Preliminary Evaluation of the Neighborhood Energy Workshop Program
The Neighborhood Energy Workshop (N.E.W.) Program is a comprehensive, neighborhood oriented energy conservation program co-sponsored by the City of Minneapolis and Minnegasco, a natural gas utility. A description of the program is contained in Brummitt (1984).
To date, N.E.W. has been offered in 55 of the 70 residential neighborhoods in Minneapolis. Over 23,000 households have actively participated in N.E.W. since 1981. This participation rate represents 24% of all 1 - 4 unit structures in the neighborhoods served, and is an exceptionally high participation rate for a residential conservation program.
Until recently, N.E.W. has been oriented primarily toward middle-class homeowners. Renters, low-income people, upper income people and those who believe their houses are already energy efficient have been less likely to attend for a variety of reasons.
The participants chosen for analysis were selected to increase their chances
of having adequate data both before and after participation, so the historical pattern of the N.E.W. project dictated that they come from only five resi¬dential neighborhoods in three different areas of the city. Though our sample matches these neighborhoods closely in terms of building age, they are significantly older than the overall Minneapolis housing stock. They also appear to be somewhat larger than average and their owners have higher than average incomes, though we did not have sufficiently well matched data for the city as a whole and for these neighborhoods, in particular, to precisely quantify these contrasts. Much of the income difference is surely due to the program's orientation toward middle class homeowners.
Despite these differences between our sample and the rest of Minneapolis,
we found that their pre-participation gas use (measured as weather-normalized total annual consumption) closely matched the rest of the city, both in
its distributional breakdown and its mean, which was 1680+447 CCF/year for N.E.W. participants and 1715+609 CCF/year for the city as a whole.
We therefore feel that this group will at least give us a useful preliminary estimate of N.E.W. effects, though it cannot be claimed to be completely representative of the Minneapolis housing stock.
A preliminary assessment of energy savings due to N.E.W. has been conducted. A random sample of 230 individuals who participated in N.E.W. before January, 1982 was drawn. During the time covered by the evaluation, N.E.W. focused heavily on no cost and low cost conservation measures, with very little discussion of major weatherization.
Gas consumption before and after participation was normalized for weather using the "Scorekeeping" model developed at Princeton University (Dutt, Fels, Goldberg and Stram, 1982). (Figure 1). This model represents the state of the art and is considered to give highly accurate estimates of weather- normalized annual consumption.
One hundred and ten participants from the initial sample of 230 were useable for the analysis. (Table 2). Of these, 92 showed decreased energy use (up to 756 CCF saved) and 18 showed increased use (up to 221 CCF increase). Absolute savings had a mean of 132 CCF/year and a standard deviation of 177 CCF/year, while relative savings averaged 7.3+9.2%. Normalized for the area of the house, savings averaged 10,617+147000 Btu/sq.ft.-yr. Regardless of which way the change in energy use is expressed, the savings are statistically highly significant (p<:-..001 for a paired sample t-test).
At present we do not have a control group against which to compare these savings. However, we do have some information on overall conservation trends. Minnegasco's annual reports for the past several years show annual reductions in gas use per residential customer for their entire service area of 2.0
to 2.7%. Northern States Power has found weather-adjusted consumption per customer to be nearly constant over the 1981-82-83 period, according to their 1983 annual report.
Assuming that the general conservation trend is 3%, the 7.31% savings for N.E.W. participants is still significantly greater (p.(.001 for a paired sample t-test). The net savings attributable to N.E.W could then be conservatively estimated at 4.3% of total annual gas use. (72.4 CCF or approximately 6250 Btu/sq.ft.-yr). It should be noted that since N.E.W. has served so many households, the program itself may be a partial cause of the observed general conservation trend.
