Building control: airtightness

The trouble with airtightness

23 September 2011

Rob Coxon discusses efforts to improve standards of testing

If you’ve ever witnessed a full building smoke test, you’ll no doubt have been impressed by the amount of air egress, made visible by the theatrical smoke. Among diagnostic methods for air leakage, the full building smoke test is the crudest but the most visual. I’ve seen modern buildings, virtually completed and approaching handover (save for failing their airtightness test), engulfed in their own conditioned air within minutes.

It’s perhaps not surprising that, after lack of thermal insulation, air leakage is the second most significant cause of energy wastage in the UK’s building stock, most of which has never had to pass an airtightness test. Taking a conservative estimate of 10% being the proportion of all the UK’s space heating load that is attributable to air leakage, I reckon this roughly equates to over 5 million tonnes of CO2 emissions a year from UK housing alone, equivalent to the total annual emissions from well over a million cars. And that’s not including all our non-residential buildings. (European Environment Agency estimates CO2 emissions from space heating in UK households at around 0.03 tonnes of CO2 per m2 floor area per year. There are an estimated 25m dwellings in the UK, with an average floor area of 69m2. Therefore, total CO2 emissions from space heating in UK homes is roughly 25m x 69 x 0.03 = 51.7 mt: 10% of this is 5.15m.)

So air leakage is a real issue. In view of the overwhelming evidence that man-made carbon emissions are driving potentially calamitous climate change, then the curtailment of excess air leakage in new buildings, through the full enforcement of Part L, is surely worthwhile.

There exists a woeful degree of laxity and non-conformance in the airtightness arena, which is undoubtedly resulting in building owners, occupants and the environment being let down. Much of this is the result of the complexity of Part L and confusion over how airtightness testing should be carried out – and by whom. Few in the building industry would dispute the fact that building control surveyors’ work is ever more challenging. This is exemplified by Part L, which, updated to take us a step closer to zero-carbon buildings, remains an unavoidably intricate instrument. Part L compliance is now a hugely important design criterion. It is vital that the new requirement to demonstrate compliance prior to commencement is enforced by building control, otherwise many non-residential buildings in particular will suffer from designed-in noncompliance that could be very costly to rectify.

There exists a woeful degree of laxity and non-conformance in the airtightness arena

Thanks to the new Part L, glass-clad edifices, requiring vast amounts of carbon-intensive HVAC, are perhaps now a thing of the past. Whether all architects or clients fully appreciate this is another matter. Airtightness testing was first introduced into Part L for large buildings in 2002. This compares to other cold climate countries that long before embraced the principle of ‘building tight, ventilating right’, and where new buildings routinely achieve air permeability rates of less than 1m3/(m2 hr)@50Pa. This is at least 10 times more airtight than the minimum standard permitted by Part L, which many of our new buildings still struggle to achieve.

Achieving airtightness is not rocket science. It requires a sensible airbarrier design, good detailing (particularly at interfaces), airtight fenestration, construction quality control and sub-contractor buy-in. Experienced testing companies can be pivotal advisors and, compared to the expense of retrospectively fixing a failing building, the cost of their engagement is modest. In spite of this, all too often builders and contractors opt to go it alone.

Meanwhile, what are the aforementioned problems in enforcement? Part L stipulates that airtightness testers should be competent, and confirms that testers registered by British Institute of Non-Destructive Testing (BINDT) are acceptable.

Here’s the first problem. BINDT has two levels of registration. The first includes all testing companies that are members of the Air-Tightness Testing & Measurement Association (ATTMA), which is effectively a technical committee, affiliated to BINDT, whose members are thus deemed competent to test any building type. These companies are all UKAS-accredited which ensures that systems and procedures are in place to maintain staff competence, equipment calibration, and calculation accuracy.

In addition, ATTMA members engage in technical cooperation and the development of best practice, embodied in ATTMA Technical Standards for testing. There are a few other companies which are UKAS-accredited but not ATTMA members, and thus not registered by BINDT. However, providing their UKAS accreditation is for airtightness testing (rather than some unrelated activity) then they can be assumed to be competent to test any building.

Meanwhile, there’s a second group of BINDT-registered airtightness testers. These are individuals (rather than companies) who have been trained and certified just to test dwellings. Numbering around 200, they can be identified by downloading a list from the BINDT website. To confuse matters, these individuals’ companies are permitted by BINDT to present themselves organisationally as being the de facto BINDT-certified entity. While in the case of sole traders or partnerships there is little harm in this, there are larger companies presenting themselves as BINDT registered when in fact this applies to very few of their employees.

