COSHH Assessments – 7 key questions


In my last post I discussed the main problems that I often encounter with how COSHH assessments are carried out in practice. These were

  • concentrating on the hazards rather than the risks
  • neglecting to include process generated substances
  • concentrating on inhalation exposures and neglecting other routes
  • lack of emphasis on controls
  • failure to consider measures needed to ensure continued effectiveness of controls

To be “suitable and sufficient” a COSHH assessment needs to address the risks associated with the use of hazardous substances and decide on what measures are needed to reduce them to an acceptable level. The best way to achieve this is to base the assessment on the work. COSHH assessments should be focused on the process or task rather than the substance.

Guidance on risk assessments typically outlines a number of key steps as illustrated in the following flow chart




Although this provides a good overview of the process it is a little vague. In particular, it doesn’t really set out what is involved in the crucial third step – evaluating significant risks.

Some time ago, I sat down and tried to work out exactly what I do when I carry out a COSHH assessment. I concluded that I ask myself a series of questions:

  1. What substances are workers (and others) exposed to?
  2. What harm can these substances cause?
  3. Is exposure significant?
  4. What is currently being done to control exposure?
  5. Is that good enough?
  6. What needs to be done to improve control?
  7. What else needs to be done to ensure that adequate control is maintained? (e.g. testing controls, air monitoring, health surveillance, training etc.)

Let’s have a look at these in a little more detail

1. What substances are workers (and others) exposed to?

  • Starting with the task or process work out what substances are present – including both those substances bought in and those generated by the process (the latter often present the most significant risks).
  • Consider who could be exposed and how – i.e. by what routes (inhalation, skin, ingestion, injection)

2. What harm can these substances cause?

  • For substances bought it should be possible to determine the hazards they can present to health by looking at the label on their containers and the safety data sheet that suppliers must provide.
  • Information on process generated substances might be more difficult to locate, but the Health and Safety Executive publications and their website are often a good place to start.

For many people this is the end of the assessment, but if you stop here you have only identified hazards and haven’t addressed the risks. You need to continue to consider the other questions.

3. Is exposure significant?

This is probably the most difficult question to answer! Bear in mind that with hazardous substances the risk can be represented by the following equation :-


So the key is to try and quantify or estimate the degree of exposure. There are various ways this can be achieved, for example

  • personal air sampling
  • use of direct reading instruments
  • semi-quantitative measurements (e.g. direct reading dust monitors)
  • visualisation techniques (e.g. dust lamp or smoke tubes)
  • observations


The methods used will depend on circumstances. Sometimes observation is enough where it is obvious that improved controls are needed.


Whatever methods are used, a judgement has to be made on whether exposure is “significant” or not. Essentially you need to ask yourself “does something need to be done to reduce exposure?”. If the answer is “yes” then exposure is significant. (This is something I’ll return to in a future post)

4. What is currently being done to control exposure?

This question may be combined with the previous one. In most cases a COSHH assessment will be undertaken for an existing process where there are likely to be controls in place. You’ll need to identify what they are.

With a new process, the initial COSHH assessment should be carried out before the process starts. In that case you’ll need to identify what control options are available.

5. Is that good enough?

This may be asked in conjunction with questions 3 and 4. Once it’s been established what controls are available a judgement needs to be made on whether they are reducing exposure to a low enough level.

6. What needs to be done to improve control?

If existing controls aren’t good enough then, clearly, improvements will need to be made. A COSHH assessment should identify what measures are needed to control exposure.

Even where exposure is below exposure limits if there are ways of improving control they should be considered. For example, if personal protection is being used, even if it is adequate to reduce exposure below exposure limits, an attempt should be made to identify alternative controls.

Also if workers are exposed to carcinogens, mutagens or asthmagens, COSHH requires that exposure should be reduced to the lowest level reasonably practicable below any relevant limit. So in such cases it is particularly important to try to identify any additional controls.

7. What else needs to be done to ensure that adequate control is maintained?

This is probably the most neglected aspect of COSHH assessments even though the Regulations are quite explicit in requiring the assessment to consider what measures are needed to ensure compliance with all the regulations.

