www.nature.

com/scientificreports

OPEN Comparative analysis of FACT

devices for optimal improvement
of power quality in unbalanced
distribution systems
Ahmed M. Elkholy 1,2*, Dmitry I. Panfilov 2 & Ahmed E. ELGebaly 1,2
This study examines the performance of asymmetric three-phase distribution systems under the
influence of FACT deives such as a static VAR compensator (SVC) and a unified power controller
(UPC). Each suggested device’s operating principle is developed in this paper in order to provide the
best model to be used in the power flow analysis. The performance of the IEEE-13 bus imbalanced
distribution model is investigated using the Newton-Raphson method. To improve network
asymmetry, an optimization analysis is carried out to ascertain how each proposed device will work.
In this paper, the mode of operation of each device is determined every hour to flow the changing
in loads according to their daily load curve. The genetic algorithm (GA) is utilized to determine the
contribution of the unified flow controller to balance the loads between phases. The multi-objective
function is constructed to combine the total system losses and the average voltage unbalance
coefficient. The paper provides an optimal dynamic technique to optimally operate the unified power
flow controller installed in the unbalanced three-phase distribution system.

Keywords Unsymmetrical distribution system, Unified Power Controller, Daily load curve, System level
control, GA optimization, Single-phase GA optimization

Unsymmetrical loading situations are one of the factors contributing to the decline in the quality of the electrical
supply. Unwanted interruptions in the operation of distribution systems caused by unsymmetrical loads or short
circuits are known as unsymmetrical situations. As a result of overheating, decreased efficiency, and shortened
lifespan of power system equipment and customer devices, voltage unbalance has a detrimental impact on distri-
bution network managers and users. Every source of unbalance individually contributes to the overall unbalance
in the distribution network that causes extra system l­ osses1,2.
Arc steel smelting furnaces, traction substations of alternating current, electric welders, single-phase thermal
electric facilities, or any strong single-phase, two-phase, or three-phase unsymmetrical users of electric power,
including domestic, are the main sources of voltage symmetry. Further detrimental effects of the unsymmetrical
operation include the lifespan of electrical equipment ,and substantial losses in distribution f­ eeders3–5.
Three-phase voltages at all buses must be fully observable for an extended period of time in order to determine
the network’s state of unbalance with accuracy. Due to associated expenditures, only a very small number of
buses in the network have monitors installed, so evaluating the network’s unbalance is made significantly more
difficult by the lack of monitoring data. In practice, monitoring device reports or consumer input are often what
alert distribution network operators to serious power-quality i­ ssues6. Unbalance may be undetected and persist
in the system because it is a “long-term” event, which is a condition that normally does not prompt the rapid
disconnecting of equipment from the supply or activation of a protection d ­ evice7. In this paper, the power flow
analysis of a standard unsymmetrical distribution system has been developed using Newton-Raphson method.
The loads included in this study are time-dependent to estimate the distribution system state periodical every
certain proper time.
The usage of controllable devices should be employed to reduce the negative effects of the asymmetric loads
on the mode of operation of the transmission line in order to guarantee the quality of electrical energy meets
the IEEE 1159.3-2019 s­ tandard8. These instrument are constructed using power e­ lectronics9,10. They respond

1
Electrical Power and Machines Department, Faculty of Engineering-Tanta University, Tanta, Egypt. 2Industrial
Electronics Department, Moscow Power Engineering Institute, 111250 Moscow, Russia. *email: ahmed_
[email protected]

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 1

Vol.:(0123456789)
 www.nature.com/scientificreports/

quickly with excellent reliability. The power transmission line operation can be adaptively controlled by these
devices using a variety of technologies.
The employment of static reactive power compensator (SVC) is one of the most straightforward and often
utilized technologies adopted to affect the operating modes of a power transmission l­ine9. They operate as a
regulated generator of reactive power that is shunt connected to the load and compensates for reactive power
during load ­periods11,12. The implementation of the instrument’ circuits is rather straightforward. However,
without their assistance, it would be difficult to symmetrically run the transmission line in terms of active powers
by controlling the reactive power flow between the p ­ hases13,14.
3–5
The authors of Refs. offer a method for balancing the three-phase transmission lines that feed asymmetrical
loads using specific instalment. To guarantee equal transmitted power per phase, this device (the Unified Power
Controller, or UPC), automatically balances the transmission line’s phase currents. The continual determina-
tion and adjustment of both the negative and zero sequence load current components are essential to the UPC’s
basic working theory. This device’s capacity to generate only pure sinusoidal currents gives it a crucial benefit.
In order to deploy such a device inside the distribution system, a suitable power flow calculation is needed to
ascertain their mode of operation.
Recent advancements in power flow optimization in electrical grids have led to the development of multi-
objective optimization algorithms. These algorithms address complex trade-offs, as demonstrated by Refs.15,16.
­Reference15 proposed a two-stage solution for optimal power flow in AC/DC grids, using a decision analysis
technique, a hybrid coding scheme, and fuzzy c-means clustering for optimization. Conversely, Ref.16 introduced
a two-step approach for the MORPD problem, using a multi-objective optimization algorithm to find Pareto
optimal solutions and integrated decision-making to extract the best compromise solutions.
Despite extensive research, significant gaps persist in the analysis of unbalanced systems optimization. The
literature lacks comprehensive exploration of optimization at both the phase and system levels. Moreover, the
impact of SVC and UPC on unbalanced distribution systems with standard load variations remains unaddressed.
The main contributions of this paper can be summarized as the following:

• Developing a suitable power-flow model of the suggested power-electronics-based devices to adjust the
three-phase unbalance.
• Developing a way to determine the contribution of UPC in unbalanced distribution network based on the
system daily loading.
• Assessing the performance of the unbalanced distribution system under the effect of SVC or UPC under
daily loading.
• Developing optimization technique using GA to control the UPC mode of operation to achieve multi-objec-
tive function which combines the minimization of both the energy loss and unbalancing voltage coefficient.
In this paper, the optimized operation of UPC has an abbreviation of UPC-GA.

Modeling of UPC and SVC for power flow analysis

This section introduces the operating concepts FACT devices used on this study. These guidelines can be used
to create the appropriate instrument models for the power flow analysis.

