be further discussed, except for the costs related to the preparatory work.

Whilst current exposure levels are expected to be lower than PO 3, 5% of the public authorities have levels close to it (69.62 ng/ m3) 142 . Therefore, it is expected that these public authorities will need to invest in alternative RMMs as a precaution, accounting for €62.5 million over 40 years, of which €16 million would be spent during the first year. Under PO 2, all public authorities are expected to invest in alternative RMMs as the exposure is higher than the limit value associate with this option. In total, these investments would represent about €5.9 billion (of which about €550 for the first year).

All POs would entail monitoring costs. These costs estimates vary depending on the number of Member States with existing limit values and their levels. The total monitoring costs would range from €240,000 to €870,000 over 40 years, depending on the PO, with PO 4 entailing the lowest costs. These monitoring costs are relatively low as public authorities would only undertake air monitoring on sample groups. The administrative costs for air monitoring would be proportionally low for the same reasons, with administrative costs ranging from € 50,000 and €180,000 over 40 years.

At all options, public authorities would save money through reduced cases of ill-health cases. With its low impact on workers’ health, PO 4 would allow public authorities to save just €690,000 over 40 years, mainly due to the reduction of male infertility cases. POs 3 and 2 would result in higher avoided costs of approximatively €3.9 and €5.3 million, respectively. The difference between the two is explained by the higher number of lung cancer cases avoided under PO 2.

The 9 public authorities which currently have no limit value for PAHs would also benefit from all POs in terms of avoided costs linked to the process of setting a binding limit value for PAHs. Setting a limit value requires resources for carrying out the necessary analysis, including impact assessments. Under all POs, the costs related to the preparatory work would be borne at EU level, representing cost savings for public authorities of €900,000 in total. In addition, those public authorities with a higher existing national limit value than the values proposed in the different POs will also benefit from the analysis carried out at EU level. Those Member States will save between €0.7 (PO 4) and €0.9 million (PO 2).

Overall, the costs to public authorities are expected to significantly outweigh the benefits under PO 2. Under POs 3 and 5, the costs to public authorities would also outweigh the benefits, but to a lesser extent. Under PO 4, public authorities would save money over the reference period.

With regard to the impacts on specific Member States, based on the distribution of companies within Member States 143 , at PO 2 the effects from closure of coking plants (100% discontinuation) may be most felt in Czechia (22% of plants), followed by Poland (19%), Romania (17%) and Greece and Spain (both 11%). A similar pattern should be observed at the 70 ng/m3 with 45% closures.

A high level of discontinuations in coal tar plants (62% at PO 2 and 34% at PO 3) may be the most disruptive in Spain, followed by the remainder of Member States with plants: Belgium, Czechia, Denmark, Germany and Poland. The outputs from this sector however feed into many industries, therefore industries that are related to coal tar distillation plants will be negatively affected, regardless of the Member State in which the plants are located.

In casting of metals, up to 2% discontinuations are predicted at PO 2, which will affect Czechia and Italy the most.

The discontinuations in graphite and carbon electrode manufacturers at the PO 2 and PO 3 (44% and 16% discontinuations, respectively), may be felt mostly in France (25%), Poland and Germany (both 20%), Spain and Austria (both 10%).

Among those Member States identified above, Germany has an existing limit corresponding to PO 3. Therefore, the impact on businesses operating in Germany should be limited. All other Member States have a limit value higher even higher than PO 4 (see annex 6 for more details) or have no limit value (Italy and Belgium).

It also means that the expected reduction in discontinuation from the transitional value provided by PO 5 will also more benefit to the Member States identified above.

These impacts on specific Member States should be treated with caution as it also depends on the RMMs currently in place in each company. In the absence of such specific data, only the distribution of companies among Member States could be used.

(e)Impacts on digitalisation

By leading to discontinuations in key sectors contributing to the production of semiconductors (e.g.in particular other non-ferrous metallurgy 144 and graphite and carbon electrode manufacture 145 ), PO 2 might hinder the digital transition of the EU, including the target under the Chips Act 146 to double the production share of semiconductors. 6.7% and 0.4% of businesses operating in other non-ferrous metallurgy and graphite and carbon electrode manufacture are estimated to be at risk of discontinuing, respectively. When only considering the share of companies using cobalt, it is about 44% of businesses that are at such a risk, for both sectors. In the absence of sufficient data, this impact could not be quantified.

