Abstract

In partnership with the Florida Solar Energy Center (FSEC), two manufactured homes were located on North Carolina A&T State University’s campus in Greensboro, NC and used in a side-by-side energy consumption comparison. One of the homes was built to the basic HUD code standard and the other was constructed with features expected to produce a home that was 50% more energy efficient.

FSEC and NCATSU began monitoring energy performance in both homes. In addition, the performance of each unit was evaluated using a DOE2 based computer energy analysis program developed by FSEC. A comparison of the performance of the units shows a measured energy savings in the more energy efficient unit of 52% for the Heating, cooling, and DHW energy use. This compares well with the energy savings predicted by the FSEC Energy Gauge program of 57%, even when accounting for the warmer than usual winter experienced during the testing period.

1.0 INTRODUCTION

As part of a project funded by the North Carolina Department of Administration - Energy Division, and as Part of the US Dept of Energy’s Building America Program, researchers in the Center for Energy Research and Technology (CERT) at North Carolina A & T State University evaluated technologies to improve the energy efficiency of manufactured housing.

The partnership effort described by this report required CERT researchers to monitor the energy use of two side-by-side manufactured housing units on the campus of North Carolina A & T State University in cooperation with the Florida Solar Energy Center (FSEC). One of the units monitored was designed and built to basic HUD code requirements [HUD, 1999] and the other was designed to use at least 50% less energy (Building America compliant).

As part of this study, both units were also analyzed using the FSEC developed ENERGY GAUGE software program. This program predicts building energy consumption using the DOE 2 analysis engine with a user friendly front end that develops DOE2 input files and models that are more appropriate for residential building systems.

In addition, in this second year, modifications were made to the solar hot water heating system in the energy efficient housing unit to help improve the performance of this system. Further, a number of the incandescent light bulbs in the energy unit were replaced with compact fluorescent bulbs. These changes were staged to allow an evaluation of the effect each of these measures had on the energy use in the homes.

The following report summarizes the results of the second year of the effort described above (the first years results were previously reported) [McGinley, 2002].

2.0 Standard (HUD CODE) and Energy (ENERGY EFFICIENT) Manufactured Home Description

Each of the two manufactured homes used in this study have 1,528 ft 2 of living area, 3 bedrooms and 2 baths. Each of the two housing units had identical floor plans. The homes were oriented on the site with the front facing east. Both houses were furnished. Exterior finishes were of medium color, with dark roofs. See Figures 1 through 3.

Each home unoccupied; however incandescent lights on timers were used to simulate occupancy loading. One of the homes was constructed using conventional HUD code provisions and the other was designed to be 50% more energy efficient. Construction differences between the two homes are listed in Table 1.

The Standard housing unit used a perimeter ducting system, while the Energy housing unit used a central trunk line. The higher thermal resistance of the energy home building envelope allows this more efficient central distribution system and a reduction in compressor tonnage, which reduces initial costs and duct losses. See Figures 1 and 2.

It should be noted that the Energy housing unit incorporated the use of a solar hot water heater, with a 66-gallon hot water tank, while the “Standard” home used an electric hot water heater with a 40-gallon tank.

Table 1 Summary of Construction of the Two Homes

*(Note that the Energy House values for Duct Leakage and House leakage were based on retests done after August 2001 repairs)

Figure 1. Floor Plan and HVAC layout for the Base HUD Code (Standard) Housing Unit (Courtesy of Palm Harbor Homes)

Figure 2. Floor Plan and HVAC Layout for the Energy Efficient (ENERGY) Housing Unit (Courtesy of Palm Harbor Homes)

3.0 Monitoring Program

A computerized data logging system was used in each house to monitor:

1. The interior temperature and relative humidity.
2. The power consumption of the whole house.
3. The power consumption of the air conditioning/heat pump compressor.
4. The power consumption of the air handler fan.
5. The power consumption of the electric resistance heat (primary heating in the standard house, secondary heating for the energy house).
6. The power consumption of water heater and electric water tank coil.
7. The exterior temperature and relative humidity, solar radiation (both parallel and at the solar panel angle), and wind speed.

The data-loggers were connected to FSEC’s mainframe computer via modem, and downloaded automatically. Data were sampled at 6 second intervals and recorded in hourly intervals.

All appliances in the home were unplugged except for the hot water heater, HVAC system and some incandescent lights. There were also a few miscellaneous devices such as the data logger, phones, and controls that show as a minor electrical load. The incandescent lights were used to simulate an occupancy load of 1.5 persons and were run on the following schedule; 500 watts of lights were on 24 hours a day 7 days a week, 500 watts of lights are switched on by timers from 4 pm to 12 pm, 200 watts of lights are switched on by timers from 6 am to 9 am.

In addition, on weekdays, there were two hot water draws of 40 gallons each, one in the morning and one in the late afternoon for each of the houses. This water draw was used to simulate an average weekly water use of a typical residence.

A comparison of the performance of the units over the period from January 2001 to March 2002 was made and reported in the first year report. This report summarized the initial poor performance of the Energy housing unit that resulted from an excessively high air-handler fan speed that significantly reduced the efficiency of the system, a very large duct leak resulting from an improperly set Y-connection coming off the main supply duct trunk line, a supply duct collar failure and a poorly sealed opening around the refrigerant line and drain between the return and supply side of the coil plenum creating a return air bypass around the coil. These items were repaired by September 2001 and “good” data were obtained from September 1, 2001 to August 16, 2002.

Both homes were on cooling only mode from September 1, 2001 through October 26, 2001 at 7:00 pm. After this time, both homes were on heating only mode until, April 16, 2002 at 2:40 pm, where they were switched over to cooling only mode again until October, 2002. It should be noted that one of the critical findings of the first year of the investigation indicated that current manufactured home set up procedures may not be adequate to ensure predicted performance of the energy efficient home units. As a result, Palm Harbor, one of the industry partners in this investigation, has instituted steps to improve installation/setup procedures.

It was also found that the standby losses in the solar hot water heater in the Energy Unit were significant and on long idle periods were sufficient to make the overall efficiency of the solar hot water heater less than the standard electric unit. To help alleviate these stand-by losses, the solar water tank piping insulation was upgraded on March 6, 2002 and its effect on the water system performance was evaluated. The solar hot water tank had a significant amount of copper and plastic tubing exposed in the original installation configuration. Additional pipe insulation was applied to all accessible pipe surfaces and the effects of this upgrade was evaluated.

On May 1, 2002, in an effort to further improve the performance of the solar hot water heater, the solar hot water tank in the energy unit was wrapped with a R5 foil bubble wrap insulating blanket over the sides and most of the top of the tank. Figure 3 shows the tank with the foil insulation and additional pipe insulation applied.

The final modification made to the Energy Housing unit was made on June 4, 2002. At this time, three of the light fixtures that were on evening 4 hour timed duration were changed from 100 watt incandescent lamps to 25 watt compact fluorescent lamps.

Figure 3. The Solar Hot Water Tank with R5 Insulating Blanket and additional Pipe Insulation Located in the Energy Efficient Manufactured Housing Unit

4.0 Results and Discussion

4.1 Energy Use Results and Discussion

The measured total energy used by each of the housing units for cooling and heating are shown in tables below. Table 2 shows the energy used for heating and cooling the Standard Housing Unit over the period of January through August in 2002. The Standard Unit data logger was struck by lighting in mid August, 2002 and all data after this point was not included since only partial data is available and comparisons of performance were not possible. Table 3 shows a similar summary of the cooling and heating energy used by the Energy Housing Unit.

Tables 4 and 5 show the energy used for domestic hot water production for the Standard and Energy units, respectively for these same periods.

Also listed in each table are the monthly energy use measured during the first phase of this investigation, January through August (2001). Note that the Energy Housing Unit data prior to August 2001 is suspect due to problems in the ducting and HVAC system, as discussed previously.

Only the cooling and heating energy, and energy used for domestic hot water production, will be discussed in this and subsequent sections since each housing unit was not occupied and was assumed to use essentially the same amount of energy for the occupancy simulation. When the three incandescent bulbs replaced with compact fluoresce bulbs in the energy unit, the reduction in energy use for lighting load was not of concern, what was being evaluated was the impact this change had on the cooling load in the housing unit.

The total cooling energy used by the Standard house from April 2002 to August 16 th 2002 was 3468 kW-hrs. The total cooling energy used by the Energy house from April 2002 to August 31 st, 2002, was 2451 kW-hrs. If it can be assumed that about 500 kW-hr would be used for the reminder of the August month in the Standard housing unit (~1/2 the 2001 values and about 2 times the 2002 value), then the Energy housing unit used approximately (1- 2451/(3468+500) x 100), or a 38.2 % less energy than the Standard unit for cooling during this time. The totals for the same period in 2001 were 3860 kW-hr (Standard) and 3127 kW-hr (Energy), a 19 % difference. You can see that there is an increased difference in energy efficiency of the two housing units in the second year of monitoring during the cooling season. This may be at least partially due to less waste heat being dumped into the energy unit by the solar hot water heater and compact fluorescent lights. This will be discussed later.

Over the first and second phase of this investigation there was only one complete heating season observed. The total heating energy used by the Standard house from November 2001 to March 2002 was 7454 kW-hrs. The total heating energy used by the Energy house over the same time period was 2199 kW-hr. Over this time, the Energy housing unit used approximately 70 % less heating energy than the Standard unit.

The total heating and cooling energy used by the Standard housing unit from September, 2001 through August, 2002 was 12,365 kW-hr (a sequential heating and cooling season). Over the same period of time, the Energy housing unit used 5194 kW-hr, a 58% reduction.

The total energy used for domestic hot water production from September 1, 2001 to August 16 th, 2002 in the Standard unit was 2810 kW-hr. The total energy used for domestic hot water production from September 1, 2001 to August 31, 2002 in the Energy (Solar) unit used 1911 kW-hr of energy. If it is assumed that the Standard unit hot water tank would used about 110 kW-hr for the rest of the August month (about ½ of that used in previous months), the Solar hot water tank in the Energy unit used approximately 34% less energy than the Standard unit.

Combining the energy used for domestic hot water production with that used for heating and cooling produced a total of 15,285 kW-hr of energy used by the Standard housing unit between September 1, 2001 and August 31 st, 2002. The Energy housing unit used a total of 7,105 kW-hr over the same period of time. The Energy unit used 53 % less energy than the Standard unit for heating cooling and production of domestic hot water over this period. As will be discussed in the next section, the improvements made on the solar hot water tank and their effects on energy use suggests that the Energy housing unit would use even less energy than the Standard housing unit with these changes in effect over a entire year.

4.2 Effect of Changes in the Solar Hot Tank and the Compact Fluorescent Fixtures

The pipe insulation on the solar hot water tank was upgraded on March 6, 2002. This increase in insulation on the hot water piping appears to have had a significant effect on the performance of the solar hot water system and appears to have reduced the stand-by heat losses in the system.

Since no hot water draws are made on the weekends, it is possible to examine how standby losses are influenced by system changes by looking at this time period specifically. The stand by losses for 18 week end days in the period of March 6 though April 30, 2002 showed that the pipe wrap has cut standby energy losses for the energy house by about 65% (an average of 2.43 kWh/ day (2001) vs. 0.85 kWh/day (2002)) over a similar period last year.

In addition, the reduction of standby losses helped the solar hot water system use less energy than the conventional electric system in the month of March. The Standard Unit used 283 kW-hr and the solar hot water system used 203 kW-hr, a 28.2% reduction. This reduction was further increased in the month of April where the standard system used 265 kW-hr and the Solar system used 146 kW-hr, a 45% reduction. It should be noted that these values represent significant reductions in energy use when they are compared to 2001 values where the solar system actually used more energy than the standard unit in March 2001 and used only 25% less than the standard unit in April 2001.

In an effort to further improve the performance of the solar hot water heater, the water tank was wrapped with a foil bubble wrap insulating blanket over the sides and most of the top of the tank.

Over the month of May, the total energy used for DWH heating was 137.8 kWh for the Energy housing unit and 249.6 kWh for the Standard housing unit. This represents a 45% reduction in energy use for the solar hot water system, about the same as the 45% reduction shown for the month of April.

A comparison of the tank losses over the weekends in months of April, 2002 and May, 2002 give a good indication of actual losses since there are no tank draws on these days. This data shows an average daily week end loss of 2.83 kWh for the Standard home and a 3.08 kWh for the Energy home for the Month of April and an average daily week end loss of 3.92 kWh for the Standard home and a 2.97 kWh for the Energy home for the month of May. There appears to be a little improvement in tank heat loss over the two periods.