UPC modeling
With the use of this device, the three phases of the transmission line’s reactive and active phase powers can be
balanced. The device transfers the energy from one phase to another by controlling the voltage provided to a
reactive element. A multi-tap transformer is used to manage the voltage, and it is linked by thyristors. This device
produces currents that are almost pure sinusoidal, which is thought to be a benefit of the t­ echnology3–5,17–20.
The general configuration of the UPC implemented in a three-phase, four-wire transmission power line that
supplies a load with three-phase impedance ( Za , Zb , Zc ) is shown in Fig. 1. In general, the three phases’ currents

Figure 1.  Connecting of the Unified Power Controller to a four-wire, three-phase power line.

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 2

Vol:.(1234567890)
www.nature.com/scientificreports/

Inb and Inc , and are out of balance, therefore the UPC supplies three currents Ipa , Ipb and Ipc and to balance the
three phases’ supply currents Ia , Ib and Ic . The current system equations are as shown in Eq. (1).
Ia = Ina − Ipa
Ib = Inb − Ipb
(1)
Ic = Inc − Ipc
In = Inn + Ipn

The UPC’s function is to compensate for the load currents’ negative and zero components. The positive com-
ponent may be changed in some circumstances to account for reactive power. Therefore, it is crucial to measure
load currents in real time in order to split them into symmetrical components of positive, negative, and zero
sequences. Next, a method for calculating and producing UPC positive, negative, and zero currents is developed.
Thus, depending on the three- or four-wire topology of the three-phase transmission line and the balancing goal,
the UPC’s number of independent control variables might range from 2 to 6­ 4. The goal of altering the power
flow in each phase is to achieve particular objectives by figuring out the right quantity of power flow in each
phase as well as balancing the power in each phase completely. The reduction of distribution system losses or
the unbalancing coefficient may be the goal. The way a UPC operates in a power flow study depends on it affects
how power is distributed throughout the three phases. If the load has three active power values ( Pna , Pnb , and
Pnc ), which may not be equal, and three reactive power values (Qna, Qnb, and Qnc ), which might not be the same :

In1 = Ina + a2 Inb + aInc

In2 = Ina + aInb + a2 Inc (2)
In0 = Ina + Inb + Inc

Pna + Pnb + Pnc = Pt

(3)
Qna + Qnb + Qnc = Qt
where, Pt and Qt are the total active and reactive power of the load respectively. The UPC makes the active Peq
∗

and reactive Qeq powers in the three phases equal in the assumption of typical balancing operation:
∗

∗ ∗ ∗ Pt
Peq(a) = Peq(b) = Peq(c) =
3
(4)
∗ ∗ ∗ Qt
Qeq(a) = Qeq(b) = Qeq(c) =
3
∗ and reactive Q∗ powers may need to be adjusted to have optimal values in accordance
The supply active Popt opt
with a specific objective function for the UPC to operate optimally (load reactive power is fully compensated):
∗ ∗ ∗
Popt(a) + Popt(b) + Popt(c) = Pt
∗ ∗ ∗ (5)
Qopt(a) = Qopt(b) = Qopt(c) =0

If the device is managed according to the overall system requirements, the mode of operation slightly is dif-
ferent. The UPC is optimally controlled by Genetic Algorithm in the system level. In the mode UPC-GA, the
active power of each phase is determined optimally. Moreover, the reactive power is generated by UPC according
to system reactive power requirements Qopt(sys) as Eq. (6).
∗ ∗ ∗
Popt(a) + Popt(b) + Popt(c) = Pt
∗ ∗ ∗ (6)
Qopt(a) + Qopt(b) + Qopt(c) = Qopt(sys)

SVC modelling
The SVC in the system underwent an analysis research, and the results indicated that its model might differ. The
SVC models used in power flow studies should be consistent with the SVC’s steady-state, fundamental-frequency,
and balanced performance. The imbalanced operation of the SVC to adjust reactive power or manage voltage
balance forces the SVC model in this study to deal with each phase s­ eparately21. In the power flow research, the
typical SVC model is depicted in Fig. 2 within its control range. Reactance XSL is used to describe the SVC char-
acteristics’ slope. Reactance is connected to the high-voltage side bus (PQ) and the auxiliary bus (PV)22. When
the auxiliary bus’s voltage is set to equal Vref and its power is P = 0, it is referred to as a PV bus. The high-voltage
side bus, on the other hand, is referred to as a PQ bus, where Q may equal any value within the control range of
its rated reactive power and P = 023.
In this study, SVC is paired with an unbalanced load under the condition that it operates to fully feed the
reactive power while having no effect on the load’s active power. In order to achieve the supply reactive power
equalling zero, the SVC creates reactive power QS VC equal to the load reactive power in each phase as follows:

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 3

Vol.:(0123456789)
 www.nature.com/scientificreports/

Figure 2.  SVC model for conventional power flow study.

Figure 3.  Single line diagram of the IEEE 13 bus test system.

QSVC(a) = Qna
QSVC(b) = Qnb (7)
QSVC(c) = Qnc

System description
The used system for this study is the standard IEEE 13 buses. This system is described in Refs.24,25, the single line
diagram is shown in Fig. 3. In this system, there are 13 buses with 8 loads. According to load flow s­ tudy24, the
devices can be installed in two buses #634 #675, because those buses are heavily loaded with three-phase loads.

Load profiles
The load profiles utilized to evaluate the performance of the proposed devices (SVC, UPC, and UPC-GA) within
the system over a 24-hour period are derived from the low voltage European s­ ystem26. These load profiles,
depicted in Table 1 and illustrated in Fig. 4, are crucial for assessing the system’s response to the devices. Spe-
cifically, the devices are installed on buses #634 and #675, where the critical load profiles are identified as nos.
#1, #2, #3, #5, #6, and #7 and shown in the Fig. 5. The goal is to ensure that SVC and UPC adhere to these load
profiles to enhance the system’s balance. A thorough evaluation of the overall system performance is essential to
determine the optimal operating modes for these devices.
The load profiles shown in Fig. 4 identified the loading every hour, the loading in MW and MW are calculated
by multiplying the load profile shown in Fig. 4 by the standard loading condition for IEEE 13-bus system. The
new loading conditions has the same ratio between the active and reactive power (the angle of the total apparent
power is the same over the whole day).
All load profiles adopted in this study are residential and commercial load profiles, that means there are wide
variations in the load profiles. The summation of all load profiles is shown in Fig. 6; that means all loads at every
hour are summed to get one point indicated on Fig. 6

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 4

Vol:.(1234567890)
www.nature.com/scientificreports/

Bus Phase A Phase B Phase C

634 1 2 3
645 4 – –
652 14 – –
675 5 6 7
611 – – 8
646 – 9 –
671 10 11 12
692 – – 13

Table 1.  The load profile for each load phase in the system.

