- Email: [email protected]
Load-shape data for residential lighting: Survey results for incandescent and compact fluorescent lamps
Energy Vol. 18, No. 2, pp. 211-217, 1993 Printed in Great Britain. All rights reserved
036@5442/93 $6.00 + 0.00 Copyright @ 1993 Pergamon Press Ltd
LOAD-SHAPE DATA FOR RESIDENTIAL LIGHTING: SURVEY RESULTS FOR INCANDESCENT AND COMPACT FLUORESCENT LAMPS BENT NIELSEN Research Association of Danish Electric Utilities, Postboks 259.2800, Lyngby, Denmark
Abstract-Understanding the load pattern created by compact fluorescent lamps is key to evaluating the impact on the utility’s system peak demand and calculating energy savings. A household survey was distributed to consumers involved in a Danish CFL programme. The primary goal was to map the daily patterns of CFL and incandescent lamp use in the residential sector throughout the year. A secondary aim was to evaluate consumers’ acceptance of the CFLs. Two CFLs were offered free of charge to each of the utility’s 240,000 residential customers. The participation rate was 90% (1.8 CFLs per home). Although this represents only 7% of the total number of lamps per home, the CFL’s long burning hours accounted for nearly 20% of the total burning time of all lamps. Total annual household electricity savings consumption for lighting due to the CFLs is approximately 15%, or 125 kWh per home. During afternoon peak hours on winter days, up to 80% of the CFLs are in use. During the morning peak hours, however, only 30% of the CFLs are in use.
INTRODUCTION Since 1988 European utilities have run a variety of programmes to promote the use of compact fluorescent lamps (CFLs), primarily in the residential sector.Is2The largest example has been the giveaway programmeby SEAS, a Danish distributioncompany that serves approximately 120,000 residential consumers on Zealand, south of Copenhagen. The utility’s peak load is about 500 MW. In the programme, which took place in early autumn 1989, each residential consumer received 2 Philips ClXs: a 13-Watt magnetic typeSL and an 1l-Watt electronic type PLC. Consumers were encouraged to put the lamps in places where they would be used during peak hours. To investigate the lighting load curve, and especially the pattern of use of the CFLs, the Research Association of Danish Electric Utilities (DEFU) conducted four surveys. The research project was financed by ELand ELSAM. THE SURVEY The background, the method of survey, and the results are reported in detail elsewhere.3 The questionnaire method was used successfully following the lighting programme by EFFO, held elsewhere in Denmark.’ The EFFO sample included 2,400 consumers and yielded a response rate of 70%. The survey determined load curves for one month (November), while in the SEAS survey the aim was to establish load curves for week-days throughout the year. This was done by mailing questionnaires four times in the winter 1989 and spring 1990 (Table 1) and assuming seasonal symmetry (e.g. autumn months correspond to spring months). In addition, the supply area and number of consumers given CFIs was much larger (50 times) in the SEAS project than in the EF’FOexperiment. Table 1. SEAS survey.
Number of questionnaires sent Number of replies Response rate (%>
Nov. 1989
Jan. 1990
Mar. 1990
May 1990
Total
600 232 39
600 273 46
1,ooO 372 37
1,ooO 299 30
3,200 1.176 37
211 EGY18:2-K
BENT NIELSEN
212
The 37% response rate was rather low (Table 1). apparently due to the large and complicated questionnaire and the fact that, unlike the EFFO consumers, the SEAS consumers did not ask for the CFLs. In addition, SEAS sent no reminders, which reduces the motivation for replying. Figure 1 shows a sample of the questionnaire, in which the consumer had to specify the number of lamps in each room of the home and the periods of lamp operation. The usage patterns of all lamps and the CFLs were mapped separately. The results were aggregated according to dwelling type: single-family houses, multi-family houses and farmhouses. This facilitates the construction of CFL load curves for any combination of dwellings. On a typical March weekday, the lighi will normally be “on” the followinghours: lndalewhha8tfdmlhatRgMbmmdwMelhenumbwoflafnps6t
A.m. 1 2 3 4 5 6 7 8 9 1011121
Roon-lS -Qf&f room c1si(dmrc room -L&f room -room Kitchen Bath
P.m. 2 3 4 5 6 7 8 9 101112
7
3
_a.-
4 -_L-
Outdoor
212 12
3
1
Fig. 1. Sample questionnaire (all lamps). Seven lamps ate in the sitting mom, four of which are lit from 6 to 8 pm. From 8 to 11 pm, only three lamps am lit. The three lamps in the childrens’ room are lit from 7 to 9 pm. Lamps lit for less than one half hour am not included.
