I was talking to an industry professional recently. When asked “Why do we measure airflow?” his response was “Because the guidance tells us to”.
For me that said it all. We know it’s important, but not any of the fundamentals that underpin it. Airflow is really rather important, and for very sensible reasons. And the guidance tells us to test. But these are two very different things.
My approach to asbestos regulations, the Approved Code of Practice (ACoP) – and all guidance – is always to ask “Why are we doing this, and why is it important?”
Only when you understand this can you make the judgement calls that the real world forces on you, when reality doesn’t quite fit what someone wrote 12 years ago, hundreds of miles away.
Let’s start with the fundamentals.
Disturbing asbestos is very dangerous because inhaling fibres has a disease risk where there is no safe limit. Removal of asbestos inevitably liberates these fibres into the atmosphere where they can be breathed in.
We carefully design methods so that we minimise how many fibres are released. We build enclosures so that any fibres that are released are contained, and can only affect a small area. We apply negative pressure to the enclosure so that if there are leaks, then clean external air leaks in – rather than contaminated air leaking out. This is achieved by drawing air out of the enclosure using Negative Pressure Units (NPUs), and scrubbing it through HEPA filters.
So, we have created a hermetically sealed bubble that we are asking workers to enter to do the work, potentially exposing them to fibres. The methods should reduce the fibre release, and the workers have respiratory protective equipment (RPE). But as there is no safe limit, and respirators are the last line of defence, we have to do more.
Helpfully, those NPUs we were using to apply negative pressure have another very useful trait. They remove contaminated air, and replace it with clean air. If the enclosure is carefully designed there will be a constant flow of clean air to all areas, so asbestos fibres are drawn away from the worker and there are no stale air dead spots. The contaminated air is being constantly diluted.
So, airflow is critical, as it underpins all of the controls:
How do we ensure we get it right?
Paragraph 395 of the ACoP states:
Before starting work in the enclosure, a thorough visual inspection and smoke test must be conducted to check the enclosure’s integrity. The filtered air extraction equipment must be tested to ensure it is achieving negative pressure and the required air change rate.
There’s a need for some interpretation first. ‘Required air changes’ is a rough measure of the dilution I mentioned above. In the UK this is eight ACH – that means you need sufficient NPU power to replace all of the air, eight times every 60 minutes.
It doesn’t quite work that way – but that’s how it is sold. Think homeopathy and that’s how much it lines up with reality. The industry standard, though, is at least 10ACH, and often many more.
The smoke tests mentioned in the ACoP are fairly straightforward: fill the enclosure with smoke, and check outside to see if any of it leaks. If not, we have a sealed enclosure.
Now the last and most crucial bit: the NPU should be tested to ensure it is achieving negative pressure, and that it is achieving those air changes.
This word ‘tested’ has been taken to mean that we should actually measure how much the NPUs are ‘sucking’. Once we have established definitively that the NPUs are pulling at least eight times the enclosure volume, then all questions are answered.
Some HSE inspectors are even insisting that NPUs MUST be tested using an anemometer or an onboard airflow meter. An anemometer is a handheld device that is essentially a fan with an electronic read-out. Place the fan near the NPU head and it will tell you how fast the air is passing through it. Get several readings, take an average, and multiply by the area of the pre-filter and we have an answer.
Well not really, because:
And while it sounds easy, even measuring the area of the pre-filter is near impossible. The guidance says you should measure height by width, but the reality is that those pesky cross hatches will reduce it somewhat. If you look closer, you will see there are vertical veins – the matting actually increases the surface area dramatically. The veins also create turbulence, causing further uncertainty.
Image – Beacon International.
In any case, all NPUs are tested by trained professionals at the service company every six months, using equipment that is kept in lab conditions. Are we really saying that a supervisor on site with a one-year-old anemometer from the van is better than that?
I have seen data where two readings were taken on site by two different people on a single NPU (moments apart). One set differed by 14%, and a second pair differed by a whopping 30%. Even the guidance provided by the HSE says that handheld anemometers typically overestimate by 10%. If we’re measuring a critical value, overestimating is not something that we should be doing.
I’m not a fan, as you can probably tell.
