e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

Contents lists available at ScienceDirect

e-Prime - Advances in Electrical

Engineering, Electronics and Energy
journal homepage: www.elsevier.com/locate/prime

Mitigation of Ethiopian industry sector power quality problems using

ultra-capacitor based dynamic voltage restorer
Beliyou Molla , Emiyamrew Minaye Molla *, Abebe Wolie Yimam , Tefera Mekonnen Azerefegn
Faculty of Electrical and Computer Engineering, Renewable Energy Research, Development, Training and Consultancy Center, Jimma Institute of Technology, Jimma
University, Jimma, Ethiopia

A R T I C L E I N F O A B S T R A C T

Keywords: Nowadays, power quality issues; specifically, voltage sag, swell, harmonics, and interruptions have become
Voltage quality significant challenges for customers in Ethiopia. These problems can lead to equipment damage, production
Dynamic voltage restorer losses, reduced trust, and increased unsold energy for utility providers. To address these issues, the imple­
Adaptive Fuzzy-PI
mentation of Flexible Alternating Current Transmission Systems (FACTS) and Custom Power Devices (CPD) is
Ultra-Capacitor
crucial. In this study, we focus on Desta Garment PLC, an Ethiopian garment industry. Due to faults in a nearby
substation, the factory’s feeder line frequently experiences voltage sag, swell, and interruptions. As a result, the
factory faces partial shutdowns and occasional hours-long interruptions. To mitigate these voltage quality
problems, we propose the use of an Ultra Capacitor-based Dynamic Voltage Restorer (DVR). The DVR has been
carefully selected and implemented with advanced control strategy which is adaptive fuzzy proportional integral
controller. We collected one year of data from the factory and conducted simulations using MATLAB, comparing
scenarios with and without the DVR. The simulation results demonstrate that the proposed mitigation technique
significantly enhances voltage sag and swell conditions. Importantly, the performance aligns with both IEEE
1250–1995 and IEEE 519–1992 standards. This study highlights the effectiveness of the Ultra Capacitor-based
DVR in improving power quality and ensuring stable operations for Desta Garment PLC. By addressing
voltage fluctuations, the factory can maintain productivity and minimize disruptions.

1. Introduction socio-economic development [6,7]. While power quality issues are a

concern globally, they are particularly exacerbated by the increasing
1.1. Motivation of the study reliance on advanced technology and power electronics-based loads that
are sensitive to fluctuations in power. To address this, various mitigation
Power quality refers to the specific attributes of electrical supply as techniques are being employed to alleviate the problem [8–11].
experienced by end-users during normal and stable operating condi­ The objective of this study is to evaluate the power quality challenges
tions. It encompasses various parameters, including: continuity of sup­ faced by Desta Garment PLC in Ethiopia, explore improved mitigation
ply, voltage and current characteristics (magnitude, frequency, techniques, and assess their effectiveness through MATLAB simulation.
waveform shape, and symmetry). Beyond being a technical concern,
power quality has significant economic implications. Poor power quality 1.2. Literature review
can lead to equipment damage, resulting in financial losses. This issue is
particularly pronounced in developing countries like Ethiopia. [1]. Various power quality improvement devices, such as DVR [12–15],
Several studies have demonstrated that inadequate power quality leads Solid-State Transfer Switch (SSTS) [16,17], Unified Power Quality
to significant financial losses in industrial sectors and high power losses Conditioner (UPQC) and others, have been developed to address power
in the power sector [2–5]. quality challenges [9,11]. These devices are categorized as custom
Voltage sag, voltage swell, harmonics, and interruptions are preva­ power devices (CPD) and flexible alternating current transmission sys­
lent power quality issues in the manufacturing industries of Ethiopia, tems (FACTS) based on their application level and rating. FACTS are
significantly impacting sector productivity and the country’s overall utilized in variable renewable energy generation and transmission

* Corresponding author.
E-mail address: [email protected] (E.M. Molla).

https://doi.org/10.1016/j.prime.2024.100612
Received 5 January 2024; Received in revised form 25 March 2024; Accepted 19 May 2024
Available online 22 May 2024
2772-6711/© 2024 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).
B. Molla et al. e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

systems, while CPDs are used at the distribution system and customer 2. System design and modeling
level. The primary goal of CPDs is to utilize power electronic controllers
in power systems to deliver high-quality power and ensure reliability for Fig. 1 shown below represents the block diagram of the proposed
sensitive equipment [18,19]. system and the detail description of the components presented in the
Among these devices, DVR stands out as a cost-effective and efficient following sections.
solution for compensating large voltage variations due to its smaller size The proposed DVR system consists of injection transformer, Voltage
and dynamic response [2,5,11,20–24,24]. Researches has shown the Source Inverter (VSI), bidirectional DC-DC converter and a UCAP energy
impact of DVR, UPQC, and SSTS on improving power quality in end storage device. The designed system can supply and absorb active power
users, distribution and transmission systems [2,5,15,18]. DVR has been from the grid during the two-mode operation of UCAP. The bi-
used for power quality enhancement of intermittent energy sources such directional dc-dc converter serves as a boost converter while the
as wind and solar [21,25–27]. Numerous studies have been also con­ UCAP is discharging to the grid and as a buck converter while UCAP is
ducted on mitigating power quality issues using FACTS or CPDs, charging from the grid system. The performance of the designed UCAP-
focusing on different energy storage devices, control techniques, and DVR is enhanced by utilizing an Adaptive Fuzzy-PI controller to deliver
application areas [2,5,11–13,15,16,19–24,28]. Energy storage devices precise and fast mitigation for power quality problems.
such as batteries, flywheels, super magnetic energy storage, and
ultra-capacitors have been proposed for storing energy to enhance
power quality [11,29]. Ultra-capacitors, known for their fast char­ 2.1. Design of dynamic voltage restorer
ging/discharging capabilities and high energy density, play a crucial
role in fully mitigating frequent power quality problems [11,11,29]. DC A) Design of Ultra-capacitor (UCAP)
link capacitors are typically used for smoothing DC voltage and filtering Choosing the number of UCAPs to be connected depends on the
purposes, while ultra-capacitors serve as energy storage units. Various factors such as terminal voltage of UCAP, the dc-link voltage, and dis­
electrode and electrolyte materials are combined in ultra-capacitors to tribution grid nominal voltage. The capacitance and Equivalent Series
achieve different capacitance, energy density, cycle-life, and cost char­ Resistance (ESR) value of the UCAP depends on the Formula (1).
acteristics [29]. Vmax
This study aims to explore better energy storage solutions and Ns = (1)
Vcell
advanced control strategies for enhancing power quality effectively.
Where, Ns denotes number of cells interconnected in series, Vmax is
1.3. Contribution of the paper and its organization voltage rating of series connected ultra-capacitor cells and Vcell is voltage
rating of ultra-capacitor cell.
To the best of the authors’ knowledge, the predominant power The number of parallel branches Np in the ultra-capacitor bank is be
quality improvement techniques utilizing Dynamic Voltage Restorers given as
(DVR) have traditionally relied on batteries for energy storage. How­ Ns ∗ Ceq
ever, batteries exhibit limitations in terms of charging/discharging Np = (2)
Ccell
speed and energy density compared to ultra-capacitors. Moreover, the
prevalent control strategies often employ conventional single-type pro­ Where, Ceq is equivalent capacitance of ultra-capacitor in farad, and
portional integral (PI) and proportional integral derivative (PID) con­ Ccell is the capacitance of each cell in farad. From the series and parallel
trollers, potentially compromising response time and precision. The cell combinations, the total number of cells NT required in an ultra-
technical ramifications of integrating DVR with advanced control sys­ capacitor is given as (3).
tems remain underexplored. NT = Np ∗ Ns (3)
Hence, this study aims to investigate superior energy storage options
and advanced control methodologies to address power quality chal­ B) Equivalent Series Resistance (ESR) of Ultra-Capacitor Bank
lenges in industrial settings. Assessing the technical impact of custom ESR is a resistance developed due to ultra-capacitor cells which are
power devices on specific distribution feeders can greatly assist utilities interconnected in series and in parallel. It is given by:
in deploying these solutions effectively, benefiting both the utility and R S ∗ Ns
its customers. While existing investigations typically focus on IEEE test RESR = (4)
Np
systems to evaluate custom power device efficacy, this study seeks to
bridge this gap by implementing an adaptive fuzzy-PI controlled ultra- Where, Rs is series resistance of each cell in ohm.
capacitor-based DVR in the real-world scenario of Desta Garment The maximum energy stored (Emax ) in ultra-capacitor is depend on
PLC’s distribution network. the equivalent capacitance Ceq [30].
The performance of the proposed system design is evaluated through
Ceq ∗ Vmax
2
simulations conducted using MATLAB/SIMULINK. The paper is struc­ Emax = P ∗ T = (5)
2
tured into five main sections:

1. Introduction: Providing an overview of the study’s motivation,

literature review, and the paper’s contributions.
2. System Design and Modeling: Detailing the design of the ultra-
capacitor bank, DC-DC converter, parameter optimization,
controller modeling, and payback period analysis.
3. MATLAB Model: Presenting the overall system model developed in
MATLAB.
4. Simulation Results and Discussions: Analyzing various scenarios and
discussing the outcomes.
5. Conclusion: Summarizing the key findings and implications of the
study.

Fig. 1. Block diagram of the proposed system.

2
B. Molla et al. e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

2∗P ∗ T Table 1
Ceq = (6)
Vmax
2 UCAP Simulink parameters.
Parameters Ratings
Where, P (kW) is load requirement and T is duration of voltage sag.
The rated capacitance (C) required to increase voltage value is: C = Ceq Rated voltage 180 V
Capacitance rated value 165 F
∗Ns
Correspondent DC series resistance 7mΩ
Percentage of discharge (%d), minimum sag voltage, and discharging Number of series capacitors 18
current are respectively calculated as follow. Initial voltage 60 V
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ Current preceding to open-circuit 10 A
Req ∗ P Operating temperature 200 c
%dmin = × 10
Vmax
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
Vmin = Req × Id = Req × P
√̅̅̅̅̅̅̅̅
P
Id =
Req

Valuable energy and maximum energy are determined depending on

the suggested duration of discharge.

