Dwelling airtightness in Ireland: where we are, and where we're going

By Gavin O Se, NSAI certified airtightness tester, certified passive house designer and BER assessor with Greenbuild

Recent issues of Construct Ireland have featured houses that have been to the very best international standards of airtightness: eg Jer Rynhart's house in Wicklow and Tim O'Donovan's house in Cork, both of which had just a fraction of an airchange under standardised test conditions.

At the same time as these super-airtight houses are being built, the latest draft Part L of the Building Regulations is proposing to lower the air permeability rate for new dwellings from its present level of 10 m³/(hr.m²) to 7 m³/(hr.m²).

I was quite disappointed ― though unsurprised ― to learn of the new maximum airtightness level. It set in process a train of thoughts, the result of which is this article, in which I will look at:

    •    Where are we in terms of airtightness and airtightness testing?
    •    Where are we going?
    •    Where do we stand internationally?

Where we are

Airtightness in dwellings is important from a number of points of view. Amongst other advantages, better airtightness can:

• reduce heating energy requirements in the dwelling

• improve air quality in the dwelling, if achieved in conjunction with a properly designed and implemented ventilation system.

• improve sound insulation – both from noises entering the house, and if your kids practise their rock band sounds inside, from exiting the house.

• Depending on the airtightness measures used and the overall construction build-up, an expertly installed airtightness layer can also act as a vapour barrier/check to reduce the risk of interstitial condensation. Specifically designed breathing walls that allow the passage of water vapour do not use the airtightness layer as a vapour barrier/check.

So far in this country, we have looked at airtightness generally only from the point of view of the first of the these items – ie the energy benefits of an airtight home, and thus the airtightness requirements have been written into Part L (conservation of fuel and energy) of the building regulations.

The first requirements for homes to be tested for airtightness came in the 2007 revision of Part L. On the face of it, the introduction of a maximum allowable airtightness level was a major advance, and even more so because an independent test must be conducted on the quality of the build.

Unfortunately, as with the BER regulations, it will be found that many properties have been completed and sold or occupied without the test being undertaken, where the test should have been. However, this is not a report on the lack of building control in this country. The standard introduced was a q50 of 10 m³/(hr.m²). (Not sure what q50 means? See 'q50 & n50' at the bottom of this post)

How difficult has that standard been to reach? Not difficult at all.

Shortly after a similar standard was introduced in the UK in 2006, a large airtightness testing company (BSRIA) issued a summary of over 300 tests they had carried out coming up to and shortly after the regulations came into force (1). The findings were that overall, 83% of dwellings tested complied with the regulations.

To determine what the standard of dwelling airtightness is in Ireland for new buildings, we have reviewed many of our own test results and requested test results from a number of other testers around the country. The resulting dataset, when trimmed to ensure like-with-like comparisons, is about 120 tests, of which 56 are from new buildings built and tested between 2008 and 2010. The average q50 for these tests is a tad under 5 m³/(hr.m²), with just two of the 56 having a q50 of more than the regulation maximum 10 m³/(hr.m²). Fifteen of the 56 dwellings would fail to meet the proposed new q50 of 7 m³/(hr.m²).

However, there is possibly a skew here, as many of the tests performed in recent years are of houses going for low energy and passive standards, where the clients understand the need for airtightness. Also, the standards they are hoping to achieve, such as the passive house standard, demand an air pressure test with more vigour than the requirements of the Irish building regulations.

Looking at the bigger picture of the entire dataset, which included tests for people who wished to deal with cold drafts in their houses, as well as those seeking to comply with building regulations, the best result was a q50 of 0.32 m³/(hr.m²), the worst a q50 of just over 23 m³/(hr.m²).

The average of all dwellings built before 2000 was a q50 of marginally above 10.50 m³/(hr.m²) (21 dwellings tested). Most interesting, of the 21 houses in that group, 13 were tested following more recent renovations with an average q50 of 10.40, while the other 8 un-renovated houses had an average q50 of just about 11 m³/(hr.m²)) – in other words, houses renovated in the last few years may not show a great improvement in their airtightness levels. Those built between 2000 and 2007 averaged a q50 of just about 7 m³/(hr.m²) (32 dwellings tested).

Therefore it is clear that a new maximum q50 level of 7m³/(hr.m²) will not be hard to reach.

