Impact of Electricity Theft on Power Quality

Arango, L.G., Deccache, E., Bonatto, B. D., Arango, H., Ribeiro, P.F., Silveira, P. M.
CERIn - Center of Excellence in Smart Grids
UNIFEI - Federal University of Itajuba
Itajuba-MG, Brazil
lucasarango10@yahoo.com.br; elciodeccache@yahoo.com.br; bonatto@unifei.edu.br; hector.arango@uol.com.br;
pfribeiro@ieee.org; pmsilveira@unifei.edu.br

Abstract Electricity theft is responsible for economic problems Other non-technical losses such as fraud, measurement and
for the electric utility due to revenue loss caused by electricity billing errors, among others, will not be considered in this
consumers that are not paying for it. The stealer has a tendency paper, which initially focus only on electricity theft [2]-[4].
to consume more energy, resulting also in power quality
problems. An increase in power demand to values greater than
the transformer rated power can result in different quality
deviations, like transformer overload, voltage unbalance and
steady state voltage drop on system buses. The objective of this
paper is to analyze, by using MatLab simulations and considering
different grid configurations, how electricity theft results in
power quality issues, specifically voltage drop in steady state.
Additionally, it is shown how the steady state voltage drop can
result in economic penalties for the electric utility when the
proper voltage exceeds the network operational standards.

Keywords Electricity theft, Impact on Power Quality, Steady

state voltage drop, Simulations, Penalties.

I. INTRODUCTION
Figure 1 Electricity theft schematic on a distribution grid..

In a regulated environment, the electric utility buys In Fig. 1,

electricity from energy generators based on energy auctions. T = Distribution Transformer (Medium to Low Voltage).
This energy supplies the network through the transmission line
borders, representing the required energy. The total energy M1, M2,..., Mn = Energy Meters.
consumed by the loads is by definition the consumed energy. C1, C2,, Cn = Energy Consumers.
The difference between the required energy and the consumed
energy represents the global energy losses.
Not all energy is actually billed. For various reasons, licit or The energy theft causes an increase in power consumption
not, some of the customers do not pay for all the energy they in a region, due to the trend of stealer consumers to increase the
absorb from the distribution network, causing what is called amount of consumption because they are having a free
non-technical losses or commercial losses [1]. Commercial energy, since detecting and applying legal penalties to them is
losses represent an important aspect for the electric utility usually time consuming for the utilities and authorities.
finance balance, representing an unbilled revenue for them.
This scenery can increase the consumption to an average
Fig. 1 represents one component of the commercial losses, demand value that exceeds the transformer nominal power and
which is the energy theft, represented by the dotted line in the then can cause power quality problems [5]-[8] in the power
picture. The main feature of theft represents diverted energy, or grid, affecting all customers and the regulated utility.
energy that bypasses the energy meter and is then not recorded
Among these problems are the transformer overload, which
by the same. Therefore, a portion of the energy is billed and the
consequently may reduce its lifetime. Moreover, there is a
other one is unbilled.
steady state voltage drop on the buses connected to this
transformer, which is the power quality issue of this paper.

The authors would like to thank ELETROBRAS, ANEEL, INERGE,

CAPES, CNPq and FAPEMIG for financially supporting this research.

978-1-5090-3792-6/16/$31.00 2016 IEEE

557
 Other effects on power quality can be noted as the voltage B. Percentage Variation in Consumed Energy with theft
unbalance and the interruptions growth, which can be treated presence
on a future work.
Considering a region where 100% of consumers pay for the
consumed electricity consumed, then:
II. THEORETICAL REFERENCE

= = (5)
A. Energy Theft Economic Model
Now, considering the same region where some consumers
When there is occurrence of electricity theft, the total pay for their energy and the other party of consumers does not
energy that is supplied from the distribution transformers is pay for it, then:
segregated in a billed energy portion and another portion of
energy that is unbilled. Therefore:
(1 )
= (1 ) + = (6)
= + (1)

The increased percentage in consumption due to electrical

where: energy theft can be modeled using (7):
E - Total electrical energy required by the electric utility.
EF - Energy billed by the electric utility.
%= . 100 = 1 . 100 (7)
EN - Unbilled energy by the electric utility.
- Stolen energy portion or theft percentage.
(1 ) - Billed Energy portion or non-theft percentage. By inserting (5) and (6) in (7) results:

Based on economic theory [9]-[10] and an electricity ( )

economic market model [11] the demand curve for the billed %= 1 . 100 = . 100 (8)
energy can be algebraically expressed as:

Therefore, the energy consumed for the new situation with

= = (1 ) (2)
energy theft is:
where:
- consumer avidity or eagerness to pay for electricity; = (1 + %) (9)
- consumer satiety for electricity;
- electricity tariff; Dividing (9) by the billing period, it results the average
power demand for the new situation with energy theft:

For the case of unbilled energy, the tariff value is zero due = (1 + %) (10)
to stealers non-payment for the electricity to the utility.
Hence, the demand curve for theft can be algebraically
expressed as: C. Voltage Variations in Steady State

= = (3) According to [12], it is established the adequate, precarious

and critical ranges for the steady state voltage levels, as well
as the individual and collective indicators of voltage
compliance, the criteria for measurement, registration and
By inserting (2) and (3) in (1) results: deadlines for compensation to the consumer, if the voltage
measurements exceed the standard regulated limits.
(1 ) Table I and II summarize the limits for steady state voltage
= (1 ) + = (4) variation levels based on [12].

558
 TABLE I CONNECTION POINTS STEADY STATE VOLTAGE LIMITS FOR T Three-phase Transformer;
RATED VOLTAGES EQUAL OR LESS THAN 1 KV (220/127V).
M1 Measurement point 1;
Voltage Service Voltage range variation according to reading
(VS) voltage TL (Volts) M2 Measurement point 2;
Adequate (202 TL 231)/(117 TL 133)
(191 TL < 202 or 231 < TL 233)/ M3 Measurement point 3;
Precarious
(110 TL <117 or 133 < TL 135) M4 Measurement point 4;
Critical (TL < 191 or TL > 233) / (TL < 110 or TL > 135)
M5 Measurement point 5;
M6 Measurement point 6;
TABLE II CONNECTION POINTS STEADY STATE VOLTAGE LIMITS FOR
RATED VOLTAGES OVER 1 KV AND UNDER 69 KV.

Voltage range variation according to reading

Voltage Service
voltage (TL) in relation to the reference voltage
(VS)
(TR)
Adequate 0,93TR TL 1,05TR
Precarious 0,90TR TL < 0,93TR
Critical TL < 0,90TR or TL > 1,05TR

D. Economic impact due to violations on Quality Goals

The problem of electricity theft is a major economic issue

for electric utilities, because it represents a significant revenue
loss. Added to this, the problem of the steady state voltage drop
to values outside of the appropriate voltage limits, results in
economic penalties to the electric utilities.
Depending on the period that the voltage is out of the
appropriate limit and the number of affected consumers, the
amount of compensation may be increased. The distribution
procedures [12] establishes by ANEEL, the Brazilian Electric
Energy Regulatory Agency explains in detail how this
Figure 2 Part of the modeled system by Matlab/Simulink.
compensation or economic penalty is calculated in Brazil.
A technical way to control the voltage within the proper
limits in a distribution system is the use of voltage regulators,
capacitors banks [5]-[8], which requires investments and the
consequent financial cost of acquisition, operation and
maintenance.
Therefore, the electric utility has to decide which situation
is affordable in terms of cost and problems recurrence. The fact
is that the energy theft contributes to electrical grid deviations
and, consequently, the electric utility, the regulatory agency,
the authorities and all customers must take actions to mitigate
this illicit act.

III. SIMULATIONS

The following computer simulations are performed to

study the impact on power quality considering a distribution
network with and without the presence of electricity theft. Fig. Figure 3 Single-line diagram of the simulated distribution network.
2 depicts part of the modeled system in Matlab-Simulink [14]
and Fig. 3 shows a single line diagram of the simulated
network. In Fig. 3 the symbols are:
The loads, lines and equipments features are represented
G Ideal generator; in Tables III, IV and V.

559
 TABLE III GENERATOR PARAMETERS. Later, there will be carried out a study of how this increase
Abbreviation Scc (MVA) Vrated (kV) f (Hz) in consumption will affect power quality on the simulated
G 13.8 60 network.

