Don't blow your air test

In these energy-conscious times, uncontrolled air leakage from buildings is not likely to pass unremarked - and this month, the new Building Regulations Part L will make air leakage testing mandatory for more building types than ever before. But, judging by typical on-site reactions to air tests, the construction industry is only just waking up to this issue.

Much of the responsibility for compliance seems to rest with contractors, but if designers produce drawings without an air barrier clearly delineated, it makes their lives very difficult. On the other hand, if a barrier has been prescribed, but no one on the project team has been given responsibility to maintain its continuity and ensure everyone involved understands its composition, the building will fail. There is no greater tip for getting a building through an air test than this.

A cursory look at Building Sciences' records reveals an astonishing difference in the failure rate between buildings tested "cold", and those that have had the benefit of prior consultation. As a building consultant myself, I'm duty bound to emphasise the benefits of professional consultancy - but the simple fact is that any attention drawn to the air barrier during design and construction is worthwhile, and the earlier and more regularly it is reviewed the better. The last thing test teams want to see when they turn up on site is a foam gang desperately trying to assemble an air barrier, as an afterthought, yet unfortunately, that is more the rule than the exception.

This a shame, not least because it's actually a harder task to fill a myriad of small holes at a late stage than spend a little more time ensuring air barrier continuity throughout the programme of works. A multitude of small holes in the internal finishes may only feed one exit hole and this doesn't shrink as the internal gaps are plugged. The flow of escaping air will be diverted and may be slightly impeded, but it's misguided to think that adding up the size of internally sealed gaps will produce a corresponding improvement in the leakage area of the building as a whole - it's like sealing every inch of a lift riser instead of capping it off at its head.

About the test

To better understand how to pass a test, it helps to know something of the procedure. It's also worth bearing in mind that procedural differences and inconsistencies are best avoided by employing an accredited testing house. Look out for members of the ATTMA (Air Tightness Testing and Measure-ment Association), who must be UKAS accredited, to ensure a high level of quality and consistency of application.

The current practice is to use fans to blow air into an enclosure and measure the flow required to hold that pressure steady. It doesn't take a genius to work out that the more flow needed to maintain pressure, the less airtight the building.

During the test, which may last up to half an hour, all internal doors must be open and all external apertures kept closed.

Once test pressure is achieved, a series of readings is taken as the pressure is dropped. These are used to generate points on a graph so that a single reading can be cross referenced with others, ensuring greater data integrity. A point that is "off" will not sit comfortably on the generated curve, and the tester will normally retake that flow measurement.

Using single point extrapolation, it is possible to get an almost instant idea of leakage as soon as the first reading is taken, but only after several readings can a truly representative trend line be produced and the leakage rate derived. The single point result gives the test team a ball-park figure for the likely outcome of the test, but if it's particularly inconsistent with the visible quality of construction it may just indicate that a door has been opened after "lock-down" and the start of the test. It's this (and site workers standing on pressure hoses) that give pressure testers the most headaches.

Doing the sums

If things go well when the fans are turned on, standard test pressure - currently 50 Pa, equivalent to an increase of 0.05% over normal atmospheric pressure - will be reached in no more than a minute or so.

Total air flow into the building is then divided by the total envelope area to provide a leakage rate in cubic metres per hour per square metre of envelope area, at 50 Pa (m3/h/m2@50 Pa).

The basic equation used to work out air leakage rates has three principle values:

• difference of pressure, inside to out

• volume of airflow from the fans, and

• An area of leakage.

As with all sound equations, it is possible to calculate any one of these values by knowing the others - and this is why so much care is taken calibrating equipment and refining procedures to accurately measure the pressure difference and airflow volume.

But nothing in life is completely straightforward and air density and pressure equalisation - both of which have a direct effect on pressure differentials and fan flows - change under the combined influences of atmospheric conditions and building design (consider stack effects in tall buildings). So the primary data must pass through a number of checks and corrections before the final answer is given.

Pass or fail?

A multitude of small holes in the internal finishes may only feed one exit hole – and this won’t shrink as the internal gaps are plugged

An air leakage rate at or under 10 m3/h/m2@50 Pa currently constitutes a pass in the UK; although the levels for some enclosures are considerably lower.

Gas-protected enclosures, such as those that house mission critical servers, need to be as airtight as possible for maximum fire-suppressing efficiency. Archives and cold stores, meanwhile, need more stable and controllable climatic conditions than ordinary storage areas, and hospital care units often have stringent air quality requirements. All of these are affected by air leakage.

A pass is usually accompanied by some very British hand clasping and pledges to buy drinks. But not every building passes, of course - and in those cases where buildings fail, the consultant should be there to help with leak identification using smoke while the fans keep the building pressurised. It is worth remembering that smoke is best used as a diagnostic tool, it doesn't prove any form of compliance in itself.

Getting up close and personal with a smoke puffer is one of the best ways to find out where air is going; the rate at which plumes are sucked away will help an experienced consultant to grade leaks with some accuracy. Larger capacity smoke generators may be useful to get smoke into places that are difficult to reach, but they are best employed as an attention grabber. It is one thing to tell a client there's a leak into a cavity and another to fill it with smoke and watch it pouring out of the eaves.

During the identification of leakage paths, the consultant's experience and the constructor's inside knowledge come together, like partners in a good detective drama. An effective partnership will usually detect the area(s) of concern in a matter of minutes, so make the most of your consultant, get someone who knows the building inside out to accompany them, and hopefully one re-test is all you'll need to pass.

How to avoid the common mistakes

So, what are the main reasons for failing? A straw poll conducted among Building Sciences' senior consultants produced the list below - but first, two pearls of wisdom discovered in the course of researching this article: 1) be suspicious about anything in the structure that is, quite simply, hard to get to and 2) watch out for any area marked up as ‘by others'. As one site manager said: "if you're not doing it yourself, you can assume it's not being done."

The main reasons for failing an air test are:

• Failure to seal the inner leaf of cavity walls, where the cavity is ventilated

• No adequate seal between brick/blockwork and steel frames

• Failure to properly collar services between conditioned and unconditioned spaces

• Unsealed, or badly sealed, vapour control layer laps; particularly around the eaves

• Poorly engineered interfaces between different materials generally

• Failure to effectively shut off/ cap ventilation systems prior to testing.

Finally, it is worth noting that although the current requirement for maximum air leakage is

10 m3/h/m2@50Pa, in order to comply with the government's carbon emissions target the Building Regulations may soon require a lower permeability. Current levels in some Scandinavian countries are lower than 3 m3/h/m2@50 Pa - so if 10 m3/h/m2@50 Pa is considered hard to achieve, as it clearly still is in the UK, well, something will have to change. 

Willem Postema is a consultant at Building Sciences.

If you've missed any of the articles in BSj's series on airtightness, they can all be found at www.bsjonline.com