There are many examples in industry where expensive control measures are installed only for them to remain unused, used infrequently or used incorrectly thereby rendering them ineffective. To overcome these problems, effective management measures need to be put into place. COSHH Regulations 8 to 12 are about the things employers can and should do to ensure the controls they implement continue to work effectively. So once appropriate controls have been identified, the assessor needs to ask:

  • what needs to be done to ensure that the controls are used properly
  • what maintenance and testing is needed to ensure that engineering controls and personal protective equipment continue to operate effectively and what auditing should be carried out to ensure that the procedures and safe working methods are followed
  • is exposure monitoring and health surveillance needed as additional checks that the controls are effective
  • what information, instruction and training is required to ensure workers know why the controls are needed, how to use them correctly, procedures for reporting faults etc.

Common problems with COSHH assessments

The Control of Substances Hazardous to Health Regulations (COSHH) are undoubtedly the most important legal requirements for occupational hygienists working in Great Britain. They’ve been around for a long time now – the first version being enacted in 1988 – yet, in my experience, there are still many organisations that haven’t completely got to grips with them. One of the main problems is that many employers do not fully understand the Regulations and what they require and often misinterpret them.

In essence, COSHH is relatively straightforward. It simply sets out a framework for the management of risks created by using hazardous substances at work. I’ve often used the following diagram to summarise the main requirements of the Regulations


Employers whose operations involve the use of hazardous substances have to undertake a risk assessment in order to decide on what controls are appropriate. The other measures are then about they need to do to ensure that the controls remain effective (they’re the management controls that I’ve discussed in a previous post).

1. COSHH is about CONTROL

The first mistake employers often make is to forget the real objective of COSHH – i.e. controlling  the risks. They put a lot of work into the assessment, often generating a lot of paperwork, but don’t follow through to implement effective control regimes. The assessment is important – it is needed to establish priorities by deciding on which are the most significant risks, but it is primarily meant to be a tool to allow the employer to develop an action plan. It’s a means to an end not an end in itself.

2. Assess the RISK not the HAZARD

The second common problem is that many employers do not understand how to undertake the assessment.  It is quite common to focus on the substances used taking information from the safety data sheets provided by the suppliers and transferring it to a standard form , in effect, producing a summarised data sheet. This often involves a lot of time and resources, and although such summarised data sheets can be a useful information tool  they are not a risk assessment.

To see why this is the wrong approach it’s useful to take a look at the wording of COSHH regulation 6 which sets out the requirements for the assessment.

‘An employer shall not carry out work which is liable to expose any employees to any substance hazardous to health unless he has

a}made a suitable and sufficient assessment of the risks created by that work to the health of those employees and of the steps that need to be taken to meet the requirements of these Regulations

b) `implemented the steps referred to in sub-paragraph a.”

The key words that I’ve highlighted are “the risks created by the work”. The substances present hazards but, as the Regulations clearly state,  the risks are associated with the use of the substances.

I’ve heard many people state that they have carried out the assessments for all their substances. This normally indicates that they haven’t got it right. It’s not the substances that need to be assessed, but the work involving their use. For example, take a simple case where solvent is used for cleaning purposes. The risks, and the necessary controls,  associated with using a few millilitres applied with a cotton bud are clearly quite different than using several litres of hot solvent in a vapour degreasing bath. It is inappropriate to carry out a risk assessment of the substance – two assessments of two quite different processes are needed.

So the assessment should focus on the process. It should also concentrate on the risk not just the hazard. Risk is often defined as “the probability of harm occurring in practice”. For occupational hygienists it is best summed up by the following equation:


So the hazardous properties are important, but they are only half the story. The key to effective risk assessment, which s often neglected, is understanding the exposure of employees (and anyone else who could be affected by the work).

3. Identify ALL hazardous substances

Most COSHH assessments I’ve seen concentrate only on the substances that are bought in by the company. Yet there can be other substances present which need to be considered such as fumes, dusts and other airborne contaminants generated by the process or reaction products and intermediates in chemical manufacturing. In many cases it is these process generated substances that present the most significant risks, but they are often neglected.

