Multi-stage voltage control in high photovoltaic based distributed generation penetrated distribution system considering smart inverter reactive power capability

Multi-stage voltage control in high photovoltaic based distributed
generation penetrated distribution system considering smart inverter
reactive power capability
Vinay Kumar Tatikayala a,⇑, Shishir Dixit b
Department of Electrical Engineering, Madhav Institute of Technology and Science, Gwalior, India
a r t i c l e i n f o
Article history:
Received 7 September 2022
Revised 28 February 2023
Accepted 27 March 2023
Available online xxxx
Keywords:
Distribution system
Photovoltaic
Reactive power
Smart inverter voltage control
Power loss
a b s t r a c t
The intermittent nature of photovoltaic (PV) based distributed generation can cause voltage control
issues. This research aims to investigate the impact of using the reactive power capability of PV smart
inverters, which can function as distributed static compensators (DSTATCOMs) during non-feed-in hours,
to address this problem. In other words, the suggested PV-DSTATCOM can be used to provide voltage con-
trol whenever there is a high demand placed on the system around the clock. This study presents a coor-
dinated multi-stage voltage control (CMSVC) strategy that utilizes both PV-DSTATCOMs and traditional
voltage control devices through a hybrid of local and centralized control algorithms. The goal is to min-
imize energy waste while maintaining a voltage that is within acceptable limits. To achieve the best
results, an improved whale optimization algorithm has been proposed for optimal optimization. To test
the proposed method, the IEEE 33 bus radial distribution system and IEEE 69 bus radial distribution sys-
tem were evaluated. According to the findings, the solution offered in this research significantly reduces
energy losses and voltage variations, demonstrating the effectiveness of the proposed method
Ó 2023 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Ain Shams Uni-
versity. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/
by-nc-nd/4.0/).
1. Introduction
The increasing use of photovoltaic (PV) based distributed gener-
ation (DGs) in low voltage (LV) grids has the potential to signifi-
cantly impact the distribution system’s operation [1–3]. To
address these challenges, Volt/VAR control (VVC) utilizing voltage
control devices presents itself as a viable solution [4–5]. Traditional
voltage control devices, including capacitor banks (CBs) and on-
load tap changer (OLTC) transformers, have long been utilized to
achieve conventional VVC. However, their frequent use results in
a shortened lifespan [6]. In contrast, smart inverter interfaced PV
generation is gaining traction due to its adaptable modes of oper-
ation, such as volt/var mode and volt/watt mode [7–9]. As a result,
utilizing smart inverters for VVC presents itself as a promising
alternative. By effectively utilizing these inverters, the LV grid
can be stabilized without causing damage to the voltage control
devices. Additionally, their adaptable modes of operation can help
lessen the negative impact of high diffusion levels of PV-based DGs
on the LV grid.
The article addresses the subject of high DG allocation and volt
VAR control. It offers helpful insights into the issues that are con-
nected with integrating large volumes of distributed generation
into electrical network as well as potential solutions to those chal-
lenges. The authors of [10] utilised the Manta Ray Foraging opti-
mization algorithm (MRFO) to optimise the capacity and
allocation of distributed generation (DG) Type I in order to reduce
power losses in radial distribution networks (RDNs). The study
provided conclusive evidence that MRFO is an effective method
for solving the issue of scattered generators. Chaotic Maps Inte-
grated Stochastic Fractal Search (CMSFS) is a revolutionary
approach that was developed by authors in the article [11] to
address the optimum distributed energy resources placement
problem in radial distribution networks. It was demonstrated that
the technique that was provided was successful in determining the
best possible solutions for the issue. The authors of paper [12]
explored the use of four bio-inspired optimization algorithms to
optimise the placement of three distributed generation (DG) units
in a power system under load uncertainties. These algorithms
included Grey Wolf Optimization (GWO), Manta Ray Foraging
Optimization (MRFO), Satin Bower Bird Optimization (SBO), and
https://doi.org/10.1016/j.asej.2023.102265
2090-4479/Ó 2023 THE AUTHORS. Published by Elsevier BV on behalf of Faculty of Engineering, Ain Shams University.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
⇑ Corresponding author at: Department of Electrical Engineering, Madhav
Institute of Technology and Science, Gwalior, Madhya Pradesh 474005, India.
E-mail addresses: [e-mail] (V. Kumar Tatikayala), shishir.
[e-mail] (S. Dixit).
Ain Shams Engineering Journal xxx (xxxx) xxx
Contents lists available at ScienceDirect
Ain Shams Engineering Journal
j o u r n a l h o m e p a g e : w w w . s c i e n c e d i r e c t . c o m
Please cite this article as: V. Kumar Tatikayala and S. Dixit, Multi-stage voltage control in high photovoltaic based distributed generation penetrated dis-
tribution system considering smart inverter reactive power capability, Ain Shams Engineering Journal, https://doi.org/10.1016/j.asej.2023.102265

Whale Optimization (WOA). The research showed that optimiza-
tion techniques are effective in solving the problem of distributed
generators, as evidenced by the study. In the article [13], the
authors advocated for the utilisation of the Bald Eagle Search
(BES) optimization method for the purpose of allocating shunt
reactive compensators (SRC) and distributed generation (DG) units
with higher usage capacities into distribution systems in order to
reduce power loss. The study provided conclusive evidence that
BES is an effective solution to the challenge posed by distributed
generators. The authors of [14] used the Cuckoo Search algorithm
to enhance the benefits of the Static Var Compensator (SVC) and
synchronous generator (SG) controller at the port of the induction
generator (IG) in order to minimise frequency and terminal voltage
variations. This was accomplished by improving the gains of the SG
controller and the SVC. The results of the investigation revealed
that the method is effective in resolving the DNR issue. The authors
of the article [15] suggested using the Artificial Hummingbird
Algorithm (AHA) to optimise the location and capacities of renew-
able distributed generators sources within a power system. The
strategy that was suggested took into account a number of differ-
ent goals, such as lowering total cost and emissions, reducing volt-
age deviation, and enhancing voltage stability, all while taking into
account the uncertainties connected with the loading and RDG
output power. To solve the problem of DNR, the authors of [16]
suggested applying the stochastic fractal search (SFS) algorithm.
It was demonstrated that the SFS algorithm was successful in
determining the best possible solutions for the problem. In [17]
and [18], researchers created a stochastic model with the purpose
of scheduling distributed generation (DG) systems that are pow-
ered by renewable energy sources and managing energy consump-
tion. A MILP solver was used in order to find an answer to the
optimization problem, which led to a solution that was both effec-
tive and efficient. In order to achieve optimal scheduling of renew-
able energy generation and energy management in modern power
systems, this study demonstrates the significance of advanced
modelling and optimization approaches.
One effective technique for managing voltage fluctuations in
distribution systems is through the use of Volt Var Control (VVC)
strategies. Some of the metaheuristic techniques used for VVC
include honey bee mating optimization based on enhanced chaotic
scheme, genetic algorithm, non-dominated sorting genetic algo-
rithm, enhanced grey wolf optimization, and particle swarm opti-
mization. While these techniques have been effective in
minimizing energy loss and optimizing VVC devices in distribution
systems, they have not considered the smart inverter’s reactive
power capability in relation to voltage management. In [19], a
honey bee mating optimization based on an enhanced chaotic
scheme was presented to estimate the active and reactive power
dispatch of DGs, reactive power compensation from capacitor
banks, and tap positions of OLTC transformers. Daily volt/var con-
trol was performed in [20], and a genetic algorithm (GA) was used
in [21] to ensure optimal VVC device operation in active distribu-
tion systems. Similarly, a non-dominated sorting genetic algorithm
(NSGA-II) was used in [22] to simultaneously optimize peak load
reduction and energy loss minimization in the distribution system.