The total energy used over the month of June, 2002 for DWH heating was 74.8 kWh for the Energy House and 192.2 kWh for the Standard home. This represents a 63% reduction in energy use with the solar hot water system (compared to the 45% difference for May). This appears to indicate the tank insulation may be having an effect on the losses in the tank. It should be noted that the solar radiation was about the same as the month of May (within 3%) but the water consumption was slightly less. These results suggest that the tank wrap may be reducing some of the heat losses.

The total energy used over the month of July, 2002 for DWH heating was 80.3 kWh for the Energy House and 200.25 kWh for the Standard home. This represents a 60% reduction in energy use with the solar hot water system. This compares well with the June reduction of 63% with about 11% less solar radiation in the month of July. This reduction and those in May and June are significantly greater than the efficiencies observed in 2001 without tank and piping insulation where energy use reductions ranged from 27% to 40%.

The total energy used over the period of August 1, 2002 through August 15, 2002 for DWH heating was 27.13 kWh for the Energy House and 85.18 kWh for the Standard home. This represents a 68% reduction in energy use with the solar hot water system. This compares well with the June and July reductions of 63 and 60%, respectively. These results appear to indicate the tank and pipe insulation is reducing the losses in the tank, particularly the standby losses and improving the efficiency of the solar hot water system.

To look at the impact of improved insulation of the solar hot water system on the cooling energy used in the Energy housing unit, the total cooling energy used for the months of March through August must be examined. To remove the effects of the outside temperature on this evaluation, a comparison of the percentage difference between the cooling energy used by the Standard home and the Energy home will be made. This comparison shows that in the months of March 2002 to August 2002 the Energy housing unit used 29% to 69% less cooling energy than the Standard housing unit. In the same period in 2001, this reduction ranged from only 3% to 48%. This suggests that the improvements in tank insulation may also have had a significant effect on the cooling load within the Energy home. However, the previously described deficiencies in the Energy Unit present in early 2001 make definite conclusions relative to this effect difficult.

In the Energy housing unit, three of the 100 watt incandescent lamps that were on the evening 4 hour timed duration were exchanged for 25 watt compact fluorescent lamps on June 4 th, 2002. This change did appear to have a small effect on the cooling load in the Energy housing unit. The relative cooling energy used by each of the housing units from June, 2002 through August 2002 showed a small change. The percentage reduction in cooling energy used by the Energy housing unit increased from about 30% to 38 percent. However, it is difficult to isolate the effects of the improvements in the solar hot water system insulation and the effects of the compact fluorescent bulbs. In any event, these effects appear to be much smaller than that produced by the hot water system changes.

4.3 Energy Analysis

The two housing units described in the previous sections were analyzed using a computer simulation program. The Energy Gauge Program (Version 1.25) developed by the Florida Solar Energy Center was used for the analysis. The Energy Gauge Program uses the basic DOE 2 energy analysis engine to provide an hourly energy use simulation for light commercial and residential structures [Danny Parker, et-al, 1999]. This program was developed to provide a simple to use interface for the DOE2 analysis program that more accurately analyzes the energy use of single and multifamily residences, and light commercial structures.

An analytical model was developed for each of the manufactured home units. These models were essentially the same with differences only in the R-values in the various building envelope components, the duct leakage values, the heating and cooling equipment and the fenestration properties. Figure 4 shows the wire model of the building envelope configuration used for the Standard Home. The Energy Unit model was similar.

The envelope leakage values were measured and these values were used in the analysis (See Table1). Table 1, and Figure 4 also show the window and door U values as well as the HVAC system properties for the unit. In addition, a uniform three-foot crawl space was assumed in the analysis of both houses. The Input Summary Sheets for each of the Energy Gauge runs are shown in Figures 4 and 5. It should be noted that the solar hot water heater was not incorporated in the analysis, a standard electrical unit was assumed in both unit’s analyses.

The analysis of each of the manufactured housing units was also repeated using the newest version of the Energy Gauge program, Version 2.0. This program was reported to have made changes in the analysis modeling and incorporated a number of “bug fixes”. The same input files were used for both set of analyses, Version 1.25 and Version 2.0.