Building America Industrialized Housing Partnership
TRIP REPORT
Clayton Homes Plant, Waycross GA
Plant visit and testing

DATE
July 20, 2001

ATTENDEES

David Beal, Neil Moyer – FSEC
George Perkins-General Manager, Wendall Morgan-Service Manager, Joel Beasley-Production Manager, Darlene Wright-Quality Control Manager, and Les Rodgers-Material Manager

TRIP REPORT DISTRIBUTION LIST

George James, Keith Bennett – DOE
Subrato Chandra, Neil Moyer, David Beal, Dave Chasar, Janet McIlvaine– FSEC
George Perkins, Tom Rehrig – Clayton Homes

INTRODUCTION

Clayton Homes, a new BAIHP partner, is currently looking at improving their duct systems and considering providing an EnergyStar product. Leakage from duct systems is recognized nationally as a major cause of energy waste, poor indoor air quality, and poor durability in site-built and manufactured housing. Supply duct leakage and inadequate return air pathways can cause significant negative pressures within the building. Negative pressures pull in outside air (which if located in the hot, humid South) that can result in severe moisture damage to building envelope assemblies. Recognizing this, BAIHP partner Clayton Homes in Waycross, GA requested the Florida Solar Energy Center (FSEC) personnel to inspect their plant, concentrating on the duct systems, and recommend improvements to the systems.

PURPOSE

This was the first visit to examine the various factory assembly processes of the Waycross GA, Clayton Homes’ plant. The Waycross plant currently produces about six floors per day and employs approximately 125 people. The primary focus was on the duct system construction processes and assemblies that might affect building performance and energy efficiency. This facility manufactures both single- and double-wide housing units with a sheet metal ductwork within the floor system or an overhead duct system made with fiberglass ductboard and flex-duct.

DESCRIPTION & FACTORY TOUR OBSERVATIONS

Duct system - The metal in-line duct is made in a continuous piece, the ends are sealed by cutting and folding the end of the duct and holding it closed with foil tape. (See figures at end of report).

This process of sealing one end requires approximately 12 feet of foil tape, 2 feet of sheet metal duct and 5-8 minutes of labor to complete.

The floor registers are placed directly on top of the supply duct. After the floor is decked, holes are cut in the floor (using a circular saw) for the riser installation. The metal trunk line is cut freehand on three sides with a utility knife and folded over. The riser is then placed in the hole, tabbed over and sealed with a foil tape. The tape is pressed into place by hand. When the air handler is installed at the factory, typical with a floor system, a plenum is installed. This is accomplished by cutting a hole in the deck and in the duct. The air handler plenum is installed into the duct using tabs and foil tape. It is screwed into the floor decking to hold it in place. (This was not observed at the time of visit, but is how it was described to us.)

As currently done, the floor metal duct assembly needs improvement. The end cap construction is somewhat time consuming, and utilizes a fair amount of metal duct and foil tape. The alignment of the metal trunk line is not very accurate, often resulting in the duct to be misaligned by as much as a couple of inches (See figures at end of report). The duct is aligned by measuring from the edge of the floor to the edge of the duct and then held in place with strapping. The strapping allows sideways movement of the duct. This makes the installation of the plenum and risers difficult at best. As the metal isn’t cleaned before taping, the tape adhesion will probably fail, allowing the over sized holes to become a leak path.

The attic supply ductwork is composed of flex duct and fiberglass ductboard. (See figures at end of report). Collars are inserted into a hole made in the boxes created with a utility knife in a freehand fashion. There is a hole cutting tool on hand, but not used for whatever reason. The collars are held in place with tabs (tape and/or mastic are not used). The supply boots are taped together and held in place with a foam sealant. The tape that is used is manufactured by the Ideal Tape Company, and is a UL-181A-B pressure sensitive foil tape. The tape had a date stamp of 10/00, which means the tape is 9 months old. According to the manufacturer, this tape has a useful shelf life of 12 months – when stored in a cool, dry place (60-80degF and 50% relative humidity). The flex duct is attached to the boot collar using two collar ties, on both inner liner and then on the outer liner. The factory personal use a tensioning tool to attach the straps.

