Printer Friendly

Flat oval duct leakage class measurement.


Unsealed duct leakage depends on the machinery used for its fabrication, material thickness, assembly methods employed, and workmanship during installation. Duct leakage tests by ASHRAE/SMACNA/TIMA (1985) and Swim and Griggs (1995) have verified that a power law relation can represent the duct leakage from longitudinal seams and transverse joints of assembled duct sections. It was also confirmed that for the same duct construction the behavior of duct leakage was almost the same under negative and positive pressure. These tests have suggested that for unsealed or unwelded ducts the joint leakage dominates, as the longitudinal seam leakage is about 10% to 15% of the total duct leakage.

There is a wide range of products and sealing methods available for ducts. A forecast of leakage class attainable by commonly used duct construction and sealing methods, based on the data obtained by Swim and Griggs (1995) and ASHRAE/SMACNA/TIMA (1985), is available in the Duct Design chapter of the ASHRAE Handbook (2009). These data do not account for the presence of fittings, a realistic spacing of rectangular and round duct transverse joints and duct-mounted components, such as access doors and balancing dampers. The leakage classes were calculated under the assumption of 0.82 joints per meter, or 25 joints per 100-ft of duct length. The studies cited above suggest that duct system leakage rates are primarily a function of the geometry of the joint and seams, the sealing used (if any), and the pressure difference between the inside and outside of the duct. Future leakage classes should be adjusted for typical rectangular and round transverse joint spacing, 1.2 m (4 ft) and 3.1 m (10 ft), respectively. Ducts possessing a lower proportion of joints would have less leakage in both the sealed and unsealed categories.

Many researchers have previously attempted to perform in-situ measurements of the effects of duct leakage on air distribution system performance and building envelope infiltration; representative studies are by Modera (1989), Yuill and Musser (1997), Proctor (1998), and Walker (1999). Most of these studies have tended to compare different sealing techniques. Performance studies by Xu et al. (2000, 2004) of air distribution systems in light and large commercial buildings have reported air leakage ratios from one-quarter to one-third of the fan-supplied airflow in constant-air-volume systems. The air leakage from the ducts - including supply and return, were reported in terms of the ASHRAE-defined leakage class with the reported values within a range much higher than the leakage classes predicted by the ASHRAE Handbook (2009) for unsealed ducts.

Leakage measurement studies by Aydin and Ozerdem (2006) on round and rectangular ducts for positive internal pressures were conducted along with a branched duct system having different duct diameters; the results were fitted to the power law model developed by J. Stratton from ETL data (ASHRAE/SMACNA/TIMA 1985) and confirmed by Swim and Griggs (1995). Apparently leakage data are unavailable in the literature for flat oval ducts. As a result the present project was initiated to study the leakage characteristics of spiral seam galvanized steel flat oval ducts.


Experiments were performed to estimate the leakage class of three cross sections of flat oval ducts. The tests conformed to requirements of SMACNA's HVAC Air Duct Leakage Test Manual (1985). The duct cross sections are listed in Table 1. In every instance the ducts were constructed from 24-ga galvanized sheet metal with spiral seams. The test section consisted of a total of six duct sections. Each duct section was 1.2 m (4ft) in length, and individual sections were connected by means of standard beaded slip couplings. Then 19 mm (3/4-in.) thick plywood caps secured with duct tape were used to seal the ends of the ducts (refer to Figure 1). Thereafter, the system was pressurized with a wet/dry shop vacuum connected to the ductwork by a combination of PVC and dryer hose tubing. The shop vacuum was capable of producing in excess of 2.5 kPa, equivalent to approximately 10-in water pressure or vacuum in the ductwork. A ball valve was used to regulate the pressure from the shop vacuum. The makeup air flow rate entering the enclosed system (which equaled the leakage rate) was determined by measuring the pressure drop across a calibrated Meriam Instruments laminar flow element (LFE), model 50MC2-2. Care was taken to ensure that sufficient straight runs of tubes were mounted at the entrance and exit of the LFE.

