Pontificia Universidad Católica de Chile
Departamento de Ingeniería Eléctrica
Benchmark Regulation and
Efficiency of Electricity Distribution
Hugh Rudnick
Bundesnetzagentur Conference
Bonn, 25-26 April 2006
Objective
Share the experience of Latin American countries that, from
1982 onwards, reformed their electricity power sectors to
establish conditions of economic efficiency and attract
private investment, replacing central planning and control
by market oriented approaches.
Competition was introduced in generation and “pseudo
competition” in transmission and distribution.
Share the characteristics of benchmark yard stick
regulation for distribution, with analytical cost models,
model company and model network concepts, used in
Chile, Argentina, Peru and others.
The importance of efficiency in distribution
Regulating the monopoly
The distribution activity- costs to consider
Analytical models
The building of an efficient company
Results of incentive regulation
Projections into the future (DEA, SFA, etc.)
Other elements of regulation
The building of the price for the final
consumer
Generation and transmission systems
Distribution systems
MgC
Tx
MgC
STx
P generation + cost Tx
+ cost STx
P generation + cost Tx + cost STx +cost Dx
Electricity cost
Industrial consumer
Chilean Pesos
7000000
6000000
5000000
4000000
3000000
2000000
1000000
0
1992
1993
1994
1995
1996
1997
Year
Energy cost
Distribution cost
1998
1999
2000
2001
Electricity cost
Residential consumer
Chilean Pesos
7000
6000
5000
4000
3000
2000
1000
0
199
1993
1994
1995
1996
1997
Year
Energy cost
Distribution cost
1998
1999
2000
2001
Market reforms: benefit final consumer
Market environment
MgC
Tx
STx
MgC
P generation + cost Tx + cost STx +cost Dx
Monopolistic activity
(stimulate efficiency through regulation)
The importance of efficiency in distribution
Regulating the monopoly
The distribution activity- costs to consider
Analytical models
The building of an efficient company
Results of incentive regulation
Projections into the future (DEA, SFA, etc.)
Other elements of regulation
Price regulation objectives for monopoly
Strategic objectives
-avoid loses of monopoly prices
-avoid individual tariff considerations
-provide
-economic efficiency signals for operation and investment
-flexibility and stability
-adequate response to market changes
-symmetry of risks and opportunities for the regulated agent
Prices must be fixed considering
-cost of service provision
-reflect to the consumer the supply cost structure
-adequate return to investment capital
-incentives for cost reduction
Distribution monopoly regulation
Requirements
-concession required to run the business
-public service activity
-obligation to serve
-open access to wires
-quality of service regulated
Rights
-concession granted for the right to use public roads and
private grounds with compensation
-realistic price regulation considering restrictions
-adequate return to investment capital, if efficient
-higher efficiency rewarded
Regulatory Models
All models to regulate industrial monopolies consider fixing
initial prices and adjusting them with time.
Po ⋅ f(P't ) + Zt = Pt
Initial Price
Annual
adjustment factor
Price in year t
Adjustment
Regulatory Models
Fixing initial prices
From “what are the costs?” to “what they should/could
be?”
Based on real costs, standardised ones, efficient ones?
Adjustment with time
Regulatory models adjust prices on a yearly bases, given
parameter changes or incentive needs
Based on inflation, efficiency increases?
Regulatory Models
Cost Plus (Cost based) Regulation
recognize real costs and determine required rent, no incentives to reduce
costs
Incentive Regulation - Price Cap
bridge between traditional regulation and deregulation – costs fixed and
reduced over time (productivity index RPI-X)
Model Company or Yardstick
Competition
possibility to compare costs of different similar companies- introduction of
concept of a model company
Model Company or Yardstick
Competition
P(q)
Average cost
Inefficient company
Model company
High efficient company
q
Quantity
‰ Comparison of different companies achieved through identifying
different “distribution areas”
‰ Models determined for each area, tariffs different for each area
Challenges
-definition of “efficient” investment and
administration
-manage information deficiencies
-regulator independent from monopoly
pressure (avoid capture)
-need to consider reliability and quality of
supply
Risks
-conflictive interests between regulator and
monopoly
-regulator manipulating model company with
other objectives
The importance of efficiency in distribution
Regulating the monopoly
The distribution activity- costs to consider
Analytical models
The building of an efficient company
Results of incentive regulation
Projections into the future (DEA, SFA, etc.)