Silica containing dust generated during cutting of paving slabs

4. Consider ALL routes of exposure

Many people carrying out assessments only focus on substances that can be inhaled. But there are other routes by which substances can come into contact with the body and be absorbed. All of these need to be considered.

Skin exposure by immersion (picture source:

Skin exposure is a common problem in industry leading to direct effects on the skin such as irritation, chemical burns and dermatitis and some substances can be absorbed through intact skin (this is something I’ve addressed in a previous post). Ingestion of hazardous substances can sometimes occur – normally due to contamination of foodstuffs, often following skin exposure. Penetration through the skin can occur by injection where workers are handling needles (e.g. healthcare workers) or where the skin has been abraded or otherwise damaged.

5. Identify controls and other management measures

The other common mistake is that employers often do not properly consider the measures needed to control the risks. They forget that controlling risk is the objective of COSHH and that the Regulations specifically require that the assessment must set out “the steps that need to be taken to meet the requirements of …. (the)…Regulations””.  This means that they need to decide on what controls are needed and what other measures might be needed (i.e. on the use, maintenance and testing of controls, air monitoring, health surveillance and information, instruction and training)

The first step in solving a problem is to recognise that it exists, but we then need to decide how to resolve it. I’ve set out some of the common problems with COSHH that I come across; I’ll consider how to avoid them when carrying out a COSHH assessment in my next post.

Key steps to control health hazards

At the end of June I was invited to make a presentation to the BOHS workshop on the control of health hazards at work. The key points made are summarised in a previous post.

I’ve finally got round to uploading the slides I used to Slideshare. These days I try not to overload my presentations with bullet points to avoid the risk of “death by Powerpoint” so I’ve added some explanatory text on the individual slides for the Slideshare version.

Reconsidering the “hierarchy of control”– Part 2

In a recent post I discussed the basis of the hierarchy of control and how it should be applied in practice. It’s something we cover when we deliver the BOHS module M103 “Controlling hazardous substances” and is also relevant to the control of physical hazards such as noise and vibration.

During the M103 course I’ll normally include an exercise where I ask the participants to list the different types of controls they’d include in the main categories (i.e. source, path and worker). They will usually include “training” in the list of worker based controls. Now, I’m not one for being controversial(!) but I usually use this as an opportunity to start a discussion as I don’t think training fits here. “Training” is something of a vague term. Of itself it won’t control exposure.


When we talk about training as a control what we are usually thinking about is the underlying measure – good working practices or safe working procedures. Different ways of carrying out a job can lead to different levels of exposure. To minimise the health risks we need to establish which working practices will minimise exposure and make sure workers are aware of them. Training will help with this, but only if we’ve identified the best methods. Training workers in poor methods will actually be counterproductive and increase exposure.

In fact training (along with information and instruction) is always needed whatever controls are introduced. Workers need to be aware of the hazards they’re working with and the associated risks. They also need to know why they shouldn’t reintroduce a substance that has been eliminated or substituted, how to use any engineering controls provided properly and, where appropriate how to maintain and test them, what good working methods and safe working practices need to be followed and how to obtain, fit, use and maintain personal protective equipment. In fact quite a lot of training is needed in most cases. So, although I don’t “training” fits into the traditional list of measures in the hierarchy, it’s an essential component of an effective control regime.

To me, training fits into a group of measures that are usually needed to ensure that the controls that have been introduced continue to work effectively. I call these the management measures  and they sit alongside the traditional hierarchy of control.

There are many examples in industry where expensive control measures are installed only for them to remain unused, used infrequently or used incorrectly thereby rendering them ineffective. To overcome these problems, effective management measures need to be put into place. They include

  • information, instruction and training to ensure workers know why the controls are needed, how to use them correctly, procedures for reporting faults etc.
  • supervision to ensure that the controls are used properly
  • maintenance and testing to ensure that engineering controls and personal protective equipment continue to operate effectively
  • auditing to ensure that the procedures and safe working methods are followed
  • exposure monitoring and health surveillance as additional checks that the controls are effective
  • welfare facilities (i.e. washing facilities, changing rooms, segregated rest rooms, separate storage for clean and contaminated clothing) may need to be provided.
  • good housekeeping and cleaning of the workplace.