An enhanced grey wolf optimization (IGWO) was devised in [23] to
study the combined effect of VVC control devices and distribution
network reconfiguration for loss minimization. A particle swarm
evolutionary algorithm was adopted in [24] to ensure that voltage
control devices in distribution systems operate efficiently, and an
ideal coordination of OLTCs and static reactive power compen-
sators was proposed in [25] to minimize overall line losses. Particle
swarm optimization (PSO) was used in [26] to set up VVC devices
in the best possible way while taking dispersed energy into
account to minimize energy loss. However, according to the litera-
ture reported in [19]-[26], the reactive power injection capability
of smart inverters has not been fully explored for voltage manage-
ment, with most studies focusing only on active power injection
from PV-based DGs. Additionally, PV-based DGs are often operated
individually [24]-[29], with little consideration given to multi-
stage coordinated voltage control strategies for loss minimization
and voltage regulation. To address these issues, optimal reactive
power dispatching from PV inverters, capacitor banks, and OLTCs
in a distribution system was accomplished in [27] to minimize loss
and voltage deviation. In [28], the energy savings from coordinat-
ing VVC devices with a solar PV inverter were calculated. In [29],
a voltage control loop was implemented in PV inverters to keep
the voltage within acceptable bounds by absorbing or supplying
reactive power. A smart inverter control strategy was suggested
in [30] for high photovoltaic (PV) penetration in distribution sys-
tems, while [31] demonstrated the significant impact of cascaded
voltage regulators in high PV-penetrated distribution networks.
In [32], benefits of VVC have been validated in real time digital
systems.
In summary, while various VVC techniques have been proposed
to estimate active and reactive power dispatch of DGs, compensate
reactive power from capacitor banks, and adjust tap positions of
OLTC transformers, little attention has been given to the reactive
Nomenclature
Indices
i notation for Bus
h notation for hour
T notation for time
N b number of buses in the network
XCB; XPV set of capacitor banks installed buses and PV-
DSTATCOM installed buses
Parameters
V min
, Vmax voltage magnitude minimum and maximum limits
respectively
DqCB
i fixed step change of capacitor bank
S max
PV;i apparent power of PV smart inverter at i th bus
Q max
PVDST;i maximum PV-DSTATCOM reactive power at i th bus
Variables
Q h
loss;i; P h
loss;i reactive power loss and active power respectively in
a branch connected to the ith bus at hth hour
Q h
grid; P h
grid reactive power and active power taken from grid
respectively at hth hour
Q h
dem;i; P h
dem;i reactive power and active power demand at i th bus
respectively at hth hour
Q h
CB;i reactive power injected by capacitor banks at ith bus
Q h
PVDST;i; P h
PV;i reactive power and active power from PV-
DSTATCOM
taph step change of on load tap changer (OLTC) transformer
at hth hour
steph
i variable Step change of capacitor bank at hth hour
V h
i voltage at ith bus at at hth hour
V. Kumar Tatikayala and S. Dixit Ain Shams Engineering Journal xxx (xxxx) xxx
2

power injection capability of smart inverters for voltage manage-
ment. Moreover, coordinated voltage control strategies for loss
minimization and voltage regulation in PV-based DGs have yet to
be fully explored. The contributions of this research paper can be
summarized as follows:
 Development of a time series model: A time series model of
synchronized VVC scheme has been developed to minimize
energy loss and voltage variations in active distribution net-
works. This model provides a framework for the coordinated
control of both conventional and cutting-edge VVC devices.
 Introduction of coordinated multi-stage voltage control
methodology: A coordinated multi-stage voltage control
(CMSVC) methodology has been suggested, which takes into
account both traditional and advanced VVC devices. The CMSVC
approach provides a hybrid of local and centralized control
algorithms to improve the effectiveness of voltage control.
 Enhanced Grey Wolf Optimization: Grey wolf optimization
(GWO) has been improved and applied to the scheduling of
the mixed-integer nonlinear programming (MINLP) problem,
without relaxation or linearization. This optimization technique
enhances the efficiency and effectiveness of the proposed
approach.
 Proposed PV smart inverter control approach: A PV smart inver-
ter control approach for local reactive power voltage (Q-V) has
been proposed, which enhances the voltage control capability of
PV-based DG systems.
 Investigation of the impact of high PV penetration: The impact
of high penetration of PV-based DG on the voltage profile in
active distribution networks has been explored. This analysis
provides insights into the effect of high PV penetration on volt-
age control and highlights the need for advanced control
techniques.