Furthermore, many of these companies (i.e. those not organisationally qualified in airtightness testing), which, incidentally now make up the majority of airtightness testing companies in the UK, are offering airtightness testing for non-residential buildings. Some are delivering commercial building testing via partnerships with ATTMA/UKAS companies, which is fine. However, others are undertaking this work themselves, which is beyond their staff’s proven level of competence.

Then there is a fourth group of testers who have no proof of competence whatsoever, either as individuals or corporate bodies, and who survive, either by making misleading claims on their publicity material, or due to a lack of understanding or concern on the building control surveyor’s part.

Should building control surveyors be concerned about airtightness tests being carried out by unqualified persons? Yes. Reputable testing companies spend small fortunes on, for example, UKAS accreditation, quality systems, staff training and equipment calibration to ensure their tests are accurate. Companies that opt not to do all this have lower costs and thus lower fees, but also lower reliability. If they prevail then the ‘good guys’ will either leave the sector or forced to abandon their standards.

Then there’s the issue of sample testing of representative dwelling types for Part L1a. Problems start with the selection of the plots to be tested. Understandably, building control surveyors tend to ask that the builder’s testing provider should determine the sample testing regime. What typically happens is that the builders will ask competing testers to offer their view of the necessary sample size per house type, along with their unit testing fees, at quotation stage.

Now, unless the tester is provided with an accommodation schedule, site plan, GA plot drawings, M&E specs and the design air-permeability figures for each plot on the site, then they cannot reliably determine the testing regime. Yet I know from experience that most regimes are prescribed by testers armed with little more than a site plan.

Moreover, less scrupulous testers are deliberately underestimating the sample size in order to make their overall quote the cheapest and thus carry favour with the builders. This practice is rife among testing companies, regardless of their status or qualifications. What building control surveyors can do to overcome this is insist that the testing provider carries it out after they’ve been appointed by the builder, and that they do so in the form of a schedule that shows all the relevant information for each plot, such as the envelope area and design airpermeability (taken from the SAP assessments).

Similarly, there’s a problem with ‘envelope area calculations’. These are a vital component of a test result. They are often complex and prone to errors, which if not rectified will corrupt the test result. Reputable companies have CAD technicians well-practised in accurately calculating envelope areas in 3D. This is then sent to the builder for checking by the architects. At the other end of the spectrum are ‘fag-packet calculations’ that go unchallenged. And if an envelope calculation is out by 30%, then so will the test result.

Then there’s the issue of consistency. Housebuilders have ample notice of which plots are to be tested. It’s therefore not surprising that builders freely concede that the mastic sealing work they’ve carried out on site has been limited to only the tested plots. Building control must insist on evidence that all plots have been built and sealed to the same standard. If remedial work is required, then this must be carried out to all similar plots, not just the one(s) that failed the initial test.

Of course, gaining this proof is easier said than done. Recognising this, DCLG has introduced the ‘plus 2 penalty’ to the airtightness testing requirements of Part L1a: the ‘results’ for non-tested plots that have to be used for the as-built SAP calculation have to be lifted by 2m3/(m2 hr)@50Pa. This has caused many builders to declare that they will be having all plots on Part L1a 2010 sites tested, rather than carry the ‘plus 2’ penalty within SAP: obviously good news for the testing fraternity, but more to the point for owners/occupants and the environment. Better still would have been 100% testing mandated in Part L1a.

Building control surveyors will ask ‘shouldn’t the airtightness testing sector be better regulated, rather than placing the onus on us?’. The answer is, of course, yes. One problem is that DCLG will never mandate within Part L precisely who can, or cannot, carry out tests.

Part L itself is a guidance document. This results in an ever-present loophole which basically states that anyone can carry out testing if they can (somehow) satisfy the building control surveyor that they are competent, their kit is fit for purpose and calibrated and their calculation software delivers accurate results.

Nonetheless, I can vouch for the fact that the ATTMA and BINDT are working to improve the quality and verifiability of testing at least among their registered testers. Our ultimate aim is to establish a system of ‘lodgement’ of airtightness test results (similar to the EPC model), which will deliver quality and reliability through disclosure. Building control, SAP/SBEM assessors and clients alike will be able to check on test results and vital data like envelope areas via a central data repository.

Sub-standard testing will be filtered out. Watch this space.

Rob Coxon is a Director of pre-completion testing company, Stroma Technology and Chairman of the ATTMA

Further information

Related competencies include: M009, T005, T044 and T051

See related content in: