BAIHP Research: A. Manufactured Housing Research Cont'd

• Building Science and Moisture Problems in Manufactured Housing
  Papers: Subrato Chandra, Danny Parker, David Beal, David Chasar, Eric Martin, Janet McIlvaine, Neil Moyer. Alleviating Moisture Problems in Hot, Humid Climate Housing. Position Paper for NSF Housing Research Agenda Workshop, UCF Feb. 12-14, 2004.
  
  Moyer, N., Beal, D., Chasar, D., McIlvaine, J., Withers, C, & Chandra, S. (2001). “Moisture Problems in Manufactured Housing: Probable Causes and Cures.” ASHRAE - IAQ 2001 Conference Proceedings, San Francisco, CA.

Figure 45 Palm Harbor HUD Code Manufactured Housing factory – production line.

Figure 46 Completed HUD Code Manufactured Home, Palm Harbor Homes.

Manufactured homes have a permanent steel chassis attached below the floor and are constructed in a factory (Figure 45) to meet a national code maintained by the U.S. Department of Housing and Urban Development (HUD). After production, homes may travel a few hundred miles, hauled by truck, before final setup. The homes are setup by placing blocks under the steel I-beams and anchoring the beams firmly to the ground. A skirting covers the blocks and steel frame in a fully setup home (Figure 46).

Manufactured homes are typically heated or cooled by a system of ductwork, which delivers hot or cold air from the air handler unit (AHU). The ductwork can be in the attic or in the belly cavity of the home. The ducts are typically made of aluminum or fiberglass trunk lines which supply air to the floor registers through in-line boots or flex ducts. The boots or ducts terminate at perimeter registers on the floor. Supply duct leaks represent one of the biggest causes of moisture problems in manufactured homes. (Figures 47 and 48). Poor design and construction leave holes at the AHU connection to the main trunk, and where the boots connect to the trunk, supply registers, end caps, cross-over duct connections, and other connection points. When the AHU blows air, some air leaks into the belly and eventually to the outside through belly board tears. This loss of air creates a negative pressure inside the house and a positive pressure in the belly. The negative pressure pulls outside or attic air into the house through cracks and crevices which connect the inside of the house to the outside or to the attic. During northern winters, this outside air is cold and dry and its entry increases occupant discomfort and heating energy use.

Figure 47 Pressure field and unintentional air flow created by supply duct leaks.

During summer in the Southeastern US, the air is consistently at or above the dewpoint of 75. If a homeowner keeps their home thermostat set below this 75 F dewpoint, the moisture laden outside air condenses as it comes into contact with the cold inside surfaces. If it condenses behind an impermeable surface such as vinyl flooring or wallpaper, serious mold, mildew, and floor buckling problems can result.

Many manufactured and site-built homes have only a single return and, therefore, very little return air transfer from the bedrooms (basically via the undercut at the bottom of interior doors). When interior doors are closed, rooms off the main body (e.g., bedrooms) become pressurized and the main body of the house depressurizes. Even though negative pressures are usually only one to three pascals (Pa) - they can cause serious problems in a home.

Researchers use a calibrated fan called a ductblaster to measure duct leakage. The ductblaster is attached to the return grill or the crossover duct opening (Figure 49) and all supply registers are masked off and the fan is turned on. Once the house ductwork reaches –25 Pa, airflow through the fan is read (in CFM). The resultant measure is the total duct leakage. In good airtight ductwork, total duct leakage (CFM@25 Pa) should be less than 6% of the homes square footage.

Figure 48 Cross section showing foundation support, crossover duct, and one type of ventilation system in a manufactured home.

A second duct leakage test measures leakage to the outside. This leakage is calculated by depressurizing the entire house to –25 Pa with a blower door, then adjusting the ductblaster flow so there is no pressure difference between the house and the ducts. This measurement is a true indicator of duct air loss to the outside and is used in energy calculations for estimating the energy loss from leaky ducts. In good duct systems, duct leakage to the outside (in CFM) is less than 3% of the home’s square footage.

Figure 49 Floor and belly area with supply ducts. These ducts supply conditioned air to all rooms through floor vents, a common duct system layout in manufactured homes.

The battery of tests run in a problem house typically includes measuring the airtightness of the house with a blower door, depressurizing the house to –50 Pa. At that time, the house to belly and belly to crawlspace pressures also can be measured. Researchers also test pressure differentials caused by AHU operation and closed interior doors. An additional measurement of duct leakage, called pressure pan, is conducted on some houses to pinpoint specific registers which might have large leaks. In this measurement the house is first depressurized to –50 Pa and all the register vents are unmasked. Then the registers are covered one by one and the pressure difference between the covered register and the house is measured. A zero reading indicates no leakage at that register. Readings over one Pa indicate a sizeable leak that should be repaired.

• BAIHP Field Visits to Moisture Problem Homes
  Papers: Moyer, N., Beal, D., Chasar, D., McIlvaine, J., Withers, C, & Chandra, S. (2001). “Moisture Problems in Manufactured Housing: Probable Causes and Cures.” ASHRAE - IAQ 2001 Conference Proceedings, San Francisco, CA.

A significant number of new manufactured houses built to HUD code and located in the hot, humid Southeast have exhibited moisture problems. Soft wallboards, buckled floors, damaged wood molding, and extensive mold growth are the most common symptoms. These problems do not respond to the standard service and repair strategies for water intrusion. (Please see Appendix B for sample problem home inspection trip reports.)

Summary of 1 st-4 th Budget Period Field Visits to Moisture Problem Homes

At the request of six manufacturers, 69 such moisture damaged homes were investigated from 1999 to the end of reporting year four (through March 31, 2003) to determine likely causes. In Year 4 alone, 18 homes were investigated by FSEC. One-time blower door, duct tightness, and pressure differential measurements were performed on all homes. Field data on ambient, crawlspace, belly and house temperatures, plus relative humidity levels were collected on a few of the homes. Recommendations and reports were prepared for the manufacturers’ service, production, and design staff. Field repairs were performed in most of these homes. A general theme was found in the houses investigated.

• Air conditioner thermostat settings (typically 68 to73 F) set below the ambient dew point.
• Negative pressures across the envelope from high supply duct leakage (CFM @25Pa >10 per 100 square feet of conditioned floor area), inadequate return air paths, interior door closures, exhaust fans, or a combination thereof.
• Inadequate moisture removal from disconnected return ducts, continuous fan operation (air handler or ventilation), inadequate condensate drainage, oversized air conditioners, or a combination thereof.
• Moisture diffusion from the ground into the house because of poor site drainage, inadequate crawl space ventilation, tears in the belly board, or a combination thereof.
• Vapor-retardant in the wrong location (i.e., vinyl or other impermeable wall or floor coverings located on the colder surfaces).

Recommended solutions provided to the manufacturers to eliminate moisture problems included:

• Maintain air conditioning thermostat settings above the ambient dew point (at least 75 F).
• Eliminate long-term negative pressures created by air handler fans or ventilation equipment.
• Tightly seal all ductwork and provide adequate return air pathways.
• Enhance moisture removal from the conditioned space by correct equipment sizing and maintenance.
• Eliminate ground source water and provide an adequate moisture barrier for the floor assembly.
• If possible, remove vapor barriers located on the wrong surfaces.

Research continues to determine if these steps will be sufficient to prevent problems even when vapor barriers are incorrectly located in homes in the hot, humid climate. Preliminary results are encouraging. One manufacturer has not reported a single new moisture problem in any of the homes produced since 2000 in a factory that previously had a significant number of problem homes. Steps taken by the factory were inclusion of airtight duct systems (a zero net-cost increase), right-sized cooling systems (a negative cost), return air ducts from all bedrooms (a cost of about $15), installation of a ground vapor barrier (no change from previous practice).

Summary of 5 th Budget Period Field Visits to Moisture Problem Homes

Figure 52 HUD Code required perforations in skirting may not allow adequate volumes of ventilation, creating higher than usual vapor pressure difference across the floor assembly even though the ground cover and belly board are in good condition.

BAIHP researchers at FSEC received fewer requests in the 5 th budget period for assistance with moisture damaged homes (Table 27), reflecting improvement of duct construction and sealing, addition of return air pathways from bedrooms, and reduction of vapor impermeable interior surfaces. Additionally, service personnel who have attended BAIHP training and participated in field work with BAIHP are more prepared to resolve problems without assistance. Service personnel report installing passive return air vents in bedrooms, providing appropriate moisture barriers, and sealing duct leaks to resolve humidity, comfort, and moisture damage call backs.

When service personnel have been unable to resolve a problem, they request assistance from BAIHP researchers who attend a service call and conduct various diagnostic tests to identify factors contributing to the moisture, comfort, or high energy bill problem. (MHRA has been providing similar services on a fee basis to the industry also.) After BAIHP researchers complete a field visit, a trip report is issued detailing the findings and recommendations, include basic building science background material.

It has been BAIHP’s experience that corrective measures from repeated moisture problem Diagnostics have been incorporated into the production process, resulting in thousands of improved manufactured homes. These are noted in Category D of Table 2.

A common problem that remains unresolved involves the combination of abundant crawl space moisture (Figure 50 and 51) and poorly vented skirting (Figure 53). In the hot-humid coastal regions, this combination raises vapor pressure across the belly to critical levels. This was evident in several of the homes visited this year. As a result of this field research, BAIHP has designed a study that will be initiated in the summer of 2004 to evaluate the moisture flow characteristics of crawl space conditions.

Northwest BAIHP Random Home Testing

SGC Random Home Testing: In 1994-1995 (prior to implementation of BAIHP), SGC staff conducted field testing of 178 SGC homes built in 1992-1993. In 1999, the first year of the BAIHP effort, staff in Idaho and Washington field-tested 49 SGC homes built in 1997-98. In 2000, analysis of field test data confirmed some improvements to home set-up procedures and air leakage control, while highlighting a need to improve duct tightness and ventilation system operation (through homeowner education.) In 2001, BAIHP staff produced an updated homeowner ventilation brochure.

In 2002 and 2003, BAIHP staff worked with Ecotope to develop a valid sample for the next round of field testing, and began to develop the field testing protocol. In 2004, Ecotope selected 105 homes from the total production for the years 2001-2002. The field testing took place in the summer of 2004. Findings from the testing include:

• Average house size is 1769 ft 2; double section homes are also getting bigger, on average. The house size is very comparable to the homes built in 1997-1998 but 20% larger than the homes in 1994-1995 study
• Houses are getting tighter, according to the blower door results. The average air leakage rate at 50 Pa is 4.2, which represents a tightening of almost 25% over the original MAP home average. The median equivalent leakage area (ELA) for double-section homes has decreased by about 12% despite a substantial increase in house size.
• Only about 20% of NEEM homes in this study contain intentional outside air inlets. This is the result of BAIHP research indicating that intentional outside air inlets are unnecessary to provide adequate fresh air.
• 2/3 of homes in the study have dedicated whole house fans and a substantial fraction of homeowners are using their whole house fans. However, a significant minority (30%) does not turn them on.
• About half of homes in the study use central cooling, with more than half of these homes using a heat pump.
• Duct systems are about 20% leakier than in the Year 1 study and about 10% leakier than in the 1994-1995 study (when the comparison is normalized by house size).
• The median supply leakage fraction is 11-13% for the homes in this sample. The duct loss translates into a heating system efficiency loss of between 10-20% overall, depending on the location of the home (west side or east side of the mountains) and type of heating equipment (heat pumps perform worse).

In 2004, BAIHP staff conducted a billing analysis on a limited number of random field study homes. The conclusions (although not statistically significant) suggest that temperature related energy use in NEEM homes remains similar to previous larger studies on cost-effectiveness. The analysis attempted to evaluate total and space conditioning energy use by HVAC system types but was limited by small sample size.

In 2004, a sub-sample of homes that are believed to represent the best case for duct tightness were selected for additional field testing. These homes include those with in-plant tested ducts and thru-rim crossover duct systems. The goal of this effort is to establish a “tightest” duct case benchmark. Field testing will be completed in 2005; report will follow.

Northwest BAIHP Field Visits to Problem Manufactured Homes

Figure 51 The downstream exit for the water draining across the site via the crawl space. Note flow pattern away from house.

Figure 52 HUD Code required perforations in skirting may not allow adequate volumes of ventilation, creating higher than usual vapor pressure difference across the floor assembly even though the ground cover and belly board are in good condition.

In offering technical support to owners of over 100,000 homes built since 1990, the BAIHP staff in the Northwest answers questions from homeowners, manufacturers, retailers and others. In The 6th budget period, staff from Washington, Oregon and Idaho responded to over 70 phone calls and conducted 27 field visits. The number of field visits to problem homes has significantly decreased over the history of the program, in large part because of manufacturers’ and installers’ increased adoption of the NEEM Super Good Cents/Energy Star (SGC/E-Star) specifications which include duct air tightness specifications (duct leakage is a major contributor to pressure and air flow related moisture problems), and the requirement that manufactured home installers be certified in Washington and Oregon.

Northwest BAIHP staff began to utilize Energy Gauge USA as a toll for evaluating high bill complaints in 2003-2004.

BAIHP staff participated in quarterly meetings of the Washington State Manufactured Housing Technical Working Group, which coordinates the certification of manufactured housing set-up crews.

While butyl duct tape is no longer allowed under current NEEM SGC/E-Star specifications, a consistent issue in the field continues to be excessive duct leakage, due in large part to failures of duct tape. These findings were brought to the attention of the NFPA-501 Manufactured Housing Standards Committee, resulting in a successful proposal to revise the duct sealing specifications to eliminate the use of duct tape in favor of better performing mastic and fiberglass mesh in the NFPA-501 standard. See a summary of supporting research findings in BAIHP Duct Data Compilation.

• Manufacturers Participating in Building Science Research

Blue Sky Foundation

Blue Sky Foundation, in coordination with FSEC, conducted an evaluation of energy efficiency and the moisture damage potential in 16 North Carolina homes in the summer of 2001. Blue Sky foundation proposed that the energy and moisture evaluation focus on the building envelope integrity, HVAC duct systems, and the moisture impact of unvented space heaters. All of the homes in the study were manufactured models located in Carteret and Craven counties, each located on the North Carolina coast. Field teams gathered additional energy and moisture information from homeowners.

Only three of the 15 tested homes recorded moisture and/or mildew problems. Because of the small sample size, the results are mostly anecdotal and would need to be evaluated within a larger data set. Planning for this is underway. Data from the summer field program as well as the final report are now on the BAIHP website (www.baihp.org) under Publications.

Cavalier Homes

BAIHP visited one Cavalier Home in Florida for a moisture damage investigation in response to home owner complaints of persistent air flow problems and floor damage. BAIHP made recommendations to correct the installation of the duct system and supply registers, repair the rodent barrier to make it air tight, do site work to reduce flooding under house, place a ground cover if site work done, increase crawl space venting, and replace damaged flooring with plywood.

Figure 53 Testing Results from Fleetwood Homes Plant in Alma, Georgia
illustrate that tape sealed ducts can result in total duct leakage under
Qn=<6%. This initial tightness, however, is often eroded by adhesive failure.

Fleetwood Homes

During the 5 th budget period, BAIHP continued to support Fleetwood’s service department making six visits to moisture damaged homes in Florida (4), Texas (1), and West Virginia (1).

Six Fleetwood homes, all in Florida, were tested for moisture and mold damage from April 2002 through March 2003, the 4 th budget period. All of the homes had damaged flooring due in part to a lack of ground cover and poor crawlspace ventilation. Damage to the floor in one home was exacerbated by a plumbing leak. Only one home had moisture damage to the wallboard material, and this home showed a history of thermostat settings below 72 F. A report for each home was submitted to Fleetwood for corrective measures. One additional high bill complaint in Cobb, Georgia was investigated during this reporting period.

In 2002, four Fleetwood factories in Southern Georgia were visited to investigate possible causes of moisture related building failures found in homes installed in hot, humid climates. The factories were located in Douglas, Alma, Pearson, and Willacootche. (Figure 53.)

Homes of Merit

In 2002, researchers performed multiple diagnostic tests on a home located in Marathon, Florida that was experiencing “mold problems.” Researchers determined that the mechanical system was significantly oversized and that the home was operating under negative pressure when the system was operational. The home’s owner exacerbated humidity problems by leaving the fan in the “on” mode. On-site relative humidity readings showed that indoor and outdoor relative humidity were the same, approximately 70%.

Figure 54 Wall assembly used in moisture transmission experiment.

Palm Harbor Homes
(See also, Palm Harbor Homes in Section I, Technical Support and Manufactured Housing Indoor Air Quality Study in Section III, Research, below).

Palm Harbor Homes, James Hardie®, and FSEC performed two separate drywall assembly tests to determine the cause of some moisture damage occurring in homes sheathed with Hardipanel. Hobo dataloggers recorded temperature and relative humidity measurements inside the assembled panels on eight different wall panel configurations. (Figure 54.)

Results determined that the unprimed, unwrapped sheathing performed best. The painted drywall assemblies allowed the greatest moisture movement - or wall assembly drying. (Table 28) The vinyl-covered drywall held moisture longest, recording the slowest drying time. Adding perforations to the vinyl reduced the drying time.

In 2002, two Palm Harbor homes with comfort problems were tested in Ocala and Okahumpka, Florida and one high bill complaint was investigated in Odessa, Florida. Duct leakage testing and infrared imaging revealed a duct disconnect near the attic crossover in the Ocala home. Inspections with the IR camera found no insulation problems in the Odessa home. Ductblaster and blower door tests revealed airtight duct and envelope systems. Other than an oversized air conditioning system, there were no obvious reasons for the high bills.

Southern Energy Homes
(See also, Southern Energy Homes in Section I, Technical Assistance.)

During Year 2001, 12 homes were field tested in the Houma, Louisiana area. Some of the homes had new moisture damage. Others were rechecks of previous moisture problems already repaired by SEH personnel. FSEC inspectors reported improper repairs and recommended additional dealer and staff training. An additional five homes were field tested in Houma during the 4th reporting period, with another home in Mississippi and one in Alabama also field tested.

During the 5 th budget period, BAIHP visited two Southern Energy Homes in Texas (1) and Kentucky (1).

• Side By Side Study Of Energy Use And Moisture Control Comparing Standard Split System Air Conditioning And A Coleman® Prototype Heat Pump, Bousier City, LA
  Research led BAIHP Researchers Dave Chasar, Neil Moyer, and Chuck Whithers
  Papers: Withers, C., Chasar, D., Moyer, N., and Chandra, S. "Performance and Impact from Duct Repair and Ventilation Modifications of Two Newly Constructed Manufactured Houses Located in a Hot and Humid Climate", Thirteenth Symposium on Improving Building Systems in Hot and Humid Climates, May 20-22, 2002 Houston, Texas.

In 2001, the BAIHP team conducted research on two homes to define how tight ducts and a prototype Coleman® heat pump (proprietary technology) affect energy use and moisture control in a hot, humid climate. FSEC, in collaboration with Fleetwood Homes, York International Manufactured Housing Division (now Stylecrest Sales), and Coleman®, monitored two nearly identical side-by-side homes in Bossier City, Louisiana. The homes contained different air conditioning systems. House A used a standard split air conditioner, while House B used the Coleman® prototype unit (a more efficient, two-speed split air conditioner).

Figure 55 Power draw over a 24-hour period, September 2, 2000.

Figure 55 shows the reduced power draw of the two-speed compressor (green, dotted line) over a 24-hour period on September 2, 2000. With the unit operating at low-speed for most of the day, the cooling energy savings were 28% when compared to the energy use in House A. Average daily cooling energy was reduced by about 12% over the monitored period. An added benefit of the two-speed air conditioner was 20% greater moisture removal on days with an outdoor dewpoint above 60 F.

Savings from Duct Repair and POS Ventilation: In addition to comparing one house to the other, the BAIHP team also compared home performance before and after ductwork and ventilation system changes were made.

To make the comparison, duct and other leaks were sealed in both houses until the two were equally airtight. The ventilation method in each home also was changed from exhaust-only to a positive pressure system (POS). With exhaust-only ventilation, bathroom fans removed stale air from the home which caused fresh air to be pulled in through the building envelope. To simulate occupant use, two bath exhaust fans were operated by a timer for three hours in the morning and six hours in the evening.

In contrast to exhaust ventilation, the POS system introduced a small amount of fresh air on the return side of the air conditioning cooling coil. A POS system was installed in each home at the same time the ducts were repaired. Subsequent monitoring looked at the effects of this alternate ventilation system. Tightening the ducts and installing a POS ventilation system resulted in an 18% and 37% cooling savings in the two homes. Only about 2% of these savings were attributable to the ventilation system change, the remaining savings are a result of duct repair.

• WSU Energy House
  Olympia, Washington
  Technical Assistance by BAIHP Contractors Washington State University Energy Program, Oregon Office of Energy and Idaho Department of Water Resources, Energy Division

Figure 56 WSU Energy House in Olympia, WA

This 2600 ft 2 home was built beyond SGC standards and incorporates Energy Star lighting and appliances. The home (Figure 56) has received significant national exposure through WSU campus and alumni newsletters, tours, the BAIHP website, and local and trade media including an article in the Automated Builder magazine and a feature by KING 5 News of Seattle.

WSU staff uses the house to try out innovative technologies and testing methods.

In 2003, BAIHP staff developed a moisture case study based on research at the WSU Energy House, published under a separate Building America project. The WSU Energy House has been monitored since 2000.  Collected monitoring data includes weather, temperature, humidity, CO 2, CO, and eight differential pressures.   Energy use data is being collected for water heating, laundry, fireplace and heating, ventilating, and air conditioning (HVAC).  Data from the house is available on the BAIHP web page (under Current Data) and has been presented to the building science, indoor air quality (IAQ) and HVAC research communities at conferences sponsored by ASHRAE, Air Infiltration and Ventilation Center (in the UK), HUD/NIST, NFPA, and BTECC. (See also Appendix D, WSU)

Working with Ecotope, ASHRAE, and the Energy Conservancy, BAIHP staff conducted “Delta Q” and “nulling” duct leakage tests in 2001.  Follow up pressure tests and analysis of test data conducted in 2002 indicate these tests are effective methods of measuring duct leakage in manufactured homes, and may be included in the upgrades to the National Fire Protection Association-501 standards for manufactured homes.

Blower door and duct leakage testing indicate very good whole house and duct airtightness (2.4 ACH50 and 61.6 CFM50 out). Tracer gas testing demonstrated that the use of a furnace-based intake damper does not change the leakage rate of the home.

In 2004, moisture problems associated with siding and trim details were eliminated using and an improved window flashing system. The adoption of this system is currently under discussion with some manufacturers, and NFPA-501.