RESULTS
Although the load curves are the major results of the study, a few other results are also of interest, including: (i) the survey appears to be reasonably representative of the Danish population, with respect to the dwelling size and number of inhabitants per dwelling (Table 2); (ii) the total number of lamps per home varied from 16 in multi-family houses to 28 in farmhouses; (iii) the electricity consumption for lighting varied from 550 kWh to 980 kWh per year per home (Table 3), calculated based on the lead curves.
Table 2. Average living area and number of inhabitants,
House type Single-family houses Multi-family houses Farm houses
Living area, mz per home
Adults+children, per home
SEASSurvey
SEASSurvey
135 81 165
1 Denmark 130 77 157
1.9+0.7 1.5+0.5 2.otO.6
1 Denmark 1.9+0.8 1.4+0.3 2.1+0.7
Consumers could install 88% of the CFLs received under the SEAS programme; an additional 2% were placed in second houses or given away. Lack of suitable fixtures was the primary barrier to use. Most of the CFLs were placed in the living room (55%) and in the kitchen (20%) (Table 4).
Load-shape
data
for residential lighting
213
Table 3. Average number of lamps per home. Single-family house
Notes:
Multi-family house
I Farmhouse
Main room Kitchen Bath Hall, etc. Other rooms Outdoor Miscellaneous
7 4 2 3 6 2 2
2 3 1 1
7 4 2 3 5 2 5
Total
26
16
28
Consumption” O
850 6.3
550 6.7
980 6.0
(a) Incandescent
5 3
1
lamps + existing CFLs.
Table 4. Use of CFLs. Type of dwelling
Single-family
Room
house
Multi-family
house
SL%
PLC%
66
38
56
9
25
14
3
2
0
1
10 4
7 7
9 11
11
10 4
14 4
12 1
3 0
3 0
18 5
11 4
100
100
100
100
100
100
SL%
PLC%
SL%
PLC%
Living room
52
53
60
Kitchen
21
19
20
Bath
0
1
HaII, etc. Other rooms
6 3
Outdoor Miscellaneous Total
Farmhouse
3
The annual operation of each CFL was high, compared to common wisdom, ranging from 1,600 hours per year in single- and multi-family houses to 1,900 hours per year in farmhouses. CFL operation declines as the number of CFLs per home increases. The average operating time of all lamps is approximately 600 hours or less. The average wattage of the incandescent lamps, which the CFLs replaced, ranged from 52 to 61 Watts. Twenty percent of the consumers said they operated their CFLs longer than the incandescent lamps they replaced. Less than 5% said they operated their CFLs for a shorter period of time. A control calculation shows, however, that there was little difference between the two groups. Both groups operated their lamps approximately 7 hours per day during mid-winter and 3 hours per day in May. The 1.8 CFLs per home represent only 7% of the total number of lamps per home. Calculations based on lamp use show, however, that the CFLs account for 20% of the total operation hours of all lamps in the home. The savings in electricity consumption for lighting due to the CFLs are thus approximately 15%, or 125 kWh per home in the SEAS area. Consumers were also asked how many CFLs they possessed prior to the SEAS give-away, and how many more CFLs they thought they could use in existing fixtures (Fig. 2). The average was 5.3 CFLs in single-family and farmhouses and 3.6 CFLs in multi-family houses. The figures include all possible CFLs, i.e. existing (on average less than one CFL per home), those from the SEAS give-away, and possible future CFLs. One percent of the consumers will not be able to use any CFLs, 12% will use one CFL, etc.
214
BENT NIELSEN
Peroentageofcollaumera 22 20 18 16 14 12 10 8 8 4 2 0 0
1 2 3 4
5 8
7
8
B 10 11 1213 14 15 18 17 18 19 2021 22 CFLslholJse Single-family houses
Fig. 2. Maximum number of CFLs useableperhome.