Figure 4.  The load profile adopted for the system.

Bus 634

Phase A
2
Phase B
Phase C
loading %

1.5

0.5

0
0 5 10 15 20 25

Bus 675
2 Phase A
Phase B
1.5 Phase C
loading %

0.5

0
0 5 10 15 20 25
Hours

Figure 5.  Load profile for the three phases at buses #634 and #675.

In the buses where the installed devices (#634 and #675), the system is light loaded in the beginning then the
system is sharply loaded to more than 2.5 MW while its first peak is happened at 10 a.m. After that the system
is sharply unloaded to reach less than 1 MW . Then, the other buses which are far from the balancing devices
are heavily loaded. As a conclusions, the system with these load profiles is able to test all possibilities that face
the devices.

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 5

Vol.:(0123456789)
 www.nature.com/scientificreports/

2.5

Active Power Loading MW

2

1.5

0.5
0 5 10 15 20 25
Hours

Figure 6.  All adopted load profile for all loads on the system.

0.5
Active Power
0.45 Reactive Power
Power Rating First Device (MW, MV A)

0.4

0.35

0.3

0.25

0.2

0.15

0.1

5 · 10−2

0
0 5 10 15 20 25
Hours

Figure 7.  Active and reactive power passing throw the first device.

UPC device rating

The rating determination of the UPC device is important for identifying the active and reactive power controlled
or supplied by this device. This device is connected in shunt with the load, so the load rating and its distribution
between phases identify the device r­ ating27. Figure 7 shows the active and reactive power for the load which is
close to the first device. This load should to balanced by the device in case of unbalanced loading. As well, Fig. 8
shows the active and reactive power for the second device. In this study, the rating of the installed device is sug-
gested to be 80% of the connected load (maximum apparent power for the connected bus)3. This percentage is
suggested based on the assumption that the installed UPC can balance the overall power fed through only one
phase to be fed equally by the three phases. Moreover, the UPC can compensate the reactive power required by
the load and the remaining system reactive power requirements. Therefore, according to Fig. 7, the rating of the
first device equals 0.452 MW. while, according Fig. 8, the second device rating equals 1.07 MW.
UPC-GA acts as a reactive power compensator device for the overall system. So, this device can supply the
reactive power for the other bus in the system according to its rating. The rating is not only identified by the con-
nected load, however for economic aspect, the management process is tuned to not overtake the selected rating
of the device. To achieve the capacitive and inductive control, the load is used as reactive element for the system.

Management and optimization methodologies

In this section the management algorithm for driving the UPC and the optimization problem that identifies the
best UPC output are developed.

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 6

Vol:.(1234567890)
www.nature.com/scientificreports/

1.4
Active Power
Reactive Power
1.2

Power Rating Second Device (MW, MV A)

1

0.8

0.6

0.4

0.2

0
0 5 10 15 20 25
Hours

Figure 8.  Active and reactive power passing throw the second device.

Management operation of UPC

The Unified Power Controller (UPC)4,5 is designed to regulate the power flow of the connected load, with the
understanding that modifications to the UPC influence other buses within the power system. The UPC’s manage-
ment operation aims to enhance system performance by minimizing power losses and improving voltage quality
across the entire grid, thereby increasing the device’s efficiency and flexibility.
In large-scale systems, a management algorithm is essential to mitigate computational complexity. This algo-
rithm employs forecasting, big data analysis, and measurement tools to estimate load based on historical data
rather than real-time calculations. Two optimization processes are employed: a full optimization process, which
is used to derive the optimal solution when sufficient processing time is available, and a lightweight optimization
process, which is applied when real-time adjustments are necessary. The full optimization process’s results are
stored in a database for future reference.
The management framework of the UPC is depicted in Fig. 9, which outlines the sequence of operations start-
ing with the collection of system data. Data collection methods vary, with some utilizing micro-measurement
units (µPMU )28 and others employing Internet of Things (IoT) ­technologies29. Despite these differences, the
timing of data processing is paramount for power system operators, as it directly impacts the optimization pro-
cess. In this study, the optimization is performed a day in advance of the operating day, ensuring that the data
is processed and ready for use.
The data collection process, which spans from the first to the third step, involves SCADA and optimization.
These stages collectively take up to a day to complete. The second step involves data handling, where minute-
based data is transformed into hourly-based data, which is more conducive to the optimization process due to
the time required to identify the optimal solution based on the objective function and constraints. The final
solution is then stored in the database.
The second stage, or the management layer, retrieves the stored solution from the database and initiates a
comparison operation. This operation compares the current load (Pn) with the forecasted load (Pn−1). Depending
on the comparison results, the management system has three potential outcomes: if the current load is less than
1.2 ∗ Pn−1, the stored solution is applied; if the load is between 1.2 ∗ Pn−1 and 1.5 ∗ Pn−1, a lightweight optimiza-
tion process is triggered, and a flag is sent to the database; if the load exceeds 1.5 ∗ Pn−1, default control is applied.
This management framework integrates distributed and centralized control strategies, as well as multiple timing
regions, to maximize the device’s efficiency.
This approach to UPC management not only enhances the overall grid but also reduces real-time computa-
tional demands, thereby improving the system’s performance and reliability.

