From Energy Guzzler to Near Zero Energy Home: Lessons from the Phased Deep Retrofit Project

December 06, 2017
Winter 2017
A version of this article appears in the Winter 2017 issue of Home Energy Magazine.
Click here to read more articles about Retrofit

Between 2012 and 2014, DOE sponsored the Phased Deep Retrofit (PDR) study in 60 all-electric Florida homes. The purpose of the study, which was conducted by researchers at the Florida Solar Energy Center, was to establish annual energy and peak energy reductions from two levels of retrofit—shallow and deep. As part of the study, one of the homes with very large starting energy use received extensive retrofits, enabling it to move from a large user to a near zero energy home. Results from the ten-home deep retrofit segment showed savings large enough (averaging a measured reduction of 39%) that some homeowners chose to go further by adding a solar photovoltaic (PV) generation system. This paper describes how the owners of the home referred to above moved over a period of four years from excessive electricity consumption to near zero energy in a hot-humid climate

Phased Deep Retrofit Project

We obtained detailed audit data for all ten homes in the deep retrofit project. These data included the size and layout of the house, insulation levels, materials, finish, and mechanicals. A blower door test was run on each home. Detailed photographs were made of the home’s exterior, appliances, mechanicals, and thermostat. All of these data were collected during the retrofits. The flow rate of showerheads was measured during the shallow retrofit, and duct testing was conducted as part of the deep retrofit.


The 60 PDR study sites are located in central and south Florida. The 60 homes in the study were built between 1942 and 2006, using various methods of construction. They average 1,777 square feet in living area, with an average occupancy of 2.6 persons. Homes were audited and instrumented during the second half of 2012. Shallow retrofits of all 60 homes were conducted in spring 2013. Deep retrofits of the 10-home subsample were conducted in summer and autumn of that year, following analysis of the six-month shallow impact at individual sites.

End-use energy data were collected to evaluate energy reductions and the economics of each retrofit phase. Monitoring included whole-house power consumption and power consumption of various end uses such as heat pump compressor, air handler and resistance heat, water heating, clothes dryer, range, refrigerator, freezer, and swimming pool pump. Several spare monitoring channels were used to pick up unconventional end uses like hot tubs and wine coolers. With all major end uses submetered, the category of “lights and other” energy was arrived at by subtracting all submetered loads from total energy use.

Monitored Data

Figure 2 shows energy end use for each of the 60 sites in 2013. The average for the entire sample is shown by the far left bar labeled “All” in Figure 2 and is illustrated in more detail in Figure 3. Average annual consumption across sites was 42.8 kWh per day, although the highly diverse end uses in this study pose a complex challenge for efficiency programs. Heating, cooling, and water heating comprised 48% of measured consumption. Difficult-to-tackle loads, such as clothes dryers and home entertainment centers, accounted for 9% of consumption, and home office, game consoles, lighting, fans, and other plug loads accounted for 23% of consumption.

Shallow Retrofits

Shallow retrofits were conducted in all 60 homes from March to June 2013. The energy reduction measures were chosen based on ease of installation. These targeted lighting (CFLs and LED lamps), domestic hot water (tank wraps and showerheads), refrigeration (cleaning of condenser coils), pool pump (reduction of operating hours), and the home entertainment center (smart plugs).

Most houses already had some energy-efficient (defined as CFL or LED) lighting. Indeed, one home already had 100% LED lighting, while six other homes had mostly CFLs and needed fewer than 20% of bulbs changed. Owners sometimes objected to lighting retrofits for some lamps, so those were not changed. A total of 55 homes were affected by the lighting retrofit. On average, 54% of lamps were replaced with CFLs or LEDs, ranging from 5% to 96% of the home’s total lighting.

Phased Deep Retrofits Electricity Use by End Use

Phased Deep Retrofits Electricity Use by End Use
Figure 1. Site energy end use for 2013 tracked against mean indoor temperature and RH for the 60 homes in the PDR study.

Electricity by End Use January 1, 2013 to December 31, 2013

Electricity by End Use January 1, 2013 to December 31, 2013
Figure 2. Average energy end use for 2013 for the 60 homes in the PDR study.

We reduced DHW energy consumption in two ways—by reducing use with low-flow showerheads and by reducing storage thermal losses by insulating tanks and piping. Many homeowners refused to have low-flow showerheads installed, so the greatest gains came from reducing storage thermal losses.