Table 1. Flat Oval Duct Cross Section and Sealing

Cross Section mm x   Seam   Joint Type    Joint Sealing   Surface Area
   mm (in.x in.)     Type                                  [m.sup.2]

356 x 152 (14 x 6)  Spiral  Beaded Slip  Sealed/Unsealed  6.48 (69.7)

381 x 102 (15 x 4)  Spiral  Beaded Slip  Sealed/Unsealed  6.42 (69.1)

559 x 152 (22 x 6)  Spiral  Beaded Slip  Sealed/Unsealed  9.45 (101.7)

Ambient pressure was measured with a Fortin-type barometer with an accuracy of [+ or -]0.25-mm (0.01-in.) of mercury. Pressure taps constructed from 6.4 mm (1/4 in.) diameter copper tubing were soldered onto the outer duct surface at the center of the second and fifth ducts in the test section. Flexible tubing was used to construct piezometer rings at both measurement locations. The piezometer rings were connected to a single electronic manometer using flexible tubing to measure the average static gage pressure in the test section. Similarly electronic manometers were used to measure the pressure drop across the LFE, and the gage pressure at the inlet to the LFE. Likewise the air temperature entering the LFE was measured using a type-K thermocouple with a measurement accuracy of [+ or -]0.25[degrees]C ([+ or -]0.5[degrees]F). The latter two measurements were performed to calculate air density entering the LFE using the ideal gas law. All gage pressure measurements were performed with a measurement accuracy of [+ or -]0.01 kPa ([+ or -]0.05 in. [H.sub.2]O).

Per Table 1, for a 'sealed' leakage test all duct joints were first sealed carefully with duct tape. For an 'unsealed' leakage test all but the center joint (between the third and fourth ducts comprising the test section) were sealed with duct tape. The spiral seams were not sealed in any manner, and there were no other duct wall penetrations, i.e., sheet metal screws were not used to assemble the duct sections. It was deemed that this would represent a worst-case leakage scenario. SMACNA (1985) notes that "helical (spiral) lock seams are exempt from sealing requirements". More rigorous sealing means such as mastic sealants or gaskets were not used on the joints, as it was required that the duct sections were to be re-used in other pressure loss tests. In performing the leakage tests the static pressure was increased gradually in steps, and thereafter the pressure was reduced by adjusting the control valve. The ducts were tested under both positive and negative pressures, which were not allowed to exceed [+ or -]1 kPa ([+ or -]4 in. wg.) in order to strictly limit any observable duct deformation.


By convention, the ASHRAE Handbook (2009) defines the leakage class 'CL' as the leakage rate in mL/s per [m.sup.2] of duct surface area, at a static gage pressure of 1 Pa. This implies the following

[C.sub.L] = [1000 Q/[DELTA][p.sub.s.sup.0.65]] (1 SI)

Similarly, the leakage class can be interpreted as the leakage rate in cfm per 100 [ft.sup.2] of duct surface area at 1in. wg. Hence

[C.sub.L] = [Q/[DELTA][p.sub.s.sup.0.65]] (1 I-P)

In Equation 1 the exponent N = 0.65 represents a mean value obtained by averaging extensive ETL data (ASHRAE/SMACNA/TIMA 1985). As defined by Equation 1, the leakage classes are dimensional quantities, and have different values depending on the units that are employed. However using straightforward units conversions, it can readily be shown that if N = 0.65 the classes are related as follows

[C.sub.[L,SI]] = 1.408 [C.sub.[L,IP]] (2)

Equation 2 is the basis for the leakage rate charts presented in the ASHRAE Handbook (2009).

In terms of the major and minor flat oval duct dimensions, 'A' and 'a', respectively, the duct perimeter length is given by Equation 3.

P = 2(A - a) + [pi]a (3)

Therefore the total duct surface area is

[A.sub.s] = PnL'/1000 (4 SI)

[A.sub.s] = PnL'/12 (4 I-P)

where 'n' is the number of duct sections in the test section, and L is the length of an individual duct section.

