Although this seems to be true, this is a dangerous generalization. There are many factors that govern airtighness of round ductwork, including roundness (any slight ovality can cause leakage at the joints) and the flatness of the seams at the joints between two ducts (protuding or convex seams can cause point leakages at the joint irrespective of the gasket type).

This is generally not possible. In systems with good airtightness, the leakage flow rates are too small for one to be able to accurately measure the flow rate over the fan.

For more information you can check-out the Webevent presentation “Duct System Leakage Testing — Methods, requirements, handing over” by P.G.Schild, available on the website www.asiepi.eu

A good starting point is the paper "stimulating better envelope and ductwork airtightness with EPBD" by Carrié et al., at the AIVC 2008 conference (www.aivc.org).

Another possibility is to check out the presentation “Including leakage in energy calculations” by Dr. J-R Millet, held at the ASIEPI web event on ductwork airtightness (http://www.asiepi.eu/wp-5-airtightness/web-events.html)

It is not presently mentioned in CEN standards. Other countries might be using the same default level as France in their energy performance calculation standards, but there is no reason to harmonize it. The leakage class 2.5*A is very much leakier than ductwork systems in Scandinavia, for example. By the way, ASHRAE’s classification system for duct airtightness has higher classes near 2.5*A.

For more information you can check-out the Information Paper P187 “Duct System Air Leakage — How Scandinavia tackled the problem” by P.G.Schild & J.Railio, on the BUILDUP website

Sweden does not use its building regulations to impose its strict requirements on ductwork airtightness. Rather it is a voluntary trade standard (the AMA VVS specifications) that has driven the market transformation to more airtight ventilation systems. This trade standard is used in virtually all building projects.

Norway is another example of a country where good products are used (predominantly round ductwork with pre-fitted gasket seals, giving leakage class C) without any explicit requirement in the building regulations, and where duct leakage is not taken into account in the energy performance calculations.  The reason for this is historic. Up until maybe 15 years ago, Norway had explicit requirements in its building trade specifications, just as Sweden, but this is no longer deemed necessary because almost all products on in the Scandinavian market are quality airtight ductwork anyway.

For more information you can check-out the following:

•  Information Paper P187 “Duct System Air Leakage — How Scandinavia tackled the problem” by P.G.Schild & J.Railio, on the BUILDUP website
•  Final technical report from project ASIEPI, Work Package 5, see www.asiepi.eu
  

The reason for this is twofold:

(1) When one uses only one test pressure, then the test results can be expressed with just one number, the “flow coefficient”. This is very simple and avoids the need for any calculations. You are suggesting testing with multiple pressures so that the results can be plotted on a log-log scatterplot, and the best-fit line is a power function with two coefficients, the “flow coefficient” and the “flow exponent”. This is unnecessarily complicated.

(2) Experience tells us that ducts with good airtightness (i.e. that complies with the requirements in Scandinavia, normally Class B or C) have a pretty uniform flow exponent. Ducts with poor airtightness can have a significantly lower flow exponent, but then again such duct systems fail the leakage requirements anyway. Thus there is no real need to measure the flow exponent.

For more information you can check-out the following:

•  Information Paper P187 “Duct System Air Leakage — How Scandinavia tackled the problem” by P.G.Schild & J.Railio, on the BUILDUP website
• Webevent presentation “Duct System Leakage Testing — Methods, requirements, handing over” by P.G.Schild, available on the website www.asiepi.eu

The reason for this is twofold:

(1) When one uses only one test pressure, then the test results can be expressed with just one number, the “flow coefficient”. This is very simple and avoids the need for any calculations. You are suggesting testing with multiple pressures so that the results can be plotted on a log-log scatterplot, and the best-fit line is a power function with two coefficients, the “flow coefficient” and the “flow exponent”. This is unnecessarily complicated.

(2) Experience tells us that ducts with good airtightness (i.e. that complies with the requirements in Scandinavia, normally Class B or C) have a pretty uniform flow exponent. Ducts with poor airtightness can have a significantly lower flow exponent, but then again such duct systems fail the leakage requirements anyway. Thus there is no real need to measure the flow exponent.

For more information you can check-out the following:

•  Information Paper P187 “Duct System Air Leakage — How Scandinavia tackled the problem” by P.G.Schild & J.Railio, on the BUILDUP website
• Webevent presentation “Duct System Leakage Testing — Methods, requirements, handing over” by P.G.Schild, available on the website www.asiepi.eu

It is important to use the correct type of screw or rivet. Avoid screws with a wide drill tip (which drill a hole in the duct wall) and blind pop rivets (which are not airtight because the inner splint falls out). The best screws are ones with a sharp tip or pressure-tight blind pop rivets (though these are very laborious to fit).
For more information see the section on screws and rivets in EN 12237 (“Ventilation for Buildings - Ductwork - Requirements for ductwork components to facilitate maintenance of ductwork systems”)

In Sweden, according to the trade guidelines VVS AMA, on-site pressure tests need only cover 10% of the duct surface area round ducts, and 20% for rectangular ducts. Denmark has similar rules. In Finland, 20% of the duct surface area is tested in the case of Class C airtightness, and 10% in the case of Class D or better.