There are four areas of worth noting concerning the attic flex systems:

•    The length of the flex duct between connection points is not being cut to proper length. Many of flex ducts observed were excessive in length. This lends itself to unnecessary bending and kinking which restricts airflow. Also, the extra length of duct also reduces the airflow through the system.

•    The extra length of flex duct displaces the area for the attic insulation – thus reducing the overall attic insulation levels.

•    The insulation and outer liner of the flex duct do not always adequately cover the flex duct collars. This may lead to condensation on the exposed metal collars in the hot, humid attic during air conditioning usage.

•    One of the flex duct-to-collar connections had the inner tie strap misaligned. This resulted in reducing the duct diameter by nearly 50%.

Although assembly of various fiberglass boxes were not observed, ceiling ducts could be made more air tight if mastic was used to assemble the systems. The system is assembled with no sealing of the collars where they go into the fiberglass duct board boxes. Sealing this junction with mastic will improve the air tightness of the system and increase the strength of the duct system. This is most important on the riser ducts (from crawl space to attic) supplying the system. These are large ducts, and subjected to high pressure, as they are closest to the air handler fan.

Return Air Pathways - The current practice for sizing of the return air pathways from the bedrooms and through utility doorways follows the HUD code procedure (Manufactured Home Construction and Safety Standards 3280.715(b)(4)). This results in the undersizing of return air pathways for two reasons:

•    Bathroom areas with bedrooms or other closed off areas are not included,

•    The airflow calculations are based on heating system flows, not cooling system.

Provisions need to be made to address these issues by increasing return air pathways. This can be done by increasing current pathways and/or adding addition pathways. The utility room acting as a return plenum is of concern. When the furnace or air handler unit pulls it’s air from the utility room, and that room can be closed off by a door, then the room acts as a return plenum. This will cause the room to go into a negative pressure with reference to outside (as well as the rest of the house). Return air will come from the house, but also from all of the other holes, cracks and openings that exist in that room. A better system would not use the utility room as a return plenum, or make the room extremely “leaky” to the rest of house to preclude any pressure difference between the utility room and the rest of the house.

The air handler or furnace blower speed used for cooling may also be a way to reduce airflow into various rooms. Typically, the cooling tap of 3 and 4 speed blowers is set on the high speed. The heat gain calculations done on the home should provide not only the model of the unit needed, but also the flow of air needed to each zone. The blower speed selection should match this calculation. (Note: as an additional benefit, when the blower speed is reduced, dehumidification increases.)

House Ventilation -All of the units inspected used positive pressure system. This is a 5 inch flex duct connected to the top of the air handler and pulls air from the exterior every time that blower operates.

Envelope - The belly board is attached as a complete assembly – no tears or holes. As the floor migrates through the factory, intentional and unintentional holes and tears are introduced. The standard for patching is with belly tape and staples. Pervious experience that the authors have had with this type of closure system has shown that it tends to fail over time. This exposes the floor cavity to the crawlspace environment. A more permanent solution needs to be found for factory and field applications.

•    Marriage line walls appear to have various irregularities that would prevent a tight closure. (See figures at end of report). Three areas were noted that would cause this:

•    The bottom plate of the end wall juts out past the floor system.

•    The roof assembly and top of end wall are not in line. A spacer has been placed on the end wall to partly compensate.

The trimming of the floor decking was out of line with the end plate. This was probably due to staples or other debris that caused the trimming router to move out of line.

Insulation of the floor assembly is done with unfaced fiberglass batt material. The walls are insulated with fiberglass batts (some in plastic wrap and some unfaced – see sidebar). It appears that the insulation is being applied correctly. The attic insulation is a mineral wool loose blown application. The installation of insulation around the attic flex duct systems is generally of concern. It is possible for large areas of ceiling to be missed or not completely insulated because of the ductwork being in the way.

TESTING

While at the plant, FSEC personnel performed some simple building airtightness and duct testing on a recently built single-wide home (Model: CU16763A, Serial NO: WHC011643GAA). (See figures at end of report). This home had an in-line metal floor duct system with a furnace.

Blower door test - A blower door test was done to determine the airtightness of the building envelope. As a general rule of thumb, typical values of building airtightness range between 0.75 and 1 CFM50 per square foot of conditioned floor area. This house has a floor area of 1216 ft² and a CFM50 of 860 or 0.7 CFM50 per square foot – a tighter home.

Duct system test - A duct system airtightness test was also completed. A duct tester was attached to the air handler unit. The supply registers were temporarily sealed off and the system was then depressurized to 25 pascals. The total and outside leakage flow components were measured. An airtight duct system would have zero leakage or both the CFM25_total and CFM25_out would be 0. Generally, acceptable values are 6% of floor area for CFM25_total and 3% for CFM25_out.

In this case, it appears that the leakage to the outside is very small and in fact passes the acceptable criteria. However, this can be misleading. The reason for the very low leakage to outside is due to the tightness of the belly board. A 3 pascal pressure difference was measured between the floor cavity and the house when the house was depressurized to 50 pascals with the blower door assembly. This means that the floor cavity can be considered within the house (as designed). Problems will occurs if and when the belly board integrity is compromised. This may result from transport, setup and/or damage at some future time. When the belly board is compromised the duct leakage to the exterior will likely approach the total leakage – which is unacceptable.

Inspection of the duct system revealed a large hole in the duct where a riser had been installed, caused by the duct tearing when the flap of duct was folded back (See figures at end of report). This was also observed in the floor area during the plant inspection. A better solution would be to cut all four sides of the duct, using a template to guide the cutting operation, or use the tabless riser shown to plant personnel. The tabless riser is screwed down to the trunk duct, and the hole is then cut out, using the riser as a guide.

A double-wide unit was also going to be tested. This proved impractical, as the crossover collars had not been cutout in the belly board. An inspection by the factory personal revealed that the crossover collar had not been installed or perhaps was installed in the wrong location. Conversations with the Service Manager revealed that this has been a problem that he has been dealing with during setup.

The double-wide unit used the utility room as a plenum. This room will experience negative pressures with reference to the outside any time the air handler/furnace operates and the door to the utility room is closed. (The problem will be exacerbated when the dryer operates.) Air will pass through the door undercut and grill as designed, however it will also enter around the ceiling light fixture, plumbing penetrations and electrical penetrations as the grill and door undercut do not provide adequate return air pathways. The electrical service panel will be a prime candidate for air entry. It is possible that condensation may be severe enough to cause rusting of metal parts and even water running down the wall.

RECOMMENDATIONS

Duct systems - Improve construction and sealing techniques. Avoid excessive runs of flex duct. Consider use of a end cap for metal systems. Consider the use of mastic where possible. Strive to reach a duct sealing standard that is 3% of the floor area of less in the factory.

Return air pathways -Consider increasing the pathways for air from the closeable rooms to the return grill (jump ducts, larger transfer grills, etc.). Placement of the return grill in the common space should be prioritized (as opposed to a utility room return).

Verify airflow requirements in cooling mode. Adjust blower if necessary.

Belly board - Improved sealing techniques that are long lasting and durable.

Insulation - Verify insulation levels around attic installed duct work.

Marriage lines - Ensure a flat surface for the mating of the sections. Consider a factory installed marriage line gasket (closed cell) that is durable and will fill the gap. A gasket of at least 1 inch thickness should be considered. Several commercial products are available that will work well.

For questions concerning this trip report please contact Neil Moyer or David Beal at FSEC, 321-638-1000 or e-mail nmoyer@fsec.ucf.edu or david@fsec.ucf.edu.