Phased Deep Retrofit Site 19: Description of Site

As can be seen in Figure 1, Site 19 was one of the highest energy users without a swimming pool having very high HVAC, water heating and dryer loads. Site 19 is a 2,554 ft2 single-story home built in 1988 with three occupants, one of whom (the daughter) is home only periodically. Construction is slab-on-grade with concrete masonry unit walls and average 10.6 foot ceilings. Windows are double glazed, some untinted with metal frames and others with solar-control glass and vinyl frames. Home airtightness as originally tested was 6.51 ACH50, which is typical of this age Florida home. Existing systems consisted of a 5 ton heat pump system with the original 1988 condenser (the compressor was replaced in 1997); a manual thermostat; a 50-gallon electric water heater; and R-19 fiberglass batt ceiling insulation.

Shallow Retrofit Measures Taken at Site 19

The shallow retrofit at site 19 was performed on April 17, 2013. It involved replacing incandescent lamps with CFL or LED lamps (there were 89 audited lamps in the home); wrapping the hot-water tank and pipes with R-3 insulation; and replacing two showerheads with low-flow showerheads. The total lighting load was changed from 3.394 kW to 1.529 kW—a reduction of 55% in lighting wattage. The achieved savings is the quantity produced for fixtures given their operating hours, which may vary substantially from the lighting load reduction. At site 19, the lighting retrofit saved about 0.65 kWh per day, or about 240 kWh per year.

The shallow retrofit also reduced daily average water-heating energy use from 13.3 kWh per day to 11.8 kWh per day for an 11% savings.

Deep Retrofit Measures Taken at Site 19

The following deep retrofit measures were taken at site 19. (1) A high-efficiency heat pump was installed (2) ducts were sealed (3) and passive return air pressure relief installed in the master bedroom. (4) A smart thermostat was installed. (5) A heat pump water heater was installed to replace an electric water heater. (6) Ceiling insulation was upgraded to R-38, using blown-in fiberglass; knee wall insulation was upgraded as appropriate. (7) Existing appliances were replaced with an Energy Star clothes washer and a low-energy clothes dryer. Deep retrofit measures began on August 26, 2013, with the HVAC system install concluded on November 18, 2013 with final appliance installations.

Cooling-Energy Savings at Site 19

The home’s existing 10 SEER 5 ton York heat pump was replaced with a 5 ton two-speed 16 SEER Carrier heat pump. Ducts were tested and sealed, and a Nest learning thermostat was installed. Normalized duct leakage averaged 0.09 Qn,out presealing and 0.05 postsealing. Figure 4 displays daily HVAC retrofit savings analysis at site 19. Postretrofit savings averaged 47% (19 kWh per day) as evaluated by regression results applied to the entire period. There is a large reduction in cooling-energy use from week pre to week post (from 76.8 kWh to 37.9 kWh per day). Figure 5 shows the change in interior temperature after the combined retrofit on August 26, 2013. The occupants maintained a slightly higher indoor temperature with the Nest learning thermostat.

Site 19 AC Energy Use

Figure 3. Site 19 AC energy use pre- and postretrofit, May–October 2013.

Site 19 Indoor and Outdoor Temperatures

Site 19 Indoor and Outdoor Temperatures
Figure 4. Average indoor and outdoor temperature May–October, 2013.

To evaluate weather-related influences, we used the pre and post daily air-conditioning data and then determined daily cooling kWh as a function of the average daily air temperature, using quadratic regressions. At a daily average outdoor temperature of 80°F, the regressions indicated that the AC used 61.0 kWh per day preretrofit and 30.8 kWh per day postretrofit, for a savings of 30.2 kWh per day savings or 47%. This represents the HVAC retrofit overall savings, including AC changeout, duct sealing, and Nest learning thermostat installation. Through examination of pre and post interior temperatures we attempted to separate out the savings attributable to the learning thermostat. In the month before the retrofit, the occupants maintained an average temperature of 75.1°F. In the month after the retrofit, the interior temperature rose to an average of 75.4°F. Differences in regression against outdoor temperatures and another regression that examined changes to cooling use against the outdoor-to-indoor temperature difference enabled us to evaluate how the learning thermostat influenced savings. At an average daily summer outdoor temperature of 80°F, the average outdoor-to- indoor temperature difference for the regression was 5.34°F pre and 4.66°F post.

Evaluating pre- and postretrofit energy consumption at the post-Nest temperature difference, the predicted pre consumption falls to 58.4 kWh per day. Thus, when controlling for changes to indoor-to-outdoor temperature, the AC retrofit and duct repair reduced consumption by 26.3 kWh per day (from 58.4 to 32.1 kWh), or 45%. The remainder of the cooling-energy savings (2.1 kWh per day) comes from the learning thermostat with an implied cooling-energy savings of ~4%. For sake of comparison, a separate evaluation of Nest thermostats alone in the PDR project showed cooling savings averaging about 10% in a sample of 22 homes, but with high variability.