Other elements of regulation
Process of electricity distribution
Investment
planning
• Identify new
system needs.
• Develop plans to
satisfy need.
• Facilitate plans
and budgets.
•Assess large
projects.
Network
construction
System
operations
System
maintenance
• Build lines,
substations, low
voltage
networks.
• Connections &
disconnections
•Follow up and
control of
system
operations
•Correct
damages &
incidents.
• Maintenance of
lines,
substations and
low voltage
networks.
• Maintenance of
protection
equipment.
Business
management
• Measurement,
meter reading
•Billing
•Contracts
•Solution
management.
Cost components in added distribution
value
Distribution Value Added
Fixed costs
• Meter reading
• Billing
• Distribution of bills
• Accounting related to
client
• Bill follow up
• Client relation
• Other fixed costs
related to client.
Investment costs, Operation & Maintenance
Capital
costs
Operational
costs
HV
High Voltage
• Feeders
• Control
equipment.
• Protection
equipment.
Low Voltage
• Distribution
substations
•Low voltage
networks
•Protection
equipment
LV
Distribution losses
Technical
losses
Other
losses
ƒStorage
ƒWorkshops
ƒLabs and tools
ƒOperational engineering
ƒTransport
ƒSecurity
ƒRent, insurance
ƒNetwork maintenance and
operation
ƒPatents and property taxes
Characteristics of distribution costs
Cost of distribution dictated by capacity requirements
(demand defines network dimensioning, thus defines
investment and operation)
Energy only important given network losses
Important economies of scope (the higher the load
density, the lower the unit costs)
Marginal cost of distribution equal to average costs
(except for high density)
Distribution characteristics
Total monthly costs semi urban area, Chile, 1988
Average value 10,2 US$/kW/month
Total cost (MUS$/Month
1800
1600
1400
1200
1000
800
600
400
200
0
0
50
100
150
Maximum demand (MW)
200
Distribution characteristics
Total monthly cost rural area, Chile, 1988)
Average values 13,8 US$/kW/month
Total cost (MUS$/Month)
700
600
500
400
300
200
100
0
0
10
20
30
Maximum demand (MW)
40
50
The importance of efficiency in distribution
Regulating the monopoly
The distribution activity- costs to consider
Analytical models
The building of an efficient company
Results of incentive regulation
Projections into the future (DEA, SFA, etc.)
Other elements of regulation
Cost structure for distribution
Investment annuity + Operation + Losses
Cost =
kWHV + kWLV
Replacement cost concept for investment.
Investment annuity considering useful life of installations and
given return rate.
Econometric models are built to represent the industry costs, with
a log-log regression between the average cost per km*kW and the
km*kW, for high voltage and low voltage.
Following examples represent Chilean distribution industry.
Distribution characteristics
Relation between unit costs HV and the product
kmHV*(kWHV+kWLV)
2
1
-1
(m$/kW/km)
ln(Cost HV per kW*km)
0
0
5
10
15
20
25
-2
-3
-4
-5
-6
-7
ln(kW*km)
Cost
HV
= km
HV
⋅ (kW
HV
+ kW
LV
)⋅ e(A⋅ln (km
HV
⋅(kW
HV
+ kW
LV
))+ B )
Distribution characteristics
Relation between unit cost LV and the product kmLV*kWLV
4
3
ln(Cost LV per kW*km)
(m$/kW/km)
2
1
0
-1 0
5
10
15
20
-2
-3
-4
-5
-6
ln(kW*km)
Cost
LV
= km
⋅ kW ⋅ e
LV
LV
(A⋅ln (km
LV
⋅kW
LV
)+ B )
25
Investment (VNR) and operation cost (Cexp) for 12 of the 39 companies considered
Physical parameters: Iength of HV and LV networks, kW demand HV and LV
Structure of yardstick competition
Industry segmentation
-companies are grouped according to similar distribution
costs, considering a maximum deviation
-segments are identified
Reference company is chosen for each segment
-cost close to the group average
-representation coverage
-existence of previous studies
Efficiency distribution costs determined for each segment
-global industry check up
Tariffs set every four years
-indexation
2004 Chilean study
Segments and model company
Area 1:
Chilectra
Area 2:
Río Maipo, Emelat, CGE , Luzandes, Puente Alto, Eliqsa, Elecda, Conafe,
Coop. Curicó.
