This article was originally published in the January/February 1997 issue of Home Energy Magazine. Some formatting inconsistencies may be evident in older archive content.


| Back to Contents Page | Home Energy Index | About Home Energy |
| Home Energy Home Page | Back Issues of Home Energy |


Home Energy Magazine Online January/February 1997

Florida House Aglow with Lighting Retrofit

by Danny Parker and Lynn Schrum

Danny Parker is a principal research scientist and Lynn Schrum is a research assistant at FSEC.

In a residential lighting retrofit, how much energy can be saved with current technology? The Florida Solar Energy Center decided to find out by retrofitting every lamp in a Miami home.
Most lighting studies focus on average lighting energy use or on how much energy can be saved by retrofitting large numbers of homes. However, at the Florida Solar Energy Center (FSEC), we were interested in finding out how much lighting energy we could save in a single house. We picked one house with high utility bills and extensive interior lighting, thoroughly monitored it, and retrofitted every light we could. The study also helped us learn what sort of monitoring is most useful, and how residents respond to efficient lighting. What We Did and How We Did It First, we instrumented the house, a 1,341 ft2 single family South Miami home. We began monitoring it in baseline condition on April 27, 1995. Our initial method of monitoring was to isolate the lighting and plug loads from other major loads. We metered the electrical use of the refrigerator, the clothes dryer, and the heating and cooling systems in addition to total household usage. We subtracted the major loads from the total to isolate miscellaneous plug loads and lighting. If only lighting was altered, we could use the differences in the miscellaneous loads before and after the retrofit to estimate lighting energy savings.

However, the household had a lot of miscellaneous plug loads, including three TVs, two VCRs, six ceiling fans, a home computer system, electric clocks, a dishwasher, and a vacuum cleaner (see Phantoms Strike Miami). In order to keep these miscellaneous loads from distorting the data, we installed individual time-of-use light loggers or plug loggers on each of the lighting fixtures in the home to establish the actual on-time of each. This let us assess how monitoring pure lighting loads compared with calculating energy savings by subtraction.

We began light logger monitoring on August 8, 1995. We didn't have enough lighting loggers for all the lamps in the house, so we had to capture the pure lighting loads by metering two groups of fixtures at different times.

We inventoried the home's lighting, and found 40 lamps on 26 switches with a total connected load of 2.5 kW (see Table 1). In general, the lighting consisted of incandescent A-lamps of various wattages.

Researchers learned some lessons about placing loggers in enclosed fixtures. This logger continued to function, even after melting in an incandescent fixture.

Table 1. Lighting Inventory Before and After Retrofit.
Lamps Before Retrofit;Replacement Lamps
Counter 60W incand. globes (2) 15W CFL (3)
Overhead 100W, 60W, 25W incand. 30W circline (1)
Under counter 20W FL tube --
Dining room
Over table 50W halogen --
Aquarium 25W incand. 11W tube fluorescent
Living room
Floor lamp 75W incand. 15W CFL
Overhead anteroom 60W incand. globe 18W CFL
Table lamp 75W incand. 22W circline
Table lamp 100W incand. 22W circline
Florida room
Overhead 60W incand. globe 15W CFL
Table lamp 100W incand. (2) 22W circline
Desktop lamp 100W incand. 15W CFL
Portable swing lamp 13W CFL --
Table lamp 75W incand. 22W circline
Overhead lamp 60W incand. --
2 table lamps 60W incand. (2) 22W circline (2)
Master bedroom
2 table lamps 60W incand. (2) 42W halogen
Vanity 55W incand. (4) 15W CFL (3)
Master bathroom
Vanity 60W incand. (3) 40W incand. globes (3)
General overhead 100W incand. globe 15W CFL 
Desk lamp 100W incand. 22W circline 
Garage door 55W incand. 42W halogen
Front porch 60W incand. globe 60W incand. globes (2)¥
Back porch 65W PAR (4)
--:  No change
PAR:  Parabolic aluminized reflector, commonly known as floodlights
¥:  Motion sensor installed
Energy Use Before the Retrofit Both the subtractive method and the logging method showed a lot of lighting energy use. From five months of baseline monitoring, we estimated annual use at over 4,050 kWh! This is high, although a recent widescale monitoring project by Tacoma Public Utilities (see Shedding Light on Home Lighting Use, p. 15) found one home using 7,400 kWh per year for lighting. More enlightening than total energy use was what we learned about where in the house lighting electricity was consumed most (see Table 2). We looked at both the average number of hours fixtures were used within each room and the fraction of the measured daily lighting energy that was used in each space. Like the Tacoma study and a 1993 phone survey by the Lighting Research Center, we found that outdoor and kitchen lighting comprised the largest fraction of total lighting consumption.
Outdoor lights offer big potential savings, but pose problems for retrofitting and monitoring. These lamps were retrofitted using a motion sensor which provided some savings in the early evening. However, the residents usually overrode the control, and left the light on all night. To add to the problems, a dead moth on the photosensor disrupted the logging.
We expected the electric lighting loads to vary seasonally with day length. An early study by Pacific Northwest Laboratories (PNL) found a 40% variation from a high in December to a low in June. The recent Tacoma study showed 30% less lighting energy use in the lighter versus the darker months.