TABLE IV TRANSFORMER PARAMETERS. 1) Theft Economic Model

Abbreviatio Sr Vp Vs R X Connecti
n (kVA) (kV) (kV) (%) (%) on It is considered a subsystem with a certain percentage of
T 75 13.8 0.22 1 6 Y-Y theft () of the total energy. Table VIII illustrates the
characteristics of the aggregated consumption for this region.
TABLE V LINES PARAMETERS.
Abbrevi Materi R X Length TABLE VIII CONSUMER PARAMETERS OF A REGION.
Lines
ation al (/Km) (H/Km) (m)
Line 1 L1 Al 0,269 934 100 a Eagerness [R$/MWh] 800
Line 2 L2 Al 0,269 934 10 T Average Tariff of Captive Consumers [R$/MWh] 600
Line 3 L3 Al 0,269 934 10
Line 4 L4 Al 0,269 934 500
Line 5 L5 Al 0,269 934 500
For the situation proposed is possible to calculate the
increase in consumption due to electricity theft in three
Source: Based on Catalogue [13].
different theft scenarios, considered by (8). Then, using (10) is
possible to calculate the average power demand in a theft
TABLE VI LOADS PARAMETERS. situation. Table IX illustrates these calculations.
Sn
Loads Abbreviation Vr (kV) fn (Hz) fp
(kVA)
Load 1 C1 15 0.22 60 1 TABLE IX DIFFERENT SETS OF ENERGY THEFT.
Load 2 C2 15 0.22 60 1 Sets E% P1 (kW)
Load 3 C3 15 0.22 60 1 1 10% 30% 78
Load 4 C4 15 0.22 60 1 2 23,33% 70% 102
3 33,33% 100% 120

For the connection cables size, it will be considered 5

meters of length. Examining Tables VIII and IX, consumers in this region
have the willingness to increase their consumption of
A. Situation 1 - Consumption Condition without Energy Theft electricity in 100% in scenario 3, 30% and 70% in the scenery
1 and 2, respectively, regarding to a scenario without energy
theft. Therefore, an additional load of 18 kW, 42 kW and 60
In situation 1, it was measured the RMS voltage value in
kW will be considered to represent the electricity theft
the system buses in a situation that there is no electricity theft,
situation in the three considered scenarios. These additional
which results are presented in Table VII:
loads are simulated in each of the load buses and it will be
studied the effect of this increase in power quality
TABLE VII RMS VOLTAGE MEASURED ON SYSTEM BUSES.
measurement for each measurement point illustrated by Fig. 3.
Voltage at Rated Voltage (V) Stationary Voltage
Points (V)
1 220 215.337 2) Impact on Power Quality
2 220 219.593
3 220 219.590 Regarding to power quality, it will be analyzed the impact
4 220 215.337 of electricity theft in the voltage quality on the buses of Fig. 3.
5 220 219.716
The additional load increase will be added in each of the four
6 13,800 13,829
system loads and it will be analyzed the voltage drop on the
system buses. Table X summarizes the voltage drop in busbar
It is possible to observe from Table VII, that the steady 1 due to electricity theft.
state voltage values for all system buses are within the proper
voltage limits as required by [12]. That is, for this situation TABLE X VOLTAGES MEASUREMENT AT POINT 1.
without energy theft, the system does not indicate power
Load Stationary Voltage with drop
quality problems related to transformer overload and steady point Voltage = 10% = 23,3% = 33,3%
state voltage drop. 1 215.337 209.807 202.537 197.201
2 215.337 214.905 214.327 213.887
B. Situation 2 - Consumption Condition with Energy Theft
3 215.337 214.979 214.497 214.126
4 215.337 214.925 214.324 213.850
First, a microeconomic analysis will be carried out of how
many electricity theft impacts on the increased consumption of
consumers in a given region. When applying the additional load of 60 kW or = 33.3%,
the voltage with a drop in points 2, 3 and 4 are still within the