Ceq(Vmax − Vmin)2
Euse full = (7)
2

Ceq(Vmax)2
Emax = (8)
2
The maximum power withdrawn from the ultra-capacitor bank for a
period of T second is[30]:
Emax
Pmax = (9)
T
C) Bidirectional DC-DC Converter
During voltage sag and swell events, the bidirectional DC-DC con­
verter will operate to mitigate the power quality. The converter act as a
buck during voltage swell and a boost during voltage sag. UCAP terminal
voltage, voltages of the grids side and dc-link voltage are the main Fig. 3. Bi-directional DC-DC converter charging and discharging controller.
determinant parameter to specify the number of UCAPs needed for grid
support as shown in Fig. 2. 2.2. Parameters design
Accordingly, for the optimal performance of the DVR, the DC link
voltage requirements taken to be 260 V. Therefore, three UCAPs are 2.2.1. Power device selections
interconnected in series each with the voltage of 60 V which acts as In the power switch devices, IGBT is extensively utilized in a high-
input to the bidirectional DC-DC converter. The specifications of UCAP frequency switching devices. It has the benefits of MOSFET and GTR.
design and bidirectional DC-DC converter with its controller model is It has also a high switching frequency, high input impedance, less
shown in Table 1 and Fig. 3 respectively. switching loss. It is selected as a switching element in the Buck-Boost bi-
In this study, the non-isolated bi- directional DC-DC converter is used directional converter. The examination within the choice of the main
as a bidirectional power flow. The current and voltage stress were circuit indicates that the highest current through the IGBT is equal to
compared in different non-isolated bi-directional DC-DC converters that through the inductance L. By overlooking the ripple current, the
[31]. The two switches S1 and S2 are complementary, which ensures highest current through the IGBT is given as:
that the current of the inductor L is unceasing within the entire process.
When S2 is switched on the ultra-capacitor charges and it will discharge Po
Is = (10)
as S1 is switched on. V1

2.2.2. Inductor current selection

In the bi-directional DC-DC converter, the inductor is designed ac­
cording to the working mode of the circuit continuous conduction mode
or discontinuous conduction mode and the demand of the ripple current.
When the DC/DC operates in the Buck continuous conduction mode, the
inductance ripple current can be conveyed as [32];
D2 V2 (1 − D2 )
ΔI = (11)
fs L
Where, D2 is the duty cycle and
D2 = V1
V2

D2 V2 (1 − D2 )
L= (12)
fs ΔI
When the DC/DC converter operates in the Boost continuous con­
duction mode, the inductance ripple current is analyzed as [32] :
Fig. 2. Buck-Boost bi-directional converter topology.

3
B. Molla et al. e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

Fig. 4. Controlling model of the proposed DVR.

D1 V1 D1 V1 IF AND Δ, THEN Kp AND Ki

ΔI = ⇒L = (13)
fs L fs ΔI With the defined rule bases, the values of the fixed Kp and Ki are
D the duty cycle in the Boost state: altered continuously according to the value of the error signal, and the
1 is D1 = 1 − V1V2 rate of-error Δ. The determination of the rule base is done utilizing trial-
Agreeing to the engineering practice, half of the current ripple within and-error approaches and is done through experimental. The Simulink
the inductor is chosen as 30 % of the rated current. Based on this, the model of adaptive fuzzy- PI controller is presented in Fig. 5
minimum inductance value will be computed as shown in (14). Fuzzy logic controller design has four states which are fuzzification,
( )
V1 V1 knowledge base, fuzzy inference and defuzzification.
L= 1 − (14) In this study defuzzification of output signal is done by utilizing
0.6 VP0 fs V2
mamdani technique. Fuzzy-PI-based controller is designed by consid­
1

ering inputs as an error, which is the change between measured voltage

2.2.3. Filter capacitor selection
and the reference voltage of DVR and its derivative while ΔKp and ΔKi as
According to the design necessities, the voltage ripple ought to be
output. Where Kp and Ki are proportional gain and integral gain
less than 1 %, in the Buck continuous conduction mode, and the
correspondingly.
capacitor ripple voltage can be given as [33]:
In fuzzy rules design, both the input and output fuzzy sets (the two
D2 V2 (1 − D2 ) D2 V2 (1 − D2 ) inputs, the voltage error and the derivative of voltage error and the two
ΔVc = ⇒C1 = (15)
8f 2 s LC1 8f 2 s LΔVc outputs, ΔKp and Ki) are described as linguistic variables. These are NB
(Negative Big), NM (Negative Medium), NS (Negative Small), Z (zero),
In the Boost continuous conduction mode, the capacitor ripple
PS (Positive Small), PM (Positive medium), and PB (Positive Big). As it is
voltage can be revealed as:
indicated in Table 2 and Table 3 below, there are seven membership
I01 D1 I01 D1 0
D1P functions for voltage error in dq-frame and for the derivative of voltage
V
ΔVc = ⇒C = = 2 (16) error in dq-frame. Therefore, there are 49 rules in the fuzzy decision-
fs C fs ΔVc fs ΔVc
making procedure and membership curves as indicated in Figs. 6, 7
and 8 respectively.
2.2.4. Voltage reference calculation method
The following proofs are considered for developing rules for Kp and
The controlling model of the proposed DVR is depicted in Fig. 4 as
Ki .
shown below. When the supply voltage is at its nominal value, the DVR
is controlled to be in standby mode. The synchronous reference frame
a) For large values of ε and Δε a large ΔKp is needed
method which is based on the instantaneous values of the supply
b) For very large values of ε and Δε, ΔKi is set to zero to evade control
voltage; used to compare and control system voltage with the reference
saturation.
voltage to generate pulses to the PWM inverter.
c) ΔKi is considered to be large for small values of ε in order to decrease
The voltage sag-swell is noticed by evaluating the error between the
steady state error [34].
dq-voltage of the supply and the dq-reference value. The difference
between the reference and the source voltage is applied to the adaptive
2.2.6. Output filter design
fuzzy PI controller to generate pulses activating the gates of VSI to yield
Equivalent circuit of a shunt passive filter per-phase in S domain is
the rated load voltage.
designed for this system and it is applied at a 400 V load side as shown in
Fig. 9.
2.2.5. Adaptive fuzzy PI controller
Applying KCL at point x:
The PI controller response is not promptly enough. It is incapability
to respond the sudden and fast variations in the error signal. To solve Vi − Vx Vx
= Ig + 1 (17)
this problem, an adaptive fuzzy PI control is used in this research. Fuzzy sL1 sC
Controller adjusts the parameters of PI gains based on error and rate of
variation of error and sends them to PI controller. The determination of Vx = IgsL2 (18)
the output control signal is done with a rule base having if-then rules in
the form of

4
B. Molla et al. e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

Fig. 5. VSI adaptive Fuzzy-PI based controller.

