Should you even be using the safety data sheet for a COSHH risk assessment? Consider what is now in the latest ACoP for COSHH.

Paragraph 74 - Where a work activity may expose employees to more than one substance hazardous to health, the employer must consider the possible enhanced harmful effects of combined or sequential exposures.
and
Paragraph 57 - The risk assessment should consider the work activity, including:
all the substances hazardous to health (including biological agents, and simple asphyxiants) arising from the work (used, produced, synthesised, created as waste or by-products, or released from processes or during accidents, incidents and emergencies);
work done by sub-contractors, at the workplace, that may exposure employees to substances hazardous to health.
and
Paragraph 67 - It may be necessary to collect information on the properties and attributes of substances hazardous to health from a variety of sources to fully inform the assessment process

This is further covered in paragraph 35, which states:
When deciding whether the substances used or produced in the workplace are covered by COSHH, employers should also consider the following:
Different forms of a substance may present different hazards, eg substances may not be hazardous in solid form but may be hazardous when ground into fine powder or dust that can be breathed into the lungs.
Nanoparticles (ie particles less than 100 nanometers) may be more toxic than larger particles of the same chemical substance.
Impurities in a substance can make it more hazardous, eg crystalline silica is often present in minerals which would otherwise present little or no hazard.
Some substances have a fibrous form which may present a potentially serious risk to health if the fibres are of a certain size or shape.
Some substances have a known health effect but the mechanism causing it is unknown, eg certain dusts of textile raw materials cause byssinosis.
Exposure to two or more substances at the same time or one after the other may have an added or synergistic effect.
Epidemiological or other data, eg reports of illness due to new and emerging agents, indicate that a biological agent that does not already appear in The Approved List of biological agents could nevertheless cause a hazard to health.
One-off, emergency situations arising out of the work activity, such as a dangerous chemical reaction or fire, could foreseeably produce a substance hazardous to health.
‘Wet work’ is one of the most frequently and consistently reported causes of irritant occupational contact dermatitis. ‘Wet work’ is the term used to describe tasks involving prolonged or frequent contact with water, particularly in combination with soaps and detergents.

On the basis of this I contend that a risk assessment for COSHH has to start with the task, then identify what chemicals are used during the task and what happens to them when they are used. You can then identify the real hazard.

Furthermore, there are thousands of chemicals that have never been assigned a hazard statement but that in use can represent a hazard to health (note the comment in the ACoP on wet work). In his latest edition of his guide to patch testing de Groot lists 4350 substances known to dermatologists as skin sensitiser, only a small fraction of which will ever have been assigned R43/H317.

In my work I see numerous risk assessments based on the safety data sheet that are simple wrong because they have not identified the real hazard.

Chris

Jorge

The problem is that the safety data sheet does not really tell you what the real hazard is and can mislead. I had a case of supposed occupational allergic contact dermatitis to a sensitiser in an epoxy impregnated carbon fibre mat. The sensitiser (tetraglycidyl methyldianiline - R43) made up 63% of the epoxy impregnant. However, when we tested the material the TGMDA was not bioavailable and was therefore not the cause of the employees dermatitis. So the risk assessment did not need to show this as a hazard for the task in question. However, in another part of the plant the mat was softened using a solvent. As a result the TGMDA became bioavailable and thus a significant hazard requiring quite different control measures. Same product, different hazards. Would you be able to recognise this from the data on the safety data sheet?

In any event, the safety data sheet is not a reliable indicator of what you have. A study by the European Chemicals Agency (ECHA), the organisation responsible for REACH, found that of 1,181 companies they inspected in 29 countries, mostly ‘downstream users’, i.e. formulators, regarding compliance with REACH, 52% were producing safety data sheets described in the study as ‘deficient’.
ECHA – REACH-EN-FORCE2

The HSE has actually cautioned its inspectors about using the safety data sheet as a reliable source of product hazard data.

Chris

While it is true that SDSs' have varying information, we have to recognise that the classification of substances & mixtures is not to the degree of perfection, but based on current toxicological etc. data, knowledge & expertise, of which skin is a part but not the only one.

It is indeed in Section 2 of the SDS that has a heading "Hazards identification" i.e. the hazard statements pertaining to the mixture/preparation are included and they should be the key ones to consider. For mixtures/preparations, Section 3 gives composition/information on ingredients

I very much doubt that using SDS & H statement and other information in from SDS's is a flawed process as long as you recognise the COSHH Principles of good control practice. The HSE's COSHH e-Tool uses the H-Statements as one of the key aspects for a limited range of activities/operations.


http://www.hse.gov.uk/ch...-classification-work.htm

Jay

"...so they should be the key ones to consider"

What if the product does not contain any chemicals that have been allocated a hazard statement, or if, as a result, the supplier does not supply a safety data sheet? Does this mean you can ignore the potential for the chemical to cause damage to health? In my work I see many situations where damage to health has occurred as a result of exposure to a chemical that has not been assigned a hazard statement/risk phrase.

Chris

So what do you do about water then. Wet work, i.e. exposure to water, is the most common cause of occupational contact dermatitis. Do you insist on a safety data sheet for water, and would it indicate water as an irritant? Yet for the dermatologist water has long been recognised as an irritant.

"An irritant is defined as any agent, physical or chemical, capable of producing cell damage. Everything can be an irritant if applied for sufficient time and in sufficient concentration. Water, being the most abundant element of the skin, is usually regarded as banal and gentle. However, the irritancy of water is beyond doubt." - From Dermatotoxicology, second edition, Zhai H, Maibach HI, CRC Press

Chris

An additional consideration is whether any one of the constituents is bioavailable. For example stainless steel will contain both chrome and nickel, two potent sensitisers. However, since they would not normally be bioavailable they present virtually no real hazard. Of course, should your process cause one or both of them to become bioavailable, then you have a real hazard to contend with. So the fact that your product may contain one or more 'nasties' does not mean automatically that these require special control measures.
Another consideration is whether the product could contain something that changes in use. For example, a machinist allergic to formaldehyde reacted to a metalworking fluid where there was no mention of formaldehyde on the safety data sheet. The fluid did contain a biocide, but this was not per se formaldehyde. It was, however, what is known as a formaldehyde releaser, i.e. in use it changed and formaldehyde was released into the fluid, hence causing the allergic reaction.
Chris