The Robo Retrofit House

Much is learned from a side-by-side study of a retrofit home and standard-built home.

March 05, 2011
March/April 2011
A version of this article appears in the March/April 2011 issue of Home Energy Magazine.
Click here to read more articles about Retrofit

A home built in the mixed-humid climate of Knoxville, Tennessee, with energy efficiency upgrades used 42% less energy than an identical standard built home in a side-by-side measured and modeled comparison. The two-story homes have three bedrooms, 21/2 baths, and 2,400 ft2 of living space. The house type, size, and dominant features were selected after a demographic housing study in the Tennessee Valley Authority (TVA) service territory determined that this was a typical house on the market from 2000 to 2008. The energy efficiency packages, all commercially available in the area, made a significant difference in energy use between the two houses. The Retrofit House has a HERS index of 68 (a HERS of 0 is a zero-energy house; a conventional new house would have a HERS rating of around 100). The identical Builder House next door has a HERS of 102.


The Retrofit House and the Builder House are both built with 2 x 4 wood frame construction and both homes were built with Kraft-faced R-13 batt insulation in exterior walls. The Retrofit House has an unvented insulated attic; supply mechanical ventilation; ducts inside the conditioned space; single hung, dual pane, low-e, gas-filled windows; Energy Star appliances; and a single heat pump compared to the Builder House, which has two heat pumps. The Retrofit House and the Builder House, which served as a control, were built side-by-side through collaboration between the Tennessee Valley Authority (TVA) and DOE. The authors, Jeff Christian and Tony Gehl of Oak Ridge National Laboratory and Rex Dockery of Conservation Services Group, studied construction costs, measured energy consumption compared to predicted energy consumption, and conducted a cost analysis using typical weather year and average U.S. residential energy consumption to learn how a standard retrofit package can save energy and money for the homeowner. We also wanted to learn whether a standard retrofit package done on a large scale could significantly reduce energy use in the region.

We used detailed data measurement at 15-minute intervals for a year to generate a computer simulation of the all-electric Retrofit and Builder Houses with typical three-person occupancy patterns and energy services (that explains the “Robo” in the title). Energy consumption for the Retrofit House is predicted to be, on average, 40 kWh/day, at a cost of $3.76/day, including the current hookup charges of $0.24 per day. By contrast, the Builder house would require $6.46/day. These energy costs are based on average 2009 local residential rates of $0.093/kWh. The Retrofit House costs were calculated utilizing the actual builder cost invoices, including the subdivision lot, 15% overhead, and profit, totaling $253,800 or $105/ft. We used the detailed actual cost data for the Retrofit House and the Builder House to determine the incremental cost of the retrofit package and to conduct a neutral cash flow analysis.

Christian points out the insulated and sealed attic that allows 100% of the HVAC system to be located within the condition envelope. (ORNL)

The Retrofit House and the Builder House

Below is a comparison of the building envelope and mechanical features used in the Retrofit House, compared to the Builder House.

Walls. The Retrofit House and the Builder House are both built with conventional 2 x 4 framing. The exterior walls of both houses were insulated with R-13 batt insulation with a framing factor of 0.23. The blower door identified excessive leakage around the patio and kitchen doors in the Retrofit House. These were tightened up, resulting in a June 9, 2009, test result of 3.43 ACH50. The reading for the Builder House, which was not tightened up, was 5.7 ACH50.

Windows. The only retrofit to the vertical envelope on the Retrofit House was done on the windows. The windows used in the Builder House were replaced with low-e, gas-filled windows. The glazing in these windows has a U-factor of 0.35 and a SHGC of 0.34, with some overhang to provide shade, compared to the glazing in the Builder House, which has a U-factor of 0.5 and SHGC of 0.58 and no overhang. There is 294 ft2 of window area in both houses, with 107 ft2 on the south, 8 ft2 on the east, 15 ft2 on the west, and 166 ft2 on the north. The windows were installed according to best practices. Both caulk and expanding foam were used appropriately to seal the window frames.