Heat Pump Water Heater at Site 19

On September 19, 2013, the water heater in the garage at site 19 was changed from a 50 gallon electric-resistance GE model GE50M06AAG to an A.O. Smith 80 gallon heat pump water heater. Evaluated over 30 days pre- and postretrofit, the change reduced water- heating loads by 67% or 5.8 kWh per day (weather normalized). This was one of the most successful energy retrofits in the PDR project, which collectively averaged 65% savings in preretrofit water-eating electricity use. Figure 6 shows the water-heating energy consumption data plotted a year before and a year after the heat pump water heater was installed at site 19. The figure shows both the impact of the shallow retrofit in mid-April 2013 (two showerheads changed to low-flow models and tank wrapped with exterior insulation) and that of installing the heat pump water heater in September 2013. Consumption prior to intervention is cut by approximately 80% (14 kWh per day to 3 kWh per day) through the combination of new showerheads, tank wrap, and the heat pump water heater.

Site 19 Water Heater Energy Use

Site 19 Water Heater Energy Use
Figure 5. Daily water-heating energy consumption a year before and a year after installation of heat pump water heater: September 2012 to September 2014.

Clothes Dryer

Electric-resistance clothes dryers were replaced with a new higher-efficiency Samsung DV457 model in eight homes, including that in site 19. Data analysis of the eight-home sample revealed that achieved energy savings of the clothes dryer were highly variable but averaged 0.60 kWh per day or 18%. Savings were lower than expected, as homeowners, unhappy with long drying times in ECO mode, opted for the less efficient but quicker standard mode. Site 19, with the largest PDR clothes dryer use (8 kWh per day), showed a savings of 26% (2.0 kWh per day) in the 60-day period following the installation of the new unit on November 18, 2013.

Change to Heat Pump Clothes Dryer

The PDR project conducted further clothes dryer research in 2015 by replacing the electric-resistance unit in eight homes, including site 19, with a new Whirlpool heat pump clothes dryer (HPCD). The dryer model (WED99HED) is designed to be approximately 40% more efficient than standard units. The dryers were matched with a Whirlpool 4.5 ft3 washer (WFW95HED). This washer has an Energy Guide label of 109 kWh per year, and an MEF of 3.2.

The HPCD is a 7.3 ft3 condensing unvented clothes dryer, similar to European models. It has both a heat pump section and a supplemental electric heating element. There are three operation modes: ECO, which uses only the heat pump, but has long dry times; BALANCED, which uses the heat pump and the supplemental heating element when needed; and SPEED, which uses them both at once for fastest drying.

As explained earlier, site 19 had started the PDR project with a standard washer and dryer but then had undergone retrofits which included the more- efficient Samsung DV457 dryer. With continued heavy dryer use at site 19, we wanted to see what impact the HPCD dryer would have on dryer energy usage.

Measured dryer baseline data were from January 1, 2014, to the install date of June 3, 2015. The post period data were from install date through mid-July 2015, a period comprising about 45 days. Table 1 summarizes the measured data for site 19 as well as PDR study averages.

Table 1. Summary Heat Pump Clothes Dryer Energy for Site 19 and the Entire PDR Study

*Original, standard dryer at Site 19 used 8.30 kWh per day prior to November 18, 2013.

Median energy savings for the eight-home sample were estimated at 312 kWh per year or about 44% of baseline consumption, with average savings of 359 kWh per year (39%). The savings for site 19 would be 35% if based on the original baseline unit, rather than on the more-efficient Samsung DV457 unit in operation in 2014. The monthly energy use for clothes drying at site 19 shows a large drop in daily dryer energy use following installation of the Samsung unit and a smaller drop after the HPCD is installed, but with increasing long-term dryer energy consumption as occupants selected less efficient dryer modes over time. A change in occupancy at the end of 2014, when the number of occupants increased to four, has a strong impact on daily average consumption. Average daily use falls from about 8.5 kWh per day to about 5.5 kWh per day with the more-efficient resistance unit, and then to about 4.5 kWh per day initially with the HPCD. However, in subsequent months the household chose more energy-intensive dryer cycles, which increased energy use.

The homeowners at site 19 were not pleased with the unvented Whirlpool HPCD, as it made the utility room very warm. In the future, we hope to substitute a vented HPCD (manufactured by LG) to see how well it will work in this home—which remains the site with highest dryer energy use—although that use has been cut by 25–30%. Occupants strongly prefer to do daily laundry with very dry cycles, even though they know that it uses a lot of energy.