 Validation of an autonomous volt/VAR droop controller: An
autonomous volt/VAR droop controller has been validated for
use in dynamic voltage control under cloud cover conditions.
This validation provides evidence of the effectiveness of the
proposed approach in real-world scenarios.
 Verification on established distribution systems: The proposed
coordinated approach has been verified on established 33 bus
and 69 bus distribution systems. This verification demonstrates
the practicality and effectiveness of the proposed approach in
real-world scenarios
The paper is organized as follows. In Section 2, we introduce
the proposed architecture for the multistage coordinated voltage
control methodology. This architecture lays the foundation for
the rest of the paper and provides a clear understanding of the
proposed methodology. In Section 3, we provide a mathematical
formulation of the research objective. This is an important sec-
tion as it outlines the objectives of our research and provides
the necessary background for the implementation of the method-
ology. Section 4 is the core of the paper where we detail the
implementation of the proposed multistage coordinated voltage
control methodology. We describe the execution of the central-
ized algorithm using an improved whale optimization technique,
as well as the proposed local control. This section provides a
comprehensive overview of the methodology and the technical
details of its implementation. In Section 5, we summarize the
conclusions and conversations of the paper. This section high-
lights the significance of our research and the implications of
our methodology for future research. Finally, in Section 6, we
provide a closing section that covers the conclusion of the paper.
This section provides a summary of the main findings and high-
lights the contributions of our research.
2. Proposed coordinated multi stage voltage control scheme
The proposed multistage coordinated voltage control architec-
ture is a novel solution that addresses the challenges of voltage
regulation in active distribution networks. As illustrated in Fig. 1,
the architecture incorporates both centralized and local control
techniques for efficient voltage control. At an hourly level, the cen-
tralized control technique assigns the slowly controlled voltage
devices, such as capacitor banks (CBs) and on-load tap changers
(OLTCs), as well as quickly controlled voltage devices like photo-
voltaic (PV)-distribution static synchronous compensators (DSTAT-
COM). However, CBs and OLTCs are not suitable for handling
sudden changes in renewable generation or load dynamics due to
their slow reaction times. Therefore, the local control scheme han-
dles the scheduling of fast-acting devices within each hour, such as
PV-DSTATCOM, which provides assistance against abrupt changes
in voltage and responds more quickly. The proposed multi-stage
coordinated voltage control strategy offers an effective solution
to the challenges of coordinated volt-VAR regulation in active dis-
tribution networks. The suggested approach ensures efficient volt-
age control by incorporating both conventional and cutting-edge
VVC devices. Moreover, it provides a better choice for practitioners
who are seeking a coordinated volt-VAR regulation strategy that
reduces energy losses and eliminates voltage violations. The effi-
cacy of the proposed architecture was verified through extensive
simulations on the established 33 bus and 69 bus distribution sys-
tem. Overall, the study presents an important contribution to the
field of active distribution network research by offering a new
and effective approach to coordinated volt-VAR regulation.