Thirty percent of the consumers indicate that they are willing to buy CFLs at their own expense at the price level at the time of the survey, 150 to 250 DKK (20 to 30 ECU). The price, however, should be lowered to approximately 40 to 50 DKK (5 to 6 ECU) to make it acceptable to the majority of consumers (Fig. 3).
willingto buy CFLa asafundonofpdce.
Percentage of amsumem
6050403020loo I 10
20
30
40 5
50
60
70
80 10
90
, 100 110 120 130 140 DKK ECU 15 PhxofCFL
Fig. 3. Consumers’ price acceptance.
Load curves Daily load curves based on the questionnaires were established for incandescent lamps and the CFLs throughout the year, assuming sinusoidal variation with the season, i.e. natural light conditions, plus a constant load. This appears to fit well with the observations. Results are given in Watts per home, number of lamps “on” per home, or as load factors (l.f.), defined as the number of CFLA “on” at a given time divided by the total number of CFLs. Figure 4 presents a typical curve for the month of May, showing lamps in use in the various rooms of the house.
Load-shape data for residential lighting
215
Number of lamps “on”
I-
MAY ,-
I-
1
I 2
3
4
5
6
7
8
9 101112131416161719192021222324hr~
Fig. 4. Usage patterns for lighting in single-family
house (all lamps).
Figure 5 at the top of the next page shows hourly load factors by month for the 1.8 CFLs installed by participating households. During afternoon peak hours on winter days approximately 80% of the CFLs are in use. During morning peak hours only 30% of the CFLs are operating. Values are substantially lower in summer. System loads may be obtained by multiplying the appropriate number of lamps and wattages by the load factor. It should be noted that the load factors will decrease if the number of CFLs per home is increased. Accuracy of the method
Obtaining reliable data on electricity consumption for lighting and particularly load curves is difficult.3e5.6 The results obtained with the method described above show 10 to 18% higher annual electricity consumption. In addition, the load curves generally show a higher lighting load in summer and on mornings than do previous curves.3 The earlier estimates were based on the “residual method” and on the well-known “sales and stock” method.5 The latter method, however, is illsuited for Danish conditions, due to the difficulty of separating lamp sales in the residential sector from those in the commercial sector. Hence we have more faith in the questionnaire method, although it has potential weaknesses. We know our samples are representative as to dwelling size and number of persons per dwelling, but could the behavior of households in our sample differ from that of the broader population? Tests have shown that the answers are not influenced by the examples provided, on how to complete the questionnaire. We have also found good agreement between the results in SEAS’s service area and the previous survey in the EFFO area (Fig. 6).
BENT NIELSEN
216
Load factors are slightly lower in the EFFO area due to a larger number of CFLs per home. The survey did not include weekends and holidays. This would have increased the cost of the experiment substantially. A rough correction has been made, but this could still result in overestimates. Other methods include direct energy-use monitoring, which has been done in one experiment.4 The cost of using this method on a large scale is prohibitive.
1.f.-
No. of CFb ‘on’ Tdd no.oi CFLa
‘D,
.
‘D,
ID.
1.0,.
‘D,
LO_
lD,
1.0,
0.0 0.0 0.4
02 OP 1
ad
24
Fig. 5. Load curves for CFLs in single-familyhouses.
Single-family
houses. November.
c.f. 0.6
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 1
2 3 4 5 6 7 6 9 E EFFO (3.7 CFWhouse)
101112131415
161716192021222324hn
n SEAS (1.6 CFLsJmuse)
Fig. 6. Comparisonof load factors in the EFFO and SEAS survey.
Load-shape
data for residential
217
lighting
Measuring the effect at the system level is not practical. Even in a project as large as the SEAS programme, with 1.8 CFLs per home, the load reduction due to the 0.22 million CFLs is 8 to 9 MW. This, unfortunately, is within the “noise level” in a system with daily peak loads (all customer classes) of 400 to 500 MW in winter (Fig. 7). Measurements from low-voltage feeders for residential areas combined with interviews and recordings at the individual household level could improve residential load data. Wednesday,
1
2
3
4
5
6
7
5
D
21 November 1990
10 11 12 13 14 15 16 17 18 19 al 21 P
29 24 lma
Fig. 7. Totat SEAS load (including non-residential).
