1、OR-05-2-2 European Projects on Ductwork Quality Johnny V. Andersson, PE Remi F. Carri, PhD Peter Wouters, PhD Member ASHRAE Tor G. Malmstrom, PhD Member ASHRAE ABSTRACT This paperpresents results from European Union projects on ductwork qual tightness. The projects were done in Sweden, Belgium, and
2、France. The tightness on a large number of ductwork installations in the three countries was compared and it was found that there is a tremendous difference in tight- ness. The answer to the question, “Why this large difference between the countries?” is most probably that Sweden has been requiring
3、tight ducts, i.e., specifiing how much they are allowed to leak at a certain test pressure, whereas in the two other countries, tightness of ductwork is normally neither required nor tested. INTRODUCTION Tight Ductwork is One Way of Saving Energy This paper describes a part of a research project fin
4、anced by the European Union. The background and the objective were as follows. There is a great concern in Europe to reduce energy consumption. In Sweden this is especially important as it has been decided in a referendum that all Swedish nuclear power plants are to be shut down within the next deca
5、de. Using ductwork with a low leakage rate is one way of using energy more efficiently. With tighter ducts there is less need to compensate for the leak flow with higher fan airflow; there is less air to heatcoollclean, etc. Typical ductwork installations in Belgium, France, and Sweden were leakage
6、tested and compared. Most of the duct installations were installed in residential and commercial buildings, mostly offices. As described, there was a tremen- dous difference in tightness between the ductwork in Sweden Christophe Delmotte and that in Belgium and France. Probably the most important re
7、ason for this difference is that tight ductwork has been required in building specifications in Sweden since 1968 and testing ductwork tightness is also normally required. One conclusion in the report (Cad et al. 1999) is that if other countries in Europe would install ductwork with tight- ness simi
8、lar to what is required in Sweden, a considerable amount of energy would be saved annually. The potential of energy savings is described in the report as: The order of magnitude of the energy savings that can be achieved by using tighter ductwork in Europe is proba- bly in the region of 1 to 10 TWye
9、ar (3.6 to 36 PJ/year). Long History of Ductwork Requirements in Sweden In Sweden, requirements on ductwork tightness and other quality aspects of ductwork have been specified as part of building specifications since 1966. Nearly all buildings and their installations are performed according to the A
10、MA spec- ification guidelines (AMA stands for “Allmn Material- och Arbetsbeskrivning,” Le., “General Requirements for Material and Workmanship,” published by The Swedish Building Center, a nonprofit organization). AMA comprises five differ- ent technical parts, each having a book with requirements a
11、nd another giving advice to the consultant on how to speci. The first author is general secretary for VVS AMA (the HVAC part of AMA). The AMA books are shown in Figure 1. The Swedish Building Center does not develop any national or intemational standards but refers to Swedish national standards, Eur
12、opean norms, and International Stan- dard Organisation (ISO) and European Committee for Stan- dardization (CEN) standards where applicable. A national or international standard is only compulsory to follow if so spec- ified; otherwise they are regarded as recommendations. In Johnny V. Andersson is t
13、echnical director at Ramboll Sverige AB, Stockholm, Sweden. Remi E Carri is the head of construction pole at CETE, Lyon, France. Peter Wouters is the head ofthe Department ofBuilding Physics, Indoor Climate and Building Services, BBRI, Brussels, Belgium. Tor G. Malmstrom is a professor of building s
14、ervices engineering at KTH, Stockholm, Sweden. Christophe Delmotte is the adjunct head of the Laboratory of Air Quality and Ventilation, BBRI, Brussels, Belgium. 02005 ASHRAE. 51 5 Tightness Class Approximate Corresponding US Leakage Class C, A (CL20) B WL7) C (CL4 D (-1 Europe all countries have, h
15、owever, agreed to follow European Norms (EN) and to withdraw their national standards when a corresponding EN is established. When such standards, or national Swedish bylaws, are missing measurable requirements in the AMA, general spec- ifications are defined by the center itself. When preparing the
16、se requirements reference is made to international docu- ments. The AMA requirements are then decided by a steering group comprising building proprietors, researchers, builders, contractors, and consultants after having been under open consideration. The AMA requirements are made binding when they a
17、re referred to in the contract between the owner and the contrac- tor. The level of the AMA quality requirements is advanced when possible by technologyprogress and found profitable for the owner on a life-cycle cost basis. Proposed increased requirements are established after they have been referre