All parts of the duct system are equally important in terms of potential air leakage. In the main duct, near the fan, the operating pressure is higher than the rest of the duct system, so leakage points there have a higher leakage flow rate. On the other hand, field studies of leaky duct systems have shown that most of the total leakage flow occurs in branch ducts near the terminals. This is because branch ducts constitute a very large surface area compared to the main duct, and branch ducts tend to have more joints between components (e.g. tee junctions, VAV-boxes, flexible ducts, plenum boxes) which creates numerous opportunities for leakage [see reference 24 in Paper P187 below]. 

For the above reason, all parts of a duct system should be subject to leakage tests. To save time, one can test select multiple test sections at random (each of approx 25 m² and never less than 10 m²), constituting a small fraction of the total duct surface area. Test sections can be isolated from the rest of the duct system by e.g. pumping up an air bladder (balloon) with a hand pump.

For more information you can check-out the following: 

• Information Paper P187 “Duct System Air Leakage — How Scandinavia tackled the problem” by P.G.Schild & J.Railio, on the BUILDUP website
• Webevent presentation “Duct System Leakage Testing — Methods, requirements, handing over” by P.G.Schild, available on the website www.asiepi.eu

In part. Leakage testing of duct systems, by means of a pressure test, is a very effective way of achieving more airtight duct systems. This type of commissioning is required throughout Scandinavia, to check compliance with minimum airtightness requirements (except for Norway, where the practice has become less common in the last 15 years). However, airtightness testing is no substitute for installing quality duct systems with good airtightness in the first place. Various field studies have shown that rectangular ductwork is generally leakier than rigid round ductwork with prefitted gasket seals. Moreover, it is well known that the airtightness of some other duct systems can degrade over time. The gasket seals on flanged metal rectangular ductwork can become more leaky over time and need to be dismantled regularly for maintenance. Field examinations have shown that taped seals on rectangular duct-board systems tend to fail over extended periods of time. In addition, the clamps required by the trade standard (UL 181), for example at the connections to flexible ductwork, can fail and their durability has been questioned.

For more information, read the Information Paper P187 “Duct System Air Leakage — How Scandinavia tackled the problem” by P.G.Schild & J.Railio, on the BUILDUP website Normal 0

Article 4 of the EPBD stipulates explicitly that attention should be paid to indoor climate conditions: “Member States shall take the necessary measures to ensure that minimum energy performance requirements for buildings are set … These requirements shall take account of general indoor climate conditions, in order to avoid possible negative effects…”.

This is also listed in the annex of the EPBD: “The methodology of calculation of energy performances of buildings shall include…

(d) ventilation;…

(h) natural ventilation;…

(i) indoor climatic conditions, including the designed indoor climate.”

Whereas in the past the major challenge was to keep our buildings sufficiently warm, nowadays the challenge is in guaranteeing reasonable comfort conditions in summer without (or with minimum) cooling energy. It is therefore important that building designers and other stakeholders understand the thermal behavior of a building and its occupants and are aware of the available alternative techniques that substantially improve the comfort in the building and significantly decrease (or even eliminate) energy consumption. For example, solar and thermal control techniques, heat amortisation and heat dissipation techniques have been proven to be extremely efficient and may decrease the cooling load of buildings up to 80 %.

This is defined in EN 13829, § 6.1.2. All exterior surfaces, plus floors, ceiling and walls to neighbouring apartments are taken into account.
However, other assumptions are used in some countries to extract indicators that better fit the national requirements of the EP-regulations (see question regarding the measurement of multi-family buildings).

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

Measurements usually show that light (e.g., timber-frame or steel) constructions are leakier than massive construction, but this is not bound to be. In fact, PassivHaus houses are often light constructions and are very airtight.

The general statement that can be drawn is that light constructions are more sensitive as the airtight layer can be more affected by poor design and workmanship than in massive constructions.

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

To some extent yes. A collection and analysis of some data collected through the ASIEPI partners has been presented at the 2008 BlowerDoor conference (Papaglastra et al., 2008). However, caution should be exercised when comparing airtightness values :
- the measurement methods are not necessarily the same. EN 13829 mentions two methods (A and B) that can lead to very different results e.g., depending on how combustion appliances are sealed; (*)
- the calculation of the volume or the envelope area that is used to normalise the leakage flow may be different for the same building, depending on the assumptions adopted nationally or even by the operator.
It would be useful to harmonise these methods to be able to reliably compare and monitor results.

