Remember that most standards don't require a specific person to assess conformity.

If you were in the business of manufacturing switchgear then you would have an independent test house assess to IEC 61439-1 and other pertinant standards - product indemntiy insurers would likely insist on it and you may need to use a notified body.

If however as you suggest you're taking a terminal box and building something with it - or even instructing a 3rd party to do this - then you've no market need to go through all the compliance hoops. However if it goes horribly wrong and someone gets hurt then you're the one liable.

In R&D environments people build this kind of stuff all the time. They need to have a safe method of working. In any case it's always a good idea before instructing an independent test house to be converse with the standards and the test forms to know what your requirements are. It's not difficult to get a copy of the standard and take a broad view - of course the specialists at the test houses know these standards intimitely and an engineer who doesn't is going to widely misinterprit bits - but that doesn't mean you cant take a dilligent approach by reference to the standards.

Standards are there to be read and applied, not to tick boxes.

Remember that most standards don't require a specific person to assess conformity.

If you were in the business of manufacturing switchgear then you would have an independent test house assess to IEC 61439-1 and other pertinant standards - product indemntiy insurers would likely insist on it and you may need to use a notified body.

If however as you suggest you're taking a terminal box and building something with it - or even instructing a 3rd party to do this - then you've no market need to go through all the compliance hoops. However if it goes horribly wrong and someone gets hurt then you're the one liable.

In R&D environments people build this kind of stuff all the time. They need to have a safe method of working. In any case it's always a good idea before instructing an independent test house to be converse with the standards and the test forms to know what your requirements are. It's not difficult to get a copy of the standard and take a broad view - of course the specialists at the test houses know these standards intimitely and an engineer who doesn't is going to widely misinterprit bits - but that doesn't mean you cant take a dilligent approach by reference to the standards.

Standards are there to be read and applied, not to tick boxes.

Keir Stitt said:

In R&D environments people build this kind of stuff all the time.

Isn't R&D and experimental use exempted from one or more of the "Directives" (e.g. EMC Directive - Electromagnetic compatibility regulations 2016)?

Keir Stitt said:

If however as you suggest you're taking a terminal box and building something with it - or even instructing a 3rd party to do this - then you've no market need to go through all the compliance hoops. However if it goes horribly wrong and someone gets hurt then you're the one liable.

There is a different approach for one-off products vs products "placed on the market" but the "essential requirements" of the legislation are to be met for both. With certain legislation, e.g. relating to explosive atmospheres, independent notified body testing is always necessary.

gkenyon said:

There is a different approach for one-off products vs products "placed on the market" but the "essential requirements" of the legislation are to be met for both. With certain legislation, e.g. relating to explosive atmospheres, independent notified body testing is always necessary.

Yes the Electrical Equipment (Safety) Regulations 2016 has scope limit for R&D. But the Health And Safety At Work Act 1974 obviously doesn't - so ultimately the former exclusion is moot because you can't really do R&D at LV voltages without observing safe working practices which typically takes you back to the very same standards.

EMC is slightly different because it doesn't typically rise H&S issues - unless it's something very high powered.

And the structure of the regulations in both cases are headline requirements and then presumption of conformity by compliance with harmonised standards. It isn't conceptually that different to how EAWR's require electical installations to be safe and we use BS 7671 as the tool to show that an installation is safe.

Keir Stitt said:

But the Health And Safety At Work Act 1974 obviously doesn't - so ultimately the former exclusion is moot because you can't really do R&D at LV voltages without observing safe working practices which typically takes you back to the very same standards.

Yes, Section 6 of H&S@W etc act, EAWR and PUWER apply.

However, none of them require a product marking or a DofC, and a standards approach is not even discussed.

The provision of a DofC is similarly not required.

Let's use the example of products rather than assemblies. Are you saying that, in a workplace, a mains-powered electronic circuit ought not be energized on a bench in controlled conditions until you've followed the entire product standard? That can't be the case. However, you can carry out the design of the circuit board (if you are having one made) considering appropriate creepage and clearance distances (designers ought to know those anyway), and you can carry out some basic electrical safety tests before you energize.

Back to assemblies, whilst consideration of provisions related to safety (protective bonding circuit, insulation/creepage/clearance distances, etc.) you wouldn't necessarily follow all aspects of the standard? You can't run temperature rise tests without energizing - so if you couldn't run temperature rise tests until you've done temperature rise tests, it's a circular problem?

Keir Stitt said:

EMC is slightly different because it doesn't typically rise H&S issues - unless it's something very high powered.

Wholly disagree based on real-world experience - with modern controls, there are safety aspects to many items of equipment that may be affected by EMC. This is most prevalent in Machinery, or appliances that, if they were in the workplace, might be classed as Machinery, but also these days in many other areas - even switching operations.

One example might be an inverter which contains switching for 'island mode' operation - if the system moves into island mode and doesn't close the 'system referencing relay', that's a potential breach of EAWR (and a safety issue). Another example could be a washing machine ('machinery' in a workplace, BS EN IEC 60204-1, and the relevant parts of BS EN 60335 for domestic use) - these are, today, pretty much all electronically controlled ... if the machine attempts to run up into spin without first emptying sufficient water and distributing the clothes, this can cause a serious mechanical incident, which at the very least will cause the whole machine to physically move, but worse can happen. Of course, you can manage this particular risk using the "no entry when operational" safety cage and procedures to help make sure the machine can't be energized when the machine is operating ...  but that doesn't mean EMC can be ignored, it can be managed differently.

Keir Stitt said:

And the structure of the regulations in both cases are headline requirements and then presumption of conformity by compliance with harmonised standards.

I disagree.