The optimization process

The genetic algorithm is a search heuristic that is based on Charles Darwin’s theory of natural selection. This
algorithm is based on the natural selection process, which selects the fittest individuals for reproduction in order
to produce offspring for the next generation. The individuals here, in this study, are the input of the objective
­function30. Unbalanced distribution system load flow (UDSLF) is used to create all individuals which are the
active and reactive power compensated by UPC-GA as Eq. (6), then the objective function is used to judge all
population. Mutation and crossover is used to ensure the evaluation and randomness of of the ­population31.
The optimization process has two version full version and light weight version. The significant differences
between the two versions is the number of the population and the number of the generation. The full optimization
process is the default one so it declared here as optimization. The input for the Optimization is 10 variables, the
first and the second are the phase a and b of the load at bus #634, while the third and the fourth are the phase a

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 7

Vol.:(0123456789)
 www.nature.com/scientificreports/

Figure 9.  Management framework for proposed system.

and b for the load at bus #675, the fifth to seventh are the reactive power of the phases of the installed device on
bus #634 and eighth to tenth are the reactive power phases for the installed device on bus #675.
The optimization is performed on the MATLAB platform using genetic algorithm (GA). The number of
population is 500 individuals and the max generation are 50, the maximum stall generation is 25, the elite
individuals are 5 and the crossover are 0.7. This type of problems needs higher number of population compared
with generation. The role of crossover to prevent the algorithm form locked on the local minima while the elite
individuals accelerate the solution ­reach32.

Objective function
The objective function is developed in this problem to decrease the power losses and to enhance the voltage
unbalance in DS. The first objective is the minimization of the sum of the power losses in the TL or feeders
while the second objective is the minimization of the sum of the unbalancing ratio in every bus. Although the
objective is the minimization of both of losses and unbalancing coefficient, the range of losses value and the
range of the voltage coefficient value are very different. Therefore, the weights of objective function should be
adjusted to make the terms as unit values by dividing the power losses by 105 but because the unbalancing ratio
is less than one no need for dividing this value at any number. In the studied system the power losses reduction
is more significant than the voltage unbalance coefficient therefore the power losses term is multiplied by 10. The
final objective form is achieved by trail and error in variety of system loading condition. The objective function
as equation is shown in the Eq. (8) where the KUn is negtive voltage compenant/positive voltage compenant , the
objective over the 24 hour as numbers is shown in Fig. 11.

10 
nbranch
Obj = ∗ min(W1 ∗ Ploss ) + W2 ∗ min(Sum(KUn )) (8)
105
1

The power losses and the unbalancing coefficient is calculated using UDSLF program, which is adopted
f­ rom25, this program is implemented in MATLAB programming language, Fig. 10 shows the step for UDSLF,
this algorithm is used to create all individuals that used in the optimization process.

Constrains
There are two type of constrains used in this optimization process, the equality constrains such as the power flow
constrains as shown in Eq. (9). The loading constrain is the three phases’ power of the bus where the device is
installed which should be constant as shown in Eq. (10). The bus voltage constrain as Eq. (11), and the burden
of the UPC reactive power as Eq. (12) are the inequality constrains in this optimization process.
Pgenrated = Pload + Ploss (9)

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 8

Vol:.(1234567890)
www.nature.com/scientificreports/

Figure 10.  Flow chart for UDSLF.

12

10

0
0 5 10 15 20 25

Figure 11.  Overall objective function for IEEE 13 bus system.


3
Pphase = P3 (10)
phase=1

vmin < vbus < vmax (11)

QUPC < Qdevice rating (12)

Results of optimization
In this section, the results of the optimization process are introduced. Figure 11 shows the objective function as
numerical value, this numerical value is calculated using Eq. (8). Figure 11 shows that the optimization process

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 9

Vol.:(0123456789)
 www.nature.com/scientificreports/

0.6

0.5

0.4

KU n
0.3

0.2

0.1

0
0 5 10 15 20 25

Figure 12.  The unbalancing ratio of IEEE 13 bus system in four cases.

140

120

100

80
kW

60

40

20

0
0 5 10 15 20 25

Figure 13.  The active power losses of IEEE 13 bus system in four cases.

finds the best solution compared to the other cases when the SVC and UPC installed without GA-based man-
agement algorithm.
However, the device can decreases the power losses in DS as shown in Fig. 13, in the light loading conditions
the saved power is not very high compared to the same case when SVC or the base system case, so the device
could be turned off or run the basic functionality. In general, DS in light loading conditions doe snot suffer from
the challenges that faces the contingencies DS. The effect of UPC or UPC-GA is more significant when the system
is heavily loaded because not only the power losses will be decreased but also the system became more stable in
terms of the voltage and transient stability. The significant improvement in system voltage stability is the value
of the unbalancing coefficient when the system is loaded in the first peak at 10 o’clock.
Figure 12 shows the unbalancing coefficient for the system under study, this curve is the sum of the KUn of
every bus. KUn is enhanced if the unbalanced loads is near to the devices locations as shown in the first peak,
but if unbalanced load is far from the devices locations, then this improvement is limited as in the power losses.
The efficiency of the transmission line (TL) is shown in Fig. 14, the efficiency of TL is calculated as in Eq.
(13) where the Pd is the delivered power to the loads and Ptotal is the total power of the all power system. The
enhancement in the TL system by UPC-GA is higher than using UPC or SVC at every hour. Nevertheless, at
light loads, there isn’t significant improvement in the efficiency.
Pd
µ(%) = (13)
Ptotal

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 10

Vol:.(1234567890)
www.nature.com/scientificreports/

0 5 10 15 20 25

Figure 14.  Efficiency of the system TL.

Figure 15.  The contraction of b­ oxplot33. (a) (m = − 019, solid vertical line), (IQR = 1.38, gray shading). A skewed
normal distribution is shown with mean μ = 0 (dark dotted line) and s.d. σ = 1 (light dotted lines). (b) Box plots
for an n = 20 sample from (a). The box bounds the IQR divided by the median, √and Tukey-style whiskers extend
to a maximum of 1.5 × IQR beyond the box. The box width may be scaled by n.

Results
In this study, the various results related to the voltage and the unbalancing coefficient need special way of illus-
tration. To illustrate the deviation of results in additional to their average, the box plot or boxplot is used. It is
a way for graphically displaying the localization, spread, and skewness groups of numerical data through their
quartiles in descriptive statistics. A box plot can also have lines (referred to as whiskers) extending from the box
to show variability outside the upper and lower quartiles. Outliers that stand out from the rest of the dataset
can be shown as separate points outside the box’s whiskers. Box plots are non-parametric because they don’t
make any assumptions about the underlying statistical distribution; instead, they show variance in samples of a
statistical population. The box-spacing plot’s in each subsection show how dispersed and skewed the data are,
which are typically summarized using the five-number summary.
As mentioned earlier, boxplot show the distribution of the data as shown in Fig. 15 where the Q1 is the data
starting, Q3 is the last data, IQR is the distance between Q1 and Q3, and σ is the standard deviation; the box
in the plot contains the half of the data, while the whiskers shows the distribution of the other half taking into
account that the whiskers long doesn’t accessed 1.5 from IQR33,34.

