Effect of mastic on duct tightness in energy-efficient manufactured homes.

INTRODUCTION

Over the last 18 years, the Northwest manufactured housing industry and the region's utilities have maintained a partnership aimed at developing and marketing energy-efficient manufactured homes throughout the Pacific Northwest. The Northwest Energy Efficient Manufactured Home (NEEM) program is a consortium of state agencies, utilities, and manufacturers that has evolved from this relationship. Over 55 utilities offer rebates on NEEM homes. Approximately 60% of all manufactured homes built currently in the Northwest meet NEEM program standards and are among the most energy-efficient manufactured homes in the United States. Quality control processes were developed and became an integral part of the program, to ensure that the homes meet NEEM, Building America Industrialized Housing Partnership (BAIHP) and EnergyStar guidelines. The US Department of Energy BAIHP program also provides technical support to the NEEM consortium and, along with a DOE State Energy Program (SEP) Special Projects Building America grant, provided funding support for this study.

As part of the quality control process, field studies of a random sample of homes were conducted on homes manufactured in 1992-93, 1997-98, and 2001-02. Observations during these field tests led the NEEM program to suspect that a significant amount of reported duct leakage was due to failure of various duct sealing tapes. Starting January 1, 2004, NEEM specifications were revised to require all central forced air duct systems to use UL 181 AM or BM rated duct mastic for all sealing (the only exception is tape on duct board assemblies). To evaluate the efficacy of mastic used to seal ducts combined with in-plant duct leakage testing in the program, field data were collected on 71 homes built after January 1, 2004, with duct systems sealed with mastic, and the data were compared to data from previous studies in the region.

STUDY DESIGN

The sample selection was one of convenience and focused on manufacturers who were also known to be duct testing at least some of their homes at the factory separate from the quarterly testing of the third-party inspectors. A separate third-party inspector also tests ducts in the 18 participating plants on a quarterly basis. Ten of the region's 18 factories (including all of the major builders) were represented in the sample. Testing on 30 homes sited in Washington was completed between March 2005 and August 2005. An additional 41 homes sited in Oregon were also tested to a similar protocol from September 2004 through March 2006. Homes tested were all sited, set up, and either occupied or ready for occupancy.

Homes were visually inspected to confirm the use of mastic and identify obvious deficiencies. Tests were performed to determine total duct leakage, duct leakage to the exterior, envelope leakage, and airflow through the systems. Airflows were determined with either a TrueFlow[TM] airhandler flow meter or duct tester pressure matching, depending on system configuration.

Prior to summarizing the results of the duct leakage tests and the blower door tests, cases that were suspect as a result of exponent values out of range were removed. A total of three blower door tests and five duct leakage tests were removed as a result of this screening.

FIELD DATA

Duct Leakage Results

All homes tested used mastic to seal the sheet metal ductwork. Table 1a shows previously reported exterior duct leakage in Super Good Cents (SGC) and Manufactured Acquisition Program (MAP) homes in the Northwest. As reported (Davis and Baylon 2004) only about 12% of the tested SGC homes built in 2001-02 used mastic to seal ducts. Results are presented with both medians and means to preserve the previous reports' approach; in 1997-98 homes, outliers greatly skewed the mean.

Table 1a. Exterior Duct Leakage (Earlier NW Manufactured Home Studies)

                       SGC Mfd Homes Built 2001-2002
                            Medians (Avgs)

Group

                       Leakage @ 25 Pa          Leakage@ 50 Pa
                      ([ft.sup.3]/min)         ([ft.sup.3]/min)

All cases                 131 (139)               192 (209)
                          n = 94

Double-section            119 (132)               180 (199)
Home                       n = 69

Triple-section            176 (174)               259 (265)
Home                       n = 22

Idaho                     127 (151)               187 (229)
                           n = 20

Oregon                    135 (134)               200 (198)
                           n = 37

Washington                115 (132)               179 (202)
                           n = 39

                       SGC Mfd Homes Built 1997-98
                            Medians (Avgs)

Group

                      Leakage @ 25 Pa        Leakage @ 50 Pa
                      ([ft.sup.3]/min)         ([ft.sup.3]/min)

All cases                 103 (151)               159 (231)
                           n = 47

Double-section             97 (157)               157 (240)
Home                       n = 34

Triple-section            144 (134)               223 (210)
home                       n = 13

Idaho                     106 (165)               168 (254)
                           n = 24

Oregon                       NA                      NA

Washington                103 (135)               159 (208)
                           n = 25

                      MAP 1992-93 (Avgs except for
                                triples)

Group

                      Leakage @ 25 Pa         Leakage @ 50 Pa
                      ([ft.sup.3]/min)        ([ft.sup.3]/min)

All cases                     (104)               (157)

Double-section                (101)               (155)
home                         n = 124

Triple-section                 122                 169
home                         n = 11

Idaho                          --                   --

Oregon                         --                   --

Washington                     --                   --

Table 1b contains exterior duct leakage values for the homes in the current study. In general, there appears to be about a 60% reduction in exterior duct leakage for the overall group compared to the 2001-02 study and an overall 43% reduction compared to the previously best reported values from the 1992-93 MAP study.

