CI RED 20th International Conference on Electricity Distribution
Paper 0552
EVALUATION OF TECHNICAL LOSSES ESTIMATION IN LV POWER DISTRIBUTION
SYSTEMS
Leonardo QUEIROZ Celso CAVELLUCCI
ANEEL - Brazil UNICAMP - Brazil
leonardoqueiroz@aneel.gov.br celsocv@densis.fee.unicamp.br
ABSTRACT
This paper proposes a methodology to estimate technical
losses in low voltage (L V) distribution systems. Its main
contribution is the development ofregress ion models able
to estimate technical losses with low levels ofinformation
about the network. Regression models are used to predict
one variable (dependent variable) from one or more
variables (independent variables). The dependent variable
is the technical loss, and the independent variables are any
data that allows characterizing a L V network. Different sets
ofindependent variables are evaluated, with respect to two
scenarios: a distribution company and the regulatory
agency. Independent variables are analyzed with respect to
relevance (given by the value of statistical correlation),
availability and auditability. An analysis ofthe regression
models and of the set of independent variables is also
presented
INTRODUCTION
Since the fifties, with the increasing popularization of the
digital computer, power flow models have been developed
to determine the status of the systems - voltages and
currents. In distribution systems it has been more common
since 1980, with some approaches that take advantage of
the radial configuration of distribution networks [1].
Since then, the usual approach to evaluate technical losses
is based on a power flow algorithm, sometimes improved
with on-line measurements at strategic points. However,
there are many Brazilian distribution companies that still do
not have a power flow algorithm available for studies of low
voltage (LV) networks. Indeed, for some municipal
distribution companies with less than 3,000 customers, it is
not economically attractive to implement some
computational tools. Also, some larger distribution
companies have poor information about their LV networks,
what turns difficulty to assess system performance.
Furthermore, the Brazilian Electricity Regulatory Agency -
ANEEL needs information about technical losses for each
Brazilian power distribution company; this is a key
information to establish reference prices for energy.
Therefore, ANEEL needs an estimation of technical losses
in power distribution companies. Additionally, it is
desirable to assess all distribution companies with the same
methodology.
The estimation of technical losses with low level of
information in LV networks has already been handled in
previous papers [2-5]. In most approaches the motivation
was the loss estimation for companies that do not have
enough information about their networks. These papers
CIRED2009 Session 5 Paper No 0552
Prague, 8-11 June 2009
Christiano LYRA
UNICAMP - Brazil
chrlyra@densis.fee.unicamp.br
proposed a variety of approaches, from simple equations to
load flow analysis in representative networks. The present
paper proposes a new approach, which relies on regression
models to estimate technical losses with low level of
available information.
AN OUTLINE OF LOSSES IN BRAZIL
Technical losses are part of the process of production-
transportation-consumption of electrical energy. Since they
cannot be eliminated, it is necessary to consider them in the
planning and design of networks. It must be remarked that
technical losses do not need to be reduced at any cost, but
optimized to the right compromise. In a regulatory analysis,
ANEEL needs to stimulate the efficiency of the electric
market. Specifically in distribution systems, companies
need to follow guide lines for quality and efficiency.
When comparing Brazilian technical losses, we would
expect they were bigger than the majority of other countries.
The first reason is the generation characteristic: the system
is predominantly hydro-generation, and the hydraulic power
plants are usually far away from the main consumption
centres - and this scenario will get worse, since the river
falls near consumption centres are almost over. So, we need
big transmission lines connecting loads to generators.
The second reason for higher technical losses in Brazilian
networks is the low level of load density. In addition to the
large size of the country, most of the population is
concentrated in the central-south regions.
Figure 1 shows the Brazilian technical and non technical
percent losses of 54 distribution companies. Losses in the
National Connected System - the HV transmission lines -
are not computed. The mean technical losses in distribution
companies seem to be in an expected level. However, Brazil
has critical non technical losses levels. This is a social-
economic problem, and even with some recent efforts, the
problem does not seem to have a short term solution.
ANEEL adopts different regulatory treatment for technical
and non technical losses. For technical losses, ANEEL
applied a recent methodology to compute segmented losses,
according to the voltage levels. In order to define acceptable
levels for non technical losses, ANEEL uses a regression
analysis with some socio-economics attributes that are
somehow correlated to non technical losses, and it is
establishing target levels for distribution companies.
It is important to have an adequate method to estimate
technical losses, both for its own sake and because non
technical losses are estimated as the difference between
total losses and technical losses.
CI R ED 20th International Conference on Electricity Distribution Prague, 8-11 June 2009

Paper 0552

Figure I : Distribution total losses of 54 Brazilian Distribution Compani es.

