Section 3.1 Proposed Methodology for the Identification Andassessment of Substances For

SEMI / IPC Feedback Regarding
European RoHS Methodology for New Substances
(April 12, 2013)

Section 3.1 Proposed methodology for the identification andassessment of substances for a potential restrictionunder RoHS2 (Page 14)

In Figure 1, we propose that Option 2 should point to Part I rather than Part II. Member States (MS) should not be allowed by by-pass Part I.

Our rationale: Part 1 methodology should be applied to all proposed candidate substances, not just those submitted by the Commission (COM). The criteria related to inclusion are technical and socio/economic, rather than political, so the methodology should not favour who initiates it but rather be based on uniform application of methods and criteria. The existing Option 2 allows substances that are not hazardous (or generate hazardous by-products at EOL) or are not used in EEE to be included in this methodology.

We propose adding text to clarify the condition for inclusion of a new substance to the methodology. We propose adding the following text directly after Figure 1:

If the proportion of a substance used in EEE is less than 50% of the total usage of the substance in the EU, then consider using REACH restrictions instead of RoHS; unless – a specific uncontrollable risk to human health or the environment has been conclusively identified that is due solely to end of life processes for WEEE.

Section 3.1.1 Draft methodology for identifying candidate substances(hazardous substances used in EEE)(Pages15 – 21)

In a number of Sections between Pages 15 – 21, we find the guidance on obtaining hazard data to be overly complicated. We propose the following alternative to be used in order of priority in determining hazard data:

Use EU harmonised classification if available
Use classifications in ECHA classification and labelling database
Is there evidence available that substance is persistent, bioaccumulative, an endocrine disruptor, sensitiser,
Use other sources of hazard data with caution, such as IARC.

Section 3.1.2.2 Inventory of substances used in EEE (Page 22)

We propose the following language be added:

The determination of substances used in EEE should focus on stakeholder contribution because this data will not be widely available from other sources (and may otherwise be inaccurate).

Additionally, we propose that a review of published literature be required to check for omissions. Do not rely on unreliable Internet sources.Also, validate all data since inaccurate estimations have been published and are misleading.

Section 3.1.3Draft methodology for the pre-assessment of candidate substances (Page 24)

Figure 5 is inaccurately listed as Figure 6 (under “Approach”).

Figure 5 illustrating the flow of information and decision should be revised as follows:

In lower right-hand side box in Figure 5: For text box, “List of substances of comparable lower priority”, we propose adding the following text: “to be assessed after higher priority substances have been completed”

Our rationale: Provisions for addressing lower priority substances are not adequately specified in Figure 5 or in the associated text.

The final text box should read:

List of substances of comparable lower priority to be assessed after higher priority substances have been completed

Eliminate the footnote in Figure 5:

=* a distinct threshold for high and low priority substances cannot be defined at the present stage. After pre-assessment of several substances guidance for future reviews can be developed.

Add in new footnote in Figure 5 as follows:

= * High priority substances are defined as having a score of 6 – 10.

Our rationale for changing the footnote: A starting point for evaluating substance hazards should be established. This can be revised as the methodology evolves.

Section 3.1.3.2 Step 2 Categorization/Prioritization of substances (Page 26)

The Table on Page 26 is inaccurately listed first as Table 7 and then as Table 5. Table 3 appears to be the accurate number.

Currently, this Table ranks hazards into only 3 tiers and this inadequately differentiates the relative hazards of the various classifications, hindering prioritization. The number of categories (tiers) should be expanded in order to differentiate between substances of significantly different hazard potential. We propose 10 categories (10 = High / 1 = Low).

Some hazards wereassigned the same score even though they have significantly different hazard properties.For instance, Category 1A carcinogens are much more dangerous than skin irritants, but both were given the same score.