Table 2
Fuzzy control rules for ΔKp.
ε/Δε NB NM NS Z PS PM PB

NB PB PB PM PM PS PS Z
NM PB PB PM PM PS Z NS
NS PM PM PM PS Z NS NM
Z PM PS PS Z NS NM NM
PS PS PS Z NS NS NM NM
PM PS Z NS NM NM NM NB
PB Z NS NS NM NM NB NB

Table 3
Fuzzy control rules for ΔKi.
ε/Δε NB NM NS Z PS PM PB

NB NB NB NB NM NM NS Z
NM NB NB NM NM NS Z PS
NS NM NM NS NS Z PS PS
Z NM NS NS Z PS PS PM Fig. 6. Fuzzy Logic Rule Viewer Model.
PS NS NS Z PS PS PM PM
PM NS Z PS PM PM PB PB
PB Z NS PS NM PB PB PB

Ig 1
= (19)
Vi S3 L1L2C + s(L1 + L2)

Substituting L1 + L2 = L, Lp = L1+L2
L1L2
the final equation transfer
function will be:
Ig 1 Fig. 7. Membership function curves of the inputs Δε.
= (20)
Vi sL(1 + s2CLp)

5
B. Molla et al. e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

√̅̅̅
Load Voltage × Load Current × 3
3 − φ kV A = (26)
1000

kV 2
Zb = (27)
kVA
√̅̅̅
VL− L (V) × IL (A) × 3
3 − φ kV A = (28)
Fig. 8. Membership function curves of the output ΔKp and ΔKi. 1000
Short circuit RMS amperes at transformer terminals are calculated by
per-unit method
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
Zt = Xt 2 + Rt 2 (29)

KVA
Isc = √̅̅̅ (30)
3 × VLL × Zt

VA = IA ∗ Zfault (31)

Fig. 9. Shunt passive filter. VL(V)

Zfault = (32)
Isc (A)
1
ωres = √̅̅̅̅̅̅̅̅̅ VL(V) × Zfault
CLp Vsag = (33)
Zt + Zfault
The value of the filter capacitor is selected by considering the reac­
tive power requirement of the capacitor. Based on thump rule, the Vmissing = Vpresag − Vsag (34)
reactive power absorbed by the capacitor is constrained to 5 % of rated
For 25MVA and 132 kV/15 kV 3-φ transformer, it’s reactance and
power [35].
resistance in per unit is calculated according to IEEE 141–1993 standard
V2 0.05 .S which has X/R and % Z value of 7 and 10 respectively.
Q= = 5% of rated aparent power⇒C = (21)
1
2π fC
2πV2f
2.2.7. DVR capacity and specification
where: V = distribution voltage, S = rated apparent power, f = operating For system nominal operating voltage of Vs of 400 V, apparent power
frequency & C = capacitance of filter capacitor of the sensitive loads (insulation room) 92.12 kVA, and having a power
From Eq. (20), substituting ωres = √1̅̅̅̅̅̅ it will be expressed as factor of pf 0.80; DVR (kVA) rating with the energy required in (kJ)
CLp
fulfilling the following specifications. The response period is assumed to
Ig 1 be 5 milliseconds, maximum 3φ voltage sag of 90 % and duration of sag
= ( ) (22)
Vi protection is around 500 milliseconds.
sL 1 + ωres2
s2
It is suggested to implement DVR technology to enhance the bus
voltage sag and restore to 100 % of the rated value. When the sag depth
At resonance frequency, s → jωsω is lower than 90 %, consequently, the enhancing voltage of DVR ought to
Ig 1 be 95 %. Considering the above assumptions, the energy required from
( )
Vi
= the DVR could be calculated as follow.
jωsω L 1 + (jωωres2
sω )2

E = 92.12 kVA ∗ 0.95 ∗ 0.8 ∗ 500 milliseconds = 35 kJ

⃒ ⃒
⃒
⃒
⃒ ⃒
⃒
⃒ For further reliability and accessibility, the DVR rating is chosen to
⃒ Ig(sω) ⃒ ⃒ 1 ⃒ be 100 kVA (50 kJ.
⃒
⃒V ⃒ ⃒
⃒=⃒
ωsω 2 ⃒
⃒ (23)
i(sω) ⃒ωsω L(1− ωres 2 ⃒⃒
⃒
2.2.8. Cost and payback time of DVR
The proposed DVR should be economical to implement accordingly.
Then, the value of the inductance can be written as:
Therefore, the payback period is calculated as shown below. Let Price of
⃒ ⃒
⃒
⃒
⃒
⃒ DVR denoted as CDV R; Price of lone voltage sag denoted as CVS;
⃒
⃒ 1 ⃒
⃒ number of sags per year NVS; payback time T, (year), then
L = ⃒ Ig(sω) ωsω 2 ⃒
⃒ (24)
⃒ωsω
⃒ Vi(sω) (1− ωres 2 ⃒ CDVR = CVS ∗ NVS ∗ T (35)
⃒ ⃒
As per the current market, the price of a DVR including maintenance
To increase the stability of the system a small amount of resistance and running cost is given by:
must be inserted in series with inductor and capacitor filters. The se­ /
lection of the damping resistance must be very small to minimize the $; 300 $; 300
+ 5% = $; 315 kVA (36)
power losses and it depends on the quality factor, calculating the kVA kVA
damping resistance at 40 quality factors: Therefore, the cost of the designed DVR for Desta Garment PLC will
2πfC be CDVR = 100 kVA ∗ $;315
kVA
= $; 31, 500
Rd = (25)
40 The cost of voltage sag, at Desta Plc. is $215/year, and by taking the
lower limit of the number of voltage sag occurrences, NVS is 37/year.
The applicable kVA size of three-phase Buck-boost transformer can
Then, the payback time will be
be estimated as:

6
B. Molla et al. e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

CDVR $; 31.500 system. This kind of sag depends on the nature of the faults and its lo­
T= = (37)
Cvs ∗ Nvs $; 125 ∗ 37/year cations. The fault close to a distribution network causes all the cus­
tomers integrated to this station will face a profound sag. The main
The average lifetime of the DVR is around 15 years; thus, the pro­
reason for this failure of the power system is the weather.
posed solution becomes very cost-effective and viable to be installed.
During the checking time, interruptions have happened throughout
the year that continued for lower and longer than 1 min. The cause of the
3. Mat lab/simulink model of the overall proposed system disruption in the industry emanates from the utility system complica­
tions. Thus, to keep sensitive and power electronic loads from destruc­
The three-phase source is transmitted from Koka and Adami Tulu at tion and loss of production, it is essential to find a resolution to this
132 kV and frequency of 50 Hz and reduced to 15 kV via a three-phase thoughtful problem. The induction motors at the factory, receipts a high
transformer which is available at the Butajira distribution substation. current at starting than the ordinary current and this current leads to a
Then for the end-user, 15 kV is provided which is further converted to drop in voltage which results in sag at the factory side. The main cause of
low voltage at 400 V by distributor transformer at the Company side. swell in the factory is during the energization of a large capacitive bank
Fig. 10 shows the existing features of the proposed system and the installed for power factor correction and switching off large inductive
components of the DVR: series transformers, controllers, VSI converters, loads.
filters, and other complementary features in the block diagram including
system parameters used for the simulation. The parameters are 15 kV
4.2. Effect of voltage sag on the industry
source voltage; 0.181 mH, 1 Ω shunt resistance, 10 mΩ series resistance
for buck-boost converter; IGBT based, 3 arm, 6 pulse, 12 kHZ inverter
Among the power quality difficulties, voltage sag is the most prom­
used. The RLC filter parameters are 1 Ω, 10 mH, 260 µF with resonant
inent of all power quality disturbances. Due to some sensitive and vital
frequency of 1200 Hz. The injection transformer has 100 kVA rating
equipment in the factory, a small voltage dip can disrupt the
with 1/10 and 50 Hz. The load voltage parameters are 400 V, 50 Hz,
manufacturing process due to the inability of these sensitive devices at a
0.1MW, 0.08 M VAR.
reduced voltage level. Thus, the malfunction or loss of these vigorous
pieces of equipment in the manufacturing procedures may lead to a
4. Simulation result and discussions complete shutdown of production; leading to major financial losses until
the problem is cleared. Such critical problems may take minutes, hours
Simulation results of the designed Adaptive Fuzzy-PI controlled or even days to be cleared and causes significant economic problems in
Ultra-Capacitor based DVR using Mat lab/Simulink environment for the factory. Some effects of voltage sag on the case study area are listed
alleviating the effects of the voltage sag, swell, and some voltage quality as follow: The tripping of dc drives, and programmable logic controllers
problems on Desta PLC company has been explained in detail. The re­ (PLCs) system which leads to a partial shutdown of production for hours
sults before and after the integration of the proposed system has been or even days prominent to substantial economic losses.
compared with the IEEE standards.
4.3. Effect of voltage swells on the industry
4.1. Voltage quality assessments on Desta PLC company
If not properly managed, voltage swell can cause a series of
To evaluate the quality of electric power provided to the company, distractive effects as that of voltage sag on industrial equipment and it
monitoring is made at the service entrance points of the company. Per- will interrupt the operation of the industrial production process. Some
year recorded data of the company is taken and the most severe of each equipment (DC drives and some control systems) in the industry, are
voltage quality problem is considered based on its magnitude and time more sensitive to voltage swell in which such problem on this equip­
of occurrence. In this study, the existing major disturbances of voltage ment’s will lead to under quality product, the full shutdown of the
quality in Desta PLC. is investigated in the case when the disturbances production process for hours or even for days till the problem is cleared,
are originated from the source and the load side. it also reduces the life span and efficiency of electronic equipment. This
Thus, the voltage variation under these conditions mainly due to the will cause a significant economic loss to the factory. The major effects on
disturbances from substation side and load or company side. The key the case study area due to a voltage swell are listed as follow: Mal­
reason of the voltage sag in the study area is the faults in the power function or damage to sensitive electronic apparatus (i.e.,

Fig. 10. Mat Lab/Simulink Model of the Overall System.

7
B. Molla et al. e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

programmable logic controllers (PLC’s) and variable speed drives

(VSD)), nuisance operation or tripping of variable speed drives, trans­
former saturation due to the magnification effects of power factor
correction capacitors and computer data interruption.

4.4. Effect of single-phase outage in the company

In the case study area, based on the annually recorded data, unpre­
dictable single- phase outage occurs due to the blow of fuse. This results
in a tripping of protective devices mechanical failure within the equip­
ment, and broken incoming power line due to lightning strike or some
other problem in the substation. During this time, the voltage profile on
the respective phase will approach zero as a result, all single-phase loads
of the industry will be affected. In numerous cases, one or more of these
single-phase load devices may trip if there is a voltage profile that comes
to zero voltage value. At this moment, partly or fully the production
process will be perturbed for hours or even for days until the problem is
cleared. Some causes for single-phase outage in the case study are
lightning strike, problem in substation and unidentified single-phase to
ground fault.