37 .50%

35 .00% o Non technical loss l-

I El Technical loss
I l-
32 .50% I I
l-
30 .00%
I-
27 .50%
l-
25.00%
I-
22 .50%
l-
20.00%
I-
17.50%
15.00% I-
I'M
12.50%
10 .00%
7.50%

5.00%
2.50%
0.00%
romut ~ N ~ ~ ~ ID ~ ro m o~ N ~ v ~ ID ~ ro m o~ N ~ v ~
~ N ~ v ~ ID ~ ro m ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ g M M M M M M M M MV V V V V V V V V V ~ ~ ~ ~ ~~
::;;

Given the available variables, we can combine them in

LV TECHNICAL LOSSES ESTIMATION
Availability of Information
This paper focuses on the estimation of technical losses in
an environment of low level information about the
network. Two scenarios are studied:
• The first scenario studies technical losses in
distribution companies with poor knowledge
about their LV networks (different degrees of
poor knowledge are considered);
• The second scenario evaluates technical losses
from a regulatory point of view; in this scenario,
the information required by the methodology
must be available for all distribution companies
and must be auditable by ANEEL.
These two scenario s were not explicitly developed, but can
be seen in the different degrees of information shown in
the sets of variables requir ed by the model. These sets were
built from the following available variables for each LV
circuit:
1. Nominal voltage: voltage of the circuit (V);
2. Minimum resistance: resistance of the best cable
(ohms/km);
3. Maximum resistance: resistance of the worst
cable (ohms/km);
4. Length: total length (metres) ;
5. Numb er of customers: number of the customers
of the circuit;
6. Consumption: total consumption ofthe customers
during the period (kWh);
7. Maximum demand: maximum coincident demand
registered at the transformer (kVA).
The data set used has information from 1082 LV
distribution networks [6].
different sets - simulating a variation of the level of
information. Table I shows the sets propo sed - they are
sorted according to the crescent level of required
information. The variables are referred by the numbers
assigned in last paragraph.
The classification according to relevance and availability is
somehow subjective, and it can vary among companies.
Variables can be classified according to their nature :
network (variables 1 to 4) and load variables (5 to 7).
Network variables are easier to obtain, since they are more
static than load variables, which can be obtained by
measurements (desired) or estimation procedures.
Variable 5 needs to be used with caution: it does not
represent any load, but has high correlation with
consumption and maximum demand (0.877 and 0.902 of
Pearson Correlation Coefficient, respectively). But it may
vary from company to company, or even for different
regions of a company; its use is recommended only for
companies that have a uniform LV market.
Some other variables were available for the study, but were
discarded. For instance, the maximum voltage drop ofeach
circuit is highly correlated to energy losses (0.8 of Pearson
Correlation Coefficient), but it is one of the most
unavailable variables and it is very difficult to audit. The
nominal power of transformers (KVA) has a good
availability, but it is influenc ed by the design criteria of
each company, and its use may lead to wrong results
among companies.
As Table I shows, sixteen sets of variables were
considered. The first (A) represents the lowest level of
information about the network - notice that no explicitly
information about load is present. Explicitly, becau se it
may carry it implicitly in design criteria, such as maximum
current, that regression methods may "learn" . The last set
(P) represents the highest level of information.