We propose the following Table as a substitute:

Classification / Scores
10=High
CMR Category 1A / 10
CMR category 1B / 8
CMR Category 2 / 4
PBT or vPvB / 10
PB / 5
Acute Toxic Category 1 / 8
Acute Toxic category 2 / 6
Serious eye damage / eye irritant (Category 1); (Irreversible effects on the eye); Reference CLP 1272/2008 / 4
Serious eye damage / eye irritant (Category2); (Irritating to the eye); Reference CLP 1272/2008 / 1
Skin corrosion / irritant (Category 1A); (exposure duration:= 3 minute, <= 1 hour); Reference CLP 1272/2008 / 3
Skin corrosion / irritant (Category 1B); (exposure duration:<= 1 hour, <= 14 days); Reference CLP 1272/2008 / 2
Skin corrosion / irritant (Category 1C); (exposure duration:<= 14days); Reference CLP 1272/2008 / 1
Skin corrosion / irritant (Category 2); Reference CLP 1272/2008
[Note: Category 3 does not exist under CLP, but does exist under GHS.] / 1
STOT-SE, STOT-RE class 1 / 8
STOT-SE, STOT-RE class 2 / 6
STOT-SE, category 3 / 1
Hazardous to the aquatic environment (Chronic Category 1) / 10
Hazardous to the aquatic environment (Chronic Category 2) / 8
Hazardous to the aquatic environment (Acute Category 1) / 10
Hazardous to the aquatic environment (Acute Category 2) / 8
Endocrine disruptor category 1 / 10
Suspected endocrine disruptor category 2
(see / 8
Potential endocrine disruptor category 3 / 6
Sensitiser (skin or respiratory) / 6

We propose adding the following text after the Table:

If a single substance has more than one hazard, only the largest hazard score is used for prioritization purposes. [For example, if a substance is both a category 1A carcinogen and a category 2 eye irritant, the hazard score is 10.]

Low priority substances are defined as having a score of 1 – 5, while high priority substances are defined as having a score of 6 – 10.

Section 3.1.3.3 Step 3 Assessment of the use in EEE (Page 28)

We propose eliminating the following bullet:

•if the substance is widely dispersive used in EEE (used in applications, components, which are part of a great variety of appliances) or used in single, specific applications, which may undergo controlled waste management processes.

Our rationale: Applications where substances are used in one type of product where these products are collected and recycled separately, are extremely rare and should not be considered an important prioritization criteria.

In the un-numbered Table: We propose dividing the “10-1.000 t/a” category into two separate categories (“100-1.000 t/a” and “10-100 t/a”) to create an additional scoring category to provide greater granularity in establishing prioritization. Our rationale: The original 10-1.000 t/a category is too broad. Separating these two categories necessitated a revised scoring scheme (high score of 4 instead of 3).

We propose a revised Table as follows:

Substances in EEE placed on the EU market / Scoring
Total quantity contained in EEE per year:> 1.000 t/a / 4
Total quantity contained in EEE per year:100 - 1.000 t/a / 3
Total quantity contained in EEE per year:10 - 100 t/a / 2
Total quantity contained in EEE per year:< 10 t/a / 1

We propose adding the following after the un-numbered Table:

Total quantities greater than 1.000 t/a (Scoring = 4) should be prioritized for risk assessment. After assessment for these high priority substances have been completed, the lower priority substances should be addressed as priorities.

Section 3.1.4 Draft methodology for the detailed assessment of prioritized/selected substances (Page 29)

The following four pages of comments represent our proposal to revise Figure 6.

Step 0

One important revision is inclusion of how “by-products” are assessed in the methodology. We propose addition of a “Step 0”:

Step 0 – Information needed about by-products which are known to be produced at EOL

Step 0 should flow into Steps 1a, 1c, and 1d.

Our rationale for including new Step 0: Under RoHS, by-products should be of concern(See Article 6.1(b)).

Steps 1a – 3

Steps 1a – 3 are collection of data from all available sources. We propose that physical properties data should be included as this information may be needed to estimate exposure levels.

Step 4a

Delete Step 4a as it provides no benefit to the methodology.

This step will not provide useful information on whether substances pose a risk. Many very hazardous substances are routinely and safely disposed of using appropriate waste disposal procedures. Risks exist only if non-standard disposal methods are used, which are evaluated in steps 4b and 4c.