Area 3:
Chilquinta, Emelari, Edelmag, Emec, Emelectric, Energía de
Casablanca, Saesa.
Area 4:
Elecoop, Colina, Edelaysen.
Area 5:
Til Til, Luzlinares, Emetal, Litoral, Frontel, Luzosorno
Area 6:
Emelca, Luzparral, Socoepa, Codiner, Coopelan, Copelec, Coelcha,
Cooprel.
The importance of efficiency in distribution
Regulating the monopoly
The distribution activity- costs to consider
Analytical models
The building of an efficient company
Results of incentive regulation
Projections into the future (DEA, SFA, etc.)
Other elements of regulation
Building the efficient company
Efficient network
-Low voltage network design
-Distribution transformer sizing
-High voltage network design
Efficient infrastructure (buildings and facilities)
Efficient management
Energy and capacity balances
Loss calculations
Use of geographical information - GIS based
LOW VOLTAGE
NETWORK DESIGN
Based on
identification of
families of cells,
electric project for
sample ones,
extrapolation to
universe
500*500 meter cells
Conductor and transformer selection
Objective function
-minimize present value of investment, losses and
depreciation.
Parameters (2004)
Value
Unit
Description
N
15
Years
Td
10
%
Return rate
Tc
0a3
%
Demand growth rate
ρE
18,55
$/kWh
ρP
5.523,36
$/kW
Study horizon
Energy price
Capacity price
Transformer selection
Proyectos Transformadores 3F de 12 kV
14000
12000
VAN [M$]
10000
8000
6000
4000
10 kVA
15 kVA
30 kVA
45 kVA
75 kVA
100 kVA
150 kVA
300 kVA
500 kVA
2000
0
1
26
51
76
101
126
151
176
201
226
Demanda Inicial [kVA]
i
i
i
INV
+
CPT
⋅
(
1
−
τ
)
−
DEP
k
k +1
k +1 ⋅ τ
VANi = ∑
k
(1 + Td )
k =0
N
Conductor selection
Proyectos Conductores Aluminio Desnudo
5000
4500
4000
VAN[M$]
3500
3000
70 mm²
120 mm²
240 mm²
300 mm²
2500
2000
1500
1000
500
0
1
26
51
76
101
126
Demanda Inicial [Amp]
N
VAN i = ∑
k =0
INVki + CPTki +1 ⋅ (1 − τ) − DEPki +1 ⋅ τ
(1 + Td )
k
HIGH VOLTAGE
NETWORK DESIGN
Based on computer
model (Peco)
Santiago high voltage
network (12 and 23 kV)
Summary of High Voltage distribution per company
Items
1.- AÉRIAL NETWORK
1.1 Km network
1.2 Poles
1.3 Structures
1.4 Electric equipment
1.5 Ground connections
1.6 Others
Unit
km
c/u
c/u
c/u
c/u
c/u
Total Aerial
2.- UNDERGROUND NETWORK
2.1 Km network
2.2 Vaults
2.3 Canals
2.4 Electric equipment
2.5 Ground connections
2.6 Others
Total Underground
Total Company
km
c/u
c/u
c/u
c/u
c/u
Quantity
Value M$
Quality of service requirements
Interruption indexes
Urban
Rural
Interruption average frequency per
transformer (FMIT)
5,0 times a
year
7,0 times a
year
Interruption average frequency per
kVA, (FMTK)
3,5 times a
year
5,0 times a
year
Total interruption time per
transformer, (TTIT)
22 hours a
year
28 hours a
year
Total interruption time per kVA, (TTIK)
13 hours a
year
18 hours a
year
Penalties
Compensations to consumers based on value of non served energy
The importance of efficiency in distribution
Regulating the monopoly
The distribution activity- costs to consider
Analytical models
The building of an efficient company
Results of incentive regulation
Projections into the future (DEA, SFA, etc.)