4.5. Mitigation of voltage sag problem

At first, the simulation is done with 50 % of balanced voltage sag

representing a voltage dip due to energization of heavy demands or
initializing of large motors in the system for the time duration of 160
milliseconds and the source voltage is reduced to less than 50 % of its Fig. 12. FFT analysis with and without DVR during symmetrical voltage sag.
rated value. This sag in voltage required to be compensated to get the
anticipated voltage level at the load side to save the sensitive equipment
as shown in Fig. 11.
The suggested DVR responds to this sag and injects the applicable
amount of missing voltage that reaches up to 50 % during the sag
occasion for mitigation. When the DVR is in operation, the voltage sag is
mitigated and the load side voltage is maintained to the standard voltage
level. It is deduced that the voltage waveform that is obtained after the
integration of DVR, is restored from 0.5p.u to 1.0p.u of the rated voltage.
This shows that the connected DVR is functioning efficiently by injecting
a controllable voltage in-phase with the source voltage at sag condition.
Under this condition, the source voltage generates harmonics about
18.72 % for the case without DVR condition, and 2 % after DVR has been Fig. 13. Source voltage during unbalance voltage sag, injected voltage by DVR,
integrated using Adaptive Fuzzy-PI controller as shown in Fig. 12. and load voltage respectively.

4.7. Mitigation of voltage swell problem

4.6. Performance under unbalanced voltage sag

In this case, the simulation is done without DVR when there is a

In the three phases, unbalanced voltage sag is introduced, with 20 %
symmetrical voltage swell in the system for the time duration of 160
voltage sag in phase A, 50 % in phase B, and 90 % in phase C from 0.02 s
milliseconds to create a voltage swell in the source voltage and the
to 0.18 s for 8 cycles of the source voltage. Then it is perceived that DVR
voltage is rise to greater than 50 % of its nominal value. This voltage
acts rapidly and supplies the appropriate quantity of voltage to enhance
swells needs to be mitigated to get the anticipated voltage level at the
this problem. Fig. 13 below shows the waveform of the source voltage,
customer demand side. To mitigate this problem, the simulation is car­
injected voltage, and load voltage.
ried out using a DVR at the load side.

Fig. 11. Source voltage during voltage sag, injected voltage by DVR and load Fig. 14. Source voltage during voltage swell, injected voltage by DVR and load
voltage respectively. voltage respectively.

8
B. Molla et al. e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

Fig. 15. FFT analysis with and without DVR during voltage sag.
Fig. 17. FFT analysis with and without DVR during single phase interruption.
The designed DVR responds to this voltage swell and effectively re­
duces the rising voltage level to acceptable values including its total
harmonics distortion as presented in Figs. 14 and 15.
The voltage swell in the source have generated a harmonic of about
7.25 % at the voltage cycle of 0.16 s. The introduction of DVR using
Adaptive Fuzzy-PI controller reduces this voltage harmonics to 1.85 %.

4.8. DVR mitigation of single phase-outage problem

In case one fuse blows fault occurs, it causes single phase-outage,

which has various effects on single phase-sensitive load on the indus­
trial equipment. Therefore, proper mitigation techniques should be Fig. 18. Source voltage during short-term interruption, injected voltage by
applied. As it is shown in Fig. 16, the voltage in phase A becomes zero DVR, and load voltage respectively.
and a sag happened for the remaining phases for a time duration of 160
ms. This single-phase outage problem and sag is required to be mitigated As shown in Fig. 18, by operating DVR, it is observed that the
to get the desired voltage level at the customer demand side. appropriate amount of voltage is injected to enhance the short-term
As depicted in Fig. 16, by using the DVR the single-phase outage and interruption problem and return the voltage at the load terminal to a
other phase’s voltage sag restored effectively. Fig. 17 presents the har­ nominal acceptable limit. In the mean time the harmonics improved
monic where the DVR reduced substantially from 28.16 % to 3.37 %, from 27.78 % to 3.4 % as presented in Fig. 19
which is acceptable under IEEE 519–1992 harmonic voltage limits
standards.
4.10. Performance under balanced sag/swell with harmonic distortion
4.9. Short term interruption
A 50 % voltage sag in the source voltage is presented in the system
from 20 msec. to 80 msec. for 3 cycles of AC mains voltage and balanced
The source voltage is reduced to less than 0.1pu for a time not
voltage swell of 50 % from 140 msec. to 180 msec. with a distorted
beyond 1 min. The disruptions primarily may occur due to equipment
supply voltage by introducing 5th and 7th harmonics in all three phases
failure in the electrical system, storms, fire, and human error.
as shown in Fig. 20 and the DVR operation reduces the harmonics and it
makes the voltage across the load to be within the nominal value.
Similarly, the total harmonics comparison is presented in Fig. 21 and the
DVR reduced it very well.

5. Conclusions

This study introduces the design of an Adaptive Fuzzy-PI Controlled

Ultra-Capacitor based DVR to address voltage quality issues at Desta Plc.
in Butajira, Ethiopia. The model presented offers Ultra-Capacitors
(UCAP) as a viable DC power source for DVR, capable of providing
Fig. 16. Source voltage during single-phase outage, injected voltage by DVR significant power quickly. It aims to assess the feasibility and stability of
and load voltage respectively. this energy storage system, enhancing power reliability and security in

9
B. Molla et al. e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

Fig. 19. FFT analysis with and without DVR during short term interruption.

