Technical Work Group Review of EUC Definitions 15 th January 2014
Technical Work Group
Review of EUC Definitions
15th January 2014
• The purpose of this analysis is to review the appropriateness of current EUC definitions for small and large NDMs.
• Band 9 should be dismissed when considering ‘bands to be merged’ as a band that has daily metered sites will always need to exist and the current boundary can not be changed.
• The data used in this analysis was taken from the Autumn collection (which is used primarily for the performance evaluation).
Analysis has been carried out at national level.
The years that have been analysed are as follows:
• 2009/10 (Gas year)
• 2010/11 (Gas year)
• 2011/12 (Gas year)
• The following slides present the analysis for 2011/12 as the results for all years are fairly consistent.
Background
Additional Analysis
• As requested by TWG (on 27.11.13) the following additional analysis has been carried out:– Band 1 has now been included in the analysis– The daily average consumption has been calculated for each
band – The daily standard deviation has been calculated for each band– An ALP has been calculated for each band
Summary of Sample Size
0
500
1000
1500
2000
2500
3000
3500
4000
1 2 3 4 5 6 7 8 9
Band
Fre
quency
Autumn2010
Autumn2011
Autumn2012
(Updated to include band 1)
Distribution of Sample within EUC Bands 2011/2012
(Updated slide)
Distribution of Sample within EUC Bands 2011/2012
Distribution of Sample within EUC Bands 2011/2012
Distribution of Sample within EUC Bands 2011/2012
Distribution of Sample within EUC Bands 2011/2012
Distribution of Sample within EUC Bands 2011/2012
Distribution of Sample within EUC Bands 2011/2012
Distribution of Sample within EUC Bands 2011/2012
58,600.00 09
58,600.00 29,300.00 08
29,300.00 14,650.00 07
14,650.00 5,860.00 06
5,860.00 2,196.00 05
2,196.00 732.00 04
732.00 293.00 03
293.00 73.20 02
73.20 -01
UpperLowerEUC
MWh
Some considerations need to be made when deciding
which bands could possibly be merged:
•Cut offs are tested for only bands 3 and above
(as agreed by DESC in Dec ’03, with a view to
mitigating summer scaling factor instability
•Upper limit of band 3 cannot be changed due to
the pricing structure (a separate pricing structure
which incorporates bands 2 and 3)
•Bands 4 and above have the same pricing
structure so merges could be possible within
these bands.
Current EUC Boundaries and considerations
Data used in analysis
• The data available that is not dependent on current EUCs:
– Daily Consumption
– LDZ
– LDZ CWV
• The first piece of analysis that was carried out was the assessment of the Winter Annual Ratio (WAR). WAR provides a quick indicator of differences within the sample. The WAR for each site has been calculated to assess how much of the annual consumption is used in the winter months (1st Dec – 31st
Mar) and how this varies within the current bands (See Box Plot).
• WAR has also been plotted on scatter plots by combining EUCs to see if there was a “step change” which indicated a different break point.
Box Plot of WAR across the EUC Bands (2011/12)
Scatter Plots of WAR by Band 2011/12
2011/12 WAR EUC BANDS 1 AND 2
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 50 100 150 200 250
AQ MWh
WA
R
WAR EUC Band 1 2011/12
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0 10 20 30 40 50 60 70
AQ MWh
WA
R
(Updated slide)
Scatter Plots of WAR by Band 2011/122011/12 WAR EUC Bands 2 and 3
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
73 173 273 373 473 573 673
AQ MWh
WA
R
2011/12 WAR EUC Bands 3 and 4
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
293 793 1293 1793
AQ MWh
WA
R
2011/12 WAR EUC Bands 4 and 5
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
732 1732 2732 3732 4732 5732
AQ MWh
WA
R
2011/12 WAR EUC Bands 5 and 6
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
2196 4196 6196 8196 10196 12196 14196
AQ MWh
WA
R
Scatter Plots of WAR by Band 2011/122011/12 WAR EUC Bands 6 and 7
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
5860 10860 15860 20860 25860
AQ MWh
WA
R
2011/12 WAR EUC Bands 7 and 8
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
14650 19650 24650 29650 34650 39650 44650 49650 54650
AQ MWh
WA
R
2011/12 WAR EUC Bands 8 and 9
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
29300 329300 629300 929300 1229300 1529300
AQ MWh
WA
R
2011/12 WAR EUC Bands 4, 5 and 6
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
732 2732 4732 6732 8732 10732 12732
AQ MWh
WA
R
Scatter Plots of WAR by Band 2011/12
2011/12 WAR EUC Bands 5, 6 and 7
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
2196 7196 12196 17196 22196 27196
AQ MWh
WA
R
2011/12 WAR EUC Bands 6, 7 and 8
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
5860 15860 25860 35860 45860 55860
AQ MWh
WA
R
From observing the WAR across the current bands, it appears that there are
clear similarities between Band 2 & 3 and Band 3 & 4.
There are also possible similarities between Band 4 & 5 and Band 7 & 8.