CONCLUSIONS
The Danish utility SEAS has conducted the largest European giveaway programme to promote compact fluorescent lamps (CFLs). From the utility’s perspective, an important finding is that cons.umers respond favorably when advised to use CFLs during the time of the utility’s system peak. As a result, the programme was effective at saving energy and at reducing utility peak load. The survey method appears to give more reliable data than previously used methods. Because consumers chose high-use sockets, installing an average of 1.8 CFLs/household resulted in a 15% reduction of household electricity use for lighting. Further savings are achievable, because the average household reports being able to use a total of more than 5 CFLs. REFERENCES 1. E. Mills, Energy Policy 19, 266 (1991).
2. E. Broend, ‘“The NESA All-Sector Programme: Results and Analyses from a Full-Scale Campaign in
3. 4. 5.
6.
1990 for the Propagation of Compact Fluorescent Lamps,” in “Proceedings of the 1st European Conference on Energy-Efficient Lighting,” E. Mills ed., Swedish National Board for Industrial and Technical Development, Stockholm, Sweden (1991). B. Nielsen and K.H. Thomsen, “Belasmingskurver f0r boligbelysning, specielt elspareprerer (Load Curves of Lighting in Homes, Especially Compact Fluorescent Lamps).” DEFU, Research Association of Danish Electric Utilities, TR 286 (in Danish), Lyngby, Denmark (1988). B. Nielsen, K.H. Thomsen, and J. Mgller, “EFFO fors0g med lysstoflamper: Slutrapport (EFFO’s Compact Fluorescent Lamp Project: Final Report),” Research Association of Danish Electric Utilities, TR 276 (in Danish) Lyngby, Denmark (1989). B. Huenges Wajer and R. Kemna, “Electricity Consumption for Lighting in the Average Dutch Household: Development of an Accounting Method,” in “Proceedings of the 1st European Conference on Energy-Efficient Lighting,” E. Mills ed., Swedish National Board for Industrial and Technical Development, Stockholm, Sweden (1991). S.Bartlett, Energy -- The International Journal 18, 171 (1993).
036@5442/93 $6.00 + 0.00 Copyright @ 1993 Pergamon Press Ltd
LOAD-SHAPE DATA FOR RESIDENTIAL LIGHTING: SURVEY RESULTS FOR INCANDESCENT AND COMPACT FLUORESCENT LAMPS BENT NIELSEN Research Association of Danish Electric Utilities, Postboks 259.2800, Lyngby, Denmark
Abstract-Understanding the load pattern created by compact fluorescent lamps is key to evaluating the impact on the utility’s system peak demand and calculating energy savings. A household survey was distributed to consumers involved in a Danish CFL programme. The primary goal was to map the daily patterns of CFL and incandescent lamp use in the residential sector throughout the year. A secondary aim was to evaluate consumers’ acceptance of the CFLs. Two CFLs were offered free of charge to each of the utility’s 240,000 residential customers. The participation rate was 90% (1.8 CFLs per home). Although this represents only 7% of the total number of lamps per home, the CFL’s long burning hours accounted for nearly 20% of the total burning time of all lamps. Total annual household electricity savings consumption for lighting due to the CFLs is approximately 15%, or 125 kWh per home. During afternoon peak hours on winter days, up to 80% of the CFLs are in use. During the morning peak hours, however, only 30% of the CFLs are in use.