A laminar flow element is constructed by breaking up the overall flow passage into numerous small-diameter channels that are in parallel, typically by inserting a honeycomb in the device. Although the total air flow through the LFE may nominally be turbulent based on its Reynolds number, the flow through individual small channels is laminar. The volumetric flow rate through a laminar duct is a linear function of pressure drop. The LFE was calibrated to measure air flow rate '[]' (in units of scfm) at standard conditions, i.e., for [] = 29.92 in. Hg (760 mm) and [] = 70[degrees]F (21.1[degrees]C). Therefore the actual makeup air flow rate 'Qa' was calculated using Equation 5, based on the ideal gas law.

[Q.sub.a] = ([[]/[p.sub.a]])([[[T.sub.a] + 273]/[[] + 273]])[] (5 SI)

[Q.sub.a] = ([[]/[p.sub.a]])([[[T.sub.a] + 460]/[[] + 460]])[] (5 I-P)

Sealed and unsealed leakage tests (ASHRAE/SMACNA/TIMA 1985, Swim and Griggs 1995) have confirmed that duct leakage can be represented by Equation 6

[Q.sub.a] = C[([DELTA][p*.sub.s]).sup.N] (6)

where 'C is a constant reflecting the area characteristics of the leakage path, '[DELTA][p.sub.s]*' = ([DELTA][p.sub.s]/[DELTA][p.sub.ref]) is the normalized static pressure differential from the duct interior to the exterior, and 'N' is a constant exponent relating turbulent or laminar flow in the leakage path. It is straightforward to solve for the constant coefficients in Equation 6 by log-linearizing the data and fitting a least squares curve through the data, such that

log [Q.sub.a] = N log [DELTA][p*.sub.s] + log C (7)

The leakage rate at a normalized static gage pressure [DELTA][p*.sub.s] = 1 is obtained by substitution into Equation 7.

[Q.sub.a]([DELTA][p*.sub.s] = 1) = C (8)

Therefore the leakage class can be calculated by Equation 9.

[C.sub.L] = [C/[A.sub.s]] (9 SI)

The leakage rate per 100 [ft.sup.2] of duct surface is proportional to duct surface area. Hence the leakage class can also be expressed as follows.

[C.sub.L] = C([100/[A.sub.s]]) (9 I-P)


Results of leakage rate measurements are plotted in Figures 2 through 4 for different duct configurations as a function of static pressure difference. The data are displayed on log-log plots. A least squares power law curve fit was employed per Equation 7, and the duct leakage coefficients, C and N, were found. The linear correlation coefficient 'R' is shown with the least squares results. For the curve fit equations in Figures 2 through 4 the coefficients are in I-P units only; the corresponding SI values were calculated using suitable units conversions based on experimentally determined values of the power law exponent N. The dual unit coefficients are given in Tables 2 and 3 for sealed and unsealed ducts. In general, the power law exponent for a duct leakage test was greater in the positive pressure mode than the negative pressure mode.
Table 2. Flat Oval Duct Leakage Coefficients with All Joints Sealed

Nominal Duct Size [A x a]       Positive Pressure    Negative Pressure

                               C mL/s          N       C mL/s       N
                           ([ft.sup.3]/min)         ([ft.sup.3]/

356 x 152 mm (14 x 6 in.)    0.22 (0.169)    1.065  1.82 (0.160)  0.675
381 x 102 mm (15 x 4 in.)    3.53 (0.554)    0.780  6.73 (0.649)  0.692
559 x 152 mm (22 x 6 in.)    2.74 (0.921)    0.918  3.75 (2.187)  1.018

Table 3. Flat Oval Duct Leakage Coefficients with One Joint not Sealed

Nominal Duct Size [A x a]      Positive Pressure     Negative Pressure

                                C mL/s         N       C mL/s       N
                           ([ft.sup.3]/min)         ([ft.sup.3]/

356 x 152 mm (14 x 6 in.)    1927 (137.5)    0.637  6483 (171.1)  0.457
381 x 102 mm (15 x 4 in.)    2141 (253.3)    0.729  8617 (220.3)  0.451
559 x 152 mm (22 x 6 in.)    4134 (272.3)    0.623  9557 (226.0)  0.437

Table 3. Flat Oval Duct Leakage Classification with All Transverse
Joints Sealed


Nominal Duct Size [A x a]  Positive Pressure  Negative Pressure

                             [C.sub.L] mL/      [C.sub.L] mL/
                             ([s.m.sup.2])      ([s.m.sup.2])
                               (cfm/100           (cfm/100
                              [ft.sup.2])        [ft.sup.2])