560
adequate voltage limits. However, in point 1 the voltage drops TABLE XV VOLTAGES MEASUREMENT AT POINT 6.
to a value within the precarious voltage range. In the cases of Load Stationary Voltage with drop
lower percentage of theft = 10% and = 23.3%, the voltage point Voltage = 10% = 23,3% = 33,3%
drop remains within the adequate voltage range. 1 13,829 13,821 13,811 13,804
2 13,829 13,821 13,810 13,801
Table XI summarizes the voltage drop in busbar 2 due to 3 13,829 13,821 13,809 13,801
electricity theft. It is possible to verify that the load increment 4 13,829 13,821 13,811 13,804
in each of the four system load points caused no voltage drop
in the point 2 beyond the adequate voltage limits.
C. Results Analysis
TABLE XI VOLTAGES MEASUREMENT AT POINT 2.
Load Stationary Voltage with drop
point Voltage
1) Economic Analysis regarding Consumption Variation
= 10% = 23,3% = 33,3%
1 219.593 219.096 218.385 217.823
2 219.593 219.152 218.561 218.112 The increase in electricity consumption depends on the
3 219.593 219.225 218.732 218.357 parameters related to the consumer tariff, the percentage of
4 219.593 219.171 218.561 218.076 energy theft and the eagerness of consumers in a given region.
For the three energy theft scenarios, varying the percentage of
theft, it was possible to observe a linearity between the
Table XII summarizes the voltage drop in busbar 3 due to
percentage of theft and the energy consumed. For larger values
electricity theft. As in the Table XI, there is no voltage drop
of theft, there was a greater increase in consumption, which
across the busbar 3 beyond the adequate voltage limits.
can be explained by the individual satisfaction in consuming
TABLE XII VOLTAGES MEASUREMENT AT POINT 3 without paying for energy. Fig. 4 shows this relationship:
Load Stationary Voltage with drop
point Voltage = 10% = 23,3% = 33,3%
1 219.590 219.173 218.560 218.077
2 219.590 219.227 218.733 218.357
3 219.590 219.152 218.562 218.115
4 219.590 219.097 218.383 217.824

Table XIII summarizes the voltage drop at busbar 4 due to

electricity theft. In the case that theft percentage is = 33.3%,
it is observed that when additional load is added to load 4, the
voltage on busbar 4 falls to a value which remains within the
precarious voltage range. For the load increment at other
system points, the voltage drops stay within the adequate
voltage limits in the three theft situations.
TABLE XIII VOLTAGES MEASUREMENT AT POINT 4.
Load Stationary Voltage with drop Figure 4 Electricity Theft Percentage versus Variation on Energy
point Voltage = 10% = 23,3% = 33,3% Consumption.
1 215.337 214.925 214.324 213.849
2 215.337 214.979 214.494 214.125
3 215.337 214.905 214.328 213.885
4 215.337 209.806 202.539 197.206 2) Power Quality Analysis

Measurement of RMS voltage at the load 1 was first

Table XIV summarizes the voltage drop in busbar 5 due to performed with the inclusion of additional load at the load
electricity theft. It is possible to verify that the load increment point 1 at t = 300ms, where it remained until the opening of the
in each of the four system load points caused no voltage drop switch at time t = 500ms. Then at time t = 600ms the switch
in busbar 5 beyond the adequate voltage limits. was closed and it was connected an additional load on the load
TABLE XIV VOLTAGES MEASUREMENT AT POINT 5. point 2, where it remained, until the switch opening at t =
Load Stationary Voltage with drop 800ms. Subsequently, the process was repeated with the
point Voltage = 10% = 23,3% = 33,3% inclusion of additional load on the load points 3 and 4 at t =
1 219.716 219.295 218.683 218.199 900ms and t = 1.2s respectively. Figure 4 represents the
2 219.716 219.348 218.856 218.479 described process:
3 219.716 219.348 218.855 218.477
4 219.716 219.293 218,683 218.197

Table XV summarizes the voltage drop in busbar 6 due to

electricity theft. Despite of the electricity theft for the three
scenarios considered, there is no voltage drop out of the
adequate voltage range at busbar 6.