Other elements of regulation
Cost reductions (Chile)
Evolution Low Voltage Cost Area 1
Pesos of July 2004
9.000
8.000
$/kW/month
7.000
6.000
5.000
4.000
3.000
2.000
1.000
0
1984
1988
1992
1996
Year
2000
2004
Loss reductions – technical and non technical
Chilectra
(Santiago, Chile)
Edesur
(Buenos Aires, Argentina)
30
26,0
25
21,9
19,8
20
%
18,8
16,1
16,2
13,6
15
13,3
12,5
12,0
10,6
10
9,3
9,3
5
0
1987
1988
1989
1990
1991
1992
1993
1994
1995
Loss reductions (Peru)
25%
21.8%
20%
13.9%
20.6%
19.7%
18.2%
17.1%
15.4%
15%
14.6%
12.4% 11.8%
10%
99
98
97
96
95
94
93
92
91
0%
90
5%
E. Zolezzi, CTE, Oct. 99
Clients per worker (Chile)
3500
Chilectra
3000
RioMaipo
2500
CGE
2000
Emelari
1500
Eliqsa
1000
Conafe
500
Emelectric
Year
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
0
1992
Clients/Worker
4000
Emelat
Litoral
4.0
80%
3.5
70%
3.0
60%
2.5
50%
2.0
40%
1.5
30%
Clientes
CE
1.0
20%
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
0%
1988
0.0
1987
10%
1986
0.5
1985
Coeficiente de Electrificación (%)
Millones de Clientes
Supply coverage (Peru)
Clients increase 77% from 1992 to 2001
Ref: PA Consulting Group
Costs as determined by parties involved
Chile results 2000 - HV cost ($/KW/year)
120.000
100.000
80.000
60.000
40.000
regulator
companies
20.000
result
0
1
2
3
4
Typical area
5
6
Returns of distribution companies (Chile)
35,000%
30,000%
CGE
25,000%
RIO MAIPO
%
CHILECTRA
20,000%
EMELECTRIC
ELIQSA
EMELARI
15,000%
CONAFE
EMELAT
10,000%
LITORAL
5,000%
0,000%
1992
1993
1994
1995
1996
1997
1998
Year
1999
2000
2001
2002
The importance of efficiency in distribution
Regulating the monopoly
The distribution activity- costs to consider
Analytical models
The building of an efficient company
Results of incentive regulation
Projections into the future (DEA, SFA, etc.)
Other elements of regulation
New methods to determine efficiency
1968 Econometric methodology (introduction of deterministic frontier
models)
1977 Stochastic frontier models and stochastic frontier analysis (SFA).
1978 Data envelopment analysis (DEA) with linear programming
technique.
-establish which companies of a sample determine the
envelopment area or efficient production frontier. The radial distance of
a company to the frontier provides the efficiency measurement.
Therefore, companies that are on the frontier (the ones that determine
it) are considered efficient, while the ones that are far from the frontier
are considered inefficient.