356 x 152 mm (14 x 6 in.)    0.034 (0.242)      0.281 (0.229)

381 x 102 mm (15 x 4 in.)    0.550 (0.801)      1.048 (0.939)

559 x 152 mm (22 x 6 in.)    0.290 (0.906)      0.397 (2.150)

Table 4. Flat Oval Duct Leakage Classification with One Transverse
Joint Unsealed


Nominal Duct Size [A x a]  Positive Pressure  Negative Pressure

                               [C.sub.L]          [C.sub.L]
                           mL/([s.m.sup.2])   mL/([s.m.sup.2])
                               (cfm/100           (cfm/100
                              [ft.sup.2])        [ft.sup.2])

356 x 152 mm (14 x 6 in.)    297.4 (197.2)      1000.5 (245.5)

381 x 102 mm (15 x 4 in.)    333.5 (366.5)      1342.2 (318.7)

559 x 152 mm (22 x 6 in.)    437.5 (267.8)      1011.3 (222.2)




Duct leakage classifications based on Equation 9 for both positive and negative pressure systems were found and tabulated. In Tables 4 and 5, duct leakage class is given for configurations with all transverse joints sealed and one transverse joint unsealed. For tests performed with all joints sealed the leakage class was found to be between 0.034 to 1.048 mL/s per [m.sup.2] (0.229 to 2.150 cfm per 100 [ft.sup.2]). Likewise, the respective values for the unsealed configuration ranged between 297.4 to 1342.2 mL/s per [m.sup.2] (197.2 to 366.5 cfm per 100 [ft.sup.2]). From these results it should be noted that for a poorly sealed or unsealed system leakage can be significant. In-situ measurements performed by Xu et al. (2000, 2004) have shown that leakage classes for duct systems in large commercial buildings with poorly sealed ductwork range from 34 to 757 mL/s per [m.sup.2].

Best design practices per ASHRAE (2009) recommend that for both flat oval and round steel ducts the predicted leakage class at a static gage pressure of 1 Pa (1 in. wg) is 4 mL/s per [m.sup.2] (3 cfm/100 [ft.sup.2]). The results found herein indicate that leakage class for sealed flat oval duct is indeed comparable, if not better, than that predicted for sealed round steel ductwork. The flat oval duct types tested are significantly less leaky than geometrically similar rectangular duct for both positive and negative pressures with proper sealing techniques (ASHRAE 2009).


Leakage tests were conducted on three flat oval ducts listed in Table 1. Tests were performed with all transverse joints sealed with both positive and negative duct pressurization. Various sealing techniques were employed. For an unsealed condition, tests were performed with one transverse joint not sealed. The leakage behavior conformed to follow the power law leakage equation developed by Swim and Griggs (1995).

For sealed conditions, the values of the power law coefficient N ranged from 0.675 to 1.065. Likewise for the unsealed conditions the values ranged from 0.437 to 0.729. Generally for an individual duct tested the value of the coefficient found for negative pressurization was less than that for positive pressurization

The leakage class results, based on experimentally determined values of N, are summarized in Tables 4 and 5. For tests with sealed transverse joints, the leakage class for the flat oval duct type considered was found to be less than the predicted values given by ASHRAE (2009) for round or flat oval ducts. However, the predicted leakage class for an unsealed system given in the ASHRAE Handbook considerably underestimates the leakage found in this test program. A typical value of 42 mL/s per [m.sup.2] (30 cfm/100 [ft.sup.2]) is predicted by ASHRAE for unsealed conditions. Leakage class values obtained herein are up to an order of magnitude greater than those suggested in the ASHRAE Handbook, but are comparable to in-situ leakage measurements reported in Xu et al. (2000, 2004). Unsealed leakage class data measured in the present study would indicate that large gaps occurred at the unsealed joint, in the absence of any sheet metal screws.

