What an SDS Doesn't Tell You for a COSHH Assessment

A safety data sheet tells you about a substance as it leaves the supplier. A COSHH assessment has to tell you what happens when your people use that substance, in your building, at your quantities and frequencies. Almost everything in that second category, the route of exposure, how long it lasts, who is nearby, and whether your controls are working, is absent from the SDS and has to be gathered by you.

Why isn't a safety data sheet enough for a COSHH assessment?

A safety data sheet and a COSHH assessment answer different questions. The SDS answers "what is this substance capable of doing?" and follows a fixed 16 section format set out in Annex II of REACH, which the UK retained as GB REACH after EU exit.

The Control of Substances Hazardous to Health Regulations 2002 ask a narrower and more awkward question: what is the risk to health created by your work with this substance, and what will you do about it. Regulation 6 requires a suitable and sufficient assessment of that risk before work starts. An assessment that simply restates the hazard classification from Section 2 of the SDS has not answered the question the regulation asks, because it says nothing about exposure.

The practical test is this. Two workshops can use the identical product, with the identical SDS, and reach entirely different conclusions. One decants 50 ml into a fume cupboard twice a month. The other sprays five litres a shift in a room with a wall fan. Same sheet, very different assessments.

What information is missing from an SDS?

The table below sets out the recurring gaps and where the missing information actually comes from.

The quantity and frequency gap

Quantity and frequency are the two variables that most often decide whether a COSHH assessment concludes "adequate control" or "not adequately controlled", and neither appears anywhere in a safety data sheet. The SDS is written once for every customer, from the sole trader buying a 250 ml bottle to the plant buying an IBC. Only you know which one you are.

The process form gap

The physical form listed in Section 9 of an SDS is the form in the container, not the form in the air. Heating a solvent, spraying a coating, sanding a cured surface or transferring a powder can all produce an inhalable or respirable fraction that the sheet never anticipated. This is the gap that most often makes an inhalation exposure route look far smaller on paper than it is in the room.

The control effectiveness gap

Section 8 of an SDS recommends control measures in the abstract, but COSHH holds you to the effectiveness of the controls actually installed. Regulation 9 requires engineering controls to be maintained in efficient working order, and local exhaust ventilation to be thoroughly examined and tested at defined intervals. An SDS cannot know whether your hood is in the right place, whether the fan has been serviced, or whether the operator leans past the capture zone.

Which exposure figures should you use, the SDS or EH40?

For work in Great Britain, the workplace exposure limits published by HSE in EH40/2005 Workplace Exposure Limits are the reference point, not whichever figure the supplier chose to print in Section 8. Safety data sheets are frequently written for multiple markets and may quote an American OSHA permissible exposure limit, an ACGIH threshold limit value, or a European indicative limit, none of which is the GB legal standard.

Two points matter when you transfer a limit into an assessment:

1. Check both reference periods. GB WELs are expressed as an 8 hour time weighted average and, for some substances, a 15 minute short term exposure limit. A task that averages comfortably below the long term limit can still breach the short term one during a five minute decant.
2. Check the notations. Skin, sensitiser and carcinogen notations change what "adequate control" means, because for carcinogens, mutagens and asthmagens COSHH requires exposure to be reduced as low as is reasonably practicable rather than merely below the limit.

Many substances have no GB WEL at all. Absence of a limit is not evidence of safety, and the assessment still has to reach a conclusion using the hazard statements, any supplier DNEL, and a control banding approach such as HSE's COSHH essentials guidance sheets.

What about substances that have no SDS at all?

Some of the most significant COSHH exposures in UK workplaces involve substances that no supplier ever sold you, so no safety data sheet exists. These are created by the process itself, and they are a common blind spot in SDS led assessments:

• Welding fume, which HSE now treats as carcinogenic regardless of the metal being welded
• Respirable crystalline silica from cutting, grinding or dry sweeping stone, concrete and engineered worktops
• Wood dust, including hardwood dust, from sawing and sanding
• Flour and grain dust in bakeries and food production
• Thermal decomposition products from heating plastics, oils or coatings
• Biological agents, which fall within COSHH but never arrive with a supplier sheet

If your COSHH register is built only from the SDSs in your filing system, these substances will be missing from it entirely.

It is also worth knowing what COSHH does not cover. Asbestos, lead and radioactive substances are excluded because they are dealt with under their own regulations, and hazards arising purely from flammability or explosive atmospheres fall under DSEAR rather than COSHH.

Where do safety data sheets go stale or thin?

Even where an SDS exists and is relevant, three weaknesses recur often enough to check every time.

Age. Suppliers must revise a sheet without delay when new hazard information emerges, and must supply the revision to recent customers, but there is no fixed refresh cycle that guarantees the copy in your folder is current. A sheet in a drawer since 2019 may predate a reclassification.

Composition ranges. Section 3 discloses hazardous ingredients above threshold concentrations, often as wide bands such as 10 to 30 per cent. That band can be too coarse to support a quantitative judgement, and non hazardous or below threshold components need not be named at all.

"No data available". Sections 9, 11 and 12 frequently carry this entry, particularly for mixtures. It is not a reassurance. It means the assessment has to lean on the hazard classification and the task rather than on the sheet.

How do you fill the gaps in practice?

Treat the SDS as the hazard half of the work and collect the exposure half deliberately. A workable sequence:

1. List the substance and the specific task together. "Isopropanol" is not an assessable unit. "Wiping down 30 benches with isopropanol, 15 minutes, daily" is.
2. Record quantity, duration and frequency for each task, from observation rather than from memory.
3. Identify the form released during the task, not the form in the container.
4. Name who is exposed, including anyone nearby who is not doing the task, and flag young workers, new and expectant mothers, and anyone with existing respiratory or skin conditions.
5. Check the substance against EH40 and record the limit, the reference period and any notation.
6. Describe the actual controls, with their test or examination dates, then judge whether they are adequate.
7. Decide on monitoring, health surveillance and training on the basis of the exposure you have just described.

Getting this right is largely a documentation problem, which is why tools that read the SDS and then ask task specific questions tend to produce more defensible assessments than a copy and paste from Section 8. Safe Foundry works this way: it extracts the hazard data from your uploaded sheet and pairs it with the activity description you provide, so the task variables the SDS never covered are captured alongside the substance data. You can see how the SDS to COSHH workflow is structured or read more about what each assessment captures.

None of this removes the need for competent judgement. An assessment still has to be made by someone who understands the process, and complex or high hazard work may warrant occupational hygiene input, including air monitoring under Regulation 10. The SDS gets you a fast, reliable start on the hazard. The exposure is yours to describe. Practical questions about scoping and reviewing assessments are covered in the help centre and FAQ.

This article is general guidance on interpreting safety data sheets and is not legal or medical advice. Duty holders should refer to the current COSHH Regulations and HSE guidance, and seek competent advice where exposures are complex or high hazard.