18、d for consideration to a large number of property owners, contrac- tors, consultants, and other interested parties. The require- ments can be updated twice a year. Proposed changes are published in AMA-nytt (AMA Navs Journal). If the designer should decide that these AMA require- ments are either to
19、o high or too low for a project he or she will add a more suitable text to the specification-under actual code and heading-and this text then takes over the AMA requirement. The need for tight duct systems has thus been identified in Sweden since the early 1960s. AMA does not address whether the sys
20、tem is installed in residential, commercial, or industrial buildings or type of joints or sealing method. The require- ments differ between round and rectangular ducts with higher demands for the round ducts as described below. The require- ments have evolved over time in conjunction with technology
21、 progress: Maximum Accepted Leakage L/(s,m2) at Test Pressure 400 Pa 1.32 0.44 0.15 0.05 AMA version 1966: Two “tightness norms,” A and B, were used to specifi the tightness. The tightness was to be spot checked by the contrac- tor; minimum tested surface was 10 m2 (100 fi*). AMA version 1972: Requi
22、rements transformed into two “tightness classes” A and B (same as the EUROVENT classes used today). Class A . See Table 1 for a comparison between the AMA, i.e., also EUROVENT, tightness classes and the leakage classes, CL, as defined by AISI/ASHRAE/SMACNA/TIMA in US. Figure I The Swedish AMA books
23、(VVS = HVAC), edition 1998. was the requirement for the complete duct system in the air- handling installation (i.e., including dampers, filters, humidi- fiers, and heat exchangers). It was advised to meet Class B when: the system was in operation for more than eight hourslday and the air was treate
24、d (cooled, humidified, equipped with high-class filters, etc.). AMA version 1983: Class C was introduced in 1983 and is required for circular duct systems with a surface larger than 50 m2. In this version of AMA tightness, the following require- ments were given: Class C should be used for round duc
25、twork with a perimeter surface larger than 50 m2. Class B should be used for round duct systems with a surface smaller than 50 m2 and also for rectangu- lar ductwork. Class A was only accepted for visible supply and exhaust ducts within the ventilated room (where leakage is less important). If the d
26、ucts were installed above a false ceiling, and thus not visible, tightness according to classes C and B was required. AMA version 1998: In this version of AMA, tightness Class C is required for circu- lar ductwork already at an area of 20 m2. Tightness Class D has been added (three times tighter tha
27、n Class C) as an optional requirement for larger circular duct systems. 516 ASHRAE Transactions: Symposia 2 1 0.8 0.4 0.2 0.1 0.08 0.04 0.02 0.01 40 80100 200 400 800 1000 2000 Static pressure difference, Ap, Pa 20 15 8 4 2 1.5 0.8 0.4 0.2 0.16 0.240.4 0.6 1.6 2.4 4 8 Static pressure difference, dp,
28、 inches of water Figure 2 Duct leakage clusscjcations (SI units). Figure 3 Duct leakage classcutions (I-P units). Further comparison between European tightness classes TIGHTNESS COMPARISON WITH OTHER COUNTRIES and the ones used in the US is presented in Figures 2 and 3. Specified Requirements are Ch
29、ecked Besides specieing classes that have to be met, AMA also requires that all ventilation and air-conditioning systems be carefully commissioned. The procedures include: Measurement and adjustment of all extracted and sup- plied airflow in the building; the result should be within +15% (including
30、the measurement error). The result is to be presented on standard AMA protocols. The duct system leakage has to be verified, normally by the contractor as part of the contract. This is undertaken as a spot check where the parts to be checked are chosen by the owners consultant. For round duct system
31、s, 1 O%, and for rectangular ducts, 20%, of the total duct surface has to be verified. In case the system is found to be leak- ier than required, the tested system shall be tightened and another equally sized part of the system shall be ver- ified in the same manner. Should this part also be found t
32、o leak more than is acceptable, the complete installation In two European Union research projects-SAVE DUCT 1999 and Airways 2002-partners from three European coun- tries-Belgium, France, and Sweden-studied ductwork installations. Although duct leakage can be a source of consid- erable problems, lit
33、tle is known about the ductwork airtight- ness status in most of the European member states. However, field experiments conducted in various countries suggest that air distribution systems are very leaky in general. Except in Sweden, low-quality ductwork is widely used and poorly installed, yielding