Attic. The attic of the Retrofit House was sealed off to form a cathedralized, conditioned space. The attic was sealed with 2 inches of extruded polystyrene (XPS) blocking along the soffits, and the entire attic envelope was insulated with spray foam topped with 2 inches of sprayed fiberglass to an R-30 value.

Space-conditioning equipment. Heating and cooling design loads for the Retrofit House were calculated using ACCA Manual J (2004). The total cooling load is 22,000 Btu per hour. The heating design load is 33,000 Btu per hour. A 3-ton 16 SEER HVAC (9.5 heating seasonal performance factor, or HSPF; 36 kBtu/hour cooling capacity) was installed in the sealed, conditioned attic of the Retrofit House. The single heat pump in the Retrofit House serves both floors with a zone control air-side system. It has an indoor ECM (electrically commutated motor) circulating fan. Two heat pumps adding up to 4.0 tons were installed in the Builder House (a 2.5-ton and a 1.5-ton heat pump, both SEER 13, with 48 kBtu/hour cooling capacity). The air handlers in the Builder House are located in the unconditioned attic and the garage.

The HVAC subcontractor selected a 3-ton system that he felt compensated for the insulation and air tightness quality typically delivered on his job site. The mechanical fresh-air system in the Retrofit House is a 6-inch duct running from a vent on the north-facing roof slope. A motorized damper is controlled with an air cycler so that on average 30 CFM is pulled into the return side of the heat pump indoor coil. (The duty cycle varied depending on the amount of time the indoor fan was on high speed, low speed, and off. In general it lasted about 18 minutes out of each hour the air handler was on.)

Water heating. The 50-gallon electric water heater in the Builder House (energy factor, or EF, of 0.91) was replaced with a 50-gallon electric heat pump water heater in the Retrofit. The hybrid unit has an estimated annual coefficient of performance (COP) field performance of 2.1, exceeding the Energy Star guidelines for water heaters. It is located in the unconditioned garage and utilizes the heat in surrounding air to heat water. A unit installed in the Retrofit House for more than a year performed closer to a measured COP of 2.2.

Appliances. Both the Retrofit House and the Builder House are furnished with appliances running under simulated occupancy, where the refrigerator and freezer doors open on a timed schedule, and the oven, dishwasher, washer, and dryer run on timed cycles based upon the day of the week. The scheduling of all these appliances is based upon occupancy profiles established by Building America. An Energy Star refrigerator (421 kWh per year, compared to 501 kWh per year in the Builder House), clothes washer (774 kWh per year versus 891 in the Builder house, using hot water; washer uses 13.5 gallons/day less of hot water than the one in the Builder House), and dishwasher using 0.5 gallons/day less hot water are part of the retrofit package.

Lighting. The Retrofit House uses 100% fluorescent lights and 695 kWh/yr in energy; the Builder House uses 95% incandescent lights, 5% fluorescent, and 2,318 kWh/yr in energy.

Floor plans. The Retrofit House and the Builder House have the same floor plan. The house is laid out on two levels; the first level has an open floor plan. The two-car garage is insulated, attached and unconditioned, and there is a conditioned space above the garage. The walls adjacent to the garage, the exterior walls of the garage, and the floor of the bonus room above the garage are insulated in the original floor plan. However, the insulation between the garage and the bonus room is done poorly. There is no blocking of uncontrolled airflow anywhere, and there is a 12-inch air gap between the floor of the bonus room and the top of the insulation installed above the garage drywall, in a floor truss area that is 18 inches deep. The second level consists of three bedrooms and two full baths. The framing for the tray ceiling in the master bedroom poses a potential insulation problem, because it requires extra measures to insulate above it evenly. In the Retrofit House this is not an issue, because the insulation layer is up under the roof sheathing, not on the attic floor.

