Reference
Publication: McIlvaine, Janet,
David Beal, Neil Moyer, Dave Chasar, Subrato Chandra.
Achieving Airtight Ducts in Manufactured Housing. Report
No. FSEC-CR-1323-03.
Disclaimer:
This report was prepared as an account of work sponsored by
an agency of the United States government. Neither the United
States government nor any agency thereof, nor any of their
employees, makes any warranty, express or implied, or assumes
any legal liability or responsibility for the accuracy, completeness,
or usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately
owned rights. Reference herein to any specific commercial
product, process, or service by trade name, trademark, manufacturer,
or otherwise does not necessarily constitute or imply its
endorsement, recommendation, or favoring by the United States
government or any agency thereof. The views and opinions of
authors expressed herein do not necessarily state or reflect
those of the United States government or any agency thereof.
Many
of the manufacturers working with BAIHP are striving to meet
the quantitative goals set by the EPA Energy Star Program
for Manufactured Homes. The program provides compliance paths
for homes with three levels of duct leakage to the outside
Qnout = 3%, 5%, or 7% (MHRA, 2001.)
Among
the earliest BAIHP data, four houses built by the same manufacture
in 1997 exemplify the achievability of duct tightness in the
manufactured housing setting. Two were standard homes used
as control homes for comparison to an “energy improved”
model and a “health improved” model (Chandra,
et. al., 1998.) Standard manufacturing methods were changed
to mastic and the 3% Qnout leakage target was easily
met (Table 4).
Table
4 Demonstration of Duct Tightness Achievability
Standard Production compared to 2 Improved Models 1997
Area
CFM25total
CFM25out
Qntotal
QnOut
Control
Home
1600
118
84
7%
5%
Control
Home
1280
126
89
10%
7%
Energy
Home
1494
51
25
3%
2%
Healthy
Home
1920
79
36
4%
2%
Repair
Improvements to Taped and Mastic
Of the 190 floors in this data set, 9 test results show improvements
from repairs made to initially leaky systems (Table 5). This
is a very persuasive example for factory personnel when conducted
on a freshly produced duct system on the factory floor. This
also shows the iterative problem solving process that BAIHP
researchers use with factory staff to foster improvement in
performance and production procedures.
For
a description of improvements implemented at a single manufactured
housing plant, see Appendix A.
Tightness
of all 9 systems was improved by BAIHP repair (Table 5). Maximum
improvement was a reduction of Qntotal from 11.7%
to 5.5%. The least improved system was tightened from Qntotal
of 8.1% to 7.1%. Qntotal was reduced an average
of 3.5, representing a 43% improvement in duct tightness.
Table
5 Duct Tightness Improvement Measurements.
House
ID#
Duct
Assembly Method
Before
Repair
Qntotal
After
Repair*
Qntotal
AL5R
Tape
11.7%
5.5%
68
Tape
11%
4.9%
84
Tape
8.0%
4.8%
85
Tape
8.1%
7.1%
88
Tape
4.5%
3.3%
14
Mastic
6.4%
2.6%
47
Mastic
6.5%
4.4%
51A
Mastic
8.0%
2.6%
52A
Mastic
6.0%
3.6%
Average
7.8%
4.3%
*All
system repairs made with mastic, not tape.
The
duct systems in Houses 14, 47, 51A, and 52A were assembled
with mastic. However, the production was marred by other problems
such as inaccurate cutting, inaccessible joints, and misalignment
of components. These problems occurred in other mastic sealed
systems which also failed to make the target Qntotal
(see Duct Tightness Data).
Unit
14 illustrates BAIHP’s iterative approach to working
with factories. Unit 14 was a section of a double wide manufactured
house containing an air handler for a split heating and cooling
system. Initially, it showed duct leakage of 10.6% (Qntotal)
on the air handler side. Researchers removed the air handler
fan and connected the duct testing equipment directly to the
duct system to quantify how much of the leak was associated
with the unit itself. This dropped the CFM25total
to 33cfm bringing the Qntotal to 6.4%, still too
high to meet the tightness target of less than 6%.
Upon
investigation, researchers and factory staff found holes in
the main trunk line that were cut much larger than the connecting
floor boots, too large to seal with standard mastic practice
in use on the production line. The holes had been left open
with no attempt to seal them. Researchers used fiberglass
mesh and mastic to build up patches over the holes. CFM25total
dropped another 29cfm bringing the Qntotal to 2.6%,
well within the target zone.
Similar
holes were found on the non-air handler side of the double
wide suggesting that this was a systemic production problem.
Improved cutting using a template to match supply boot dimensions
was recommended.
A
subsequent factory visit a month later found only one out
of nine tested floors did not meet the Qntotal
= 6%. Qntotal measurements fell from the original
10.6% to: 2.3%, 2.7%, 3.9%, 3.6%, 4.4%, 4.5%, 4.5%, 5.8%,
and 6.4%.
This
result is exactly the desired effect of the BAIHP approach:
Introduce
Building America Industrialized Housing Partnership and
systems engineering
Build collaborative relationship with industry partners
to solve problems
Set goals with factory managers
Quantify
duct tightness resulting from existing practices
Identify opportunities for improvement
Report findings and make recommendations
Assist with implementation when needed (e.g. training, problem
solving for production process)
Reevaluate to assess progress until goals are reached
Implement factory quality control procedures, including
pressure testing, to sustain success
Commonly
Implemented Improvements
BAIHP researchers have observed that certain aspects of duct
system fabrication must be addressed, in addition to the change
from tape to mastic, in order to consistently deliver a substantially
leak free duct system. Data compiled here show that manufacturers
have consistently achieved duct tightness goals by implementing
correct mastic application, improved cutting precision, better
dimensional coordination between duct system components as
well as between ducts and the house. The following list represents
steps commonly taken by manufacturers to achieve their duct
tightness goals:
All systems
Train supervisors and line workers on air flow concepts
Systemize the duct assembly process
Use circle cutters and templates for standard duct cutouts
Use mastic in a form that fits with the production process
(tubes/buckets)
Seal joints with a “pinky” size bead of mastic
Seal the duct to the house air barrier (ceiling drywall
or floor deck)
Seal joints in return and supply plenums
Institute quality control measures, such as pressure testing
ducts during production, in addition to visual inspection
Floor
System (Typically Sheet Metal) Improvements:
Replace
tape with mastic and, when needed, fiberglass mesh
Ensure the supply plenum is well attached and thoroughly
sealed to the trunk duct with mastic
Select trunk duct dimensions that allow for slight misalignment
of floor register cutouts without creating a failed joint.
Improve alignment of floor register cutouts with trunk and
branch ducts
Provide adequate attachment area (tabs on all components
should be fully bent)
Use templates to cut holes in trunk ducts for floor risers
Fully bend tabs on floor risers
Ceiling System (Typically Duct Board and Flex Duct
Improvements)
Use circle cutter to cut hole in duct board for flex duct
collar
Provide
adequate attachment area around collar cut outs to allow
all collar tabs to be fully bent.
Lay generous bead of mastic (about the size of smallest
finger) into corner where the flange meets the main body
of the collar.
Press
collar into duct board hole. Mastic should spread.
Spread mastic to completely cover collar flange and beyond
edge of collar flange about ½”.
Slide inner line of flex duct onto collar past the positioning
ridge and secure with a plastic zip tie trimmed with a tensioning
tool.
Slide outer lining up as close to the collar flange as possible
and secure with a plastic zip tie trimmed with a tensioning
tool.
Remove excess flex duct, this can result in kinked supply
runs.