Where we're going

If builders aim only for the new level of 7 m³/(hr.m²) there is not likely to be much of a change from current building practice. Some more caulking of window-boards, attic hatch closers and pipe penetrations may get the least committed builder by.

The authors of the building regulations may justifiably reply that the level of 7 m³/(hr.m²) is a “back stop” value. That is, it is the very worst permissible level, and that in order to comply with the new Part L as a whole the actual levels of airtightness will need to be substantially lower. Don't forget that relative to a reference dwelling, new dwellings will have to show an overall energy reduction of 60%. It's a fair point.

Except for the elephants in corner.

The elephant in the corner – Building Control & Compliance

On the face of it, therefore, the new regulations will no doubt result in better overall dwelling airtightness levels than we have seen to now, if only marginally.

However, until building control is given more teeth, and until airtightness test result sheets become part of building regulation compliance documentation, then it is likely that many dwellings that should be tested will remain untested. This situation is repeated all over Europe, and is not just an Irish phenomenon.

The other elephant in the corner – which method?

The standard used for airtightness testing in Ireland is the European standard – EN13829. That this should be an elephant in the corner may sound a little bit strange — maybe it's more of a tapir.

The problem here is that the document outlines two main ways to prepare a building for a test – method A and method B. The methods differ in the treatment of known openings in the envelope – ie are known openings left as they are, or are they closed up, or are they sealed? The difference in the result of a house tested with method A could be easily many times worse than with Method B.

The TGD L does not give definitive guidance on which method to use, and despite the fact that most testers use Method B (or something very similar), I'm sure we do not all follow the same check-list of items to close and which to seal.

For more on this significant issue, please see 'Which Method?' below.

How do we compare?

It may be helpful then to see how we stand in relation to other countries. There are perhaps a few questions in one here.

1. Do other countries have compulsory airtight limits as we do, and what are the limits? And if they have a compulsory limit, do they have compulsory testing?

2. How do the actual results of tests in other countries compare to the types of results we reported on above?

Apart from Britain, whose situation most of us are familiar with and on whose regulations much of our own is based, it can be eye opening to look at other countries.

A report by ASIEPI(2) from 2008 looks at airtightness regulations and testing regimes over 13 countries (Ireland was not included in the report).

It also looks at how airtightness is dealt with in the local equivalent of Deap (The Dwelling Energy Assessment Procedure — the software used to calculate BERs). The findings were that ten of the 13 countries looked at had airtightness regulations and of that ten, six had minimum requirements for new dwellings(3).

However, at that time, none of the countries covered in the report had a national testing regulation – in some cases, such as Denmark and Belgium, local authorities or regions may specify it.

Also, existing dwellings were only covered in two countries, and at that, only for extensive modifications.

When we attempt to compare the regulations in those countries that impose airtightness limits to each other and to our own regulations, we quickly come back to the issues raised under 'Q50 & N50' and 'Which Method?' below. 

Some countries report in Q50 (eg Ireland, Belgium), others in n50 (eg Germany), some even in w50 (eg Denmark). This is not a big obstacle, as average values could be worked out to standardise the result to allow comparisons, eg the average volume to area ratio, or floor area to envelope areas(4). More crucial is that some countries specify method A (eg Belgium, formerly Germany), others method B (Germany now), others a mish mash (France, arguably Ireland).

Looking at France, for example, the building preparation method is a hybrid, the results are reported not at 50 Pa but at 4 Pa, and the ground floor is excluded from the envelope measurements.

Thus, cross country comparison of the regulations, and indeed of the reported results is practically impossible, unless we standardise the data. Nevertheless, the figures in the regulations for most of the other countries would seem to indicate that the houses are required to be more airtight than Irish houses  - eg for standard houses, Norway and Germany require an n50 of three or less using method B.

Briefly, we will examine here two other small north European countries here – Denmark and Belgium.

Denmark
The Danish standard is a w50 of 1.5 lps/m². In other words, at first glance the requirement reads nothing like our regulation. However, as it is a w50, we know that the m² relates to the floor area. 1.5 litres per second equates to 5.4 m³/hr. As the Danes also use method B, we can directly compare these results, if we also know the envelope area of the buildings in question. As an example, lets take simple two scenarios, a 120m² bungalow and a 200m² two story house.