Anyone familiar with the British Control of Substances Hazardous to Health Regulations (COSHH) will recognise them, as they form the basis of regulations 8 to 13

We can modify the “hierarchy of control” model to incorporate them as illustrated in the following diagram. The management measures form a separate list running in parallel with the traditional list of controls.


Once appropriate prevention strategies, engineering measures, work practices and PPE have been identified, employers need to decide which of the management measures should be introduced as part of a control regime to ensure that the controls continue to work effectively at reducing the risk to an acceptable level.

What’s a thorough examination and test?


Companies that have installed local exhaust ventilation systems have to undertake a “thorough examination and test” (TExT) at least once every 14 months. The objective of the TExT is to find any significant defects and to have them remedied to regain control. In most cases they will employ an external organisation to carry out these tests so that they can meet their legal obligations. There are a large number of companies offering this service – sadly not all of them do a proper job. Sometimes this is due to a lack of competence but it may also be because the test engineers are under pressure to test systems quickly to maximise the number of tests they can carry out during the day. Unfortunately that leads to corners being cut and too many testing companies carry out a minimal, rather than a thorough test.

One of our consultants was visiting a client recently to carry out ventilation testing on a system used to control fume generated during brazing operations in a plumbing shop. It consisted of 8 individual flexible arm captor hoods connected to a main duct from the fan (see photograph below). As I’ve discussed in a previous post, captor hoods are rarely effective at controlling contaminants and other types of hood are usually preferable. The ductwork design, with no attempt at balancing, means that it would be difficult to achieve the same velocity in each branch.  However, in this case, the design would probably be adequate providing management ensured that the users positioned the hoods are carefully over the plume of fume generated during brazing and the appropriate capture velocity could be achieved in each leg.

The system had previously been tested by another company and our consultant asked to see the report. He needed this so he could compare his results with those from the previous test so that he could determine whether there was any deterioration in the system’s performance. The report he was given (see below) was a good example – of an inadequate test report.


Guidance on what should be done during a TExT is provided in the Health and Safety Executive’s document HSG258 Controlling airborne contaminants at work: A guide to local exhaust ventilation. This advises that a TExT will normally involve three stages:

  1. A thorough visual examination to verify the LEV is in efficient working order, in good repair and in a clean condition.
  2. Measuring and examining the technical performance to check conformity with commissioning data. Even on a simple system this would normally include determining hood capture and face velocities and duct velocities and measuring static pressure at behind all hoods and across the air cleaner (where present) and fan.
  3. An assessment to check the control of worker exposure is adequate. At the very least, this will involve the use of smoke tubes or a dust lamp.

Of course it isn’t always possible to carry out all the tests recommended in HSG258 due to design, safety and practical considerations. Nevertheless a proper TExT, even on a simple system, will involve a number of different measurements and tests.

For the multi branch system, as in this case, I’d expect a TExT to involve, as a minimum, the following:

  • a thorough visual examination
  • a smoke test on each hood as a way of checking on how effective they were at capturing the fume generated by the brazing process
  • capture velocities measured on each hood at the point further from the hood where the contaminant could be generated
  • face velocities measured at each hood
  • static pressures measured behind each hood
  • measurement of the duct velocity in the main branch (it may not be necessary to measure duct velocities in each branch
  • the static pressure across the fan
  • checking the fan to ensure that it is rotating in the correct direction

The test carried out by the engineer from the previous company involved only a visual examination and measurement of duct velocities. According to the report, an assessment of control had been made from these results. Not only was the scope of the testing inadequate, the report states that the duct velocity had been measured using a vane anemometer. This is an inappropriate instrument for this type of test. I suspect the engineer had placed the anemometer at the entry to the duct inside the hood. Not only is this not the best way to measure duct velocity, It would be difficult to position the anemometer properly to measure the flow into the duct.