Table 1b. Exterior Duct Leakage (Mastic Group)

                    SGC Mfd Homes Built after
                      January 1, 2004 Medians
                              (Avgs)

Group            Leakage @ 25 Pa    Leakage@ 50 Pa
                 ([ft.sup.3]/min)  ([ft.sup.3]/min)

All cases              51 (56)           73 (80)
(Washington and        n = 66
Oregon)

Single-section        53.0 (53.0)       85.0 (85.0)
home                    n = 2

Double-section         49 (42)           64 (71)
home                    n = 41

Triple-section         62 (65)           88 (82)
home                    n = 21

Air-handler flows in the current study averaged 1145 cfm. Duct leakage to the exterior at 25 Pa was normalized to system flow. The average leakage to the exterior at 25 Pa/[ft.sup.3] per minute was 5.0%.

Duct leakage to the exterior at 25 Pa was normalized to the home's conditioned floor area. In this measure, the leakage was equal to about 3.4% of the floor area over the entire sample. The result, in Table 2, is compared to previous studies and shows an improvement of more than 50% in the mean percentage.

Table 2. Leakage to Exterior Normalized by Conditioned Floor Area

Study                                                Mean  Median
                                                       %       %

Exterior duct leak @ 25 Pa/[ft.sup.2] of house area   3.4     3.0
(Built after 1/1/2004, 66 homes)

Exterior duct leak @25 Pa/[ft.sup.2] of house area    7.9     7.5
(2001-02 homes, 89 cases)

Exterior duct leak @25 Pa/[ft.sup.2] of house area            5.9
(1997-98 homes, 49 cases)

Exterior duct leak @25 Pa/[ft.sup.2] of house area    7.2
(1992-93 homes, 150 cases)

Table 3 shows the total duct leakage as tested at the factory compared to the leakage to the exterior as tested after set up in the field for the subset of sites where data for both were available. The impact of in-plant testing is to improve the field-observed exterior duct leakage by about 30%. Even with this small sample size, this difference is statistically significant.

Table 3. Total Leakage at Factory Compared to Leakage to Exterior on
Site

          Total Leakage at Factory @ 50  Leakage to Exterior on Site @
               Pa ([ft.sup.3]/min)           50 Pa ([ft.sup.3]/min)

In-Plant               126.0                           70.0
Test
(n=37)

No                      n/a                            92.8
In-Plant
Test
(n=29)

House Tightness

Northwest manufactured homes have gotten tighter over the past 14 years as can be seen in the blower door results summarized for previous studies in Table 4a. As seen in Table 4b, this trend has continued with the current study. In the current study, the minimum [ACH.sub.50] is 1.31 and maximum [ACH.sub.50] is 8.94. Only 11 cases out of 68 have [ACH.sub.50] over 5.0. The nominal program standard was reduced from 7.0 [ACH.sub.50] to 5.0 [ACH.sub.50] effective January 1, 2004. The standard deviations in most categories are very similar to the 2000 and 2004 studies. These studies show less scatter than the original MAP results, which should be viewed as an indicator of successful quality control. The two single-wide units tested have high leakage rates for both their duct systems and envelopes.

Table 4a. Blower Door Results ([ACH.sub.50]) (Earlier NW Manufactured
Home Studies)

            SGC Mfd  Homes 2001-02        SGC Mfd Homes 1997-98

Group       # of   [ACH.sub.50]  Std.  # of   [ACH.sub.50]  Std.
            Cases    Average     Dev.  Cases    Average     Dev.

All          93 *          4.16  1.02     49          4.76  0.95

Double         66          4.30  1.03     36          4.90  0.99
wide

Triple         24          3.84  0.94     13          4.40  0.72
wide

Idaho          19          4.59  0.96     25          4.63  0.81

Oregon         33          4.36  1.13    N/A           N/A   N/A

Washington     41          3.89  0.89     24          4.90  1.08

                 MAP 1992-93

Group       # of   [ACH.sub.50]   Std.
            Cases    Average      Dev.

All         157      5.50         1.87

Double      127      5.50         1.90
wide

Triple       12      4.92         1.22
wide

Idaho        32      6.12         1.55

Oregon       48      5.43         2.10

Washington   62      5.36         1.77

Table 4b. Blower Door Results (ACH50) Mastic Group

               SGC Mfd Homes Built after January 1, 2004

Group        # of Cases  [ACH.sub.50] Average  Std. Dev.

All              68              3.82            1.42
Single wide       2              6.14            1.35
Double wide      44              3.88            1.46
Triple wide      20              3.42            1.18

DISCUSSION

The selection of homes in this study was not random, but it did include 10 of the region's 18 manufacturers and all of the major manufacturers in the NEEM program. Based on this limited sample, indications are that the revision to the specifications starting in January 2004 requiring the use of mastic to seal duct systems has produced a significant improvement in duct tightness over all previous samples in the region. Duct leakage to the exterior after set up was reduced by 43% over the next best reported value in the region.