For the bungalow, a likely envelope area would be 345m² (10mX12mX2.4m). For the two story house, a likely envelope area would be about 405m2(10mX10mX5.1m).

To compare the Irish and Danish regulations for these two house types, it may be best to present the results in a table. The numbers shown are the maximum allowed airflow using the blower door machine to meet the regulations, per hour (m³/hr):

Maximum Test Airflow	Ireland	Denmark
Bungalow 120m2 floor, 345 m2 envelope	3450 m3	648 m3
Two Story 200m2 floor, 405 m3 envelope	4050 m3	1080m3

It is clear the Danish regulation is significantly more stringent.

Belgium
The airtightness regulation is governed by the local regions rather than on a national level, but they agree on a q50 maximum limit of 12 m³/(hr.m²). At first glance, this is not as stringent as Ireland's regulation of a q50 maximum of 10 m³/(hr.m²).

However, in Belgium the test is performed using Method A (eg open fireplaces are not blocked, non-closeable vents are not sealed for the test), and thus if such a standard were imposed on Irish houses the vast majority of them would fail.

However, it would only be possible to compare directly the results of houses that do not have, eg fireplaces or open vents. Thus comparison is not handy.

Passive houses
All certified passive houses must achieve an n50 value of less than 0.6 air changes per hour, based on a relatively standardised German-centric floor area calculation. Allowing for the fact that almost certainly in various countries the floor area will be calculated slightly differently, we nevertheless can have confidence that passive house blower door test results are pretty much comparable across Europe, and thus the author feels it is important that publications such as Construct Ireland continue to report airtightness results using the n50 value as well as the q50 value.

It will be noted that an n50 of 0.6 airchanges will amount to a q50 result of probably no more than 0.75 m³/(hr.m²) depending on the shape of the house. Evidently this is far beyond the current and proposed Irish regulations. And yet people manage it.

Conclusions

To conclude then, we can see that we are ahead of the posse in having both an airtightness regulation, and a specified testing regime. We have also seen from the brief analysis of the result set that the airtightness levels are improving – at least in tested houses, and it would be expected that the proposed lowering of the maximum level of air permeability will also promote better airtightness in Irish construction.
 
Unfortunately, there are also a number of obstacles and uncertainties.

The maximum required level of airtightness is generally higher than other countries, and allows for really quite leaky houses to be built. The few builders that remain in work will not quake in their boots at the prospect of achieving an air permeability q50 of 7 m³/(hr.m²).

Also, until the requirement for testing is enforced, or better still, until the building community at large understand the importance of good airtightness and from the inside out wish to achieve it, it is likely that we will continue to have buildings that are relatively more leaky than many of our European neighbours.

Finally, the test methodology has to be fully spelled out, so that we are sure that tests are being carried out uniformly across the country. This could be done in the form of specific extra notes in the TGD L, or perhaps we could, as part of a voluntary testing association create code of conduct or the like, or perhaps better still, more testers would join a scheme such as the NSAI or INAB schemes and thus have independent verification of the quality and manner of the testing being undertaken.

Appendix 1: q50 & n50

When reviewing airtightness results you will frequently come across the terms 'q50' or 'n50'. You may also see the phrases 'air permeability' and 'air change' and 'air leakage'.

The blower door machine moves air into or out of a building. When it does so, it creates a difference in the air pressure between the inside and the outside, which is measured in Pascals (usually written Pa). For both the q50 and the n50, the result is expressed as the amount of air measured moving through into or out of the building when the pressure difference between the inside and the outside is 50 Pa. Hence the 50 part.

The EN13829 standard lists three main ways to understand an airtightness result. In simplified form: 

Measurement Name	Measurement is based on	Unit of Measurement
q50	Square metres of external envelope area	m3/(hr.m2)
n50	Volume of the building	ACH (Air changes per hour)
w50	Square meters of floor area	m3/(hr.m2)

The q50 reading is also called the 'air permeability' rate, and the n50 called the 'air change' rate.

The term 'air leakage' seems to be used interchangeably for any of the above, and also to refer to the actual rate in cubic metres, of airflow at 50 Pascal. Therefore, treat it with caution.

In this country, it is the q50 result which is required to show compliance with the Part L requirements, and for Deap calculations. For the passive house standard it is the n50 value which is required.

Note that even though the same EN13829 standard is in use, in different countries different methodologies may be in use to measure the external surface areas, volumes or floor areas. Thus for example, the air volume for a (German regulation influenced) passive house blower door test n50 will not be the same as the type of gross internal volume that many testers in Ireland will typically use.

The w50 based on floor area is not really used in Ireland.

Appendix 2: Which Method?

The relevant section in the 2008 and the draft 2010 editions of the Technical Guidance Document for Part L of the Building Regulations is unchanged, as follows:

2008 Edition (pg21)

1.5.4.1
Air permeability can be measured by means of pressure testing of a building prior to completion.

The procedure for testing is specified in IS EN 13829:2000 “Thermal performance of buildings: determination of air permeability of buildings: fan pressurization method”.

Additional guidance on testing procedure is given in Sections 2 to 4 of the BSRIA Guide “Airtightness testing for new dwellings” and CIBSE Technical Manual TM 23 “Testing Buildings for Air leakage”.

Permeability is calculated by dividing the air leakage rate in m3/hr by the envelope area in m2

The performance is assessed at 50 Pascals pressure difference. It has been empirically determined that for dwellings generally the permeability at 50 Pascals pressure difference is approximately 20 times the air change rate at normal conditions. Guidance on appropriate extent of testing is given in Paragraph 1.5.4.3.

IS EN 13829 is the main document referenced.

As we have seen, there are various ways to report the airtightness test results. The TGD Part L refers to 'air permeability' and thus we can be confident that the result we need to report is the q50 value.

However, EN13829 allows for two main ways to prepare the building envelope for a blower door test – called method A and method B.

These might be best understood in a table, as briefly summarized below:
 

Building Preparation	Method A 'Test of Building in use'	Method B 'Test of Building envelope'
Trickle vents, Slider vents	Close	Seal off
Windows, External Doors	Close	Close
Other Adjustable openings	Close	Close
Non- Closable vents	Leave as is	Seal off
Mechanical Ventilation System	Seal off	Seal off
Chimney	Close if damper, otherwise Leave as is	Seal off

To see how the choice of method can affect the result, we can quote from an older test by Overy & Associates. Although the test was conducted prior to the adoption of the EN13829 standard, a house was tested with all openings open but with chimney dampers closed (as per method A), and with openings sealed (as per method B).

The result of the first test was just on seven air changes per hour (ACH), the result of the second was about 5.5ACH. The test was also repeated with no fire dampers closed (as in many Irish houses there are none to close), and the result was a little over 8ACH.

It will be noted that neither the TGD L, nor the Deap help files, specify which method should be used, and therefore conceivably method A or B could be used.

However, if we look at the other documents referenced we may get a better indication of which method to use.

TM23 is a document from CIBSE in Britain, from 2000. At the time that TM 23 was written EN13829 was still going through the procedure of being passed, and so is referenced as a proposed standard 'pr13829' in TM23.

TM23 does not make a direct reference to which building preparation method should be chosen. Instead TM23 gives its own advice on building preparation for a test (pg7) (5):

“Where appropriate, combustion appliances must be switched off and any open flues and air supply openings temporarily sealed”, also
“-all external doors, windows, ventilation openings are closed.
- all mechanical ventilation systems are turned off and sealed”

Looking at the table outlining the method A and method B differences above, we see that the TM23 has a bit of both – seal fireplaces as per method B, but just close ventilation openings, rather than seal them.

The other document referenced is one from BSRIA, Airtightness Testing For New Dwellings (6). BSRIA is a commercial body that operates in Britain (and in Ireland via Anderson Mechanical Services). It is not clear to the present author how a commercial body is given such a prominent reference in the building regulations, but that is for another day. This BSRIA document references the English building regulations Part LA1 and on page 5 states:
“The airtightness tests need to be carried out by an independent, qualified (or accredited) person in accordance with the procedure set out in ATTMA Technical Standard 1...”

Furthermore, the 2006 edition of the English building regulations (page 22) also state that the test procedure should be carried out as per the ATTMA guidelines.