Table 2. Daily Electric Lighting Energy by Use
Location Daily kWh % of Total Average On
Outdoors 3.7 33.6 10.1
Kitchen 2.6 23.1 8.0
Garage 1.3 12.1 13.4
Study 1.1 9.8 6.8
Dining table
fixtures 0.6 5.3 5.3
bedroom 0.5 4.6 4.3
Bathroom 0.3 2.5 1.3
Aquarium 0.2 2.0 14.7
Bedroom 0.2 2.0 1.8
Florida room 0.2 1.9 1.1
Living room 0.2 1.7 0.9
bathroom 0.1 0.8 0.5
Hallway 0.1 0.1 1.0

In Miami, the shortest day is 10.6 hours and the longest is 13.7 hours, a 23% difference in available daylight hours. We found greater loads during the month of December due to holiday lighting (see The Electric Bill That Stole Christmas). The residents took a ten day vacation in July, which combined with longer days to reduce lighting use in that month. Eliminating these exceptional months, the variation in use between June and November was 24%.

Making the Switch Over a couple of days in early December 1995, we changed the lighting system in the home to efficient lamps and fixtures. We installed 27 new lamps or controls. We timed the installation to coincide with the winter solstice, so we could obtain similar seasonal data before and after the change. In most cases, we installed CFLs in frequently used interior fixtures, and motion sensor controls with PAR halogen lamps in exterior lights. We used incandescent halogen bulbs for infrequently used and hard-to-fit fixtures.

The connected household lighting load dropped from 2.5 to 1.1 kW--a reduction of 56%. In order to examine savings, we continued to log energy use for another six months.

Project installers were aided by the wide array of bulbs now available for retrofit purposes. This 5-inch triple tube 15W CFL worked well for several tight corners.
Not Perfectly Smooth Sailing Before ordering equipment, we examined and measured each of the fixtures and lamps that we wanted to alter. This made for a relatively trouble-free installation. As usual, the need for very short CFLs was apparent during the retrofit. Fortunately, the new triple-tube 15W CFLs were small enough to fit in many of the tight fixtures. The exterior-lighting motion sensors proved to be more difficult to install than we had anticipated because we needed to cut holes in the porch eaves for the control boxes. Still, it only took us a day to retrofit the whole home. 

Phantoms Strike Miami

Monitoring whole-house demand and subtracting the major loads is a good way to calculate changes in lighting use. However, miscellaneous loads affect the result. Since the first retrofit home was completed, we have started monitoring and retrofitting two more homes. In a field evaluation of one of them, we metered miscellaneous appliances with a sensitive digital power analyzer. These other loads, particularly the ones that were on all the time, were surprisingly large. While other homes probably don't have the same constant load as this one, it shows how other can account for a big part of the bill. Auditors attempting to monitor lighting loads need to be careful not to count these loads as lights.