Bus voltage results

This section presents an in-depth analysis of the voltage divergence across the three phases of the system, which
is a critical parameter for assessing the quality of power distribution. The divergence of voltages for Phase A, B,
and C is illustrated in Figs. 16, 17, and 18, respectively. Each figure encapsulates the four distinct cases under
study, beginning with the baseline system configuration without any additional devices, followed by the inclusion
of SVCs at buses #634 and #675, the subsequent addition of UPCs to the same buses, and finally, the optimal
control of UPCs using a GA in the presence of SVCs.
Voltage divergence, defined as the deviation of bus voltage from the unit value, is a key indicator of the system’s
voltage stability. The phase A voltage divergence (VD) across the four cases is elucidated in Fig. 16. It is observed
that as the system evolves from the baseline to the UPC-GA configuration, the Interquartile Range (IQR) of the
VD decreases, and the range of VD becomes predominantly positive, with a mean value of approximately 0.06 for
the UPC-GA case. The introduction of UPCs notably enhances the voltage values compared to the SVC configura-
tion, demonstrating the potential of power electronics devices to improve voltage balance in unbalanced systems.

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 11

Vol.:(0123456789)
Vol:.(1234567890)
Scientific Reports |
Voltage Deviation p.u. Voltage Deviation p.u. Voltage Deviation p.u. Voltage Deviation p.u. Voltage Deviation p.u. Voltage Deviation p.u.

(2025) 15:2672 |
0.02
0.04
0.06
0.08
0.02
0.04
0.06
0.08
0.05
0.05

0
0
0
0.1
0
0.1
-0.05
0.05

-0.05
0.05

650 650 650 650 0

650 650
632 632 632 632 632
632
633 633 633 633
www.nature.com/scientificreports/

633 633
634 634 634 634 634
634
645 645 645 645 645 645
646 646 646 646 646 646
671 671 671 671 671 671

Bus Id
Bus Id
680 680 680 680 Bus Id 680 680
684 684 684 684 684 684
611 611 611 611 611 611
652 652 652 652 652 652
692 692 692 692 692 692
675 675 675 675 675 675

0.02
0.04
0.06
0.08
0.02
0.04
0.06
0.08
0.05
0.05

0
0
0
0.1
0
0.1
-0.05
0.05

-0.05
0.05

650 650 650 650 650 650

632 632 632 632 632 632
633 633 633 633 633 633
634 634 634 634 634 634
645 645 645 645 645 645
646 646 646 646 646 646
671 671 671 671 671 671

Bus Id
Bus Id
Bus Id

680 680 680 680 680 680

684 684 684 684 684 684
611 611 611 611 611 611
652 652 652 652 652 652

https://doi.org/10.1038/s41598-024-57331-4
692 692 692 692 692 692
675 675 675 675 675 675

Figure 17.  Boxplot illustrating the voltage divergence of phase B across four cases.

Figure 18.  Boxplot illustrating the voltage divergence of phase C across four cases.
Figure 16.  Boxplot illustrating the voltage divergence of phase A across four cases.

12
www.nature.com/scientificreports/

0.04 0.04

0.03 0.03

kU n
0.02 0.02

0.01 0.01

650
632
633
634
645
646
671
680
684
611
652
692
675

650
632
633
634
645
646
671
680
684
611
652
692
675
0.04 0.04

0.03 0.03
kU n

0.02 0.02

0.01 0.01
650
632
633
634
645
646
671
680
684
611
652
692
675

650
632
633
634
645
646
671
680
684
611
652
692
675
Hour Hour

Figure 19.  Boxplot illustrating the Unbalancing ratio across four cases.

Similarly, the enhancement in phase B’s voltage divergence is revealed in Fig. 17, where the addition of UPCs
leads to a reduction in the IQR, aligning with the trend observed in phase A. This reduction is accompanied by a
significant number of bus voltages converging towards unity, indicating improved voltage stability. The pattern
is consistent in Fig. 18 for phase C, where the IQR further decreases, resulting in a reduced VD. These findings
underscore the effectiveness of UPCs in mitigating voltage divergence and enhancing system stability.
The figures provided show the voltage deviation rather than the voltage value, with the deviation being the
difference between the voltage value and the nominal value, which is 1 per unit (p.u.). In these figures, values
that tend towards zero indicate greater voltage stability. The visual clarity of the voltage deviation between the
UPC and UPC-GA configurations is not immediately apparent, however, when the unbalancing ratio is consid-
ered, it becomes evident that the UPC-GA introduces fewer unwanted symmetrical components to the system.

Unbalancing ratio
This section presents the unbalancing ratio (KUn) for the system under study, as depicted in Fig. 19. The standard
value for KUn in the IEEE standard (std)35 is not permitted to exceed 2% , whereas in the Russian ­standard36, it
should not exceed 3%. The implementation of UPCs has been observed to reduce KUn from more than 4% in the
base case scenario and to values below the standard in the buses where SVCs are installed. However, the instal-
lation of UPCs in bus #675 does not entirely prevent all loading cases, particularly those exceeding 24 h, from
surpassing the standard value of 3%. This is due to the inherent unbalance in the grid-supplied voltage. When
UPCs are subjected to system-level control via a Genetic Algorithm (GA) with the objective of minimizing the
overall system unbalancing ratio, the UPCs effectively maintain the unbalancing ratio below the standard value
of 3%, with the exception of bus #611, which lacks three-phase loads. It is concluded that the use of UPCs without
GA enhances the unbalancing ratio; however, it does not guarantee an increase in the overall system unbalancing
ratio, thereby increasing the likelihood that one or more buses may exceed the standard value.