A control group is advantageous in that it allows one to (1) compensate for residual imperfections in the weather-normalization method, (2) take into account savings that would take place even without N.E.W., due to energy awareness and financial presures to conserve, (3) regress savings on various characteristics to correct for atypical traits of N.E.W. participants compared with the general population, and (4) correct for self-selection bias, the possible tendency for conservation program participants to be predisposed to conserve. The "Scorekeeping" program is quite reliable for weather normalization, so item 1 is not a major problem. Adjusting N.E.W. savings for an assumed 3% general conservation trend allows some degree of correction for item 2. Analysis of participant characteristics shows that the sample group is similar to the general population in pre-participation gas use and differs only slightly in house size, but may differ in age of house and family income, so potential problems for item 3 are not wholly eliminated. Self-selection bias is not addressed by our currect methodology. In summary, a control group would be necessary to produce a very tight estimate of savings from N.E.W. However, we believe such an estimate is unlikely to differ much from our figure of 7.3% for direct program effects or 4.3% for incremental savings above what might have occurred without participation.
Further analysis of our energy savings data also sheds some light on how the savings were achieved. Over half of the participants showed savings in the 0 to 9% range (Figure 2), which is consistent with expected savings from no-cost and low-cost improvements. Thus the mean savings from N.E.W. appear to come not from a few people doing a lot of major work to their houses, but from many people implementing the N.E.W. recommendations. The Scorekeeping model also allows a rough quantitive assessment of whether savings are coming from decreasing thermostat settings, improving the thermal integrity of the building envelope or reducing domestic use (water heating, dryers, cooking, etc.) This analysis showed that 80 to 90% of the 7.3% savings came from improving the building envelope (beta decreased from 0.1997 + 0.0857 to 0.1785+0.0472, the only parameter change that was statistically significant), as would be expected sealing bypasses, insulating the rim joist, caulking and other low-cost measures taught in the workshops. About 10 to 20% of the savings came from reducing domestic use (alpha decreased from 1.054+0.459 to 0.982+0.387) which would result from such N.E.W. recommendations as installing shower flow restrictors and reducing the water heater setting. Only about 5% of the savings came from lowered thermo¬stat settings (tau decreased from 59.37+5.70 to 59.19+5.41). These houses apparently were already using low thernEstat settings—(a reference temperature of 59 probably correlates with indoor temperatures in the mid-sixties), so there was very little room for additional savings from this factor.
Note how much lower these reference temperatures are than the conventionally accepted value of 65'F.
The N.E.W. program costs about $80 per household to deliver. Assuming an average savings of 4.31% or 72.4 CCF gives an average annual dollar savings of $42.43 (at $0.586 per CCF). The simple payback for the program is thus conservatively estimated at 1.9 years. Since most of the retrofits involved have a life span of 5 to 50 years, a two year payback is clearly acceptable.
PARTICIPATION REACTION TO THE PROGRAM
Participants are very enthusiastic about N.E.W. Miller Research conducted a survey of N.E.W. participants for Minnegasco. Forty three precent of
the 224 participants surveyed rated N.E.W. as outstanding and 49% rated it as good. Fifty percent rated it as very important in causing them to make specific improvements they made and an additional 29% rated it as important.
OTHER BENEFITS OF THE PROGRAM
There are other benefits from N.E.W. besides direct energy savings. The goal of N.E.W. is not just to save energy, but to do it cost-effectively.
Through N.E.W. we have dissiminated information about Princeton's House Doctor approach and the importance of attic bypasses to 23,000 citizens, most of whom would have been unlikely to encounter it otherwise. We have been able to impress upon people that wall insulation has a better payback than reinsulating the attic and should be a higher priority, a fact that few Minneapolis residents are aware of before they attend N.E.W. Getting information like this out to the public helps to insure that conservation dollars will be spent wisely, on the most efficient retrofits possible. Equally important, N.E.W. has created a grass-roots organizing network that can now be used to implement more extensive energy programs such as Operation Insulation, as well as other City programs not related to energy that require citizen participation.
Full report (PDF)
Preliminary Evaluation of the Neighborhood Energy Workshop Program