Fig. 21. FFT analysis with and without DVR during harmonics injection.

Declaration of competing interest

The authors declare that they have no known competing financial

interests or personal relationships that could have appeared to influence
the work reported in this paper.

Data availability

Data will be made available on request.

Fig. 20. Source voltage during voltage sag and swell with harmonic distortion,
injected voltage by DVR, and load voltage.
Funding

distribution networks. A bidirectional DC-DC converter is designed to The authors do not have funding agency to be acknowledged for this
transfer power from UCAP to the DVR inverter’s dc-link, meeting system study.
requirements. An Adaptive-Fuzzy Logic based control system adjusts PI
gains under various operating conditions to regulate the DVR. Simula­
tion using Mat Lab/Simulink software validates the proposed model. In Availability of data and materials
practice, the DVR is typically installed either at the substation or
customer side; here, it is located on the low voltage side of the step-down The collected data and simulation file can be accessed as required at
transformer at Desta Plc. The system offers rapid responses to voltage any time.
and power fluctuations and it is also suitable for micro-grid or low
voltage distribution grid applications. References

Ethical approval [1] T.H. Meles, Impact of power outages on households in developing countries:
evidence from Ethiopia, Energy Econ. 91 (2020) 104882.
[2] S. Agalar, Y. Alper, Power quality improvement using STS and DVR in wind energy
This study is conducted in accordance with research ethics and
system, Renew. Energy 118 (2018) 1031–1040.
respecting both human and/ or animal issues. [3] N. Abas, S. Dilshad, G.S. Member, Power quality improvement using dynamic
voltage restorer, IEEE access 8 (2020) 164325–164339.
CRediT authorship contribution statement [4] R. Pal, S. Gupta, T. Dvr, Topologies and control strategies implicated in dynamic
voltage restorer (DVR) for power quality improvement,” Iran, J. Sci. Technol.
Trans. Electr. Eng. 3 (2019) 581–603.
Beliyou Molla: Conceptualization, Data curation, Investigation, [5] A.L.I. Moghassemi, S. Padmanaban, S. Member, A novel solar photovoltaic fed
Methodology, Writing – original draft. Emiyamrew Minaye Molla: TransZSI-DVR for power quality improvement of grid-connected PV systems, IEEE
access 9 (2021) 7263–7279.
Writing – review & editing, Validation, Supervision, Conceptualization, [6] L.T. Abdisa, Power outages, economic cost, and firm performance: evidence from
Methodology. Abebe Wolie Yimam: Writing – review & editing. Tefera Ethiopia, Util. Policy 53 (June) (2018) 111–120.
Mekonnen Azerefegn: Writing – review & editing. [7] F. Carlsson, E. Demeke, P. Martinsson, T. Tesemma, Cost of power outages for
manufacturing firms in Ethiopia: a stated preference study, Energy Econ. 88 (2020)
104753.