CIRED2009 Session 5 Paper No 0552

CI R ED 20th International Conference on Electricity Distribution Prague, 8-11 June 2009

Paper 0552

Table I: Sets of the seven variables. Table II: Pearson Correlation Coefficient.
Variables Variable Original Transformed
Set 1 2 3 4 5 6 7 1 0.246 0.276
A ./ ./ ./ 2 -0.326 -0.420
8 ./ ./ ./ ./ 3 -0.013 -0.040
C ./ ./ ./ ./ 4 0.632 0.693
D ./ ./ ./ ./ ./ 5 0.675 0.722
E ./ ./ ./ ./ 6 0.769 0.850
F ./ ./ ./ ./ ./ 7 0.773 0.854
G ./ ./ ./ ./ ./
./ ./ ./ ./ ./ ./
Regression Models
H
I ./ ./ ./ ./ Once the sets of variables are defined, models ofregressors
./ ./ ./ ./ ./ were developed to make the inference between variables
J
and energy losses. Regression models are used to predict
K ./ ./ ./ ./ ./
one variable (dependent variable) from one or more
L ./ ./ ./ ./ ./ ./
variables (independent variables) [7].
M ./ ./ ./ ./ ./ The regression models investigated are:
N ./ ./ ./ ./ ./ ./ • Linear regression (LR);
0 ./ ./ ./ ./ ./ ./ • Robust regression (RR); and
P ./ ./ ./ ./ ./ ./ ./ • Artificial Neural Networks (ANN).
Even though LR and RR are linear models, while LR uses
In linear regression analysis, transformations of variables the least squares method to fit the data, RR uses an
are recommended due to linear regression requirements, iteratively reweighed least square - which is less sensitive
such as linearity and normality. For this reason, the to outliers.
following transformations were attempted (for independent The ANN used is a Multi-Layer Perceptron (MLP) [8],
and dependent variables): logarithm, square root, square with the Levenberg-Marquardt optimization rule to update
power and inverse. Transformed sets of variables will be weights. ANN, especially ofthe MLP type, is widely used
represented by t; i. e., At will represent A set with in regression analysis, due to its characteristic ofuniversal
transformed variables. Thus, we have 32 different sets of function approximation.
independent variables (A to P and At to Pt), and 2 sets of After many simulations, the ANN configuration chosen
dependent variables (losses and transformed losses, Y and was one hidden layer with 10 neurons, and 5000 iterations
Y), composing 64 different sets of variables (including were used as maximum iterations stop criteria. To train
independent and dependent variables). regression models, the 1082 LV distribution networks were
Independent variables were analyzed with respect to divided in 757 for training and 325 for test.
relevance for energy losses (given by the value of its
statistical correlation). The correlation analysed was the REGRESSION ANALYSIS
Pearson Coefficient, the Spearman's and the Kendall's
Rank. The first gives a measure of linear relationship Because of the impossibility of presenting here all the
between each variable and the energy loss, varying from -1 results (there are 192 results to be shown: 3 regressors
(perfectly opposite correlated) to 1 (perfectly correlated). executed 64 different sets of dependent and independent
The null relationship is indicated by a (zero). Table II variables), Table III presents only the ten best regressions
shows the Pearson Correlation between each variable and found, ordered from the first to the tenth. All ofthem were
given by ANN. The Root Mean Squared Error - RMSE is
energy losses (for original and transformed variables).
Since we know that there are nonlinear relationships in the measurement of goodness of fit used. Column "TLD"-
Total Loss Deviation - gives the percent difference
energy loss functions, the Kendall's and Spearman's
coefficients were used. They are non parametric methods, between total loss of test set calculated by regressors and
and their measurement of relationship is given as a rank. total real loss. "Set X" refers to input variables
Both ranks corroborated the results of Table II. (independent), while "Set Y" refers to output variables
The correlation coefficient is a measure of relationship. (dependent).
Therefore, in spite of being a good indication, it cannot be As expected, ANN found the best results. But it is
interpreted isolated. Additional analysis, as always important to highlight that linear regression obtained good
recommended in statistics studies, need to be done. results too. Table IV shows the mean performance of the
Actually, the availability and auditability were also three regressors.
considered. Indeed, it does not matter if a variable has a Another assessment of performance can be presented by
direct relation with losses if it is not available and the data sets. Table V presents a comparison between the
auditable (for instance, the maximum voltage drop). 32 independent variable sets.