Attic Ambient and Floor Temperatures

Figure 1. Attic temperatures were measured on one of the hottest days in 2009.

Energy Use in the Two Houses

The average local electricity rate in 2009 for the area was $0.093/kWh. (The national average in 2009 was $0.12/kWh.) The Builder House, of the same size and with the same orientation and

Figure 1 shows the ambient and floor surface temperatures of the attics of the Retrofit House and the Builder House, respectively, on one of the hottest days in 2009. The ambient attic temperature of the Builder House fluctuates more than that of the Retrofit House because of the vented construction used in the Builder House; this construction is typical in the mixed-humid climate of Knoxville. The ambient temperature in the sealed, conditioned attic of the Retrofit House fluctuates much less from attic floor temperature when the interior ambient air temperature in both houses is set at 76°F. The lower attic temperatures in the Retrofit House provide conditions that are close to ideal for installing an HVAC unit in the attic space, as opposed to in a vented attic and or an unconditioned garage.

In the summer of 2009, the maximum hourly average temperature in the insulated and sealed attic of the Retrofit House was 82°F—6°F higher than the thermostat set point for both houses, 76°F. The maximum hourly average temperature in the attic of the Builder House during the same period was 131°F. There is no dedicated conditioned HVAC supply or return serving the attic in the Retrofit House.

In January, the coldest month of the heating season, the minimum temperature in the attic of the Retrofit House was 69.5°F. This is only 1.5°F below the thermostat set point of 71°F for the houses. The average temperature in the attic of the Retrofit House was 72.6°F. This is only 1.5°F higher than the thermostat set point. In January, the minimum temperature in the conventional ventilated attic of the Builder House was 21°F, and the average attic temperature was 45°F.

Figures 2 and 3 show the energy consumption for water heating in the two houses throughout the day on June 20, 2009. The Retrofit House water heater clearly uses less energy to recover from little to no standby heat loss throughout the day; while the peaks in consumption for the water heater in the Builder House show heavy usage to recover from standby heat loss, as well as heavy consumption right after showers.

The energy factor (EF) of the standard electric water heater in the Builder House is 0.91, and the COP of the heat pump water heater in the Retrofit House was estimated at 2.1. The hot water demand is generated in both houses with simulated-occupancy water usage. The average daily demand for hot water in the Retrofit House is 52 gallons, compared to 66 gallons for the Builder House; this difference is due to the more water-efficient clothes washer and dishwasher in the Retrofit House.

Water Heater Energy Use on June 20, 2009

Figure 2. The peaks in consumption for the water heater in the Builder House show heavy use to recover from standby heat loss, as well as heavy consumption right after showers.

Total Water Heater Energy Consumption on June 20, 2009

Figure 3. Typical total energy savings from water heating in the Retrofit House was 63% compared to the Builder House.

Energy Savings in the Two Houses

Our monthly energy consumption values are based on a combination of measurements and modeling. Since the houses have simulated occupancy from July 1, 2009, to February 1, 2010, the loads were estimated using the Building America Benchmark modeling procedure. The simulated data predict a 50% savings for the Retrofit model over the Builder model during the heating season, and a 38% savings during the cooling season. Installing 100% CFLs in the Retrofit House results in a 64% energy savings for lighting compared to the Builder House, which has 95% incandescents and 5% fluorescents. The Retrofit House would use only two-thirds as much energy as the Builder House.

We studied the energy savings that would be realized by installing individual upgraded technology and equipment in the Retrofit House, compared to standard technology and equipment in the Builder House. With all the upgraded technology and equipment installed in the Retrofit House, Energy Gauge predicts a savings of $986 per year. The greatest savings based on individual technologies would be realized by placing the ducts in the conditioned space. Further savings would be realized by, in order of the amount saved, installing the higher-efficiency heat pump, installing all CFLs, installing the hybrid water heater, and sealing the attic to achieve a tighter envelope. In reality, these features are all integrated and the results of installing any one feature cannot always be considered in isolation. This demonstrates the critical importance of locating the HVAC system in conditioned space.