Conclusions
The influence of three power electronics-based devices (SVC, UPC, and UPC-GA) on the quality and power
losses of an unbalanced three-phase system were evaluated in this study. For the purpose of determining the
operating modes of power systems with asymmetric loads, mathematical models for power electronics devices
are developed for use in the standard system (IEEE-13 bus system). It is explained how to calculate and analyze
the operating modes for each device in the phases of a power system with asymmetric loads using a universal
power flow model. There have been evaluations of the efficiency of using power electronics devices at particular
points in the power system where they are installed for overall loss and voltage imbalance reduction indicators.
The devices effects on mode of operation have been evaluated during the variation of loads during one day
(daily load curve). The paper has developed a methodology for UPC and SVC management to be adapted with
the dynamic variation of system loading every hour. A proper multi-objective function has been developed to
determine the mode of operation of UPC device. The objective is related to the minimization of both losses and
voltage unbalance. A comparison between base case, SVC, UPC and UPC-GA has been conducted. The results
have proved that UPC-GA improves the system performance according to the developed optimized parameters.
Nevertheless, it has been recommended to switch off the UPC device in certain case because its very slight effect
at light loading. The balancing of the system has been evaluated using the value of the unbalancing coefficient.
The results developed in box plot form have proved the effectiveness of using the proposed devices to improve
the voltage balance. The UPC-GA provides the most significant effect in reduction both overall system losses
and voltage unbalance over daily loading.

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 13

Vol.:(0123456789)
 www.nature.com/scientificreports/

At the final of this study, if the UPC is required to be installed on unbalanced DS, the following recommen-
dations should be considered:

• In the light loading conditions, the UPC could be shutdown or works in the base case without activating GA
because there no major differences form the base case and when the system is equipped with UPC
• The device UPC-GA relieves the contingency of the system and increase the capacity of the TL as shown in
the efficiency curve in Fig. 14
• In this study using UPC-GA guarantees keeping the KUn within the standard range
• Using system controlled UPC-GA decreases the possibilities of having violation in the std value for the KUn
not only for the local buses but also for the overall system
• Using UPC-GA increases the reliability of power system because the possibilities of having shutdown due to
std value violation is decreased

Data availibility
In this article, we would like to highlight the importance of data availability in scientific research. We understand
that data sharing is crucial for advancing scientific knowledge and for promoting transparency and reproduc-
ibility in research. For this reason, we would like to inform our readers that the data used in this study is available
through a software package published on Zenodo with a URL (https://doi.org/10.5281/zenodo.7763421) or DOI
“10.5281/zenodo.7763421”. This package contains all the necessary data and code required to replicate our results
and ­analyses37. We encourage our readers to access and use this package to validate our findings and to conduct
further research in this area. However, we understand that some readers may encounter difficulties when run-
ning the code or accessing the data. Therefore, we invite anyone who experiences any problems to contact us via
email at “[email protected]”. We will be happy to provide any assistance necessary to ensure
that our data is accessible and usable to all interested researchers. In conclusion, we would like to reiterate our
commitment to data sharing and transparency in scientific research. We believe that making our data available
to others will not only benefit our own research but will also contribute to the advancement of knowledge and
the development of new research avenues in the field.

Received: 15 March 2023; Accepted: 18 March 2024

References
1. Liu, Z. & Milanovic, J. V. Probabilistic estimation of voltage unbalance in MV distribution networks with unbalanced load. IEEE
Trans. Power Delivery 30(2), 693–703. https://​doi.​org/​10.​1109/​TPWRD.​2014.​23223​91 (2015).
2. Jayachandran, M., Reddy, C. R., Padmanaban, S. & Milyani, A. H. Operational planning steps in smart electric power delivery
system. Sci. Rep. 11(1), 17250. https://​doi.​org/​10.​1038/​s41598-​021-​96769-8 (2021).
3. Astashev, M. G., Panfilov, D. I., Chasov, A. V. & Rachitov, P. A. Toward the calculation of balancing devices for three-phase three-
wire power-transmission lines. Russian Electr. Eng. 93(6), 368–373. https://​doi.​org/​10.​3103/​s1068​37122​20600​25 (2022).
4. Panfilov, D. I., Astashev, M. G. & Chasov, A. V. Balancing modes of operation of three-phase power-transmission lines by semi-
conductor power regulators. Russian Electr. Eng. 93(6), 361–367. https://​doi.​org/​10.​3103/​s1068​37122​20600​62 (2022).
5. Chasov, A. V., Panfilov, D. I., Astashev, M. G., Rashitov, P. A. & Mostovoi, D. V. Devices for balancing operation regimes of three-
phase four-wire power-transmission lines. Russian Electr. Eng. 93(6), 374–380. https://d ​ oi.o
​ rg/1​ 0.3​ 103/s​ 10683​ 71222​ 06004​ 9 (2022).
6. Karimi, A., Nayeripour, M. & Hassanzadeh, M. E. Novel distributed active and reactive power management approach for renew-
able energy resource and loads in distribution network. Iran J. Sci. Technol. Trans. Electr. Eng. 43(S1), 439–459. https://​doi.​org/​10.​
1007/​s40998-​018-​0126-9 (2019).
7. Cetin, A., Ozen, C., Kaya, K., Bayatmakoo, A.: Comparison of International Voltage Unbalance Standards Results for Different
Power Quality Measurement Points: A Study on MV grid in Turkey. In: 2022 11th International Conference on Power Science
and Engineering (ICPSE), pp. 15–20. IEEE, Eskisehir, Turkey (2022). https://​doi.​org/​10.​1109/​ICPSE​56329.​2022.​99354​57. https://​
ieeex​plore.​ieee.​org/​docum​ent/​99354​57/
8. Committee of the IEEE Power. IEEE Recommended Practice for the Transfer of Power Quality Data. IEEE (2003). https://​doi.​org/​
10.​1109/​IEEES​TD.​2004.​94416.https://​ieeex​plore.​ieee.​org/​docum​ent/​12633​28/ Accessed 2023-02-05
9. Daratha, N., Das, B. & Sharma, J. Coordination between OLTC and SVC for voltage regulation in unbalanced distribution system
distributed generation. IEEE Trans. Power Syst. 29(1), 289–299. https://​doi.​org/​10.​1109/​TPWRS.​2013.​22800​22 (2014).
10. Alanazi, M., Alanazi, A., Almadhor, A. & Memon, Z. A. Multiobjective reconfiguration of unbalanced distribution networks using
improved transient search optimization algorithm considering power quality and reliability metrics. Sci. Rep. 12(1), 13686. https://​
doi.​org/​10.​1038/​s41598-​022-​17881-x (2022).
11. Nayeripour, M., Mahboubi-Moghaddam, E., Narimani, M. R. & Waffenschmidt, E. Secure and reliable distribution feeder recon-
figuration in the presence of static VAR compensator. Iran J. Sci. Technol. Trans. Electr. Eng. 44(1), 293–308. https://​doi.​org/​10.​
1007/​s40998-​019-​00243-1 (2020).
12. Muhammad, Y. et al. Design of fractional evolutionary processing for reactive power planning with FACTS devices. Sci. Rep. 11(1),
593. https://​doi.​org/​10.​1038/​s41598-​020-​79838-2 (2021).
13. Semensato, M.: Application of the Ideal Compensation Method in Unbalanced Distribution Network Considering Harmonics. In:
2019 IEEE PES Innovative Smart Grid Technologies Conference - Latin America (ISGT Latin America), pp. 1–6. IEEE, Gramado,
Brazil (2019). https://d ​ oi.o
​ rg/1​ 0.1​ 109/I​ SGT-L
​ A.2​ 019.8​ 89544​ 9. https://i​ eeexp
​ lore.i​ eee.o
​ rg/d
​ ocume​ nt/8​ 89544​ 9/ Accessed 2023-01-15
14. Zhang, Y., Wang, H. & Xie, Y. An intelligent hybrid model for power flow optimization in the cloud-IOT electrical distribution
network. Cluster Comput. 22(S6), 13109–13118. https://​doi.​org/​10.​1007/​s10586-​017-​1270-0 (2019).
15. Li, Y., Li, Y., Li, G., Zhao, D. & Chen, C. Two-stage multi-objective OPF for AC/DC grids with VSC-HVDC: Incorporating deci-
sions analysis into optimization process. Energy 147, 286–296. https://​doi.​org/​10.​1016/j.​energy.​2018.​01.​036 (2018).
16. Zhang, M. & Li, Y. Multi-objective optimal reactive power dispatch of power systems by combining classification-based multi-
objective evolutionary algorithm and integrated decision making. IEEE Access 8, 38198–38209. https://​doi.​org/​10.​1109/​ACCESS.​
2020.​29749​61 (2020).