2003 DEA and Distribution Value Added
SEVERAL VARIABLES CONSIDERED IN RESEARCH
Variables
Valor Neto retornable total
Valor Neto retornable alta tensión
Valor Neto retornable baja tensión
Costos de explotación**
Costos de explotación alta tensión
Costos de explotación baja tensión
Energía comprada
Energía vendida total
Energía vendida en alta tensión
Energía vendida en baja tensión
Número de comunas
Número de clientes
Longitud total de líneas
Longitud de líneas en alta tensión
Longitud de líneas en baja tensión
Potencia total coincidente
Potencia total coincidente en horas de punta alta tensión
Potencia total coincidente en horas de punta baja tensión
Número de trabajadores
Remuneraciones anuales
Bienes muebles e inmuebles
Costo por compra de energía y potencia
Costos por pérdidas de energía y potencia***
* : Miles de pesos del 31 de Diciembre de 1999
** : No considera los costos por compra de energía y potencia
*** : De acuerdo a precio monómico definido por la CNE
Sigla
VNRT
VNRAT
VNRBT
CEXPLT
CEXPLAT
CEXPLBT
ECOMP
EVENDT
EVENDAT
EVENDBT
NCOMU
NCLTS
KMT
KMAT
KMBT
KWT
KWAT
KWBT
NTRAB
REMUN
BMI
CCOMP
CPERD
Unidad
m$*
m$*
m$*
m$*
m$*
m$*
kWh
kWh
kWh
kWh
Km
Km
Km
KW
KW
KW
m$*
m$*
m$*
m$*
VARIABLES UTILIZED IN DEA MODEL
Inputs
VADT
KM T
ENFAC
NTRAB
SALARIO
min
Outputs
(1),
(1),
(2),
(2),
(2),
(2),
(3)
(3)
(3)
(3)
(3)
EVEND
KWT
NCLTS
(1),
(1)
(1),
(2),
(3)
(2),
θp
n
subject to: θ p xip −
∑λ
x ≥0
i = 1, 2 ,...,m
ykj ≥ 0
k = 1, 2 ,...,s
j ij
j =1
n
− ykp +
∑λ
j
j =1
λj ≥ 0
j = 1, 2 ,...,n
where θp is the efficiency of the electrical distribution company under evaluation
VAD EFFICIENCY RESULTS
Empresa
EMELARI
ELIQSA
ELECDA
EMELAT
EMEC
CHILQUINTA
CONAFE
EMELCA
LITORAL
CHILECTRA
RIO MAIPO
COLINA
TIL TIL
PUENTE ALTO
LUZANDES
PIRQUE
EMELECTRIC
CGE
COOPELAN
FRONTEL
SAESA
EDELAYSEN
EDELMAG
CODINER
ELECOOP
EDECSA
COOP CURICO
EMETAL
LUZLINARES
LUZPARRAL
COPELEC
COELCHA
SOCOEPA
COOPREL
CREO
Cod.
SEG.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
39
Modelo 1
θ
67.09
60.54
79.71
46.97
35.30
60.86
54.66
56.39
41.53
100.00
84.96
96.39
53.60
85.89
100.00
56.89
26.63
49.59
22.32
19.46
29.08
44.66
58.07
23.92
30.42
53.18
28.91
16.21
21.11
21.57
17.77
21.72
38.39
50.01
61.06
Modelo 2
θ
70.56
80.68
98.85
43.83
38.79
45.65
62.88
95.86
58.98
100.00
100.00
70.68
53.61
85.89
100.00
100.00
44.32
68.04
49.44
37.13
59.69
55.15
61.39
71.49
98.61
100.00
81.29
96.27
50.76
100.00
30.92
47.50
79.17
60.96
100.00
Modelo 3
θ
83.13
100.00
99.96
87.51
65.51
51.42
100.00
95.92
58.98
100.00
100.00
95.69
55.24
85.89
100.00
100.00
45.83
100.00
49.81
42.52
70.16
57.01
100.00
71.77
98.61
100.00
81.96
96.27
55.14
100.00
33.92
47.50
79.17
60.96
100.00
The importance of efficiency in distribution
Regulating the monopoly
The distribution activity- costs to consider
Analytical models
The building of an efficient company
Results of incentive regulation
Projections into the future (DEA, SFA, etc.)
Other elements of regulation
Other elements
Transmission and sub transmission networks
-efficient replacement value
-efficient model company
-centralized transmission expansion
-tariffs set every four years, indexation
Competitive gas supply wit no price regulation
-natural gas competing with bottled liquefied gas
-only security and quality regulated
-market power supervision
Hugh Rudnick
Department of Electrical Engineering
Pontificia Universidad Católica de Chile
Santiago, Chile
Email [email protected]
http://www.ing.puc.cl/power/