561
 amount of the consumption increase due to electricity theft of
consumers evaluated in a subsystem based on the consumer
eagerness, tariff, and the percentage of energy theft.
Power quality problems on the simulated network was
observed when the average demand value was greater than the
transformer rated voltage. The voltage drop phenomenon in
steady state due to the increase of the additional load resulted
from the theft, has different values depending on load point and
measurement point.
It was verified that for more distant loads, transformer are
more sensitive to voltage drop. Line impedance affect this
phenomenon. Voltage drop beyond the adequate voltage limit
leads to economic and financial problems for the electric utility
that should reimburse its customers through regulatory
Figure 5 RMS Voltage on Load 1 Theft Situation.
penalties to be paid.
The same additional loading process is executed to load 2, The idea of using a theft detection equipment through the
as shown in Fig. 6: voltage level variation could be interesting and can be better
analyzed in future works.

REFERENCES

[1] ANEEL, Technical Note n 106/2015 - SGT/SRM/ANEEL,

Methodology for Regulatory Treatment for Non-Technical Electrical
Energy Losses (in Portuguese), Brasilia-DF, Brazil, April 22, 2015.
[2] C. A. S. Penin, "Combat, prevention and optimization of electrical
energy commercial losses" (in Portuguese), Ph. D. Thesis, Polytechnic
School from the University of Sao Paulo, Sao Paulo-SP, Brazil, 2008.
[3] T. B. Smith, Electricity Theft: A Comparative Analysis, Energy Police,
Elsevier, 2003.
[4] S. Amin, G. A. Schwartz, A. A. Cardenas, S. S. Sastry, Game-theoretic
models of electricity theft detection in smart utility networks: Providing
new capabilities with advanced metering infrastructure. IEEE Trans.
Control Systems, pages 66-81, 2015.
Figure 6 RMS Voltage on Load 2 Theft Situation. [5] IEEE, Standard 1159 IEEE Recommended Practices for Monitoring
Electric Power Quality, March, 2009.
The voltage drop phenomenon in steady state at load 1 (C1) [6] M. H. J. Bollen, Understanding Power Quality Problems: Voltage Sags
in an energy theft situation is more sensitive than for load 2 and Interruptions, New York, IEEE Press Series on Power Engineering,
2000.
(C2), because the first is being located at a greater distance [7] R. C. Dugan, M. F. McGranaghan, S. Santoso, H. W. Beaty, Electrical
from the transformer. Thus, it is easy to verify that the impact Power Systems Quality, 2nd. Ed., McGraw-Hill, 2003.
of line impedance substantially affect the bus voltage drop at [8] A. B. I. Baggini, Handbook of Power Quality, John Wiley & Sons Ltd.,
the point where the load is connected. In other words, a 2008.
network with the same line and load features, presents different [9] G. A. Jehle, J. P. Reny, Advanced Microeconomic Theory, 2nd ed.,
Addison Wesley Longman, 2000.
voltage drop results through the increment of an additional [10] J. D. Martin, J. W. Petty, Value Based Management, Harvard Business
load, because of the distance between the load and the School Press, Boston, 2000.
transformer, which affects de value of the voltage drop. [11] H. Arango, J. P. G. Abreu, B. D. Bonatto, B. D., C. M. V. Tahan, N.
Kagan, M. R. Gouvea, Modeling the Influence of Power Quality on the
According to Fig. 5 and 6, when a large load is added at Creation of Market Value. in Proc.: 13th International Conference on
once, a voltage sag can be observed. This voltage sag can be the Quality of Power (ICHQP 2008), Wollongong, New South Wales,
critical to sensitive loads. When smaller loads are gradually Australia, Sept. 28 -Oct. 1st, 2008.
added, the voltage sag is too small to have significant effect. [12] ANEEL, Proceedings of Electrical Energy Distribution in the National
Electrical System PRODIST Module 8 Electric Power Quality (in
In case of electricity theft, it is difficult to say if the load Portuguese), Jan., 2016. Available in: http://www.aneel.gov.br.
increment would be at once or gradually. For this reason, the [13] General Cable, Low Voltage Cable Catalogue (in Portuguese), Sao
impact of the voltage sag will not be analyzed in this paper. Paulo-SP, Brazil, May, 2016. Available in:
www.generalcablebrasil.com.
[14] MATLAB version 7.10.0. Natick, Massachusetts: The MathWorks Inc.,
IV. CONCLUSION 2010.

This paper has discussed the impact of electricity theft on

power quality. From the results, it was possible to get the exact

562