(*) In Belgium, additional specifications for the measurement of envelope airtightness have been published (http://www.epbd.be/go/airtightness-measurement)

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

There are three major ways to estimate this impact:
- the simplest is to evaluate the infiltration losses based on a rule of thumb established by Drubul in 1988 (*) and suggested by Kronvall (**) in 1978. The rule says that the infiltration airflow rate in air changes per hour may be determined by dividing the n50 value by an empirical coefficient that lies between 10 and 30. In practice, the empirical coefficient is often set to 20 (i.e., the infiltration airflow rate is equal to the airtightness at 50 Pa divided by 20);
- an intermediate approach is to estimate the infiltration airflow rate based on the empirical model proposed in ISO 13790, annex G;
- a more detailed approach is to perform and hourly simulation of the airflow rates, based on a pressure network code such as that described in EN 13465.
Once the infiltration airflow rate is known, the calculation of the energy losses is straightforward.
In national regulations, one of those three methods is commonly used in the EP-calculation.

(*) Drubul C, Inhabitant’s behaviour with respect to ventilation, Technical note 23, Air Infiltration and Ventilation Center, 1988
(**) Kronvall J, Testing of houses for air-leakage using a pressure method, ASHRAE trans. Vol 84 no 1 1978

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

Energy wastage due to envelope leakage has been estimated in various countries through numerical simulations:
- in Belgium and in Germany, the energy impact has been calculated to be about 10% of the energy performance level for individual houses;
- estimates based on simulations on 9 real buildings in France lie between 0.5 and 15 kWh-pe/m2 per year (depending on climate, ventilation system type, and building configuration), with an average of 6.6 kWh-pe/m2, between “default” and excellent airtightness

For comparison purposes, in Belgium, France and Germany, the impact of good envelope airtightness is similar to that of solar collectors for domestic hot water.

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

The SAVE-DUCT project has shown that there are large discrepancies between countries on this subject (the project report is available through AIVC at www.aivc.org). Analyses performed on measurements results have shown that Belgian and French ductwork systems were typically 3 times leakier than Class A whereas Swedish systems commonly complied with Class B (i.e., 3 times tighter than Class A). The main reason is that the Swedish quality system (Boverket) imposes a rather pragmatic testing scheme which has encouraged the systematic use of ductwork with pre-fitted seals that guarantee good airtightness. This type of ductwork is very much used in Nordic countries where similar conclusions can be drawn.
This topic will be discussed in more detail during the next ASIEPI WP 5 Web Event in October 2009.

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

The airtightness measurement of single apartments is usually performed the same way as in individual houses. Therefore, there is no specific protocol to balance the pressure between the apartment under test and other apartments. Different protocols may lead to very different results.

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

To our knowledge, the UK is the only country that has made testing mandatory. This has been in force since 2002 for large buildings and extended to most buildings in 2006.

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

Measuring or evaluating the airtightness of multi-family buildings is challenging as there may be technical and practical difficulties that prevent from pressurizing the whole building. These difficulties include for instance building configuration (e.g., exterior hall ways that do not allow pressurisation from a single point), pressurisation fan size, stack effect, or cost.
Therefore, the measurement of the whole envelope of multi-family buildings is rarely performed in practice. However, several countries use alternate schemes to overcome these problems, for instance:
- in France, in the framework of the BBC-Effinergie label, the measurement can be performed by apartment. The basic idea is to test a sample (e.g., of 3 units for a building of 30 apartments or less) and to make a weighted average of the results to estimate the global airtightness.
- in Norway, tests on single apartments can be used. It is seen as a simple practical measure that also encourages airtightness between apartments, which is good to avoid noise and cross-contamination.

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

There is little information on the subject at the moment. A study referenced in a paper written by Erhorn et al. and presented at the 2008 AIVC conference shows airtightness values of 31 “PassivHaus” houses at commissioning and 2 year later. The average n50 at commissioning and 2 year later were 0.37 and 0.46 ach at 50 Pa, respectively.

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

There exists a certification scheme in Germany which is not compulsory to perform tests.
There exists a compulsory authorisation process in France for technicians who perform test on BBC-Effinergie buildings.
To our knowledge, there is no other certification or authorisation scheme in European countries.

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

There is no up-to-date document at this time.

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12

Now in Europe, many countries have adopted the n50 value (i.e., the leakage flow divided by the volume) for their EP regulation while others (e.g., Belgium, France, UK) use the envelope area normalisation. The rationale behind this latter choice lies in the fact that the volume is not needed for an energy performance calculation. Using the n50-value should require a precise definition of the way the volume is calculated, which to our knowledge, is not the case in any European country. On the other hand, the envelope area is usually well-defined in regulations.
More generally, the indicator used may be different depending on the application. The volume normalisation seems well-appropriate to pollutant transfer applications as the volume is a necessary input. Similarly, the exterior envelope area indicator seems well-appropriate to energy use applications. The floor-area normalisation may be interesting for comparison purposes as it is often more difficult to have access to the volume or envelope area data than the floor-area.

Anwered by: Rémi Carrié and Gaëlle Guyot (CETE de Lyon)
Date: 2008/12/12