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 14

Vol:.(1234567890)
www.nature.com/scientificreports/

17. Ahmed M. Elkholy, Panfilov, D.I., Astashev, M.G.: The Influence of Transmission-Line Parameters and Its Load on Losses and
Power Quality at the Load-Connection Point. Russ. Electr. Engin. 94(6), 403–411 (2023). https://​doi.​org/​10.​3103/​S1068​37122​
30600​44. Accessed 2024-02-15
18. Elkholy, A.M., Panfilov, D.I., El Gebaly, A.E.: Evaluation of Zero and Negative Sequence Currents Influence of Asymmetric Load
on the Power Losses and Quality in Distribution Networks. In: 2023 IEEE 24th International Conference of Young Professionals
in Electron Devices And Materials (EDM), pp. 1110–1115. IEEE, Novosibirsk, Russian Federation (2023). https://d ​ oi.o​ rg/1​ 0.1​ 109/​
EDM58​354.​2023.​10225​208. https://​ieeex​plore.​ieee.​org/​docum​ent/​10225​208/ Accessed 2023-08-31
19. Elkholy, A.M., Panfilov, D.I., ELGebaly, A.E.: Identifying the Rating of the Unified Power Controller According to Unbalanced
Distribution System Requirements. In: 2023 IEEE 24th International Conference of Young Professionals in Electron Devices And
Materials (EDM), pp. 960–965. IEEE, Novosibirsk, Russian Federation (2023). https://​doi.​org/​10.​1109/​EDM58​354.​2023.​10225​
130. https://​ieeex​plore.​ieee.​org/​docum​ent/​10225​130/ Accessed 2023-08-31
20. Panfilov, D.I., Elkholy, A.M., ELGebaly, A.E.: Evaluation of the Influence of AC Voltage Regulators, SVC and Unified Power
Controllers on Power Losses and Quality in Unsymmetrical Distribution Systems. In: 2023 IEEE 24th International Conference
of Young Professionals in Electron Devices And Materials (EDM), pp. 950–956. IEEE, Novosibirsk, Russian Federation (2023).
https://​doi.​org/​10.​1109/​EDM58​354.​2023.​10225​221. https://​ieeex​plore.​ieee.​org/​docum​ent/​10225​221/ Accessed 2023-08-31
21. Elgebaly, A.E., Panfilov, D.I., Krasnoperov, R.N.: Control Algorithms of Discrete Reactive Power Compensators with PI and Fuzzy
Logic Controllers. In: International Conference of Young Specialists on Micro/Nanotechnologies and Electron Devices, EDM, vol.
2020-June, pp. 311–317. IEEE Computer Society, Tanta University, Faculty of Engineering, Tanta, Egypt (2020). https://​doi.​org/​
10.​1109/​EDM49​804.​2020.​91534​92. https://​www.​scopus.​com/​inward/​record.​uri?​eid=2-​s2.0-​85090​82961​4&​doi=​10.​1109%​2FEIC​
onRus.​2019.​86570​58&​partn​erID=​40&​md5=​b83ee​9d458​86206​8f8a9​724b5​d8278​b2
22. Ambriz-Perez, H., Acha, E. & Fuerte-Esquivel, C. R. Advanced SVC models for Newton-Raphson load flow and Newton optimal
power flow studies. IEEE Trans. Power Syst. 15(1), 129–136. https://​doi.​org/​10.​1109/​59.​852111 (2000).
23. Duman, S. Solution of the optimal power flow problem considering FACTS devices by using lightning search algorithm. Iran J.
Sci. Technol. Trans. Electr. Eng. 43(4), 969–997. https://​doi.​org/​10.​1007/​s40998-​019-​00199-2 (2019).
24. Baughman, M.L., Dugan, R.C.: IEEE 13 Node Test Feeder. The Institute of Electrical and Electronics Engineers, Inc, 1–9 (2009)
25. Aviles Cedeño, J.: A three-phase unbalanced load flow solver for large-scale distribution power systems. Master’s thesis, Delft
University of Technology, Mathematics and Computer Science, Mekelweg, Netherlands (August 2017). http://​resol​ver.​tudel​ft.​nl/​
uuid:​0d750​fa1-​b349-​4459-​8ba7-​5f2a3​bbf0c​87 Accessed 2023-01-24
26. Baughman, M.L., Dugan, R.C.: IEEE European LV Test Feeder. The Institute of Electrical and Electronics Engineers, Inc (2009).
Accessed 2022-12-28
27. Rafi, F. H. M., Hossain, M. J., Rahman, M. S. & Taghizadeh, S. An overview of unbalance compensation techniques using power
electronic converters for active distribution systems with renewable generation. Renew. Sustain. Energy Rev. 125, 109812. https://​
doi.​org/​10.​1016/j.​rser.​2020.​109812 (2020).
28. Elkholy, A. M., Taha, I. B. M., Kamel, S. & El-Nemr, M. K. Grid synchronization enhancement of distributed generators using an
adaptive phase-locked loop tuning system. Electronics 11(5), 702. https://​doi.​org/​10.​3390/​elect​ronic​s1105​0702 (2022).
29. Hussien, M.G., Elkholy, A.M., Mansour, D.-E.A., Akl, M.M., Numair, M., Soliman, I.A.: Performance Enhancement of Wave Energy
Converters Integrated with DC Grids Using Energy Storage Systems. In: 2021 22nd International Middle East Power Systems