Recommended duct installation practice includes the use of sheet metal screws to secure the joints. Care should be taken to install the screws no more than 152 mm (6 in.) apart, working from the minor axis side to the major axis, such that the metal is drawn together. Thereafter a sealant is further applied at each joint. Failure to do so results in predicted leakage classes for unsealed or poorly sealed systems that exceed those in the ASHRAE Handbook. Hence careful attention should be given to sealing techniques and construction methods, which in turn can significantly reduce duct leakage. It is recommended that all joints in flat oval duct systems should be properly sealed, as this yields very low leakage classifications.


A = major duct dimension, mm (in.)

a = minor duct dimension, mm (in.)

[A.sub.s] = duct surface area, [m.sup.2] ([ft.sup.2])

C = power law coefficient, mL/s ([ft.sup.3] /min)

[C.sub.L] = leakage class, mL/[s-m.sup.2] ([ft.sup.3]/min- 100 [ft.sup.2])

L' = length, m (ft)

N = power law exponent, dimensionless

n = number of duct sections

p = perimeter, mm (in.)

p = absolute pressure, mm Hg (in. Hg)

Q = flow rate, L/s ([ft.sup.3]/min)

R = linear correlation coefficient

T = temperature, [degrees]C ([degrees]F)

[DELTA][p.sub.s] = differential static pressure, Pa (in. wg)

[DELTA][p*.sub.s] = normalized static pressure, Pa (in. wg)

[DELTA][p.sub.ref] = reference static pressure difference. 249 Pa (1 in. wg)


a = actual

s = static

st = standard


ASHRAE/SMACNA/TIMA. 1985. Investigation of Duct Leakage. ASHRAE Research Project 308 (ETL Testing Laboratories, Cortland, NY; Report No. 459507).

ASHRAE, 2009. I-P & SI Handbook-Fundamentals, Chapter 21, "Duct Design". Atlanta: American Society of Heating. Refrigerating and Air-Conditioning Engineers, Inc.

Aydin, C. and B. Ozerdem. 2006. Air Leakage Measurement and Analysis in Duct

Systems. Energy and Buildings 38: 207-213.

Modera, M.P., 1989. Residential Duct System Leakage: Magnitude, Impacts and Potential for Reduction. ASHRAE Transactions 95(2): 561-569.

Proctor. J.P. 1998. Verification Test of ASHRAE Standard 152P. ASHRAE Transactions 104(1B): 1402-1410.

SMACNA. 1985. HVAC Air Duct Leakage Test Manual. Sheet Metal and Air Conditioning Contractors National Association, Chantilly, VA.

Swim, W.B. and E.I. Griggs. 1995. Duct Leakage Measurement and Analysis. ASHRAE Transactions 101(1): 274-291.

Walker. I.S. 1999. Distribution System Leakage Impacts on Apartment Building Ventilation Rates. ASHRAE Transactions 105(1): 943-950.

Xu. T.T., R.F. Carrie, D.J. Dickerhoff, W.J. Fisk, J. McWilliams, D. Wang and M.P Modera. 2004. Performance of Thermal Distribution Systems in Large Commercial Buildings. LBNL-44331, Lawrence Berkeley National Laboratory, Berkeley, California.

Xu, T.T., M.P. Modera and R.F. Carrie. 2000. Performance Diagnostics of Thermal Distribution Systems in Light Commercial Buildings. LBNL-45080, Lawrence Berkeley National Laboratory, Berkeley, California.

Yuill, G.K. and A. Musser. 1997. Evaluation of Residential Duct-Sealing Effectiveness. ASHRAE Transactions 103(2): 264-271.

D.C. Gibbs

Associate Member ASHRAE

S. Idem, PhD


D.C. Gibbs is a mechanical engineer with BWSC, Inc., in Nashville, TN. S. Idem is a professor in the Department of Mechanical Engineering at Tennessee Tech University, Cookeville, TN.
COPYRIGHT 2010 American Society of Heating, Refrigerating, and Air-Conditioning Engineers, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2010 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Gibbs, D.C.; Idem, S.
Publication:ASHRAE Transactions
Article Type:Report
Geographic Code:1USA
Date:Jul 1, 2010
Previous Article:Measured and predicted pressure loss in corrugated spiral duct.
Next Article:The economics of energy savings performance contracts.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters