Building America Industrialized Housing Partnership
TRIP REPORT
Destiny Industries / Oakwood Homes
Moultrie, GA Plant Inspection

DATE
April 2&3, 2002

ATTENDEES
David Beal, - FSEC / BAIHP
Warren McDonald, Robert Maier - LaSalle Air Systems
Randall Mace - Oakwood Homes Corporate Engineering Coordinator
Various staff of Oakwood Homes of Moultrie, GA.

TRIP REPORT DISTRIBUTION LIST

George James, Keith Bennett - U.S. DOE
D. Beal, N. Moyer, D. Chasar, J. McIlvaine, S. Chandra - FSEC
Mike Lubliner - Washington State University, Energy Office
Jerry Brooks, General Manager - Destiny Industries / Oakwood Homes, P.O. 2947, Moultrie, GA. 31776-2947
Mark Ezzo, Vice-president of Product Engineering, Randall Mace, Engineering Coordinator - Oakwood Homes Corporation, P.O. Box 27081, Greensboro, N.C. 27425-7081
Phillip Hathcock, Regional Vice-president of Manufacturing - HBOS Manufacturing, 590 Palmer Rd, Rockwell N.C. 28138
Warren McDonald, Production Manager - LaSalle Air Systems, 5030 Great Oak Dr, Bldg #3, Lakeland, FL. 33815-3122

PURPOSE

Oakwood homes is a new member of the Building America Industrialized Housing Partnership (BAIHP). This was the first visit to inspect the factory and review the duct assemblies of the Destiny Industries / Oakwood Homes Moultrie, GA plant. The Moultrie facility manufactures single and double-wide housing with sheet metal ductwork in floor systems. A small percentage of their houses use overhead systems made with fiberglass ductboard and flex-duct.

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 withing 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.

Oakwood Homes has recently instituted a corporate policy to address their duct systems and make them leak free, both to aid energy efficiency and to reduce moisture related service calls in the hot, humid coastal regions. Oakwood requested assistance from BAIHP partner LaSalle Air Systems, who in turn recommended that Oakwood join the BAIHP team. This visit served to familiarize Oakwood with the BAIHP team, and to allow BAIHP team members LaSalle Air Systems and the Florida Solar Energy Center gauge how they can best help Oakwood. Assistance through BAIHP’s partner Washington State University is available for Oakwood plants in the Pacific Northwest.

DESCRIPTION

Destiny Industries / Oakwood Homes Moultrie, GA plant manufactures single and double-wide housing. The plant is capable of producing approximately 12 floors a day, but is currently running under full capacity and producing 5 to 7 floors a day. The majority of home produced have a sheet metal floor duct, with a small percentage (less than one home per day) built with an over head duct system. The floor duct is an extruded metal duct measuring 12" by 5". Most register boots are installed inline, with a few offset registered installed with a side take-off collar and a short run of flex-duct.

The floors are built right side up, with the duct installed over the insulation, then the floor trusses are installed and the duct is fastened to the floor trusses with straps. The floor is then decked, and holes are cut in the floor for the register boot installations. A hole is cut into the duct for the boot, and the boot is installed with tabs and tape.

A small percentage of their houses use overhead systems made with fiberglass ductboard and flex-duct. A brief time was spent observing the tools and techniques used to manufacture overhead systems. The factory had the necessary tools (hole cutter and strap tightening tool) to build an adequate overhead system, and appeared to be knowledgeable as to their use. The visiting team decided to wait to try and introduce mastic to this system until the factory had it on hand and was familiar with it’s use.

OBSERVATIONS

It was immediately apparent that there was great opportunity for improvement in the floor system. All cutting in the duct was being done free hand, no templates or hole cutters being employed. The holes made for the floor registers were cut on three sides, with the resulting flap folded back and the register boot installed, tabbed over where there was contact, and taped. The hole for the plenum was made by cutting a large “X” in the duct, folding back the flaps, then installing the plenum. Both of these methods resulted in sloppy holes that didn’t fit the fitting cleanly, and large amounts of tape being used to seal gaps (when the gaps were discovered).

Exacerbating the problem was a fairly consistent mis-alignment of the duct. When the floor duct was installed in the floor and strapped down, it was rarely exactly where it needed to be, often shifted to one side or the other by several inches. This mis-alignment, combined with the use of a duct that is 12" wide, registers that are 10" wide, a plenum that is 11" wide, and cross-over collars that are 12", causes there to be large gaps where the duct and the fittings do not meet.

Further problems showed up due to the critical sizing of the various components of the duct system. Even when the duct was aligned correctly, there was installation problems due to the corner of the duct trunk not being completely square. When the fittings met the duct there was a gap along the sides of the duct, as the duct didn’t meet the fitting squarely. This was most apparent when the 11" wide plenum was installed into the 12" wide duct. The fittings used to do take-off ducts were 5" collars installed into a 5" high duct. This resulted in a hole cut into the duct that was oval, and the collar not able to be tabbed over for the full circumference of the fitting. The 12" cross-over collar was fit into a hole that was cut free hand, again resulting in an oval shaped cut that didn’t allow for full tabbing of the fitting.

Additional opportunities were found where the end caps were installed. The current method uses screws and tape to make the seal on the end caps.

At this point, due to factory floor construction scheduling, it was convenient for the BAIHP team to try several different duct construction methods. Partner LaSalle Air Systems had brought along two of their register boots, a tab-over type and a screw-down type. One home was built, with one half of the home using screw-down boot, and the other half using LaSalle’s tab-down boots. The trail installation also included the use of mastic on the duct systems.

TESTING

After observing several floors in the factory and building the trail duct systems, testing of the duct systems was undertaken. Duct system testing uses a calibrated fan (duct blaster) to depressurize the duct system to a specific pressure, in this case 25 pascals (pa). The amount of air in cubic feet per minute (CFM) needed to achieve this pressure is determined, yielding a CFM25_total flow number (cubic feet per minute at 25 pa). For a duct system to be considered to be “substantially leak free” by BAIHP, the CFM25_total must be a number that is less than 5% of the air handler fan flow. Assuming 400 cfm per ton fan flow or 1200 cfm fan flow for a 3 ton unit (typical for a 1500 ft² building), and using the desired criteria of 5% of fan flow, the target CFM25_total would be less than or equal to: (3 ton * 400 cfm/ton) * 0.05 = 60 CFM25_total. As testing in the factory was done on half houses (double wide tested one floor at a time) the desired CFM25_total number would be 30 or less.

The first floors to be tested was the home built just before the test home was constructed. This was a 14' X 56' (per side) double wide. The CFM25_total on the air handler unit (AHU) side was 80, the non-AHU side had results of 48. This was a 1568 ft2 home. Assuming a 3 ton A/C unit, with a fan flow of 1200 CFM, this represents a CFM25_total (80 + 48 = 128) of more than 10% of the fan flow.

The next house to be tested was the trial floor. This was a 1456 ft2 double wide. The AHU side employed the screw-down boots and had a CFM25_total 47. The non-AHU side used tab-over boots and had a CFM25_total of 25. Again, assuming a 3 ton A/C unit, with a fan flow of 1200 CFM, this represents a CFM25_total (47 + 25 = 72) of 4% of the fan flow. This was an encouraging result, but the team felt that they could do better. Examination of the screw down boot revealed that due to a layer of insulation installed between the floor trusses and the duct the duct was approximately 1" lower than the LaSalle screw-down boot was designed for. This made the installation difficult, and also allowed there to be leaks where the boot did not sit firmly on the duct. 

The results of testing the trail floors, compared to the testing of a standard production floor did not convince the production staff of the factory of the need to stop using foil tape and start using mastic to assemble their duct system. The staff felt that if the tape was properly installed it would provide adequate results. The BAIHP team members believe that foil tape will fail unless applied according to manufacture’s instructions. These instructions are typically ignored, as there is a requirement for surfaces to be clean and free from oil. The duct material is coated with oil when manufactured, and the duct manufacturing machine gets even more oil on the duct material. It is unlikely that there would be adequate oil cleaning in the factory, and cleaning solutions might impact employee health by subjecting them to high levels of volatile organic chemicals. The BAIHP team felt that it was critical to convince the factory personnel of the merits of mastic compared to tape. To this end, several older homes were found on the factory’s back lot and inspected and tested.

The first older house inspected was an 18 month old double wide that had remained on the factory’s lot since construction. The duct system showed very poor workmanship, with large gaps in the duct system never taped, and to a lesser extent, tape failure. The CFM25_total of the AHU side of this home was 200. The staff still felt that proper tape installation would be sufficient, and proceeded to re-tape the home. After this effort, the home was retested and yielded a CFM25_total of 70. This relatively poor result started to convince the staff that tape was not the answer. The inspection continued with a 18 month old single wide that also had remained on the factory lot since construction. The same problems were evident, poor workmanship and tape failure starting. Both of these houses had never been occupied, and no heating or cooling had occurred to help the tape fail. After a brief search, a repossessed unit was found to inspect. This home had been occupied for approximately one year, and then repossessed by the factory. Here the inspection reveled sever problems. There was significant mold growing along the marriage line ceiling. The duct system’s tape was failing at an alarming rate, as well as poor workmanship in the duct system. Mold growth along marriage lines has been observed by BAIHP team members in the field, and is most often the result of several causes, all related to duct leakage. The mold growth, combined with the obvious evidence of tape failure, convinced the factory staff that foil tape most likely will fail when improperly applied to a metal duct system.

The BAIHP team and factory staff returned to the floor department and assembled a final floor duct system. The desire was to take all of the techniques developed during the trip and build a very tight duct system. the factory staff felt that the screw down register boots offered a better sealing option than the tab-over boots. The system was fabricated using the screw-down register boots. The plenum was installed correctly, and sealed with several tubes of mastic (extra mastic was needed due to the non-90⁰ corner of the trunk duct, resulting in the side of the plenum not fully contacting the duct. The resulting gap was sealed with mastic. The cross-over drop-out was also installed and sealed with mastic. After construction the half home’s duct system was tested, yielding a CFM25_total of 50. The team was not satisfied with this number, so further inspection and testing was carried out. First, the AHU was removed and the ducts retested. This resulted in a CFM25_total of 32, or showing that the AHU represented a CFM25_total approximately 20. This result was still not as tight as desired, so a through visual inspection of the duct system was undertaken. This inspection reveled that the cross-over collar was not well sealed, and there were several gaps under the screw-down boots (again, due to the duct’s non-90⁰ corner ). After resealing all of the visible duct leakage the home was retested and had a CFM25_total of 20. Although this result was not a good as hoped for, it represents a “substantially leak free” duct system.

CONCLUSIONS AND RECOMMENDATIONS

All of the observed problems in the duct system stem from two problems, poor workmanship and materials, and the use of a duct system that is 12" wide. The workmanship problems are easily rectified with worker and QC training, combined with a more rigorous inspection protocol. Specific items to be addressed are as follows:

      •    Change over to a 14" wide duct. The 12" duct requires very careful alignment, and even then is not without problems when attempting to install an 11" plenum, 12" cross-over collars, and 10" register boots.

      •    Install all plenums, register boots and collars with mastic, DO NOT use tape.

      •    Use top take-off fittings for doing offset registers. The current 5" side-take-off collar installed into a 5" high duct is leaky, and time consuming. A top-take-off collar goes into the top of the duct, screws down, and is easily sealed with mastic. LaSalle Air Systems can provide this fitting.

      •    To correctly employ a screw-down register boot it must be redesigned by LaSalle Air Systems to fit in Oakwood’s floors, as it currently is designed for ducts without a layer of insulation on top. There was some indication that LaSalle’s tab-over register boot could be taller also.

After the obvious problems are solved by the production crew at Oakwood, further testing needs to be done to insure that the duct tightness criteria are being met by the new practices. Also worthy of investigation is the sizing of return air pathways from bedrooms. Typical HUD-code homes have inadequate return air pathway from bedrooms, causing sustained negative pressures in the house when bedroom doors are closed.

For questions or comments on this trip report, please contact the author, David Beal at 321-632-1433 or via email at david@fsec.ucf.edu.

<< Southern Energy Training, Plant Visit & ES (App D)

Table of Contents

Clayton Homes >> Plant visit and testing (App D)