Although WEEE should be separately collected in the EU, it is not always recycled separately and there is no requirement to do so. The process used for recovery of metals is smelting and the feedstock to these processes includes WEEE and many other materials including ores and refinery wastes. Therefore there will never be additional costs or complexity for WEEE that contains any specific substance. Recycling processes in the EU must use BAT and comply with the Industrial Emissions Directive and so hygiene equipment is installed to comply with this legislation and to collect substances emitted from any potential feedstocks used. A few types of material from WEEE are recycled using dedicated processes such as phosphors from lamps but these processes are designed for recovery of specific materials and to comply with emissions legislation. Their complexity and cost are not affected by the presence of a hazardous ingredient. Therefore step 4a will not be useful to prioritise any substances that are used in EEE. Furthermore, any restrictions must be based on proven uncontrollable risk as this is the requirement for REACH restrictions. Where emissions can be controlled, there is no risk.

Step 4b

In the text box, we propose elimination of:

Evaluation of the risk for workers during WEEE management

In its place, we propose the following:

Evaluation of human health risks at all phases of the life cycle, but taking into special account EOL

We also propose elimination of:

Q: Is the exposure of workers > risk values or do other reasons for concern exist?

We also replacement of this deleted text with the following:

Q: Does the substance pose a risk to human health?

Our rationale for these changes: All life cycle phases should be considered, but take special account of EOL. We also propose that the limited scope of worker risks be replaced with the broader scope of human health risk.

Section 3.1.4.9 Step 4b) Evaluation of the risk for workers and neighboring residents(Page 35)

The current language does not provide adequate guidance as to how to conduct a risk evaluation.

We propose deleting the following text:

The information collected in previous steps (e.g., toxicological reference values, endpoints of concern, exposure data) will be considered to describe the potential risk. Exposure levels above reference values indicate that there is cause of concern. Objectives:

A qualitative risk characterization
If appropriate data are available a quantitative assessment may be performed.
If appropriate data are available it will be examined if there is an unacceptable exposure of workers in WEEE
If appropriate data are available it will be examined if neighbouring residents are at risk (e.g., persistent or volatile properties of substances)

We propose adding the following text in place of the deleted text above:

The risk to human healthdepends on whether the level of exposure to people who use or recycle the equipment or live near to recycling sites are exposed to levels of the substance, or to by-products from recycling that pose an uncontrollable risk. This is determined from:

If available, actual known levels that cause harm such as NOAEL (for all of the identified human health hazard classifications).
Alternatively use estimated published levels such as DNELs.
If no estimated values exist, then determine whether these can be estimated from data available for very similar substances, i.e. substances that are structurally and functionally similar with only small changes in carbon chain lengths. Chemical speciation also needs to be addressed (chromium). This should only focus on exposure (and chemical properties as a proxy) and not toxicity.

A possible definition of “similar substances” that is based on the definition used for the QSAR (quantitative structure activity relationship) methodology could be:

• The substances have common functional groups such as carboxylate groups, alkyl chains, specific metal ions (with the same valence state, etc.) and do not have different groups to each other.

• The carbon chain lengths of substances must be similar, i.e. plus/minus 4 carbon atoms of the substance in question.

• The substances are likely to give the same decomposition by-products.

And

The level of exposure in order of priority:

Actual measured levels or
Levels estimated only from vapour pressure, amount available at location of consumer (general public) or at recycling site environmental fate, from known values for very similar compounds, etc.

Mention again that these substances may be used in other non-electronic consumer products (clothes, home products, etc).

Prioritise substances with known levels that cause harm and actual measured levels of exposure over substances with greater uncertainty. Priority may be raised if synergistic effects from mixtures exist.

A human health risk exists if the above shows that:

The level of exposure by a substance or by-product (actual or estimated) at any site is one tenth of a NOAEL, DNEL, OEL (occupational exposure limit), or equivalent values.

By-products should be included in this Step because substancesare often burned at EOL and this may yield by-products (e.g. dioxins and furans). This is a requirement of Article 6.1(b).