Conference (MEPCON), pp. 77–83. IEEE, Assiut, Egypt (2021). https://​doi.​org/​10.​1109/​MEPCO​N50283.​2021.​96862​94. https://​
ieeex​plore.​ieee.​org/​docum​ent/​96862​94/ Accessed 2022-12-05
30. Nozaripour, A. & Soltanizadeh, H. Sparse representation based on modified genetic algorithm for classification. Iran J. Sci. Technol.
Trans. Electr. Eng. 46(3), 819–827. https://​doi.​org/​10.​1007/​s40998-​022-​00499-0. (2022).
31. Zareb, M., Nouibat, W., Bestaoui, Y., Ayad, R. & Bouzid, Y. Evolutionary autopilot design approach for UAV Quadrotor by using
GA. Iran J. Sci. Technol. Trans. Electr. Eng. 44(1), 347–375. https://​doi.​org/​10.​1007/​s40998-​019-​00214-6. (2020).
32. ElGebaly, A.E., El-Wahab Hassan, A., El-Nemr, M.K.: Reactive power compensation by multilevel inverter STATCOM for railways
power grid. In: S., S. (ed.) Proceedings of the 2019 IEEE Conference of Russian Young Researchers in Electrical and Electronic
Engineering, ElConRus 2019, pp. 2094–2099. Institute of Electrical and Electronics Engineers Inc., Electrical Power and Machines
Engineering Department, Faculty of Engineering, Tanta University, Tanta, Egypt (2019). https://​doi.​org/​10.​1109/​EICon​Rus.​2019.​
865705​ 8. https://w​ ww.s​ copus.c​ om/i​ nward/r​ ecord.u
​ ri?e​ id=2-s​ 2.0-8​ 50634​ 50488​ &d
​ oi=1​ 0.1​ 109%2​ FEICo ​ nRus.2​ 019.8​ 65705​ 8 &p ​ artn​
erID=​40 &​md5=​b83ee​9d458​86206​8f8a9​724b5​d8278​b2
33. Krzywinski, M. & Altman, N. Visualizing samples with box plots. Nat. Methods 11(2), 119–120. https://​doi.​org/​10.​1038/​nmeth.​
2813. (2014).
34. McGill, R., Tukey, J. W. & Larsen, W. A. Variations of box plots. Am. Statistician 32(1), 12–16. https://​doi.​org/​10.​1080/​00031​305.​
1978.​10479​236 (1978).
35. Std, I. E. E. E. IEEE Recommended Practice for Monitoring Electric Power Quality (1995). https://​doi.​org/​10.​1109/​IEEES​TD.​2019.​
87964​86.http://​ieeex​plore.​ieee.​org/​xpl/​freea​bs_​all.​jsp?​arnum​ber=​51540​67
36. Standard, R.S.: Electric energy. Electromagnetic compatibility of technical equipment. Control and monitoring of electric power
quality in the public power supply systems. STD 33073-2014, Federal Agency for Technical Regulation and Metrology “Ross-
tandart”, Presnenskaya embankment 10-2 (IQ-Block) Moscow, Russia, 123112 (2014). https://w ​ ww.r​ st.g​ ov.r​ u/p
​ ortal/e​ ng/ Accessed
2023-01-26
37. Elkholy, A. IEEE13-optimization: The Final Version. Zenodo (2023). https://​doi.​org/​10.​5281/​ZENODO.​77634​20.https://​zenodo.​
org/​record/​77634​20 Accessed 2024-02-15

Acknowledgements
We would like to express my deepest gratitude and appreciation to the Egyptian Knowledge Bank for their invalu-
able support in facilitating our research endeavors. The financial assistance provided by the Egyptian Knowledge
Bank through their agreement with Springer has been instrumental in enabling us to access crucial research
materials and publications necessary for our work. We acknowledge and thank the Egyptian Knowledge Bank
for their vision and commitment to advancing research and education in Egypt, and for their efforts to promote
access to high-quality research resources for students, scholars, and researchers. We would also like to extend
our heartfelt thanks to the team at Springer for their unwavering dedication to the advancement of academic
research, and for providing us with the resources and tools needed to conduct our work with excellence.

Author contributions
All authors contribute equally to the manuscript.

Competing interests
The authors declare no competing interests.

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 15

Vol.:(0123456789)
 www.nature.com/scientificreports/

Additional information
Correspondence and requests for materials should be addressed to A.M.E.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives
4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in
any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide
a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have
permission under this licence to share adapted material derived from this article or parts of it. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and
your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​
licen​ses/​by-​nc-​nd/4.​0/.

© The Author(s) 2025

Scientific Reports | (2025) 15:2672 | https://doi.org/10.1038/s41598-024-57331-4 16

Vol:.(1234567890)