Equipment left on included the aquarium pump, a couple of portable phones, and a security system. Items that had notable phantom loads--power use that takes place when the equipment is off--were the two TVs, a VCR, a portable stereo, and a laser printer (see HE July/Aug '96, p. 42, Off is a Three-Letter Word). 

We didn't meter the clocks or the flashlight charger, but they may have added a bit more. The total constant load was 88 W, which works out to 771 kWh per year. This equals about 4% of the house's total annual electricity consumption. 

Miscellaneous Loads
100-gal salt water aquarium pump 41 W
Entertainment center: 36-in TV, VCR, decoder 160 W-205 W, depending on screen; 18 W quiescent
Portable phone #1 2.1 W in use, 1.6 W standby
Portable phone #2 2.0 W in use, 1.2 W standby
Security system 15 W
Monitor & computer 115 W
Laser printer 250 W active, 6 W quiescent
TV 115 W on, 5 W off
Portable stereo 7 W on, 2 W off
Toaster oven 460 W toasting
Espresso maker 360 W on

Converting the lighting fixture over the dining room table proved an insurmountable challenge. This is an attractive fixture that is used to illuminate the table's centerpiece. Its dimmer controls a low-voltage miniature 50W halogen PAR. This fixture is frequently left on for long periods of time--an average of 9.4 hours per day--and is seldom dimmed.

At first, we planned to provide a motion sensor control to turn off the fixture when no one was present. However, the switch wiring was in a solid concrete wall, making installation difficult, to say the least. We couldn't install a CFL in the fixture because the residents desired continuous dimming. CFL dimming ballasts that use conventional dimmers have since become available.

There was a second dining table in the Florida room, a glassed-in porch common in our state. The lamp over this table was less frequently used and proved easy to retrofit. We replaced its 60W incandescent globe with a 15W CFL globe.

In the monitoring, we encountered difficulties using the light loggers on outdoor fixtures. A significant number of false positives were recorded on exterior fixtures during the day. At first, we believed this could be avoided by using the logger's built-in sensitivity adjustment to dull the photometric element so that it operated only when the lights were on. However, the sun can be very bright--obviously brighter than the researchers' anticipations! Fortunately, we were able to correct our data--on days when the lights appeared to have gone off in the morning and back on at midday, we were fairly sure that the sun was interfering with the loggers. Another outdoor light logger became ineffective when a dying moth, attracted to the fixture at night, fell onto the photosensor.

We recommend other sensing methods, such as clip-on current transducers, for anyone attempting to monitor outdoor fixtures. Transducers do not depend on light to show whether a fixture is on or not; when it senses current passing through lamp wiring, it records that the light is on.

Deck the halls with CFLs...
The particular type of logger we used has a status window to indicate whether the logger is sensing light. To set up the equipment properly, one must be able to see this status window. However, once the loggers are placed inside fixtures, status windows are often hidden. We found that an inspection mirror helped us see around corners to make sure sensitivity adjustments were correct.

We also learned about the danger of placing the loggers too close to lamps. By accident, we melted one of the loggers in the 180W incandescent kitchen drum fixture. Amazingly, the logger continued to take useful measurements for the entire period. Even so, such circumstances could create a fire hazard. 

The Electric Bill That Stole Christmas

In our study, we found much higher lighting energy use during December, thanks to the holidays. December had 27% more lighting use than November. We wanted to know just how much energy those Christmas lights use. Using a digital power analyzer, we measured the wattage of five strings of commercially available Christmas lights. Each was classified as a Decorative Lighting String by UL and was manufactured in China or Thailand.

The electricity use of the bulb types varied by a ratio of 46:1 between the highest and the lowest. In general, the miniature incandescent W2 bulbs use only a fraction of the power of the more traditional candelabra base bulbs. Since a typical Christmas tree may have at least 100 bulbs, and the house may have twice as many outdoor strings, connected decorative lighting loads could vary from a low of 60 W up to 3,000 W, depending on the types of holiday lighting system that are chosen.