Harmonized standards are used by the legislation that originally implemented the 'Directives' as a means of achieving conformity ('presumption of conformity'). Most are self-declare (although, costs associated with tests often mean at least some type tests are carried out by independent laboratories, which is a good thing). But, even the 'Directives' permit a Technical Construction File Route, and this is applicable to Products Placed on the Market, as well as products that are a 'one off'. This is 100 % necessary, because when you are working with innovation, there may be no product standard ... or at least no harmonized standard.

Basic (no electronics) UK plugs and socket-outlets do not use Harmonized standards either - they use BS 1363-1 and BS 1363-2, which are not Harmonized. These are mandated by the Plugs and Sockets (Safety) Regulations ... but that doesn't prevent them carrying the CE and/or UKCA marks.

H&S@W etc Act, EAWR, and PUWER are different. There is no 'route to conformity'. In general (although there are some absolute duties) products used at work have to be safe, considering the use to which they are put ('fit for purpose'), 'so far as reasonably practicable'. Guidance from the HSE does, however, say that standards can be used to help demonstrate you did what was 'reasonably practicable' ... but for lab test rigs, as above, application of the full standard is not always practicable. There are, additionally British standards (that are not Harmonized) which are generally applicable.

I would further caution that the use of a standard can only demonstrate 'safe SFARP' in a workplace, with no further measures, if the standard considered all the elements of risk pertaining to that particular workplace. In selecting a product for a particular workplace or task, it is up to the duty holder to assess the risks and procure appropriate workplace equipment. Manufacturers do provide products which go "over and above" the requirements of standards to address particular "residual risks" that arise in certain workplaces. In addition, standards may contain options for ratings, durability, or approach to a particular safety-related function, but don't contain "go/no-go" tests for any conceivable aspect of mis-use or fault.

For these particular reasons, I would NOT say that 'conformity with standards' alone demonstrates 'conformity with workplace health & safety legislation for the use of products'.

gkenyon said:

You can't run temperature rise tests without energizing - so if you couldn't run temperature rise tests until you've done temperature rise tests, it's a circular problem?

When doing R&D you need a mind to the risk, so of course you don't know if you meet temperature rise until you do the test - but you do have clear knowledge of where the e-stop and CO2 fire extinguisher are. So of course it would be impossible to fully meet the standard but the risks arising from this are mitigated.

gkenyon said:

Wholly disagree based on real-world experience - with modern controls, there are safety aspects to many items of equipment that may be affected by EMC.

Where products have risks related to EMC then the safety standards will typically have their own EMC requirements over and above those in the relevant EMC directive standard. I've had this before where a product e.g. meets EMC radiated immunity to 10V/m 80MHz-1GHz in accordance with the appropriate standard under EMC directive but the product standard requires 30V/m - which is a bit like comparing the impact protection on a fork lift truck to a formula one car.

gkenyon said:

For these particular reasons, I would NOT say that 'conformity with standards' alone demonstrates 'conformity with workplace health & safety legislation for the use of products'.

100% agree. One thing that i often find problematic is checkbox compliance. But unfortunately the wording of the Electrical Equipment (Safety) Regulations 2016 isn't helpful -  however the presumption of conformity is rebuttable and so it can be demonstrated that notwithstanding compliance with a standard that the overall product does not conform because e.g. it's dangerous for reasons not captured by the standard.

Keir Stitt said:

When doing R&D you need a mind to the risk, so of course you don't know if you meet temperature rise until you do the test - but you do have clear knowledge of where the e-stop and CO2 fire extinguisher are. So of course it would be impossible to fully meet the standard but the risks arising from this are mitigated.

Or, alternatively, remove sources of fuel and restrict human access to the test area when the equipment in energized. The point, though, is that products clearly don't need to conform to a standard to be used in a workplace - alternative approaches are available.

Similarly, a product doesn't need to conform to a harmonized standard to be placed on the market for much of the 'CE marking' related legislation - an alternative assessment of meeting essential requirements is possible for most of the legislation, and documented in a technical file that can be made available to the regulator if requested (and the technical file referenced on the DofC alongside the legislation that conformity is claimed against).

Keir Stitt said:

Where products have risks related to EMC then the safety standards will typically have their own EMC requirements over and above those in the relevant EMC directive standard. I've had this before where a product e.g. meets EMC radiated immunity to 10V/m 80MHz-1GHz in accordance with the appropriate standard under EMC directive but the product standard requires 30V/m - which is a bit like comparing the impact protection on a fork lift truck to a formula one car.

So ... if you haven't tested the product in development you shouldn't use in a development environment? No, you carry out a risk assessment and take suitable precautions in the test environment to prevent the risk, rather than using the standard ... which is where we started the discussion - 'observing safe working practices which typically takes you back to the very same standards' ?

I know this statement was made regards Electrical Equipment (Safety) Regulations, but from a safety perspective, the essential requirements of the Electromagnetic Compatibility Regulations may also be relevant.

Keir Stitt said:

however the presumption of conformity is rebuttable and so it can be demonstrated that notwithstanding compliance with a standard that the overall product does not conform because e.g. it's dangerous for reasons not captured by the standard.

Agreed.

However, I think 'Dangerous for reasons not captured by the standard' is a point that requires further discussion. It doesn't mean that the presumption of conformity is misplaced, or requires "rebuttal' unless the product is specifically designed or sold for use in a particular workplace with peculiar risk. One reason that standards are good for manufacturers is that they can bound the use in the "Scope".

It's not an easy landscape for those responsible for workplace health & safety.

Thank you for coming back on this ... I just thought it worthwhile to draw these points out a little for other (and future) users of the Forum.