Comments on draft BSI standard for vaping products

I provided some comments on the draft British Standards PAS 54115, Manufacture, importation, testing, and labelling of vaping products (closing 30 Nov). […]

Make your views known: if you aren’t at the table, you’re probably on the menu

I provided some comments on the draft British Standards PAS 54115, Manufacture, importation, testing, and labelling of vaping products (closing 30 Nov).  A PAS is an industry-led Publicly Available Standard, brokered but not imposed by the British Standards Institute, and designed to quickly set norms in an industry.  I’m broadly pretty positive about this, not least as it could give a basis for more rational regulation if the EU directive Article 20 is annulled in court, which it should be.  However, I think the current draft PAS leans towards costly overkill in some areas – we need to keep the likely risks in proportion and remember the damaging and distorting effects of subjecting these products to greater testing burdens than cigarettes.  There are many worthwhile proposals, but the excesses have been my focus in comments.

What I hope this does is open a conversation – that would be the one that was never had while the EU directive was being drafted.

Overall comment

Comment: The PAS should not attempt to incorporate the requirements of the TPD – many of which have no evidence base and are counterproductive for both health and the functioning of the single market, and may be overturned following challenge in the Court of Justice of the European Union.  The PAS should only set standards that a proportionate and likely to create material benefits for users that justify the costs.  The PAS as drafted does meet tests of proportionality – yet it would impose high and disruptive costs and tend to favour the large companies in the market at the expense of SMEs. 

The PAS does not adopt a proportionate risk based approach to regulation. The big health impacts arise from people switching from smoking to vaping products, not by marginal improvements in vaping products.  To the extent that the latter compromises the former, the PAS will be harming health.  This could arise if the effect of the PAS was to drive good products and small firms out of the market simply because of the weight of compliance. The process of PAS development should include an impact assessment. If firms are going to be put at risk, then there needs to be a good reason. 

The PAS has a tendency to demand excessive testing and toxicological risk assessment, when the tests themselves would say little useful about the product.  Rather than trying to measure toxic emissions, it would be better and cheaper to address the causes of toxicity ‘upstream’.

Remedy. 1. Embrace the concept of proportionality – this is vital if these products are not going to be subject to regulatory ‘barriers to entry’ that distort competition with tobacco products (that do not require any of this testing) and favour large companies within the category. 

2. Conduct a limited impact assessment before finalising the PAS and be more aware of what sort of business models are at risk if the PAS becomes a de facto standard. 

3. Focus on tackling precursors to health and safety risk rather than make arbitrary measures of vapour. 

– Specify liquids to pharma standard and flavours to food standard;

– Require devices to be capable of use in a food processing environment and specify materials that can or cannot be in contact with liquids;

– place constraints on the operating regime – or provide warnings about running too hot or too dry.  It is these abnormal operating conditions that cause most toxic emissions – it would be better to address these directly, than indirectly through testing vapour. 

Comments on some specific sections

4.6.5  Toxicological Risk Assessment for flavourings

All flavourings including natural extracts should be additionally subject to a toxicological risk assessment (TRA)

Comment: A toxicological risk assessment for all flavourings is hugely excessive and burdensome, and will involve large scale duplication and across the industry.  This would potentially close down most of the flavours industry and may have harmful effects of removing low-volume niche flavours from the market.This is an area where the PAS needs to be subject to a proportionality test and impact assessment. 

Remedy: Toxicological risk assessment should be considered strictly on a risk basis – where it is justified by some form of novelty or anticipated risk. Otherwise a simple chemical analysis should suffice. Some sort of listing arrangement could be used to validate flavours across the industry as a whole – it is essential to avoid large and unnecessary compliance burdens, especially given the structure of this industry

4.8 Toxicological Risk Assessment for ingredients

A TRA should be carried out by a competent toxicology specialist taking into account every ingredient used in the manufacture of e-liquids.