Cost of the Retrofit

The detailed invoice level costs data on the Retrofit House and the Builder House were used to generate the incremental costs for each of the energy-efficient technologies beyond those found in the Builder House. The retrofit package was evaluated on the assumption that the homeowner borrows money for a ten-year term at 6% interest. Overall, the incremental cost of the entire retrofit package is $9,111. After figuring in federal and TVA incentives (available as of April 2010), the cost drops to $7,111, with a positive cash flow to the homeowner of $35 per year.

The hybrid water heater was installed in the Retrofit House in March 2010 at a cost of $1,498. The cost of the electric water heater in the Builder House was $278, resulting in an incremental cost for the hybrid water heater of $1,220.

The cost of the 294.5 square feet of regular double-glazed windows in the Builder House was $3,702.99. The cost of the upgraded double-glazed, low-e, gas-filled windows used in the Retrofit House was $3,952.07. The incremental cost used for these windows is $250, or $0.85/ft2. The window retrofit has a positive cash flow of $30 without incentives.

The same contractor and the same brand of Amana heat pump were used in both the Builder House and the Retrofit House. The Builder House was equipped with a 2.5-ton 13 SEER 7.7 heating seasonal performance factor (HSPF) unit in the attic, servicing the upper level; and a second 1.5-ton 13 SEER 7.7 HSPF unit with 10 kW of resistance backup located in the unconditioned garage servicing the main level. In the cooling season, the attic unit used 70% of the cooling energy in the Builder House for three of the hottest months of the year—July, August, and September 2009. The unit that was located in the worst environment, a hot attic, was called upon to provide most of the house’s cooling. The HVAC contractor was asked to keep very good cost records for these installations. His cost to the builder for installing the two units was $7,143.75 for the Builder House.

For the Retrofit House, the Manual J calculation showed that a 2- to 2.5-ton unit was the right size for the single heat pump to be located in the insulated and sealed attic. However, the HVAC contractor felt that a 3-ton unit was more appropriate, based on his experience with the quality of construction in the area. This unit had a SEER 16 and a HSPF of 9.5. The supply and return duct system layout was very similar in both houses, except that in the Retrofit House a return trunk line had to run through a bedroom closet on the upper level to the ceiling of the hallway on the first floor; and a supply trunk had to connect the unit in the attic to the supply duct system located between the two levels. This second large trunk also ran through a corner of the bedroom closet. Motorized dampers, a zone control board, and a 6-inch ventilated air duct connected to the return plenum of the unit were also needed in the Retrofit House. The HVAC contractor found that his expenses were “about” the same for these two systems. The invoiced cost for the HVAC in the Retrofit House was $7,143.75—exactly the same as the invoiced cost for the HVAC in the Builder House.

Placing the ducts inside the conditioned space produced the largest return on investment, followed by replacing two heat pumps totaling 4.0 tons of capacity located outside the conditioned space in the Builder House with a single 3-ton zone-controlled unit located inside the conditioned space in the Retrofit House. The incremental cost of placing the ducts inside the conditioned space was assumed to be zero. The modeling results from Energy Gauge were used to predict that getting the HVAC system from almost completely outside the conditioned space to 100% inside it would save 3,921 kWh—or 37% of the total energy savings to be realized by this retrofit package.

Insulating and sealing the attic has the largest first cost. The cost of the foam insulation and the 2 inches of Spider used to cover the foam in the Retrofit House was $4,000. The result was that R-30 was installed under the roof sheathing and on the gable walls, compared to R-30 of blown-in fiberglass installed on the floor in the conventional vented attic in the Builder House. The original design of the Retrofit Home specified 1 to 3 inches of foam, to control moisture and air seal the attic; this foam was to be covered with the less-expensive spray-applied fiberglass. After repeated attempts by the builder, the fiberglass would not build up to R-30. So 6 inches of mostly low-density foam were installed, followed by 2 inches of spray-applied fiberglass. The fiberglass provides a flame-retardant cover and adds R-8 to the assembly. The incremental cost of $5,916 was calculated by soliciting several foam quotes for R-30 alone on an attic sheathing and gable area totaling 1,972 square feet and similar to that of the Retrofit House, at a cost of $3/ft2. This cost includes the cost of sealing the soffit, gable, and ridge vents and working in a more confined space than would be the case in a real retrofit application.