INTRODUCTION Since 1988 European utilities have run a variety of programmes to promote the use of compact fluorescent lamps (CFLs), primarily in the residential sector.Is2The largest example has been the giveaway programmeby SEAS, a Danish distributioncompany that serves approximately 120,000 residential consumers on Zealand, south of Copenhagen. The utility’s peak load is about 500 MW. In the programme, which took place in early autumn 1989, each residential consumer received 2 Philips ClXs: a 13-Watt magnetic typeSL and an 1l-Watt electronic type PLC. Consumers were encouraged to put the lamps in places where they would be used during peak hours. To investigate the lighting load curve, and especially the pattern of use of the CFLs, the Research Association of Danish Electric Utilities (DEFU) conducted four surveys. The research project was financed by ELand ELSAM. THE SURVEY The background, the method of survey, and the results are reported in detail elsewhere.3 The questionnaire method was used successfully following the lighting programme by EFFO, held elsewhere in Denmark.’ The EFFO sample included 2,400 consumers and yielded a response rate of 70%. The survey determined load curves for one month (November), while in the SEAS survey the aim was to establish load curves for week-days throughout the year. This was done by mailing questionnaires four times in the winter 1989 and spring 1990 (Table 1) and assuming seasonal symmetry (e.g. autumn months correspond to spring months). In addition, the supply area and number of consumers given CFIs was much larger (50 times) in the SEAS project than in the EF’FOexperiment. Table 1. SEAS survey.
Number of questionnaires sent Number of replies Response rate (%>
Nov. 1989
Jan. 1990
Mar. 1990
May 1990
Total
600 232 39
600 273 46
1,ooO 372 37
1,ooO 299 30
3,200 1.176 37
211 EGY18:2-K
BENT NIELSEN
212
The 37% response rate was rather low (Table 1). apparently due to the large and complicated questionnaire and the fact that, unlike the EFFO consumers, the SEAS consumers did not ask for the CFLs. In addition, SEAS sent no reminders, which reduces the motivation for replying. Figure 1 shows a sample of the questionnaire, in which the consumer had to specify the number of lamps in each room of the home and the periods of lamp operation. The usage patterns of all lamps and the CFLs were mapped separately. The results were aggregated according to dwelling type: single-family houses, multi-family houses and farmhouses. This facilitates the construction of CFL load curves for any combination of dwellings. On a typical March weekday, the lighi will normally be “on” the followinghours: lndalewhha8tfdmlhatRgMbmmdwMelhenumbwoflafnps6t
A.m. 1 2 3 4 5 6 7 8 9 1011121
Roon-lS -Qf&f room c1si(dmrc room -L&f room -room Kitchen Bath
P.m. 2 3 4 5 6 7 8 9 101112
7
3
_a.-
4 -_L-
Outdoor
212 12
3
1
Fig. 1. Sample questionnaire (all lamps). Seven lamps ate in the sitting mom, four of which are lit from 6 to 8 pm. From 8 to 11 pm, only three lamps am lit. The three lamps in the childrens’ room are lit from 7 to 9 pm. Lamps lit for less than one half hour am not included.
RESULTS
Although the load curves are the major results of the study, a few other results are also of interest, including: (i) the survey appears to be reasonably representative of the Danish population, with respect to the dwelling size and number of inhabitants per dwelling (Table 2); (ii) the total number of lamps per home varied from 16 in multi-family houses to 28 in farmhouses; (iii) the electricity consumption for lighting varied from 550 kWh to 980 kWh per year per home (Table 3), calculated based on the lead curves.
Table 2. Average living area and number of inhabitants,
House type Single-family houses Multi-family houses Farm houses
Living area, mz per home
Adults+children, per home
SEASSurvey
SEASSurvey
135 81 165
1 Denmark 130 77 157
1.9+0.7 1.5+0.5 2.otO.6
1 Denmark 1.9+0.8 1.4+0.3 2.1+0.7
Consumers could install 88% of the CFLs received under the SEAS programme; an additional 2% were placed in second houses or given away. Lack of suitable fixtures was the primary barrier to use. Most of the CFLs were placed in the living room (55%) and in the kitchen (20%) (Table 4).
Load-shape
data
for residential lighting
213
Table 3. Average number of lamps per home. Single-family house
Notes:
Multi-family house
I Farmhouse
Main room Kitchen Bath Hall, etc. Other rooms Outdoor Miscellaneous
7 4 2 3 6 2 2
2 3 1 1
7 4 2 3 5 2 5
Total
26
16
28
Consumption” O
850 6.3
550 6.7
980 6.0
(a) Incandescent
5 3
1
lamps + existing CFLs.