The comparison between the homes that received in-plant duct testing and those that did not showed a distinct improvement in overall performance with an in-plant quality control step. Indeed, about half of the benefit from the change in specifications and the use of duct mastic seem to be attributable to the in-plant testing. This study suggests that in-plant testing is essential to achieving the benefits of the improved duct tightness and installation specifications.

The [ACH.sub.50] of the homes averaged 4 and was well below the revised NEEM program standard of 5 [ACH.sub.50]. The significant reduction in duct leakage to exterior contributed to the reduced [ACH.sub.50].

Both duct tightness and overall house tightness for this group of manufactured homes was significantly better than values reported for site-built homes in Washington State. As reported (Hales et al. 2003), 29 site-built homes constructed after 1995 with ducts outside conditioned space averaged 7.1 [ACH.sub.50] and with average duct leakage to the exterior of 406 [ft.sup.3]/min at 50 Pa.

Leakage to the exterior tested after setup averaged 74% of total duct leakage as tested in the factory. This represents a significant variance in the region to the suggested 50% found in the referenced EPA manual. The 50% EPA assumption may be overly optimistic and result in field tests indicating leakier ducts than would be expected from in-plant tests using the EPA assumptions.

CONCLUSIONS

The use of mastic to seal ducts in new manufactured homes produced a 43% reduction in duct leakage to the exterior based on field tests of 71 NEEM homes. This is a significant improvement over previously reported duct leakage for manufactured homes in the Northwest that primarily used duct tape for sealing. Duct testing homes at the factory as a quality control measure was associated with a roughly 30% increase in tightness within the group of mastic-sealed homes. The combination of mastic for sealing and in-plant duct testing for quality control together promise to optimize duct tightness.

ACKNOWLEDGEMENTS

This work is sponsored, in large part, by the US Department of Energy (DOE), Office of Energy Efficiency and Renewable Energy, Building Technologies Program under cooperative agreement number DE-FC26-99GO10478 and DOE SEP Special Projects Building America grant number DE-FG26-04R021589. We appreciate the support and encouragement of our program managers, Mr. George James, Mr. Ed Pollock, and Mr. William Haslebacher. This support does not constitute DOE endorsement of the views expressed in this paper. The authors would also like to thank Tom Hewes of Oregon Department of Energy and David Baylon of Ecotope for their contributions and comments on this work. Finally, the Oregon field work was conducted in large part by Al Rust of the Oregon Department of Energy and the authors would like to thank him for his efforts and his attention to detail.

REFERENCES

Davis, B., and D. Baylon. 2004. Summary of SGC manufactured home field data (2001-02). Prepared for Northwest Energy Efficient Manufactured Homes, Northwest Energy Efficiency Alliance, Idaho Department of Water Resources Energy Division.

Hales, D., A. Gordon and M. Lubliner. 2003. Duct leakage in Washington State residences: Findings and conclusions. ASHRAE Transactions 109(2):393-402.

BIBLIOGRAPHY 

Baylon, D., B. Davis, and L. Palmiter. 1995. Manufactured home acquisition program: Analysis of program impacts. Prepared for Bonneville Power Administration under Contract No. DE-AM79-91BP13330, Task Order #71945.

Chasar, D, N. Moyer, J. McIlvaine, D. Beal, and S. Chandra. Energy Star manufactured homes: The plant certification process. Proceedings of ACEEE Summer Study, American Council for an Energy Efficient Economy. Washington, DC: American Council for an Energy Efficient Economy.

Davis, B, A. Roberts, and D. Baylon. 2000. Summary of SGC manufactured home field data (1997-98 sitings in Idaho and Washington). Prepared for Idaho Department of Water Resources-Energy Division.

Energy Star Labeled Manufactured Homes: Design, Manufacturing, Installation, and Certification Procedures, EPA 430-B-04-005, p. C.3. US Environmental Protection Agency.

Palmiter, L., T. Bond, I. Brown, and D. Baylon. 1992. Measured infiltration and ventilation in manufactured homes. Prepared for Bonneville Power Administration under Contract No. DE-AM79-91BP13330.

DISCUSSION

Matthew Friedlander, Vice President of Engineering, Renewaire LLC, Madison, WI: Please clarify why the factory-tested duct systems displayed better leakage rates?

David Hales: Our assumption was that the factory-tested duct systems had better leakage rates because if a system failed the duct test at the factory it was repaired until it met the test specification before it was shipped. Testing each system at the factory increases the probability that the systems actually meet the specifications when the units are shipped. When only a sample of systems are tested at the factory, some failures will be missed, and those systems will leave the factory with excessive leakage.

David Hales is a building science and energy specialist with the Washington State University Extension Energy Program, Spokane, WA. Bob Davis is an associate at Ecotope, Portland, OR. R. Brady Peeks is an energy analyst with the Oregon Department of Energy, Salem, OR.