Comment: What is the purpose of an expensive TRA if there is already a requirement that the flavours are food grade and the nicotine and diluents are pharmaceutical grade? If none of the ingredients are toxic, what is there for a toxicological risk assessment to find? Leachables should be addressed by specifying device standards – for example excluding toxic metals.  Thermal breakdown products are an artefact of the operating regime and become problematic when devices are operated in unrealistic conditions.  Again, this has been proposed without a sense of proportionality or estimation of the regulatory impact. It is important to consider what effect this would have on the businesses in the market if it became a de facto standard, rather than optional.  

Remedy The ideal proportionate approach should focus on setting standards that avoid or control the precursors to the formation of toxins, not on measuring toxins in the emissions. This means specifying pharmaceutical grade nicotine and diluents, food grade flavours and standards for devices equivalent to those used in food processing.

4.10.5  Emissions testing for liquids

E-liquids should be tested for the production of inhalation toxicants produced during the vapourization process.

Comment. This is exceedingly burdensome for no obvious return in health and safety terms. The formation of toxins in the vapour depends on what is in the initial liquids, the manufacturing specification of the device and – most importantly – the operating regime.  These issues can be more pragmatically addressed ‘upstream’ by specifying standards.  This is an area where proportionality is necessary and an impact assessment is required to judge the damage it would do to small firms if this becomes a de facto standard. 

Remedy. There is no reason to have extensive emissions testing, which will produce largely arbitrary results at great cost. The real focus should be on addressing the precursors to the formation for toxins in the emissions – standards for liquids, stability of device surfaces in contact with liquids and, if necessary, by limiting the operating parameters to prevent especially hot or dry use – or by warning that the products should not be used in that way. 

5.5 Toxicological risk assessment (TRA) of emissions for devices

Comment: An excessive and disproportionate way of addressing risks arising from the device.  Some test of proportionality is necessary here and a proper impact assessment done to understand the damage that such a burdensome regime might to do SMEs in the sector. It is difficult to see what this would achieve that could not be more directly realised by setting standards for device construction.

Remedy: It would be better to address risks from devices ‘upstream’ by setting standards for the materials and construction rather than testing emissions. Why not prohibit toxic metals from construction and have a white list for approved materials. A good analogy might be to ensure that it would be safe to use in the food industry.

5.7.2  Batteries

Batteries and chargers should be further tested to ensure they can function safely during use (see Annex G for test methods).

Comment. Batteries and chargers are now pervasive thanks to mobile technologies.  Surely this issue should be addressed generically rather than having a specific testing regime in the PAS. If practical there would be valuable consumer benefit in being able to interchange chargers with mobile phones.  Given there is a process to standardise mobile phone chargers, it would be sensible where possible to align with that.  

Remedy. It would be better simply to refer to compliance with BS EN 62133:2013 on secondary cells and batteries containing alkaline or other non-acid electrolytes. Safety requirements for portable sealed secondary cells, and for batteries made from them, for use in portable applications and the Low Voltage Directive 2006/95/EC. Unless there is some insurmountable technical reason that prevents it, the PAS should align with emerging standards for a common mobile phone charger – if necessary as an option.

Comment on contribution by the Association of Directors of Public Health 

Comment. The Association of Directors of Public Health has provided unsubstantiated opinions on unrelated issues that have no bearing on the PAS.

Remedy. Delete or ignore contribution from ADPH.

Download Post as PDF

25 thoughts on “Comments on draft BSI standard for vaping products”

  1. I can perfectly see the reasoning for your comments and remedies.

    While it does look like most of the draft regulations/guidelines in the PAS are aimed at compliance with the TPD I personally dont believe that its coming from a H&S point of view which the TPD does. I think this the PAS is all about quality which you have also addressed in the batteries comment.

    I’m rather pleased that the industry has brought this forward. Especially the TRA aspect, even if its a bit “late” as Dr Farsalinos has been doing a lot of work similar but through different avenues to my comprehension of the PAS.

    Yes, a lot of the ingredient testing etc will be costly but when it has been done once on each product formulae then it may not need to be tested again unless the formulae changes or a raw material formulae changes. I that this may reduce long term costs regarding testing especially as the draft suggests a time frame for testing and TRA’s that can surely be affordable and practicable for SMEs. Also, the testing regime builds upon what some companies are doing already to confirm that raw materials especially flavourings are free of nasty chemicals.

    Yes, “upstream” testing can be done but a large proportion of flavourings are food grade and meant for ingestion so the “upstream” suppliers would not see the benefit of carrying out the inhalation tests. Also, “upstream” testing will not reflect the final e-liquid matrix.

    I am expecting more common sense from you Clive to counter my opinions.

    1. This points to the idea of testing things once and making decision about whether to allow them if there are grounds for concern, not having everyone testing everything repeatedly. It is surely easier to make one ‘upstream’ decision on Diacetyl than to have everyone try to measure it?

      As it happens I think the risks are grossly overstated – given just how large the exposure was in ‘popcorn lung’ victims, but given the controversy, it may be better reputationally for the category if the issue was dealt with by banning it – if there are reasonable alternatives.

      I think there is a good case for being aware of the risks, but large amounts of repetitive routine testing wont achieve that.

      1. Clive,
        I totally agree – this is common sense, it seems to me.

        Btw, I have lodged a complaint with IPSO about the recent article in the Daily Mail/ Mail Online, which reported results released by the Japanese Health Ministry, purporting to quote a study that found that there were 10 times the levels of carcinogens in e-cigarettes as in tobacco cigarettes. This is NOT what the study found, hence is inaccurate, misleading and, therefore, potentially dangerous to health. I’d welcome anyone else making a similar complaint about this article, to add pressure to get a correction/ retraction published by the Mail Online. Here is a link to Dr Farsolinos’ discussion of this matter, if anyone is interested in supporting my complaint:

        1. Totally agree the way that the Japanese study has been ‘reported’ is scandalous.

          While it may well involve a ‘translation’ problem , on the face of it, the story seems very ‘strange’ what kind of self respecting scientist would announce/publish a research paper, and then in the same breath, say to the press :’BTW I have a unpublished result, that radically contradicts the research paper that I have just published’

        2. In an initial response to my complaint (and those of other people?), The Mail has now changed the headline of the offending article to “High Levels of Toxins found in some e-cigarettes: Japan research” and added a footnote to say that “An earlier version of this article stated that Japanese research had found that e-cigarettes contain 10 times more carcinogens than tobacco cigarettes. While the research did find high levels of formaldehyde in one brand of electronic cigarette, we would like to clarify that not all e-cigarettes contain 10 times more carcinogens than tobacco cigarettes. The study supports existing evidence that e-cigarettes are much less dangerous than tobacco products.”
          A classic example of far too little, much too late. So I’m still pursuing this with the Independent Press Complaints Commission. Anyone wanna join me?

      2. It doesn’t take a great deal of time or testing to “blacklist” compounds that are known to cause problems. With such a blacklist in place – and being added to, subtracted from – the “state of the art” is easily achieved and maintained. It should not require repeated testing of individual juices, but would require adherence to the standards and blacklist. Spot checks by an authorising body would then confirm compliance. Much simpler and a lot less expensive/time consuming.

  2. I am not in any way an expert here but I see what could be another argument in the section about toxicology 4.8.
    When we use food quality ingredients which are passed as safe for food production do those food safety tests include the resultant chemicals that are created when such ingredients are combined and cooked together? Does any public health standard test this?
    If not (as I suspect because if they did every chef’s special in the UK would need separate testing) why should e-cig vapour output be a special case and be tested when the safe food quality ingredients are combined and cooked?

    1. Woodsy,
      I’m no expert but I have read Dr Farsolinos blog and am therefore aware that the presence of oil in any flavourings added to the PG or VG exposes the vaper to some risk of getting lipid pneumonia.

      Because both oil and base are safe to eat (together or separately) does not mean that they are safe to vape. There may be other risks in vaping food-grade and/or pharmaceutical grade ingredients, alone or in combination, of which I am not aware.

      There is therefore – in my opinion – a need for some research into such things and some dissemination of any necessary warnings to both e-liquid manufacturers and to vapers.

  3. Clive and Woodsy, regarding 4.8, food safe does not necessarily equate to inhalation safe so this is important. Take for example Diacetyl which has recently been a bit of a scare in the vaping community and highlighted by the BBC. Diacetyl which ives a buttery/creamy flavour is (apparently) safe as a food flavouring but not so when inhaled above certain concentrations. It turns out that although present in many liquids it was mostly below the levels found in cigarettes so could still qualify as a reduced harm product, but as it can be screened for and removed from the recipe, there’s no need for it. And as Irish Lass pointed out, oils which are safe for ingestion but not inhalation. There are many unknowns when dealing with inhalation of flavouring chemicals that need to kept an eye on and tested for if they are in the future found to be unsafe- in the interest of maximising harm reduction.

  4. On the issue of testing, I don’t agree that there is no need to test emissions.
    One of the most useful works done so far on the emissions of carbonyls from e-cigs ( showed not only a difference based on the base liquid composition, but also a significant difference between flavoured and unflavoured products. This work does have some significant issues, one of the most annoying of which is that it did not analyse the liquids before the testing of emissions; it also had quite significant variation in measurements. It is, however, strongly suggestive that flavouring may have an impact on the emissions of a liquid.

    We can also expect that different designs of atomiser will have different safe working power ranges, based on their ability to wick the e-liquid, which it is entirely valid to inform consumers about. Without testing the emissions how can this range be identified?

    Requiring a method to prevent dry or overly hot use would require all devices to be electronic, instantly removing all mech mods from the market. The only working implementation of this technology of which we are aware is also subject to patent, so this would effectively become a monopoly market.

    If it were not for the time pressure imposed by the implementation of the TPD (and, since most of the work on this will have been done before the TW court case is decided, we can expect some form of national legislation, even if it is no longer mandated by the TPD) it would be much, much better to wait for more basic science to better inform the process. Sadly, if there is to be any kind of standard in place to inform implementation (both at the national and European level) this time is not available.
    This leaves us the unpalatable choice of doing nothing, or doing something based on limited knowledge.

    In a more general sense, it is already acknowledged that diacetyl and acetyl propionyl should not be used in e-liquids, yet testing done by Dr. Farsalinos ( that they were present in a significant number of products. Setting a limit is therefore largely meaningless without testing to demonstrate that it is, in fact, absent.

    The PAS is likely to generate large amounts of data which will fill the gaps in the underlying knowledge of the effects of flavourings (and other ingredients) on the emissions, on a scale which would not be possible in a research setting. This is not an optimal solution by any means, but will at least have the effect of beheading the zombie argument that ‘We don’t know what is in them’. It also places the burden of generating this data on the industry, as opposed to the current laudable (but largely unsuccessful) attempt to conduct this research by voluntary contributions from users and industry.

    The testing for batteries is not overly complicated and does not require much specialised equipment, and should be an easily accessible way for manufactures to confirm that they have complied with the design requirements of BS EN 62133:2013, and to ensure that the current ongoing reputational damage to the industry due to battery failures is minimised. The standards for mobile phones are almost certain to be aimed at devices with greater unit cost than ecigs, so it is questionable if the standards will be readily transferable.

    1. Tom
      re ” working implementation of this technology of which we are aware is also subject to patent”, do you have details?
      It sounds a bit unlikely that, electronic controls over battery power/discharge rates, protections against short circuits, et cetera , could be subject to ‘A’ patent, surely this is fairly generic stuff ?- Am thinking of all the lithium manganese, battery powered drones and the heaps of other similar toys that are around these days.

  5. I was referring to the specific example of detecting dry/over heating coils. The only device (of which I am aware) which has a means of detecting if the coil is overheating (due to either setting too high or dry wick)is the new DNA chipset ( Given that it really IS a groundbreaking technology, it is not surprising that Evolv are keen to keep a tight grip on it.

    1. Isn’t the device that detects dry or overheating coils the human user? Isn’t the need to build this into the device really a need created by laboratory testing rather than actual human use?

      1. I agree, which is why I don’t think that “place constraints on the operating regime” is a useful direction of travel. the safe operating range of a an atomiser will be dependent on it’s design, not the power supply (be it a mech mod, or VW device) that is connected to.

        1. But what is safe on an electronic power controlled and limited device – like, say, an eGo Twist or Vision Spinner – may be anything but on a mech with stacked batteries, or one of the 200 watt mods now available. And, vice versa. What a knowledgeable and experienced vaper can achieve with sub-ohm coils on a carefully constructed setup could conceivably cause catastrophic malfunction on a Twist or Spinner. Or clones thereof (which is probably more likely).
          The notion that no “top” could either overheat juice or draw too much current from a battery setup is an attractive one, from a naïve consumer’s viewpoint, I’d suggest.

      2. On dry hits and temp sensing. The actual user as a tyro or naïve user may not have a benchmark to understand the burning of wick when no juice is present. The fact that it has taken so long for researchers to even begin to understand it bears testimony – and they are learned people with, one assumes, the nous to work it out.

        If a given setup is incapable of producing high enough temperatures to produce the nasty compounds, no matter how much juice is available at the coil, then at least one variable in the equation is rendered null and void. Beginning switchers no longer need to know what a “dry hit” is – they do not occur. With temp sensing, the instruction “when vapour starts to tail off, replenish the liquid supply” actually is accurate. At the moment, it isn’t – a dry hit can produce a lot of vapour, but it’s horrid.

        1. Dave – isn’t this a bit like learning to ride a bike – a few grazed knees might be expected, but no lasting damage, and then you’ve learnt to control the temperature and avoid dry hits? Things like formaldehyde become risky when inhaled over years / decades – the odd dry hit isn’t going to do lasting damage while the user learns feedback.

          Are you saying you think that some sort of temperature control or other control over the operating regime should be built in to pre-empt user feedback responses? I gather it is possible to do this… but should that be a product feature or a mandatory requirement? I suggested this once and a large loud Greek exploded! But I’m less convinced it is necessary now, having understood the dry hit thing a bit better.

          Grateful for your views.

    2. Tom
      Hi The stuff on the Evolv website looks more like parts/components for a complete system, rather than a end user product- are these DNA chipsets available in ‘complete packages’ yet?

        1. realise that you do not want to do ‘adds’ but can you give a idea of the right search terms

  6. There is also VIR being developed by Imeo – the Golden Greek which is due for release very shortly.
    Also stories of Chinese clones of the EVOLV chip, as yet unconfirmed.
    However, as with not smoking the filter, users learn quickly not to let wicks dry out and can quickly learn to
    Spot the clues that more juice is needed to keep a device in optimal range.
    I am a big fan of temp regulation, but I’m wary of pushing it forward as a solution to a speculation, rather than a
    Measured problem. To my way of thinking a system for improving juices is far more important as we move forward.

  7. Can’t get a reply button on your 1:23 post, Clive.

    TO me, the temp control thing is a bit like seatbelts in cars and crash hats on bikes. Neither are necessary as long as everyone knows how to drive or ride properly, and yet we have them. I’ve been vaping now for well over five years and had never had any success with dripping at all – an endless chain of dry hits or soaked coils, the latter of which only disappoints, whilst the former just bloody hurts. And I’m not naïve or dense. With temp sensing, I’m enjoying dripping – no dry hits as such, and no scalded tongue from overheated juice. More to the point, I’m getting a much more pleasant experience. I have the feeling that a lot of folks have been put off ecigs because the experience has not been good – and because they have been propagandised by formaldehydegate and diacetylgate and so on. This technology is reassuring, at the least, and I do know that DH is more than aware of it.
    *I’m* not saying it should be mandated – any more than I think either seatbelts or crash hats should have been – but I have the feeling that someone will…

Leave a Comment

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Scroll to Top