Table 4. Use of CFLs. Type of dwelling
Single-family
Room
house
Multi-family
house
SL%
PLC%
66
38
56
9
25
14
3
2
0
1
10 4
7 7
9 11
11
10 4
14 4
12 1
3 0
3 0
18 5
11 4
100
100
100
100
100
100
SL%
PLC%
SL%
PLC%
Living room
52
53
60
Kitchen
21
19
20
Bath
0
1
HaII, etc. Other rooms
6 3
Outdoor Miscellaneous Total
Farmhouse
3
The annual operation of each CFL was high, compared to common wisdom, ranging from 1,600 hours per year in single- and multi-family houses to 1,900 hours per year in farmhouses. CFL operation declines as the number of CFLs per home increases. The average operating time of all lamps is approximately 600 hours or less. The average wattage of the incandescent lamps, which the CFLs replaced, ranged from 52 to 61 Watts. Twenty percent of the consumers said they operated their CFLs longer than the incandescent lamps they replaced. Less than 5% said they operated their CFLs for a shorter period of time. A control calculation shows, however, that there was little difference between the two groups. Both groups operated their lamps approximately 7 hours per day during mid-winter and 3 hours per day in May. The 1.8 CFLs per home represent only 7% of the total number of lamps per home. Calculations based on lamp use show, however, that the CFLs account for 20% of the total operation hours of all lamps in the home. The savings in electricity consumption for lighting due to the CFLs are thus approximately 15%, or 125 kWh per home in the SEAS area. Consumers were also asked how many CFLs they possessed prior to the SEAS give-away, and how many more CFLs they thought they could use in existing fixtures (Fig. 2). The average was 5.3 CFLs in single-family and farmhouses and 3.6 CFLs in multi-family houses. The figures include all possible CFLs, i.e. existing (on average less than one CFL per home), those from the SEAS give-away, and possible future CFLs. One percent of the consumers will not be able to use any CFLs, 12% will use one CFL, etc.
214
BENT NIELSEN
Peroentageofcollaumera 22 20 18 16 14 12 10 8 8 4 2 0 0
1 2 3 4
5 8
7
8
B 10 11 1213 14 15 18 17 18 19 2021 22 CFLslholJse Single-family houses
Fig. 2. Maximum number of CFLs useableperhome.
Thirty percent of the consumers indicate that they are willing to buy CFLs at their own expense at the price level at the time of the survey, 150 to 250 DKK (20 to 30 ECU). The price, however, should be lowered to approximately 40 to 50 DKK (5 to 6 ECU) to make it acceptable to the majority of consumers (Fig. 3).
willingto buy CFLa asafundonofpdce.
Percentage of amsumem
6050403020loo I 10
20
30
40 5
50
60
70
80 10
90
, 100 110 120 130 140 DKK ECU 15 PhxofCFL
Fig. 3. Consumers’ price acceptance.
Load curves Daily load curves based on the questionnaires were established for incandescent lamps and the CFLs throughout the year, assuming sinusoidal variation with the season, i.e. natural light conditions, plus a constant load. This appears to fit well with the observations. Results are given in Watts per home, number of lamps “on” per home, or as load factors (l.f.), defined as the number of CFLA “on” at a given time divided by the total number of CFLs. Figure 4 presents a typical curve for the month of May, showing lamps in use in the various rooms of the house.
Load-shape data for residential lighting
215
Number of lamps “on”
I-
MAY ,-
I-
1
I 2
3
4
5
6
7
8
9 101112131416161719192021222324hr~
Fig. 4. Usage patterns for lighting in single-family
house (all lamps).
Figure 5 at the top of the next page shows hourly load factors by month for the 1.8 CFLs installed by participating households. During afternoon peak hours on winter days approximately 80% of the CFLs are in use. During morning peak hours only 30% of the CFLs are operating. Values are substantially lower in summer. System loads may be obtained by multiplying the appropriate number of lamps and wattages by the load factor. It should be noted that the load factors will decrease if the number of CFLs per home is increased. Accuracy of the method
Obtaining reliable data on electricity consumption for lighting and particularly load curves is difficult.3e5.6 The results obtained with the method described above show 10 to 18% higher annual electricity consumption. In addition, the load curves generally show a higher lighting load in summer and on mornings than do previous curves.3 The earlier estimates were based on the “residual method” and on the well-known “sales and stock” method.5 The latter method, however, is illsuited for Danish conditions, due to the difficulty of separating lamp sales in the residential sector from those in the commercial sector. Hence we have more faith in the questionnaire method, although it has potential weaknesses. We know our samples are representative as to dwelling size and number of persons per dwelling, but could the behavior of households in our sample differ from that of the broader population? Tests have shown that the answers are not influenced by the examples provided, on how to complete the questionnaire. We have also found good agreement between the results in SEAS’s service area and the previous survey in the EFFO area (Fig. 6).