Onboard flow meters are better, as they measure internally from a static position. But just like the hand versions, they are subject to being knocked and damaged by the rough and tumble of life on site.
Even without these concerns, does this testing guarantee we’re achieving what we want?
Not necessarily. You may have ample airflow from your NPUs, but if there is air getting in somewhere else (a leak in the enclosure, big or small) you won’t have adequate negative pressure. Similarly, while we may be drawing out and cleaning lots of air, a poorly-designed enclosure will leave pockets of still-air that’s not cleaned at all.
By testing a feature of good enclosure design, rather than the outcome we’re trying to ensure, we may have gone down an unnecessary rabbit hole. And a warren that is costly, time consuming and potentially very misleading.
Let’s look again at what we are trying to achieve:
Impose 5Pa of negative pressure
Good airflow is essential for negative pressure, but in isolation it tells us nothing about whether we have achieved it, so how can we test for it?
There is a ready reckoner for 5Pa of negative pressure. If you balance the enclosure (air in and air out is roughly equal), and the flap deflection on your main air lock is around 250mm, you will have achieved approximately 5pa. There are too many vague terms in this for my liking, but as it stems from an HSE research paper, it does have some authority.
You can however measure negative pressure easily, by using a magnehelic differential pressure gauge (often known by the company that makes a lot of them: a ‘Dwyer Gauge’). These are inexpensive, very simple devices, yet they accurately state what the pressure is. Simply tape one to the side of the enclosure and it will compare pressure inside and out. And what’s more this doesn’t leave you with a single reading taken at the time of the smoke test – it gives you constant monitoring.
No dead spots
Remember we need to ensure there are no dead-spots, and therefore no highly contaminated pockets of air. Simply measuring how impressive the airflow is at the NPU tells us nothing about how successfully we are cleaning and diluting the air. On the other hand, a smoke test, done well, tells us quite a bit about this. You can actively see the smoke moving and can redesign the enclosure to make sure there are no dead spaces.
Eight air changes per hour
We must also achieve eight ACH. Dilution is a complex mathematical problem; just having eight times the enclosure volume in NPU power in no way predicts that we’ll reach the necessary figure. It also doesn’t mean we’ll be refreshing the enclosure air once every 7.5 minutes – dilution just doesn’t work like that.
Eight ACH means removing 13.33% of the air and replacing it with clean air every minute. If we assume that there is perfect mixing within the enclosure (very difficult to achieve), some of that air will be dirty and some will be clean. Very roughly, with some rounding, we get a ratio of dirty to clean air pattern that looks like this:
| Elapsed time (minutes) | ‘Dirty’ air (%) | ‘Clean’ air (%) |
| 1 | 86.7 | 13.3 |
| 2 | 75.1 | 24.9 |
| 3 | 65.1 | 34.9 |
| 7.5 | 34.3 | 65.7 |
These figures show that after 7.5 minutes, the first of your supposed ‘air changes’, you are only down to about a third of the original ‘dirty’ air. In fact, only after about 30 minutes – or four ‘air changes’ – are you anywhere close to getting rid of it.
So let’s go back to the smoke test. When you can no longer see any smoke, you’re getting an indication that you’re heading towards a complete change of the air. But again, you’ll never actually get there, because that’s not how dilution works.
Nearly all contractors use a minimum of 10 ACH, and often quite a bit more. Is it tremendously important that you are achieving 8.2, 9.7, 10.4 or 15.1 ACH per hour? I would argue that it’s not, really. In fact my position is firmly that nominal ACH is not a useful test of the effectiveness of our controls at all.
Having NPUs that are capable of drawing air to the tune of about 10x the total enclosure volume is the means by which we achieve all that we want. But the only way to test we’ve actually met the requirements in a balanced enclosure is:
Why am I going over this? Well, it’s all too easy once we have guidance, to focus on meeting the letter of that guidance. But as I said at the start, it’s important to ask why we’re doing the tests we do, and what we’re actually trying to achieve.
I’m hoping to stir up a discussion about the fundamentals of enclosure airflow. You might agree or disagree, but please get in touch to tell me why, so we can all learn and get better at what we do.
Tweet Share on Facebook Share on Linkedin