CIRED2009 Session 5 Paper No 0552

CI RED 20th International Conference on Electricity Distribution Prague, 8-11 June 2009

Paper 0552

Table III: Best regression results. Table IV: Comparison of regressors performance.
SetX SetY TLD RMSE Parameter LR RR ANN
Yt 1.90°A> 723.43 TLD 4.83°A> 17.11°A> 3.18°A>
Yt 1.89% 906.92 RMSE 1829.72 1956.17 1735.74
Yt 2.31°A> 920.14
Yt 1.32°A> 978.42 T abl e V C om panson 0 f'mpu t ser s per ormance.
Yt 2.43°A> 1108.16 Mean Mean Mean
1184.27 Set RMSE Set RMSE Set RMSE
Yt 3.61°A>
Y 4.30°A> 1279.25 J 1546.65 I 1755.54 Ht 1838.67
Yt 4. 14°A> 1315.71 K 1593.61 Mt 1769.29 Ot 1890.81
Y 4.70% 1315.79 E 1664.85 G 1783.88 D 1961.57
Yt 3.39°A> 1346.97 F 1665.58 Gt 1785.89 C 1981.28
Yt 3.18°A> 1383.98 P 1666.12 It 1789.22 Dt 2125.09
v, 1.01°A> 1406.56 H 1666.27 Kt 1791.39 Ct 2145.01
Yt 1.09°A> 1415.60 Jt 1669.89 Pt 1807.84 8 2173.15
L 1688.50 Et 1820.24 A 2203.08
Note that the lack of information about load, absent in M 1700.78 0 1823.98 8t 2219.97
variables A to D, may explain the worse performance of Lt 1731.30 Nt 1824.43 At 2249.17
these input sets. On the other hand, the maximum N 1737.97 Ft 1826.35 - -
information about loads, present in sets M to P, does not
seem to be mandatory: only two sets that contain them (P REFERENCES
and M) are present in the ten best sets. Other inference is
about transformations: sets with transformed variables [1] D. Shirmohammadi, H.W. Hong, A. Semlyen and
have worse performance than sets with the original G.X. Luo, 1988, "A compensation-based power flow
variable - although Table III has a lot of transformed sets method for weakly meshed distribution and
in the best results. It is important to emphasize that Tables transmission networks", IEEE Trans. on Power
IV and V contain mean results, and are "polluted" with Systems, vol. 3, 753-762.
other aspects. [2] J.J. Grainger and T.l Kendrew, 1989, "Evaluation of
technical losses on electric distribution systems",
1Oth International Conference on Electricity
CONCLUSION
Distribution. CIRED, vol.6, 488-493.
This paper presented a methodology to estimate technical [3] R. Nadira, S. Benchluch and C. A. Dortolina, 2003,
losses in LV distribution systems. The aim is to predict "A novel approach to computing distribution losses",
energy losses with low level of information. Although the Trans. and Dist. Conf. and Exposition, IEEE PES,
crescent uses of load flow tools, this methodology is vol.2, 659-663.
applicable to those companies that have low level of [4] H. Lasso, C. Ascanio and M. Guglia, 2006, "A
information. They can apply directly the methodology or Model for Calculating Technical Losses in the
submit their data to a pre-processing step, grouping Secondary Energy Distribution Network", Trans. &
networks in representative clusters and applying the Dist. Con! and Exposition: Latin America. TDC '06.
methodology to one network that represents each cluster. IEEE/PES, 1-6.
The methodology is also indicated for the regulatory [5] J. R. C. Orillaza, R. Del Mundo, and l A. C. Miras,
agencies, which need to estimate losses of all companies 2006, "Development ofModels and Methodology for
using the same methodology. the Segregation of Distribution System Losses for
Three regression methods were tested. While ANN Regulation", TENCON 2006, IEEE Region 10
achieved best results, LR showed that it can be used due to Conference, 1-4.
its simplicity. An analysis of variable sets showed that it is [6] L. M. O. Queiroz, C. Cavellucci and C. Lyra, 2008,
important to use load variables, even considering the "Methodology of LV Distribution System Network
difficulty to audit it. But while maximum coincident Generation for Planning Purposes" (in Portuguese),
demand is hard to audit, the energy consumption is more XL SBPO, SOBRAPO.
auditable. [7] J. F. Hair, R. E. Anderson, R. L. Tatham and W. C.
Black, 1998, Multivariate Data Analysis, Fifth
Acknowledgments Edition, Prentice-Hall.
[8] S. Haykin, 1999, Neural Networks: A Comprehensive
The authors would like to acknowledge the support given Foundation, Second Edition, Prentice Hall.
by the Brazilian National Research Council (CNPq).

CIRED2009 Session 5 Paper No 0552