PO 3 would also cause discontinuations in these two sectors, but on a lesser scale (8.8% and 16.2% of businesses using cobalt and operating in other non-ferrous metallurgy and graphite and carbon electrode manufacture, respectively). With its transitional value, PO 5 would mitigate this impact on digitalisation as businesses operating in both sectors would have more time to plan their investments and find alternative RMMs, which is expected to reduce the number of discontinuations.

PO 4 is not expected to result in impacts on digitalisation.

(f)Impacts on EU Strategic Autonomy

Among those sectors subject to international competitiveness, the structure of two sectors would be significantly impacted under PO 2: coking plants (100% of businesses expected to discontinue) and steel and iron foundries, ferroalloys (10%). This is likely to lead to greater dependence of the EU on its main foreign competitors, hampering its objective of strategic autonomy.

While the structure of the steel and iron foundries, ferroalloys market should not be significantly impacted under PO 3, 45% of coking plants are estimated to discontinue. Therefore, PO 3 would also have an impact on the EU’s strategic autonomy as sectors dependent on coking plants production (e.g. coal tar) will be forced to increase their imports from outside the EU. The transitional period provided by in PO 5 is expected to result in less discontinuation, which should limit the impact of a limit value at the level of PO 3. However, the magnitude of this mitigation cannot be quantified.

As PO 4 is not expected to entail any discontinuation, the impact on the EU’s strategic autonomy is likely to be zero.

6.2.3.Environmental impacts 

(a)Direct impacts on environment

At all POs, businesses will need to invest in alternative RMMs. To limit the exposure to workers, these RMMs will need to better enclose the different processes, resulting in lower levels of PAH emissions in the air, and therefore positive impacts on the environment. This impact is all the more positive as some individual PAHs have been identified to have persistent, bio-accumulative and toxic chemicals properties, as well as very persistent and very bio-accumulative properties 147 . Although the impact could not be quantified, the number and level of enclosure of the alternative RMMs will be proportional to the strictness of the limit values.

(b)Indirect impacts on environment

The success of the Green Deal will depend among others on the availability of key essential products. The infrastructure needed to achieve climate neutrality by 2050 relies on metal products, such as steel, aluminium and silicon. These metals are essential for manufacturing solar panels, wind turbine blades, electric vehicles and other sustainable forms of transport, building energy efficient buildings, etc. Furthermore, some recycling of metals is dependent on carbon and graphite electrodes. A disruption in any sectors relevant for the above-mentioned products or metals in the EU would therefore compromise the success of the EU Green Deal.

PO 2 would entail an estimated high number of discontinuations in the following sectors: coking plants (100%) and steel and iron foundries, ferroalloys (10%). Furthermore, some sectors are composed of a few companies using cobalt for their production. If PO 2 would not significantly impact these sectors as a whole, it would significantly impact the share of businesses using cobalt: coal tar distillation (0.2% of the entire sector / 61.9% of businesses using cobalt for their production), other non-ferrous metallurgy (6.7% / 44%), graphite and carbon electrode manufacture (0.4% / 44%) and manufacture of aluminium (0.2% / 30%). All these discontinuations are expected to result in the disruption of essential products for the EU Green Deal.

PO 3 is also expected to result in discontinuations in the following key sectors: coking plants (45% of the entire sector / 45% of businesses using cobalt for their production), other non-ferrous metallurgy (1.3% / 8.8%), coal tar distillation (0.1% / 33.6%) and graphite and carbon electrode manufacture (0.15% / 16.2%). While the impact of PO 3 is less, the significant share of discontinuations in the coking plants sector could result in disruptions in essential supply chains, hindering the EU Green Deal’s objectives.

The transitional value provided for by PO 5 would apply to all the above-mentioned sectors and is expected to reduce the number of possible discontinuations. Indeed, businesses would have more time to plan their investments or to find innovative RMMs that would allow for complying with the final limit value of PO 5 (which is equivalent to PO 3). Therefore, and although it is unquantifiable, the impact of PO 5 on the Green Deal should be limited.

As no discontinuation is expected under PO 4, the impact on the Green Deal should be zero.