The “duct” velocities given in the report are all around 10 m/s. That would be adequate for conveying fume, but would the contaminants be captured effectively? No face or capture velocities are provided. If the hoods were used as receptors, and were positioned very carefully above the plume of contaminants, there was a chance that the system would achieve adequate control. But this would need to be verified by observations of the process and, ideally, a smoke test. Neither was done.  It’s probably more likely that the arms would be used a capture hoods. In that case it would be necessary to check that the capture velocity was adequate. With moveable hoods of this type the best approach would be to determine the distance from the hood face where the minimum acceptable capture velocity could be achieved and include that in the report so that the user could be informed. There is nothing in the test report to suggest any of these important measurements were done.

In my view, the original report was inadequate. More time was needed to carry out a proper thorough examination and test. No doubt their company charged the client considerably less than we did – but, as with most things in life, you get what you pay for.

Even when an outside company is employed to carry out TExT the legal obligation to ensure that the test is thorough remains with the owner of the system. So they need to ensure that they’re using a competent company who will carry out the test properly in accordance with HSE’s guidelines.  Its important for employers to ensure that they employ a company who will do a proper job.

Key steps to control exposure to health hazards at work

Last week I travelled over to Derby to attend a workshop organised by Mark Piney, the new BOHS “topic lead” on control of exposure. The objective was to explore the key issues on control that BOHS could be working on, and to try to decide on priorities.

The workshop started by exploring the experiences and perspectives of occupational hygienists and related professionals from different sectors. As part of this I was asked to make a short presentation giving the perspective an occupational hygiene consultant. These are the key points I made during my presentation.

According to the International Occupational Hygiene Association (IOHA),

‘Occupational Hygiene is the discipline of anticipating, recognising, evaluating and controlling health hazards in the working environment with the objective of protecting worker health and well-being and safeguarding the community at large.’


To me the main objective is the control of health hazards. The other aspects (anticipation/recognition and evaluation) are important steps in a process leading to control. They are a means to an end, rather than an end in themselves.

Unfortunately, most people, including, in my experience, many health and safety professionals, haven’t heard about occupational hygiene. Where they have their knowledge is usually minimal and their impression is that occupational hygienists are all about sampling and measurement (i.e. evaluation). As consultants we’re often called in to carry out sampling for clients who often don’t expect us (and in some cases don’t want us) to advise on the control of exposure. Nevertheless, as committed professionals we inevitably will comment on control as we usually find there are ways in which exposure can be reduced when we undertake a consultancy assignment.

To me, to achieve and ensure effective control there are a number of key steps, which are illustrated in the following diagram.


First of all employers need to be aware that they have a risk to health. Without this  there is no reason to introduce controls, which, after all, is likely to appreciable involve time and money. Yet there are many situations in industry where significant health risks are unrecognised and, therefor, remain uncontrolled. Once the risk is recognised, employers need to be committed to resolving it and arrange or take effective actions.

Once the need to control a risk is recognised it’s important that employers think carefully about what controls are appropriate – i.e. they need to be properly specified. Too often personal protective equipment (PPE) will be introduced as the “easy” solution. But PPE is rarely as effective as people believe and, if it is used and managed properly, it will not be as cheap as expected. PPE should be a last resort. Other types of controls will usually be more effective and often cheaper in the long run. Where engineering controls (such as containment or local exhaust ventilation) is to be introduced it’s important that they are designed competently and this will normally require engaging external contractors. But employers need to ensure that the designers are given a proper, detailed specification. The truth is that this is rarely the case.

Powder weighing - incorrectly worn mask
Too often, employers rely on PPE to control exposure. There's usually a better solution available.

All controls should be properly designed to suit the process and the associated risks. This is particularly a problem with local exhaust ventilation, where “off the peg” systems are often installed which are  and not suitable for the application or where the designers don’t understand the risks resulting in inappropriate and ineffective designs.

Even where the controls are well designed they need to be properly implemented. At the very least this will involve verifying that they achieve effective control and meet the design specification. With engineering controls they need to be properly commissioned. Too often this doesn’t happen.

Once the controls have been implemented they need to be properly managed to ensure that during use they continue to do the job they were designed to do. This will require training, supervision, maintenance, testing, audit and review.

A trained and experienced occupational hygienist will have the knowledge and skills to assist employers throughout this process.