ATTMA is the Airtightness Testing and Measurement Association in Britain, and it produces a very good document on testing procedures. Until October of this year the document was called Attma TS1, or Technical Standard 1 (7). Page 14 of this document deals with preparing the building for the test, and it is the first place we come across reference to method B as per EN 13829.

This document states (pg 15):

“Trickle vents, smoke vents and all passive ventilation systems should be closed but not artificially sealed. Permanently open uncontrolled natural ventilation openings should be temporarily sealed.

Mechanical ventilation and air conditioning systems should be turned off. These systems should be temporarily sealed to prevent air leakage through the systems during the test. “

No direct mention is made of chimneys here. However, in essence it agrees with the TM23 document regarding trickle vents and closable wall vents.

Thus finally, we can arrive at an understanding of what the building regulation Technical Guidance Document is trying to advise – in short, use method B of the EN13829 standard, but only close rather than seal trickle vents and other closable not mechanical ventilation points, as per TM23 and the ATTMA standard.

Certainly that is my understanding, and that of several others testers I have contacted. However, it is also clear that not every tester has been following this reasoning – many others seal the trickle and closable vents in strict compliance with method B of EN13829, but contrary to TM23 and the ATTMA standard of the time.

This entire situation is complicated somewhat by the release in October 2010 of a new version of the ATTMA (8) document, which states:

“Background trickle ventilators, passive ventilation systems and permanently open uncontrolled natural ventilation openings should be temporarily sealed.”

The reason for this is that this change is specifically mentioned in the newest English Part L regulations. Whether this trickle vent change extends to sliding vent covers or not, is not entirely clear.

Submissions have been made to the Department of the Environment on the basis of the latest draft regulations, but it may also be no bad thing if testers across the country simply agreed on a test methodology so that we can all be sure that we are comparing like with like.

Acknowledgements

Mention should be made of, and thanks are due to, the other testers in Ireland and abroad  who contributed test results and/or useful discussion for this article:

Ireland

Christ Spoorenberg – www.ecoscan.ie
Chris Montague – www.betechnologies.ie
Paul Overy – www.overy.ie
Mark Shirley – www.2eva.ie
Michael Kiernan – www.rebelenergy.ie
Mark Forkin – www.nutechrenewables.com
Marcus Casey – www.premierenergyrating.ie

International

Belgium – Kris Bruneel -  www.bbouwadvies.be
Denmark - Jørgen Ladegaard  - www.termografi-el.dk
Norway - Hans Ananiassen – www.trykktesteren.no
France -Eric Rio – www.lorient-diagnostic.com

References

1. Airtightness Testing for New Dwellings, BSRIA, 2006. First published in October 2006 in House Builder magazine. Appears to no longer be available via the BSRIA website.

2. Synthèse du Questionnaire ASIEPI – État de Pratiques Européenes Concernant L'étanchéité à L'air de L'enveloppe et des Réseaux, prepared by CETE de Lyon for ASIEPI, August 2008. Downloaded from www.asiepi.eu

3. The countries with a maximum airtightness requirement were – Germany , Czech Republic, Denmark, Spain, Norway, Netherlands.

4. An example of something similar is found in Richtlinie Für Luftdurchlässigkeitsmessungen Bei Minergie und Minergie-P Bauten, Minergie, Switzerland, 2007. Downloaded from www.minergie.ch. For this Swiss low energy standard, to try to get over the q50 versus n50 reporting problem, Minergie has introduced a standardised n50 value, based on a area/volume ratio of 0.8.

5. Testing Buildings For Air Leakage - TM23:2000, CIBSE, London 2000

6. Airtightness Testing for New Dwellings, BSRIA 2007. Downloaded from www.bsria.co.uk

7. “Technical Standard 1 (Issue 2) – Measuring Air Permeability of Building Envelopes, Airtightness Testing and Measurement Association (ATTMA); download from www.attma.org

 8.Technical Standard L1 (October 2010 Issue) – Measuring Air Permeability of Building Envelopes (Dwellings), Airtightness Testing and Measurement Association (ATTMA); downloaded from www.attma.org

This article is by:

Gavin O Se
GreenBuild Energy Rating & Building Information Services Ltd T/A GreenBuild (www.greenbuild.ie)
Fortchester, Inch, Gorey, Co. Wexford.
Company No: 436513
Directors: G Ó Sé, J Uí Shé