Christmas Light Energy Consumption
Description No. of Bulbs  Measured Watts Watts/Bulb
Small clear bulbs, indoor/outdoor, 36W 100 34 0.34
Small midget globe bulbs, indoor, 10W 50 11 0.22
Small colored/clear bulbs, indoor/outdoor, 36W 100 33 0.33
C 71/2 5W bulbs, indoor/outdoor, 125W 25 128 5.1
Large 10W C bulbs, outdoor, 500W 50 504 10

How Do You Like It? Sometimes, to reduce wattage, we reduced effective illuminance, with the approval of the homeowners. For instance, one occupant thought the bare 60W incandescent globes on the master bathroom vanity were too bright and asked for something with a softer light for the retrofit. Based on the light logger data for this fixture, we knew that the vanity lighting was not typically left on for long periods of time. Thus we substituted conventional 40W incandescent globes to meet both objectives of reduced wattage and lower light output. In other cases, the light output from the newer fixtures actually increased; we communicated with the residents to ensure that they were satisfied with the changes.

The occupants responded positively to most of the changes. However, a CFL globe with a magnetic ballast was unacceptable for the Florida room dining table due to its annoying start up flicker. (This same lamp worked fine in the garage.) Similarly, the ten year old boy who frequently uses the second bathroom was at first surprised by the half second required for the electronically ballasted CFLs in the vanity lighting to start up. The mother preferred the retrofitted kitchen task lighting and saw no effective difference in the other fixtures. The father noticed no real change in lighting quality, in spite of his stated preference for a well-lit home (see They Like It, They'll Pay, and It Works).

The family was dissatisfied with the outdoor-lighting motion sensors until they realized that they could override the controls to keep the lights on. This proved to be a weak link in the overall retrofit. The new front-porch lighting was higher wattage than before, and the motion sensors were usually overridden during evening hours.

Similarly, the outdoor-lighting retrofit in the rear included a porch light with a motion sensor that was frequently overridden. The occupants left the light on from midnight until morning after the retrofit, although they hadn't always left it on before. The motion sensor retrofit saved little electricity except in the early evening hours.

Therefore, we recommend that retrofitters install CFLs or other more efficient lighting in outdoor fixtures rather than depending solely on motion sensor control. We could have obtained added savings of nearly 1 kWh per day by substituting two 15W CFLs for the two 60W incandescents lighting the front porch.

One place where motion sensors might have been appropriate was the den, but we did not install them there. The father frequently leaves table lamps on for many hours in the den with no one in the room.

The Bottom Line The new lamps and controls cost $405 retail. Paying a retrofitter to do the changeover would have made the project far less cost-effective, but we didn't include labor costs because the retrofit is an easy

We estimated savings from the retrofit in two ways. In the first method, we compared the metered lighting and plug loads from June 20 to December 10, 1995 (before the retrofit) with the loads from December 13, 1995 to June 20, 1996 (after the retrofit).

Figure 1. Average whole house lighting and miscellaneous energy demand before and after the retrofit. The subtractive method of determining savings assumes that the change is entirely caused by the lighting retrofit.

We found an average 6.8 kWh per day change over the period. Most savings were in the hours between 7 am and midnight and were highest between 6 pm and 10 pm (see Figure 1). We witnessed a 40% reduction in the metered lighting and plug loads, which we calculated to be a 61% reduction in the pure lighting load. This works out to 2,500 kWh per year, or about $200 at 8¢/kWh.

The second method we used for estimating energy savings was more traditional. We knew the wattage change for each fixture we retrofitted. We also knew the average daily use of each fixture, thanks to the lighting loggers. We multiplied the wattage change by the hours per day the fixture was on.