The BOHS is a small organisation with limited resources, but it has a committed and enthusiastic membership. Nevertheless it has role to play in improving the control of health hazards in the workplace.

In my experience employers need straightforward, clear and simple guidance to help them minimise health risks. A good example of this are the control guidance sheets produced by HSE as part of the COSHH Essentials control banding tool including a good number of sheets aimed at specific processes.


I think BOHS could take these as a model, collating good practice from the industries where our members work and producing simple guidance that our members, and others could use to help our employers and clients. Remembering that effective control isn’t just about the design of the “hardware”, we should also aim to produce simple guidance covering the management of controls.

We could also look towards working with HSE promoting their guidance and making it more accessible. The COSHH essentials sheets are also meant to reflect good practice, and as this changes they can become out of date. Again, perhaps we could find a way of working with HSE so that we can contribute to the review and modification of the guidance so that they continue to reflect current good practice.

COSHH Essentials

COSHH Essentials is a tool developed by the UK Health and Safety Executive. It’s a “control banding” technique which was originally intended to help small companies, without access to expert help, determine what controls are needed to control exposure to the hazardous substances they use or handle.

It was originally published as a “hard copy” manual in 1999, an online version appearing a few years later. The hard copy is no longer available.

In essence COSHH Essentials is a mathematical modelling tool which estimates exposure. The user enters some basic information on the substance being used (“R phrases”, boiling point or “dustiness”  and the amount used). The tool uses these details to determine which “control approach” is appropriate and the user is directed to simple control sheets relevant to the risk and the type of process.

It is a very simple tool that can work in straightforward situations where a single substance or “preparation” is being handled. However, it has some very serious limitations.

The following slides provide a more detailed outline of the process and a couple of examples to show how it works in practice

The two examples used in the presentation illustrate some of the strengths and weaknesses of the approach.

In the first scenario, where a low hazard powder is being weighed and mixed with water to form a slurry the only significant risk is inhalation of airborne dust and the tool works quite well. The specified control – local exhaust ventilation – seems appropriate. The scenario is based on a real situation where I recommended installing an extracted booth, and one of the control sheets generated by the tool provides a suitable design.

However, I think the conclusions and guidance provided for the second example – cleaning rollers with solvent – illustrate some of the problems with COSHH Essentials. Here there are two significant risks – inhalation of solvent vapours and skin contact with the liquid solvent. The tool is particularly poor where the latter is an issue. It doesn’t attempt to assess skin exposure other than directing the user to personal protective equipment (PPE) wherever skin or eye hazards are flagged up by the R phrases. Use of PPE should always be the last resort. Preventative measures and engineering controls are always preferable and should be considered first. This applies to skin contact as well as inhalation risks. In fact, protective gloves are rarely as effective as the user perceives and their introduction can sometimes be counter productive due to this. So, for this scenario I would have looked at finding an alternative solution such as using an applicator that minimised contact. Gloves might still be needed, but as a backup rather than the primary control.

My second concern with the output from the model for the second scenario is that it recommends enclosure of the process. Personally, I think this is over cautious. I think that local exhaust ventilation (LEV) would be adequate here. Enclosure would require automation and also LEV and, although quite feasible, this approach would be relatively expensive and I wouldn’t consider it to be “reasonably practicable” where only a few litres of solvent are used per day. This problem occurs because the usage category “medium” is specified where litre quantities of a substance are used. There is likely to be a big difference in risk between using 4 or 5 litres of a substance and 500 litres – yet in both cases the user is directed to classifying the quantity used as “medium”. I think the category is too broad. With this particular scenario, it would be better to consider the quantity as “small”. The tool would then specify “engineering controls” – effectively LEV – as the appropriate control approach. This seems much more reasonable to me. However I think that a user is most likely to classify use as “medium”. In this case COSHH Essentials is over cautious. That’s better than underestimating the risk, but could lead to more expense than really necessary.

In both the scenarios only a single substance was being used for relatively short durations. Where highly toxic substances, such as carcinogens, are used, or the process is more complex, particularly where there is more than one material or where substances are generated during the process, the use of COSHH essentials is unlikely to be appropriate.