10
B. Molla et al. e-Prime - Advances in Electrical Engineering, Electronics and Energy 8 (2024) 100612

[8] Y. Pal, A. Swarup, B. Singh, A review of compensating type custom power devices [30] S. Hazra, S. Bhattacharya, Hybrid energy storage system comprising of battery and
for power quality improvement, in: 2008 Jt. Int. Conf. Power Syst. Technol. ultra-capacitor for smoothing of oscillating wave energy, in: 2016 IEEE Energy
POWERCON IEEE Power India Conf. POWERCON 2008, 2008. Convers. Congr. Expo., 2016, pp. 1–8.
[9] M. Farhoodnea, A. Mohamed, H. Shareef, A comparative study on the performance [31] X. Pan, H. Li, Y. Liu, T. Zhao, An overview and comprehensive comparative
of custom power devices for power quality improvement, in: 2014 IEEE Innov. evaluation of current-fed isolated bidirectional, IEEE Trans. Power Electron. 35
Smart Grid Technol. - Asia, ISGT ASIA 2014, 2014, pp. 153–157. (2020) 2737–2763.
[10] V.A. Boicea, Energy storage technologies: the past and the present, Proc. IEEE 102 [32] M.R. Banaei, S.G. Sani, Analysis and implementation of a new SEPIC- based single
(11) (2014) 1777–1794. switch buck-boost dc-dc converter with continuous input current, IEEE Trans.
[11] M. Farhadi-Kangarlu, E. Babaei, F. Blaabjerg, A comprehensive review of dynamic Power Electron. 33 (2018) 10317–10325.
voltage restorers, Int. J. Electr. Power Energy Syst. 92 (2017) 136–155. [33] H. Yang, Y. Dong, W. Li, X. He, Average-value model of modular multilevel
[12] A. Benali, Power quality improvement and low voltage ride through capability in converters considering capacitor voltage ripple, IEEE Trans. Power Deliv. 32
hybrid wind-PV farms grid-connected using dynamic voltage restorer, IEEE access (2016) 723–732, https://doi.org/10.1109/TPWRD.2016.2555983.
6 (2018) 68634–68648. [34] A. Luo, Fuzzy PI based direct output voltage control strategy for the STATCOM
[13] K. Chandrasekaran, V.K. Ramachandaramurthy, Electrical power and energy used in utility distribution system, IEEE Trans. Ind. Electron. 5 (2009) 2401–2411.
systems an improved dynamic voltage restorer for power quality improvement, Int. [35] D. Zhou, H. Wang, F. Blaabjerg, Reactive power impacts on LCL filter capacitor
J. Electr. POWER ENERGY Syst. 82 (2016) 354–362. lifetime and reliability in DFIG grid-connected inverter, IEEE Energy Convers.
[14] A.I. Omar, S.H.E. Abdel, E.E.A. El-zahab, M. Algablawy, Z.M. Ali, An improved Congr. Expo. (2018) 4094–4101.
approach for robust control of dynamic voltage restorer and power quality
enhancement using grasshopper optimization algorithm, ISA Trans. 95 (2019)
110–129, https://doi.org/10.1016/j.isatra.2019.05.001.
[15] M. Manikandan, S. Member, Power quality enhancement using dynamic voltage Emiyamrew Minaye Molla was born in 1986 in Gondar,
restorer (DVR) -based predictive space vector transformation (PSVT) with Ethiopia. He received his BSc., MSc. and PhD degrees in Elec­
proportional resonant (PR) -controller, IEEE Access 9 (2021) 155380–155392. trical Engineering in 2010, 2015 and 2020 from Arba Minch
University and Jimma University Ethiopia and National
[16] G. Sebastian, et al., A novel PSO based fuzzy controller for robust operation of
solid-state transfer switch and fast load transfer in power systems, IEEE Access 10 Taiwan University of Science and Technology, Taiwan
respectively. He has been with Jimma University Institute of
(2022) 37369–37381.
[17] M. Faisal, M.A. Hannan, P.J. Ker, Review of solid state transfer switch on, Technology starting from 2010. Currently he is leading
Elecronics 9 (2020) 1–21. renewable energy, research, development and training center
[18] E. Hossain, S. Member, M.R. Tür, Analysis and mitigation of power quality issues in under Jimma University Institute of Technology. His-research
distributed generation systems using custom power devices, IEEE access 6 (2018) interests include power quality analysis and mitigation, micro
grid renewable energy sources and energy storage systems,
1–18.
[19] M.B. Latran, A. Teke, Y. Yoldas, Mitigation of power quality problems using load forecasting. energy management and Energy modelling.
distribution static synchronous compensator : a comprehensive review, IET power
Electron. 8 (2015) 1312–1328.
[20] F.H. Gandoman, et al., Review of FACTS technologies and applications for power
quality in smart grids with renewable energy systems, Renew. Sustain. Energy Rev. Abebe Wolie Yimam received the B.S Degree in Electrical/
82 (September 2017) (2018) 502–514. Electronic Technology from Adama University in 2007, and his
[21] E.M. Molla, C.C. Kuo, Voltage quality enhancement of grid-integrated pv system M.S Degree in Electrical Power Engineering from Arba Minch
using battery-based dynamic voltage restorer, Energies 13 (21) (2020).
University in 2013. Now, he is a Ph.D. Student/Researcher in
[22] Y. Naderi, S.H. Hosseini, S. Ghassem Zadeh, B. Mohammadi-Ivatloo, J.C. Vasquez, MOBI Research Center (Department of Electrical Engineering
J.M. Guerrero, An overview of power quality enhancement techniques applied to
and Energy Technology), Vrije Universiteit Brussel, at Brussels,
distributed generation in electrical distribution networks, Renew. Sustain. Energy Belgium. His-current research interest includes Renewable
Rev. 93 (May) (2018) 201–214. Energy Technologies, Wind Power Generations, Power System
[23] E.M. Molla, C.C. Kuo, Voltage sag enhancement of grid connected hybrid PV-wind
Optimization, AI Application for Power System, Powers System
power system using battery and SMES based dynamic voltage restorer, IEEE access Optimization, Distribution Generation and Power Electronics.
8 (2020).
[24] M. Bajaj, Design and simulation of hybrid DG system fed single-phase dynamic
voltage restorer for smart grid application design and simulation of hybrid DG
system fed single-phase dynamic voltage restorer for smart grid application, Smart
Sci. 8 (2020) 1–15.
[25] A. Darvish Falehi, M. Rafiee, LVRT/HVRT capability enhancement of DFIG wind
turbine using optimal design and control of novel PIλDμ-AMLI based DVR, Sustain. Tefera Mekonnen Azerefegn received his first degree in
Energy, Grids Netw. 16 (2018) 111–125. Electrical /Electronic Technology from Adama Science and
[26] A. Darvish Falehi, H. Torkaman, Robust fractional-order super-twisting sliding technology University, Ethiopia in 2009, MSc in power system
mode control to accurately regulate lithium-battery/super-capacitor hybrid energy Engineering from Bahirdar University in 2014, and PhD in
storage system, Int. J. Energy Res. 45 (13) (2021) 18590–18612. Power system engineering from TH Koln University, Germany,
[27] A. Darvish Falehi, H. Torkaman, Promoted supercapacitor control scheme based on in 2022. His-work experience comprises teaching, research and
robust fractional-order super-twisting sliding mode control for dynamic voltage Industrial consultancy and has been working in Jimma Uni­
restorer to enhance FRT and PQ capabilities of DFIG-based wind turbine, J. Energy versity since 2009 onwards. He published 20+ papers in
Storage 42 (2021). prestigious international journals. His-research interests
[28] A.I. Omar, S.H.E. Abdel Aleem, E.E.A. El-Zahab, M. Algablawy, Z.M. Ali, An include Modeling and Simulation of Sustainable and Renew­
improved approach for robust control of dynamic voltage restorer and power able Energy Systems, energy economics and policy, power
quality enhancement using grasshopper optimization algorithm, ISA Trans. 95 systems dynamics, power electronics, power quality, and
(2019) 110–129. power system protection.
[29] M. Farhadi, O. Mohammed, Energy storage systems for high power applications, in:
IEEE Ind. Appl. Soc. - 51st Annu. Meet. IAS 2015, Conf. Rec 89, 2015.

11