Unfortunately, the dead moth threw off the data for the outdoor rear fixtures only two days after the light logger was set up. As previously described, the motion sensor control of the front porch lighting was largely overridden and few savings were observed there. Thus, this method of estimating the savings did not account for any change in the outdoor lighting from the retrofit. Light logger data was not available for two altered fixtures due to a project oversight. Also, we logged most of the fixtures during summer and early fall, missing the high-use period in the middle of winter.

This method, therefore, gave us a very conservative estimate of lighting energy use. Nevertheless, it showed a 47%, or 5.2 kWh per day, reduction in lighting energy use, amounting to 1,900 kWh per year.

Depending on the estimation method, the simple payback on this retrofit was between two and three years, for a simple rate of return of about 40%.

What This Means to Everybody Else Generally, in a residential retrofit, substitution of CFLs for incandescent lamps is recommended for fixtures that are used more than three hours per day. This recommendation is based on the relative economics of installing CFLs against the produced rate of savings. For instance, a 15W CFL substituted for a 60W incandescent lamp will produce a savings of 49 kWh per year when burned for three hours a day. At 8¢/kWh and a $15 lamp cost, the simple payback is 3.8 years, for an attractive simple rate of return of 26%. At two hours per day burn time, the numbers are still fairly attractive (5.7-year payback and an 18% rate of return). However, when lamps are used for short periods of time, the economics rapidly deteriorate. For instance, for a fixture used an average of only half an hour per day, the payback time increases to 23 years and the rate of return drops to 4%.

However, it can be difficult for auditors to concentrate on heavily used lighting fixtures, since it's hard to know which fixtures are used more than two hours per day. This article was adapted from Danny Parker and Lynn Schrum, Results from a Comprehensive Lighting Retrofit, FSEC-CR-914-96, available from the Florida Solar Energy Center, 1679 Clearlake Road, Cocoa, FL 32922-5703. Tel:(407)638-1000.In our study the individual fixtures were metered, but most retrofitters will not have the benefit of such information for the homes they deal with. Fortunately, data from other studies provide insight into the typical hours that fixtures are used in various rooms.

The usage in our house was similar to the typical patterns found in the Tacoma study and other studies that preceded it. All of these emphasize the need to address outdoor lighting energy consumption. A simple rule of thumb, which our results bear out, is that all incandescent lights outdoors, in kitchens, and in living rooms are good candidates for replacement. Of course, circumstances differ in individual homes, and this should only be used as a guide to provide a better match to actual needs.

We sought the maximum possible savings, so we chose not to concentrate on high-use fixtures, but to install CFLs wherever they could fit. Had we followed the above advice (changing lighting only outdoors and in the kitchen and living room), we would have saved about 3.3 kWh per day, 30% of the household's lighting energy use, at a cost of only $168 for 11 CFLs. This strategy would have missed lucrative opportunities in the garage and den, but overall performance still would have been good--a $96 annual savings and a 1.7-year payback.

Do the Savings Persist? We examined the lighting retrofit six months after installation. We found all but one of the lamps still in place, although some table and floor lamps had been moved. One halogen bulb had broken when a bedroom lamp fell. However, these changes probably won't change the savings by much. A greater issue is long-term persistence. Even with a 10,000 hour life, the most frequently used CFLs--and those providing the best economics--will burn out and need replacement in two to four years. We'll keep you posted. As for us, we're already off duplicating the above research in two additional homes.

This article was adapted from Danny Parker and Lynn Schrum, Results from a Comprehensive Lighting Retrofit, FSEC-CR-914-96, available from the Florida Solar Energy Center, 1679 Clearlake Road, Cocoa, FL 329922-5703.
Tel: (407)638-1000.



 | Back to Contents Page | Home Energy Index | About Home Energy |
| Home Energy Home Page | Back Issues of Home Energy |

Home Energy can be reached at:
Home Energy magazine -- Please read our Copyright Notice



  • 1
  • NEXT
  • LAST
Email Newsletter

Home Energy E-Newsletter

Sign up for our free monthly

Harness the power of

Get the Home Energy