3. Problem formulation
In this work, our primary objective is to minimize active and
reactive energy losses, as defined in Eq. (1)
OF ¼ XT
h¼1
Xh
i¼1 P h
loss;i þ Q h
loss;i
 
ð1Þ
3.1. System operational limits
 Active and reactive power equilibrium limits
P h
grid  XNb
i¼1
P h
loss;i  XNb
i¼1
P h
dem;i þ X
iXPV
P h
PV;i ¼ 0 ð2Þ
Q h
grid  XNb
i¼1
Q h
loss;i  XNb
i¼1
Q h
dem;i þ X
iXCB
Q h
CB;i þ X
iXPV
Q h
PVDST;i ¼ 0 ð3Þ
 Distribution network voltage magnitude limits
V min  V h
i  V max ð4Þ
 Onload tap changing (OLTC) transformer settings limits
ah ¼ 1 þ taph  Dtapstep
100 ð5Þ
here,.taph tap min; ::::; 1; 0; 1; :::::tapmax
 
 Capacitor banks (CBs) limits
Q h
CB;i ¼ step h
i  Dq CB
i ; iXCB ð6Þ
steph
i  0; 1; :::::stepmax
f g
V. Kumar Tatikayala and S. Dixit Ain Shams Engineering Journal xxx (xxxx) xxx
3

 Reactive power limit of PV smart inverter
Q h
PVDST;i ¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
S max
PV;i
 2
 P h
PV;i
 2
r
; iXPV ð7Þ
Q max
PVDST;i  Q h
PVDST;i  Q max
PVDST;i ð8Þ
4. Implementation of proposed multi stage coordinated voltage
control methodology
To ensure optimal performance of the distribution system,
appropriate settings for VVC devices and PV-DSTATCOM are com-
puted hourly using a centralised control technique, as detailed in
section II. However, despite these efforts, voltage breaches can still
occur due to unforeseen changes in load or generation. To address
this issue, a local control scheme was developed that relies on the
Volt/VAr droop characteristics of PV-DSTATCOM and obtains reac-
tive power every five minutes. Fig. 2 provides a visual representa-
tion of the proposed coordinated approach that leverages both
centralised and local control techniques to ensure effective voltage
regulation and loss minimization. By incorporating these strate-
gies, the distribution system can operate at its highest level of effi-
ciency, even in the face of unexpected fluctuations in voltage.
4.1. Centralised control algorithm: Improved whale optimization
algorithm (IWOA)
The article utilizes an improved whale optimization algorithm
(IWOA) as a metaheuristic algorithm for a centralized control algo-
rithm. Metaheuristic algorithms are often employed when tradi-
tional analytic methods are not feasible, or when the problem at
hand is highly complex, nonlinear, or involves a large number of
variables and constraints. Unlike analytic methods that rely on
mathematical equations and assumptions, metaheuristic algo-
rithms are specifically designed to search for global optima, which
may not be attainable through analytic methods. Furthermore,
metaheuristic algorithms tend to be more resilient and adaptable
than analytic methods, as they can effectively handle incomplete
or noisy data and are applicable to a broad range of problems.
In this study, IWOA is utilized to implement a centralized con-
trol algorithm, whereby each humpback whale’s position in the
algorithm functions as a search agent [33]. By continuously updat-
ing these search agents, the whale optimization algorithm can
identify the best solution to the global optimization problem.
i. Encircling prey: Humpback whales have the ability to detect
and circle their prey when they are hunting. Because the
precise position of the best possible design in the search
region is not known, the WOA algorithm operates under
the presumption that the current best candidate solution is
either the target prey or is extremely near to the optimum.
These behaviours are represented by the equations that
follow
D
! ¼ E:
!X pðitÞ  X
!ðiterÞ ð9Þ
X
!ðiter þ 1Þ ¼ X p
ƒ!ðitÞ  R
!: D
! ð10Þ
Where it is the current iteration. The vectors X
! and X p
ƒ!, which
reflect the location of the whale and its prey, respectively. It is fea-
sible to derive the coefficient vectors R
! and E
! using the following
formula
R
! ¼ 2e:r1
!  e ð11Þ
E
! ¼ 2r2
! ð12Þ
Because of the iterations involved in the process of managing
exploitation and exploration, the exploration rate ‘e’ drops from
two to zero throughout the course of the procedure. The factor is
expressed as e = 2–2.iert/itermax.
ii. Spiral updating position and bubble-net assault approach: A
spiral-shaped path and a diminishing circle are both used
by humpback whales to swim around their prey. The likeli-
hood of selecting either the spiral model or the shrinking
encircling mechanism to update the position of whales dur-
ing optimization is taken to be 50% in order to model this
simultaneous behaviour.
X
!ðiter þ 1Þ ¼
X p
ƒ!ðiterÞ  R
!: D
! E:
! X
!pðiterÞ  X
!ðiterÞ ifprob < 0:5
D
ƒ!ehk  cosð2pkÞ þ X
!pðiterÞ ifprob  0:5
8
<
: ð13Þ
Where prob is a random number in [0,1].
Fig. 1. Framework of proposed coordinated multistage voltage control methodology.
V. Kumar Tatikayala and S. Dixit Ain Shams Engineering Journal xxx (xxxx) xxx
4