6.3.1,4-dioxane 

Each BLV option was set in relation to the OELs based on a function derived in the RAC opinion 148 which establishes a correlation between both. As such, a worker exposed to levels below an OEL (e.g. 7.3 mg/m³) is assumed to be also compliant with the corresponding BLV (45 mg HEAA in urine/g Creatinine). Therefore, for each policy option, the BLV is not assumed to bring additional benefits and adjustment costs than those related to its corresponding OEL. Conversely, BLVs will result in additional monitoring costs as the procedure is different than for monitoring compliance with an OEL.

6.3.1.Social impacts 

(a)Impacts on workers’ health

Only PO 2 would result in decreased health burden compared to the baseline. At this level, no worker would get ill following exposure to 1,4-dioxane. The monetised benefits resulting from the decreased health burden would account for up to €3.2 million over 40 years.

Table 20: number of avoided cases of ill-health and monetised health benefits over 40 years (compared to the baseline)

(b)Impacts on employment

Under PO 2, 6 companies active in the following three sectors, with 140 workers in total, would discontinue (a) manufacture of basic chemicals, fertilisers and nitrogen compounds, plastics and synthetic rubber in primary forms, (b) manufacture of paints, varnishes, and similar coatings, printing ink and mastics, and (c) manufacture of other chemical products. The associated social costs resulting from discontinuances would account for €13 million.

No impact on employment is expected under PO 3.

6.3.2.Economic impacts 

(a)Impacts on businesses, including SMEs

With regard to adjustment costs, PO 2 would result in the discontinuation of 5 small- and 1 medium-sized enterprises, active in the manufacturing of basic chemicals, fertilisers and nitrogen compounds, plastics and synthetic rubber in primary forms, manufacturing of paints, varnishes, and similar coatings, printing ink and mastics, and manufacturing of other chemical products. As shown in table 21, they represent a very small percentage of the enterprises with exposed workers. No discontinuation is expected under PO 3 as the evidence gathered indicates that existing RMMs are sufficiently effective to ensure compliance.

Table 21: number of discontinuations, including per company size and total discontinuation costs

The total costs in additional RMMs for those companies not discontinuing under PO 2 would be of €19.2 million over 40 years, mainly supported by businesses operating in the manufacturing of basic chemicals, fertilisers and nitrogen compounds, plastics and synthetic rubber in primary forms, manufacturing of paints, varnishes, and similar coatings, printing ink and mastics, and manufacturing of other chemical products 149 . 

According to estimates 150 , companies would have to make investments of about €53 million in additional RMMs during the first year to comply with PO 2, in particular small- (€35 million) and medium-sized enterprises (€ 7.6 million). Recurrent costs for the same PO would account for -€34.2 million. This negative value is due to the lower recurrent costs linked with the new RMMs, compared to the baseline 151 .  

According to the evidence gathered, the current workers’ exposure levels to 1,4-dioxane are already below PO 3. Therefore, no adjustment cost is expected under this option.

Some companies are expected to check that they are complying with PO 3 (e.g. by being at 10% or less of the OEL as foreseen in the UNE-EN 689 Standard on Workplace Atmospheres 152 ). However, a more significant number of companies is expected to check compliance under PO 2 than PO 3.

Monitoring costs (both for OELs and BLVs 153 ) for PO 2 would account for about €118.8 million over 40 years, of which 72% would be borne by larger companies. Under PO 3, businesses are expected to spend about €56.85 million to check compliance during the same period, of which about 67% would be at charge of large enterprises.

Table 22: total estimated monitoring costs in € million over 40 years by policy option and size of company

The administrative costs associated with the monitoring campaigns vary between the POs. Under PO 2, companies would need to spend over 40 years about €7,1 million for the organisation of the additional monitoring campaigns compared to the baseline, against €4 million under PO 3. Small-sized enterprises would bear about 70% of the total administrative costs related to the organisation of monitoring campaigns over 40 years.

Table 23: total estimated costs of administrative burden of additional monitoring by policy option and size of company over 40 years

In total, the compliance costs represent a small part of the turnover, or even negligible for PO 3. Only under PO 2 would they represent up to 0.47% for small enterprises operating in the chemical sector and which has an industrial use of 1,4-dioxane as a solvent and generate it as a by-product,.

Only PO 2 would have a positive impact on reducing the number of ill-health cases and would therefore allow businesses to achieve cost savings of €1.3 million over 40 years. These estimates only relate to local irritation in the nasal cavity cases, and are therefore likely to be underestimated.