In risk assessment, a margin of error is normally used when comparing exposure levels with relevant levels that cause harm. Margin of errors are used because exposure is often measured in a lab setting and extrapolation to humans is uncertain. A margin of error of one tenth is standard.

Section 3.1.4.10 Step 4c) Evaluation of the risk for the environment (Page 35)

The current language does not provide adequate guidance as to how to conduct a risk evaluation.

We propose deleting the following text:

Environmental concentrations near EEE processing plants will be compared with Predicted No Effect Concentrations (PNEC) in order to evaluate a potential risk to the environment.

The following information structured as described below is required:

Environmental exposure
Monitoring data: remote regions, biota

A risk characterization for the environment with focus on WEEE management will be performed.

We propose adding the following text in place of the deleted text above:

The risk to the environment depends on whether the level of exposure to the environment from manufacture, use of the equipment or recycling the equipment exposes the environment at any location to levels of the substance, or to by-products from recycling, that pose an uncontrollable risk. This is determined from:

Ideally - actual known levels that cause harm (for all of the identified hazard classifications)
Alternatively use estimated published levels such as PNECs
If no estimated values exist, then determine whether these can be estimated from data available for very similar compounds.

A possible definition of “similar substances” that is based on the definition used for the QSAR (quantitative structure activity relationship) methodology could be:

• The substances have common functional groups such as carboxylate groups, alkyl chains, specific metal ions (with the same valence state, etc.) and do not have different groups to each other.

• The carbon chain lengths of substances must be similar, i.e. plus/minus 4 carbon atoms of the substance in question.

• The substances are likely to give the same decomposition by-products.

And

The level of exposure in order of priority:

Actual measured levels or
Levels estimated only from vapour pressure, water solubility, amount available at location of consumer or at recycling site environmental fate, from known values for very similar compounds, etc.

An environmental risk exists if the above shows that:

The level of exposure by a substance or by-product (actual or estimated) at any site is one tenth of a PNEC or equivalent values.

By-products should be included in this Step because substances are often burned at EOL and this may yield by products (e.g. dioxins and furans). This is a requirement of Article 6.1(b).

In risk assessment, a margin of error is normally used when comparing exposure levels with relevant levels that cause harm. Margin of errors are used because exposure is often measured in a lab setting and extrapolation to the environment is uncertain. A margin of error of one tenth is standard.

New Section on Page 36

We propose adding a new Step (4d). We question whether RoHS is the appropriate regulation to address all substances used in EEE. To address this, we propose adding the following Section header and text:

3.1.4.11 Step 4d) Evaluation of Restriction as Only Viable Control

If other legislation or better enforcement of existing legislation is available to control risk, then options under these other schemes should be pursued instead of RoHS.

Section 3.1.4.11 Step 5) Evaluation of the availability of substitutes and of informationon their hazard properties (Page 36)

We propose providing addition clarity to the discussion of substitutes. We propose adding the following text after “socio-economic impact analysis.” and before “Sources of information:”.

For ALL possible substitute materials and designs, answer the following questions:

Are substitutes commerciallyavailable to all electrical equipment manufacturers now or will they be within 3 years* in sufficient quantities to replace all known uses and likely future uses of the substance? (Note; availability is a permitted criteria for exemptions)

* 3 years is suggested assuming a sufficient transition period of 4 years as was the case with the original RoHS Directive.

Are substitutes commercially available to all electrical equipment manufacturers now that are no less reliable or are they likely to be available within 3 years* for all main uses and likely future uses of the substance (Reliability is a criteria acceptable for exemptions)

* 3 years is suggested assuming a sufficient transition period of 4 years as was the case with the original RoHS Directive.

For all possible substitutes AND their recycling by-products – do all of these pose no risk or significantly less risk (i.e. at least ten times less) to human health and to the environment than the substance that they would replace? (Use the same methodology to determine risk as for steps 4a and 4b) – BUT, if insufficient data is available to determine whether risk is the same or greater then it is not safe to restrict substance until more R&D on alternatives is completed.

R&D of alternatives is already required under REACH.