Daily Average Energy Consumption 2011/12
Band 1
0
20
40
60
80
100
120
140
01
/10
/11
01
/11
/11
01
/12
/11
01
/01
/12
01
/02
/12
01
/03
/12
01
/04
/12
01
/05
/12
01
/06
/12
01
/07
/12
01
/08
/12
01
/09
/12
Date
KW
h
Avg energy Std Dev
Band 2
0
200
400
600
800
1000
1200
1400
01
/10
/11
01
/11
/11
01
/12
/11
01
/01
/12
01
/02
/12
01
/03
/12
01
/04
/12
01
/05
/12
01
/06
/12
01
/07
/12
01
/08
/12
01
/09
/12
Date
KW
h
Avg Energy Std Dev
Band 3
0
500
1000
1500
2000
2500
3000
3500
4000
01
/10
/11
01
/11
/11
01
/12
/11
01
/01
/12
01
/02
/12
01
/03
/12
01
/04
/12
01
/05
/12
01
/06
/12
01
/07
/12
01
/08
/12
01
/09
/12
Date
KW
h
Avg Energy Std Dev
Band 4
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
01/1
0/1
1
01/1
1/1
1
01/1
2/1
1
01/0
1/1
2
01/0
2/1
2
01/0
3/1
2
01/0
4/1
2
01/0
5/1
2
01/0
6/1
2
01/0
7/1
2
01/0
8/1
2
01/0
9/1
2
Date
KW
h
Avg Energy Std Dev
Daily Average Energy Consumption 2011/12Band 5
0
5000
10000
15000
20000
25000
01
/10
/11
01
/11
/11
01
/12
/11
01
/01
/12
01
/02
/12
01
/03
/12
01
/04
/12
01
/05
/12
01
/06
/12
01
/07
/12
01
/08
/12
01
/09
/12
Date
KW
h
Avg Energy Std Dev
Band 6
0
5000
10000
15000
20000
25000
30000
35000
40000
45000
50000
01
/10
/11
01
/11
/11
01
/12
/11
01
/01
/12
01
/02
/12
01
/03
/12
01
/04
/12
01
/05
/12
01
/06
/12
01
/07
/12
01
/08
/12
01
/09
/12
Date
KW
h
Avg Energy Std Dev
Band 7
0
10000
20000
30000
40000
50000
60000
70000
80000
90000
100000
01
/10
/11
01
/11
/11
01
/12
/11
01
/01
/12
01
/02
/12
01
/03
/12
01
/04
/12
01
/05
/12
01
/06
/12
01
/07
/12
01
/08
/12
01
/09
/12
Date
KW
h
Avg Energy Std Dev
Band 8
0
20000
40000
60000
80000
100000
120000
140000
160000
180000
200000
01
/10
/11
01
/11
/11
01
/12
/11
01
/01
/12
01
/02
/12
01
/03
/12
01
/04
/12
01
/05
/12
01
/06
/12
01
/07
/12
01
/08
/12
01
/09
/12
Date
KW
h
Avg Energy Std Dev
ALPs 2011/12
0
0.5
1
1.5
2
2.5
3
01/1
0/2
011
31/1
0/2
011
30/1
1/2
011
30/1
2/2
011
29/0
1/2
012
28/0
2/2
012
29/0
3/2
012
28/0
4/2
012
28/0
5/2
012
27/0
6/2
012
27/0
7/2
012
26/0
8/2
012
25/0
9/2
012
Date
ALP
Band 1
Band 2
Band 3
Band 4
The ALP has been calculated by calculating the sum of the energy
for the day (by band) and dividing that by the average of the sum
for the whole year (by band).
ALPs 2011/12
0
0.5
1
1.5
2
2.5
3
01/1
0/2
011
31/1
0/2
011
30/1
1/2
011
30/1
2/2
011
29/0
1/2
012
28/0
2/2
012
29/0
3/2
012
28/0
4/2
012
28/0
5/2
012
27/0
6/2
012
27/0
7/2
012
26/0
8/2
012
25/0
9/2
012
Date
ALP
Band 3
Band 4
Band 5
Band 6
ALPs 2011/12
0
0.5
1
1.5
2
2.5
01/1
0/2
011
31/1
0/2
011
30/1
1/2
011
30/1
2/2
011
29/0
1/2
012
28/0
2/2
012
29/0
3/2
012
28/0
4/2
012
28/0
5/2
012
27/0
6/2
012
27/0
7/2
012
26/0
8/2
012
25/0
9/2
012
Date
ALP
Band 5
Band 6
Band 7
Band 8
Intercept Analysis
The next piece of analysis carried out was based on the cwv (x)
intercept across the current EUC bands i.e. what the cwv is when
demand (y) is zero – and how this varies across the
bands.
To do this, regressions were calculated by:
• aggregating demand at LDZ level
• using the LDZ cwv
• Mon – Thu (excluding holidays)
Box plot of cwv (x) intercept 2011/12
From observing the cwv intercept across the current bands, it appears that
there are similarities in the relationship between energy consumption and cwv
for Bands 1, 2, 3 and possibly Band 4.
Intercept Analysis
Updated Recommendations / Conclusions
• No strong evidence of better break points from the data
• Daily average consumption shapes suggest the following:
– Band 3, 4 and 5 (and possibly band 2) look very similar
– Band 7 and 8 (and possibly band 6) look very similar
• ALP shapes also suggest that bands 2 & 3 and bands 7 & 8 look very similar
• Possible scope to rationalise Bands 3 to 5 and 6 to 8
• Simpler solution is use of more aggregation in modelling
• More complex change is to alter EUC Bands and/or reduce number of bands
TWG views now invited