(b)Impacts on competitiveness

POs 2 would have a limited impact on cost and price competitiveness with six businesses obliged to discontinue, partly or totally, due to the high compliance costs. However, such discontinuations would not have any impacts on the relevant sectors 154 , directly impacting 0.0003% of the companies operating in these sectors (6% of the companies with exposed workers to 1,4-dioxane). No discontinuation is expected under option 3. Therefore, no impact on consumers, both in terms of price or diversity of supply, is expected, including at the most stringent PO.

With regard to the impact on international competitiveness, the additional costs per company as a percentage of turnover are relatively limited. Furthermore, most of the companies impacted by the different POs are SMEs, which are usually operating at national scale and are therefore not subject to relocation. Therefore, no impact on international competitiveness is expected, even under PO 2.

Even at the most stringent PO 2, the share of the compliance costs compared to average annual R&D expenditure would not exceed 0.0025% 155 . Therefore, no impact on capacity to innovate under all POs is expected.

(c)Impacts on the single market

All Member States currently have a binding national limit value above PO 2 or no binding OEL (FI, IE and PT have an indicative OEL and the other Member States have a binding OEL). Therefore, the latter should reduce the regulatory divergences between Member States, resulting in a more level playing field in the internal market. The impact would be similar under PO 3 as 26 Member States would need to revise a more stringent or set a binding limit value for 1,4-dioxane.

Despite the establishment of a more level playing field, the impact on businesses would be limited as a significant share of them are SMEs. Therefore, it is expected that they rather operate at national scale than operating in several Member States.

(d)Impacts on public authorities/Member States

Under PO 2, all Member States would need to adapt their existing OELs, resulting in transposition costs of €810,000. The transposition costs related to OELs for PO 3 would be of €750,000, as a lower number of Member States would need to adapt their existing limit values (25 156 ). The transposition costs are likely to be slightly underestimated as the analysis does not take into account the transposition costs for BLVs.

Avoided costs of ill health for public authorities are expected only under PO 2 as PO 3 would not result in a reduced number of ill-health cases. Public authorities are expected to save €2 million over 40 years.

Furthermore, Member States are also expected to save money related to the process of revising a national limit value, as the work will be done at the EU level. The costs saved related to OELs would amount up to between €1.4 (PO 2) and €1.5 million (PO 3). These costs saved are expected to be even higher if we also take into account BLVs.

The total benefits for public authorities would outweigh costs under any POs.

Under PO 2, all Member States would need to change their legislation, which would affect all companies producing 1,4-dioxane in these countries, especially businesses in Italy, Germany, Spain, France and Greece as they share more than a half of all companies in the EU with workers exposed to 1,4-dioxane. There are no companies in Cyprus with workers exposed to 1,4-dioxane.

(e)Impacts on digitalisation

No PO is expected to have an impact on digitalisation.

6.3.3.Environmental impacts 

(a)Direct impacts on environment

The alternatives RMMs that businesses are expected to put in place under PO 2 are likely to have a positive impact on the environment due to the increased enclosure of processes (e.g. closed system and partially closed system). This impact could not be quantified. No direct impact under PO 3.

(b)Indirect impacts on environment

No impact on the green transition is expected under PO 3. PO 2 would contribute to the zero pollution ambition for a toxic free environment by better enclosing the different processes, as described in the section related to the direct impacts on environment.

6.4.Welding fumes

6.4.1.Social impacts 

(a)Impacts on workers’ health

As mentioned in introduction of section 6, PO 2 (inclusion of welding fumes in Annex I) would result in an additional 1% reduction in excess risk 157  compared to the current trend 158  for five years. This additional 5-years reduction in excess risk, which is due to the additional RMMs in place,  would result in more than 1,600 lung cancer cases avoided over 40 years compared to the baseline, representing between €997.5 (method 1) and €510 million (method 2) health benefits (see table 24).  

Table 24: number of avoided cases of ill-health and monetised health benefits over 40 years (compared to the baseline)

(b)Impacts on employment

In the absence of data on current exposure levels to welding fumes, neither possible discontinuations nor impacts on employment could be quantified. However, as welding processes are often the only technically feasible option for joining metals, the impact on discontinuation, and therefore on employment, is expected to be limited or negligible.

6.4.2.Economic impacts