In the Retrofit House the foam and fiberglass were installed in the attic without the top-floor drywall in place. However the HVAC unit had been installed prior to insulation, and working around it added to the labor cost. In a real retrofit, this work would be staged so as to install the insulation prior to setting the replacement HVAC unit and ducts in the attic. The incremental cost of sealing and insulating the attic was 65% of the total incremental cost. Sealing all envelope leaks in the top-floor ceiling and placing the ducts and indoor heat pump unit inside the conditioned envelope reduced the whole-house ACH50 from 5.8 in the Builder House to 3.43 in the Retrofit House. This accounted for 45% of the whole-house retrofit package savings.

The Builder House depends on the bathroom exhaust fans for ventilation. Typically people who live in these houses do not run the bath exhaust fans unless they are taking a shower or a bath. However, in both research houses the mechanical ventilation is controlled at an average of about 30 CFM. This decision is based on the fact that both of these simulated occupancy houses can be argued to have similar indoor air quality. Energy usage is simulated; indoor air pollutant generation is not. The energy penalty for running the bath exhaust fan 24/7 and pulling a measured 30 CFM from the Builder House is 470 kWh per year.

Model Retrofit?

Based on measured data from almost 100 sensors and a computer simulation of the Retrofit House, the energy for this all-electric house is predicted to cost only $3.76 per day. By contrast, energy for the Builder House would cost $6.46 per day. Based on a full year of measured data with the houses operated under simulated occupancy, the Retrofit House used an average of 40 kWh per day. The approximately $10,000 incremental cost of the retrofit package described in this article, assuming that windows, heat pump, and water heater are in need of replacement, has a positive cash flow to the homeowner. With the base house being a typical home built in 2005–2008, and local electric rates more than $0.02/kWh less than the U.S. national average, the 42% whole-house savings realized in this project should be exceeded in retrofits of most other homes that were constructed prior to 1990.

Jeff Christian and Tony Gehl are researchers in the Buildings Research and Technologies Integration Center (BTRIC) at the Oak Ridge National Laboratory (ORNL). Rex Dockery is an energy retrofit designer and evaluator with Conservation Services Group.

For more information:

Floor plans for both the Retrofit House and the Builder House and more details of the study are shown in Christian, J.E., and T. Blazer. 2,400 ft2 Retrofit House That Has 42% Energy Savings Compared to 2008 Builder Standard in the Mixed-Humid U.S. Climate, ORNL Draft Report, June 2010.

Jeff Christian, Tony Gehl, Philip Boudreaux, Joshua New, Rex Dockery. Tennessee Valley Authority’s Campbell Creek Energy Efficient Homes Project: 2010 First Year Performance Report July 1, 2009–August 31, 2010, ORNL/TM-2010/206, November 2010. This information on the Retrofit House, the Builder House, and a third test house built alongside them, can be downloaded at

Discuss this article in the Best Practices (Residential) group on Home Energy Pros!

Add a new article comment!

Enter your comments in the box below:

(Please note that all comments are subject to review prior to posting.)


While we will do our best to monitor all comments and blog posts for accuracy and relevancy, Home Energy is not responsible for content posted by our readers or third parties. Home Energy reserves the right to edit or remove comments or blog posts that do not meet our community guidelines.

Related Articles
SPONSORED CONTENT What is Home Performance? Learn about the largest association dedicated to home performance and weatherization contractors. Learn more! Watch Video