Trends in Energy is a bulletin of residential energy conservation issues. It covers items ranging from the latest policy issues to the newest energy technologies. If you have items that would be of interest, please send them to: Trends Department, Home Energy, 2124 Kittredge St., No. 95, Berkeley, CA 94704.

Energy Fitness programs strive for measurable energy savings for consumer and utility alike by installing lighting, water, and appliance efficiency measures in the customers' homes. They also refer residents to other utility conservation programs. Energy Fitness programs are difficult to evaluate because they usually operate on tight budgets, they don't save large amounts of electricity per customer, and they tend to rely heavily on compact fluorescent lights. Tight budgets mean little is left over for expensive surveying or monitoring, and small per capita savings mean that even monitoring may not reveal much. The dependency on compact fluorescents is a problem because energy savings from compact fluorescents are less predictable than other residential demand-side management technologies.

In its first three years of operation, the Energy Fitness program of New England Electric Systems (NEES) installed over 196,000 compact fluorescents in more than 37,200 homes (see Table 1). Five vendors used a "neighborhood blitz" approach to reach traditionally difficult-to- serve customers living in urban, low-income neighborhoods. Trained field staff installed all measures at no cost to participants. (See "Energy Fitness: Canvassing Urban Neighborhoods,"HE Mar/Apr '92, p. 27)

Program vendors estimated that as many as 98% of customers who were at home when field staff went door-to-door in their communities participated in Energy Fitness. For cost and safety reasons, field staff only went out during weekdays between 8:30 a.m. and 4:30 p.m. Roughly half of targeted customers were not home during these hours; therefore the final penetration rate was about 50%.

Magnetically ballasted compact fluorescents accounted for over 75% of projected program savings. Additional measures included cleaning refrigerator coils and air-conditioner filters, water heater wraps for customers with electric water heaters, and weatherization measures for customers with electric space heat. Staff also installed water- saving measures such as low-flow showerheads and faucet aerators, regardless of water heating fuel.

How Much Did The Package Save?

We tried to measure savings with information from three sources: vendor reports, surveys, and billing data. Vendors collected information about participants and measures they received. Surveys, collected later during evaluations of the programs, focused on whether measures that had been installed were still there and how customers felt about them. Billing data came from the NEES customer database.

Surveys: Phone Calls vs. House Calls
The data yielded a lot of information about the types of customers that we were serving, the types of measures that they received, and the quality of the retrofits. By comparing the wattages of installed compact fluorescents with the wattages of the incandescents that they replaced, we were able to see whether vendors were following manufacturers recommendations. We found that, in general, vendors had a tendency to use compact fluorescents to replace higher wattage, and therefore brighter, incandescents than recommended. This was due to a low availability for some of the higher wattage compact fluorescents, and incentives which encouraged vendor field staff to install as many as possible.

Two surveys of Energy Fitness program participants identified a number of "free riders" (those who would have installed energy efficient lighting anyway), the degree of customer satisfaction, and the percentage of compact fluorescents that had been removed. A survey of 1989 participants, completed in August 1990, found no measurable free ridership. NEES performed a second survey in late 1991, contacting participants from 1989, 1990, and early 1991. This extensive study found 1.4% free-ridership. Customer satisfaction with Energy Fitness was very high in both surveys.

During visits to participants' households for the first survey, NEES staff found that residents had removed many compact fluorescents, the equivalent of 20% of the installed wattage. The second survey found the average removal rate for all compacts installed throughout the three years of the program to be 32%. (It is important to note that when our interviewers telephoned participants, they learned of only half of the removals which were revealed by later, on-site visits). When asked why they had removed the compact fluorescents, customers most often said that it was because they were too dim.

Electric Bills
Our billing analyses consisted of a statistically based comparison between two complete years of billing history, one before and one following installation. We corrected for other changes in energy use by comparing a participant group with a control group for the same time period.

We performed the first billing analysis for a group of 983 participants who received installations in 1989. Net annual savings per household were almost 295 kWh. The second billing data analysis, developed in early 1992, encompassed 2,234 participants, chosen from those who received installations in the Spring of 1990. Net annual savings per participant dropped to about 143 kWh.

Which Evaluation Should We Believe?

In both evaluations, the Energy Fitness program produced statistically significant energy savings. Yet the savings measures differed between the two years and had large confidence level intervals. We conclude that all that could really be stated is that yes, Energy Fitness saves electricity. The participants actually showed similar savings in both years. But in 1990 the control group increased consumption and in 1991 it decreased consumption. When these effects were added to the participant savings from the program, it drove the estimates of program savings in opposite directions. This may indicate a problem with the way in which comparison group members were chosen.

Other factors also made it difficult to determine precise savings. Energy Fitness was designed to save only a small percentage of participants' electricity consumption, already low for this group. This consumption was also affected by the economy, weather, and many other variables whose effects may not have been effectively captured in the control group. Furthermore, lighting is directly controlled by the customer (as opposed to automatic devices like a thermostat or refrigerator), and anybody who can screw in a light bulb can remove a compact fluorescent.

Putting aside savings, the most most important finding of the evaluation was that recipients removed large numbers of compact fluorescent lights. Removals appear to have been greatest for lights installed in the first year of the program, and to have decreased in the following years. While removals increase with age, it should be noted that the mix of compact fluorescents delivered shifted substantially towards higher wattages over the three years as more became available. Because higher wattage compact fluorescents are brighter, we may expect to find that participants have not removed those installed in 1990 and 1991 as often as those installed in 1989.

At NEES we have not used electronically ballasted compact fluorescent lamps because they tend to have high total harmonic distortion, which can interfere with the operation of other appliances, especially those using computer chips. However, feedback from Conservation Services Group and other vendors convinced us that magnetically ballasted compact fluorescents often do not provide satisfactory service because they tend to flicker when turned on and do not immediately achieve full brightness. We are currently working with manufacturers who are developing electronically ballasted compact fluorescents with low total harmonic distortion.

All this is to say that a sophisticated evaluation of a successfully administered Energy Fitness program can yield ambiguous savings. In this case, looking at actual billing data complicated savings as much as it clarified them. However, comprehensive evaluation can also offer valuable information in addition to load reductions and energy savings. For Energy Fitness and other demand-side management programs that install compact fluorescents, it is clear that program designers must pay careful attention to all stages of he installation process, or risk large numbers of removed lamps.

-Chris Granda

Chris Ganda is a senior analyst in the Demand Planning department at New England Electric Systems.

                                    1989      1990     1991     Total
_____________________________________________________________________
Energy Fitness Participants        2,577    15,179   19,459    37,215
Numbers of compact                   5.8       5.5        5       5.3
  fluorescent lamps per
  participant
Total compact fluorescents        14,850    83,102   98,086   196,038
Types installed (percentages of total) 
       9 Watt Twin Tubes           1,022     3,607      548     5,177
                                    (7%)      (4%)     (1%)      (3%)
       9 Watt Quad Tubes           5,012    19,745   17,165    41,922
                                   (34%)     (24%)    (18%)     (21%)
       13 Watt Quad Tubes          4,794    31,905   33,214    69,913
                                   (32%)     (38%)    (34%)     (36%)
       16 Watt Single-piece lamps  2,212    12,797   24,271    39,280
                                   (14%)     (15%)    (25%)     (20%)
       22 Watt Quad Tubes          1,810    15,048   22,888    39,746
                                   (12%)     (19%)    (23%)     (20%)
Displaced watts per participant      280       260      240       250
Percentage found removed             46%       38%      25%       32%