Summary of Retrofits

Considering all the retrofit measures, we were able to estimate that total electricity use at the previously high-consuming home at site 19 was cut by 47% or nearly half. Figure 7 shows site 19’s total daily energy use from September 2012 through June 2014. Figure 8 shows site 19’s energy use broken down by end uses before intervention and after the completion of all retrofits. The left bar, showing an average daily energy use of 72 kWh, represents the home during the first year of monitoring (September 1, 2012 to August 31, 2013). During this first year, no energy measures were implemented, except for the shallow retrofit on April 17, 2013, and a few deep retrofit measures installed on August 26, 2013. Energy use at site 19 was originally dominated by heating, cooling, and water heating which made up 68% of measured consumption.

The middle bar in Figure 8 shows the postretrofit average daily energy use, which dropped to 46 kWh for a savings of 38%. Reduction to heating, cooling, and water-heating usage accounts for most of the savings (26.8 kWh per day). The more difficult loads to address, such as clothes drying, home entertainment, lighting, fans, and other plug loads now make up a far greater share of total remaining energy use. The large amount of dryer energy use remaining at site 19 is apparent. The right bar represents the entire sample of PDR homes. These homes had an average floor area of 1,777 square feet, making them 30% smaller than site 19.

Whole House Power at Site 19

Whole House Power at Site 19
Figure 6. Change in daily electricity use at site 19: 2012–2014 as retrofits installed.

Site 19 and Average PDR Home Electricity by End Use 2012–2016

Site 19 and Average PDR Home Electricity by End Use  2012–2016
Figure 7. Measured annual electric end use before retrofits (left), after retrofits (middle), and average preretrofit home in entire PDR study (right).

Installation of 10kW PV System at Site 19

After seeing the large reductions to overall energy use at site 19, the homeowner decided to install a 10kW PV system to bring the home close to net zero average per year. The PV system consists of 36 280W modules feeding a pair of 5kW inverters with a rated efficiency of 97.0%. Each inverter has two strings with nine modules.

Figure 9 shows the average 24-hour performance of the 10kW PV system between April 30, 2015 and April 30, 2016. Home total energy use is averaging 44 kWh per day against daily PV power production of 31 kWh, leaving a net energy usage of 13 kWh per day for an 82% savings over the preretrofit average of 72 kWh per day.

PV Performance

PV Performance
Figure 8. Evaluated over a year, the PV output made up 82% of the total energy requirement for site 19.

Our results show that successful near zero energy solutions and reductions to utility peak electrical demand can be obtained from homes even judged to be guzzlers, and in a hot-humid region.

Our paper highlights the progress of site 19 in the PDR project. This home went from one of the highest energy consumers before intervention at 24,483 kWh per year to a measured consumption of 12,862 kWh per year, a 47% reduction. This was accomplished through a series of retrofits that were individually submetered and evaluated: efficient lighting, increased roof insulation, heat pump water heater, duct sealing, high-efficiency heat pump with smart thermostat, and heat pump clothes dryer. Measured consumption during the warm months from April to August dropped from 87 kWh per day in 2013 to 46 kWh per day in 2014. Encouraged by these results, the homeowner installed a 10kW PV system in April 2015. Since then, the home has had an average net electricity use of 13 kWh per day after accounting for PV system output (31 kWh per day). This represents an 82% reduction toward zero energy.

Danny Parker and David Chasar are research scientists at the Florida Solar Energy Center.

Karen Sutherland and Eric Martin, also from the Florida Solar Energy Center, contributed to this article.

Learn More

Osser, R., K. Neuhauser, and K. Ueno. Proven Performance of Seven Cold Climate Deep Retrofit Homes. Lowell, Massachusetts: Building Science Corporation, June 2012.

Parker, D. “Very Low Energy Homes in the United States: Perspectives on Performance from Monitored Data.” Energy and Buildings 41, no 5 (2009): 512–20.

Parker, D., et al. “Measured Results of Phased Shallow and Deep Retrofits in Existing Homes.” In Proceedings of the ACEEE 2014 Summer Study on Energy Efficiency in Buildings, 1, 261–76. Washington, D.C.: ACEEE, 2014.

Parker, D., et al. Phased Retrofits in Existing Homes in Florida Phase I: Shallow and Deep Retrofits. Golden Colorado: NREL, February 2016.

Rosenbaum, Marc, “Thriving on Low Carbon,”.

Sutherland, K., et al. “Measured Retrofit Savings from Efficient Lighting and Smart Power Strips.” In Proceedings of the ACEEE 2014 Summer Study on Energy Efficiency in Buildings, 9, 357–69. Washington, D.C.: ACEEE, 2014.

Thousand Home Challenge, 2015. “Thousand Home Challenge: Transforming America’s Energy Stock,” Affordable Comfort Institute.

  • 1
  • NEXT
  • LAST
Click here to view this article on a single page.
© Home Energy Magazine 2020, all rights reserved. For permission to reprint, please send an e-mail to
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 Insulated, Air-Sealed Drapes Learn more! Watch Video