34、 leakage rates typically 30 times greater than those of EUROVENT 2/2 Class C systems. Typical problems were found to include: inadequate ductwork component selection, insufficient sealing work at installation, ill-fitted components, worn tapes, and has to be leak tested and tightened until the requi
35、re- ments are fulfilled. The contractor covers the cost for these additional tests. AMA also specifies other requirements on the ductwork such as material standards, installation, joints, hangers, strength, etc. The obvious result of this combination of specifications and control is that ventilation
36、 ductwork in Sweden has substantially less leakage when compared to some other countries in Europe. . physical damage during inspection or maintenance work. In addition, in some cases, ducts were found to be completely disjointed. All of these airtightness deficiencies, along with other problems oft
37、en reported (dim systems, inad- equate design, absence of commissioning, poor maintenance, etc.), show the lack of attention paid to those systems. ASHRAE Transactions: Symposia 51 7 10.0 10.0 n 9.0 9.0 8.0 E g 8.0 2 7.0 2 7.0 0 6.0 $ 6.0 a 5.0 % 5.0 5 4.0 5 4.0 a a m 3.0 o 3.0 cn 3 1.0 flwible + pl
38、enum ai registem - i. r 2 a 2.0 + 2.0 3 1.0 Mass A 0.0 Class A 0.0 1 2 3 4 5 6 7 8 9 101112131415161718192021 1 2 3 4 5 6 7 8 9 101112131415161718192021 Duct system Dud system Figure 4 Ductwork tightness measurements done in Figure 5 Ductwork tightness measurements done in Belgium. France. Measuring
39、 the Tightness Belgium In the framework of the SAVE-DUCT project, the airtightness of 42 ductwork systems was measured in France (21) and Belgium (21). In Sweden, nearly all installations are leak-tested at commissioning as part of the AMA require- ments. Therefore, a randomly selected sample of 69
40、Swedish control measurements was collected, most of these represent- ing office building duct systems. The measurements and literature review performed within the SAVE-DUCT project suggest that duct systems are very leaky in Belgium and in France. Conversely, the instal- lation of high-quality airti
41、ght (Classes B, C, or better) systems prevails in Sweden and is used in residential, commercial, and industrial installations. One reason for this lies most likely in the fact that the need for tight systems has been identified in Sweden since the early 1960s. This has resulted in a series of qualit
42、y requirements now detailed in the AMA 98 guideline (1998). These requirements are binding when they are referred to in the contract between the owner and the contractor. AMA, with its long history, is today used in practically all building and installation contracts in Sweden. The results of the 2
43、1 systems measured are represented in Figure 4. The sample included 12 ductwork systems in non- residential buildings, 5 in multi-family buildings, and 4 in single-family houses. The majority of the systems have rather bad airtightness; only four systems fulfill the Class A require- ment and one sys
44、tem reaches Class B. A distinction is made between rectangular ductwork, circular ductwork, ductwork where a plenum is used for the connection at the registers, and concrete ductwork. In this sample, rectangular ductwork is on average about seven times leakier than circular ductwork. The positive ef
45、fect of the use of circular ductwork seems to be partly lost if the registers are connected to the ductwork with plenums (which were found to be very leaky). France Figure 5 gives an overview of the 21 French systems involved in the study and the results of the measurements. The sample included eigh
46、t ductwork systems in nonresidential buildings, nine in multifamily buildings, and four in single- family houses. In Belgium and France, the multi-point ductwork pres- surization method was used, i.e., the ductwork was pressur- ized at different pressure stations to calculate the leakage characteris
47、tics of the systems. The test was performed at pres- sures in the range of 50% to 150% of the operating pressure of the ductwork. Sweden The airtightness of most of the systems did not meet Class A; only one system reaches Class B. The systems airtightness in office buildings seems to be much better
48、 than that in multi- family buildings and Single-familY houses. In Sweden, the measurement procedures are normally performed at a test pressure of 400 Pa (1.6 in. of water) regard- less of the operating pressure of the duct system. A pressure range of 200 to 1000 Pa (0.8 to 4 in. of water) is, howev
49、er, allowed ifnecessary due to measurement equipment or size of tested duct system. Ducts with perimeters larger than 6.4 m (20 ft) shall be tested at 200 Pa (0.8 in. of water). The one- point measurement method with a flow exponent of 0.65 is used (as in ASHRAE 2001). As already mentioned before, the airtightness of new Swedish installations for air distribution has to be checked at commissioning. This means that many measurement data are available from Swedish installations. Therefore it was not necessary to perform additional measurements in the frame- work of the SAVE-DUC