BENT NIELSEN
216
Load factors are slightly lower in the EFFO area due to a larger number of CFLs per home. The survey did not include weekends and holidays. This would have increased the cost of the experiment substantially. A rough correction has been made, but this could still result in overestimates. Other methods include direct energy-use monitoring, which has been done in one experiment.4 The cost of using this method on a large scale is prohibitive.
1.f.-
No. of CFb ‘on’ Tdd no.oi CFLa
‘D,
.
‘D,
ID.
1.0,.
‘D,
LO_
lD,
1.0,
0.0 0.0 0.4
02 OP 1
ad
24
Fig. 5. Load curves for CFLs in single-familyhouses.
Single-family
houses. November.
c.f. 0.6
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 1
2 3 4 5 6 7 6 9 E EFFO (3.7 CFWhouse)
101112131415
161716192021222324hn
n SEAS (1.6 CFLsJmuse)
Fig. 6. Comparisonof load factors in the EFFO and SEAS survey.
Load-shape
data for residential
217
lighting
Measuring the effect at the system level is not practical. Even in a project as large as the SEAS programme, with 1.8 CFLs per home, the load reduction due to the 0.22 million CFLs is 8 to 9 MW. This, unfortunately, is within the “noise level” in a system with daily peak loads (all customer classes) of 400 to 500 MW in winter (Fig. 7). Measurements from low-voltage feeders for residential areas combined with interviews and recordings at the individual household level could improve residential load data. Wednesday,
1
2
3
4
5
6
7
5
D
21 November 1990
10 11 12 13 14 15 16 17 18 19 al 21 P
29 24 lma
Fig. 7. Totat SEAS load (including non-residential).
CONCLUSIONS
The Danish utility SEAS has conducted the largest European giveaway programme to promote compact fluorescent lamps (CFLs). From the utility’s perspective, an important finding is that cons.umers respond favorably when advised to use CFLs during the time of the utility’s system peak. As a result, the programme was effective at saving energy and at reducing utility peak load. The survey method appears to give more reliable data than previously used methods. Because consumers chose high-use sockets, installing an average of 1.8 CFLs/household resulted in a 15% reduction of household electricity use for lighting. Further savings are achievable, because the average household reports being able to use a total of more than 5 CFLs. REFERENCES 1. E. Mills, Energy Policy 19, 266 (1991).
2. E. Broend, ‘“The NESA All-Sector Programme: Results and Analyses from a Full-Scale Campaign in
3. 4. 5.
6.
1990 for the Propagation of Compact Fluorescent Lamps,” in “Proceedings of the 1st European Conference on Energy-Efficient Lighting,” E. Mills ed., Swedish National Board for Industrial and Technical Development, Stockholm, Sweden (1991). B. Nielsen and K.H. Thomsen, “Belasmingskurver f0r boligbelysning, specielt elspareprerer (Load Curves of Lighting in Homes, Especially Compact Fluorescent Lamps).” DEFU, Research Association of Danish Electric Utilities, TR 286 (in Danish), Lyngby, Denmark (1988). B. Nielsen, K.H. Thomsen, and J. Mgller, “EFFO fors0g med lysstoflamper: Slutrapport (EFFO’s Compact Fluorescent Lamp Project: Final Report),” Research Association of Danish Electric Utilities, TR 276 (in Danish) Lyngby, Denmark (1989). B. Huenges Wajer and R. Kemna, “Electricity Consumption for Lighting in the Average Dutch Household: Development of an Accounting Method,” in “Proceedings of the 1st European Conference on Energy-Efficient Lighting,” E. Mills ed., Swedish National Board for Industrial and Technical Development, Stockholm, Sweden (1991). S.Bartlett, Energy -- The International Journal 18, 171 (1993).