Design Optimization Method and Performance Analysis of

90MW Grid-Connected Solar PV system in Cambodia

Author’s name

Author affiliation(s)

City & country location(s)

Email address(es)

Abstract

PV facilities on a large scale are reported to be capable of lowering the cost per watt of regular

power installed. A new technique for computing the best contour of big PV plants is given in this

article. The optimization method takes into account the system's effects based on the quantity,

kind, and layout of components at the installation site. The plant generates enough energy to meet

the needs of the family, with any excess energy being fed into the utility grid via net metering. A

performance analysis for an annual generation is conducted using the available data for the solar

PV plant site location. Solar irradiation and performance ratio offer the technical requirements. In

the computations to be performed by the suggested system, high accuracy of energy generation

should be used. Experimental data from a 90MW PV plant in Cambodia was used to validate the

high-accuracy feature. The results suggest that by using the optimization method, the cost of

energy generated can be lowered. As a result, the advantage achieved over the PV module's

operating life time can be maximized. The model's correctness is determined by the model's

hierarchical structure level, model measurement accuracy, model identification, structure accuracy,

measurement accuracy, simulator algorithms metrics, and calibration parameter accuracy. The

article is built on the PV power model structure's physical ranking, which allows the system to be

viewed as a high-fidelity model.

1. Introduction
The grid-connected solar PV model investigates a variety of environmental energy generation

alternatives. Environmental issues, as well as increased electricity consumption, are driving forces in the

adoption of renewable energy. The difficulty for developing and already developed countries is to

generate electricity continually. The constant rise in population and electrical consumption has resulted in

the emission of environmentally hazardous gases. [1]. The goal of this article is to show the model-based

techniques for a grid-linked solar PV system in both commercial and domestic use.

Non-conventional energy sources, such as wind, solar, rain, tide, and geothermal, are

abundant but not exhaustible. Non-conventional energy sources produce clean energy and hence

help to keep the environment clean. [1]. In this article, the solar photovoltaic cell is discussed. In

Cambodia, the installation of a solar PV plant is being examined in terms of at-home electronic

device consumption. The plant generates enough energy to meet residential demand, with any

excess energy being supplied back into the utility system via net metering. A yearly generating

performance study is performed using the available data for the solar PV plant site location. The

technical specifications details are provided by solar irradiation and performance ratio.

There is a corresponding structural and integrated research and design on how the solar

photovoltaic system is installed. The potential solutions for reducing the problems that affect big

PV system installation and electric network performance when power fluctuations occur are

examined. A test has been constructed to demonstrate the process of designing a 90MW grid-

connected solar PV system in Cambodia [2]. Finally, this article develops a standard procedure in

developing organizational large-scale grid-linked solar PV systems as applied in electrical

engineering.
Figure 2. The solar farm 90MW in Krokur district Pursat province Cambodia.
 2. Methodology

Solar photovoltaic (PV) technology turns solar radiation into dependable electricity [3]. Each module has

a specific number of solar cells based on its power rating. Because solar cells are made of silicon

semiconductors, they produce clean and dependable electricity [1]. A new technique for the excellent

installation of extensive PV plants is conferred, and the optimal contour of the large=scale PV plant is

evaluated. The impact of power rating and operating voltage range is considered during the design

process. The optimization method takes into account the huge PV plants' concurrent design ebb features.

This has an impact on energy generation and capital maintenance expenses. The devised approach takes

into account changes in PV output power during the installation of big PV systems.

The total PV modules that have to be built in the PV plant are evaluated concerning the

PV plant power appraisal. In this optimization method design, an algorithm that detects the whole

optimal distance of adjacent PV blocks is proposed. The algorithm makes exploration of the

effects of the PV plant configuration possible [2]. High accuracy is required in the evaluation of

energy production. The design results show that using the proposed optimization method can lead

to a low cost of the generated energy by the extensive PV plant. This enables the fiscal advantage

that results all along the practical life duration of the PV module to maximize.

3. The Optimization Method and Optimization results

3.1 Modeling the PV Plant

To begin, a new method is added to the system to modify the existing one. This establishes a link between

the tilt angle and the shading impact, which should be considered in flat roof solar PV systems. PV

systems are allocated in PV inverters for the power generated to be injected into the electric grid via a

transformer and a cable at a common coupling point [3]. The total modules to be equipped in the PV plant

are computed concerning the PV plant power rating. The power rating Plant, nom (MWp ) is stated by the

plant engineer as shown in the equation below [2]. Where E is energy (kWh), A is total Area (m 2), r is

solar yield (%), H is annual average solar radiation, and PR is the performance ratio.

E = A * r * H * PR (1)

These PV modules are allotted in sets, with each set being linked to a particular PV

inverter. Every set has NP PV strings comprising of Ns PV modules serially connected [4]. A

minimum number of modules (Nsmin) and maximum modules (Nsmax) computation is done

corresponding to the inverter DC input energy point voltage gauge. The following diagram

shows the PV module and a single-index equivalent model of the PV cells [5].

VOC, max and VM, max represent the maximal open-circuit voltage and maximal PowerPoint

developed at the PV component terminals in considering the solar irradiance and temperature

changes prevailing within the plant installation site [1].

For a solar cell, the IV equation is given by;

I =I L −I 0 e
Vt (2)

nkT
V t= (3)
q

Where nkT 0.026 V and n is the ideality factor that varies depending on the operating point. The

product of current and voltage should give the power that will be produced by the cell.
 V

P=V I L −V I 0 e
Vt (4)

The differential of power with respect to voltage is given as;

[ ]
V V
dP 1
=I L − V I 0 e V −I 0 e V t t
(5)
dV Vt

dP
VM, max occurs when =0
dV

VM VM
1 V V
(6)
V M ,max I 0 e −I 0 e =I L t t

Vt

Rearranging and simplifying;

VM VM
1 V V IL
V M ,max e −e = t t

Vt I0

( )
VM
Vt VM IL
e −1 =
Vt I0

VM
Vt
V
Vt (I
+ln M −1 =ln L ⁡
I0 ) ( )
V M =V t ln
( ) (
IL
I0
−ln
VM
Vt
−1 ⁡
) (7)

V OC =V t ln
( )
IL
I0
(8)

∴ V M =V OC −ln
( VM
Vt
−1
) (9)

The equation above is implicit and converges with iteration.

 3.2 The Optimization Algorithm

The optimization procedure is used to compute the optimal values for the dummy variables. The

flowchart of the optimization procedure is shown below. The optimization technique generates a list of

design variables for each PV module and PV inverter combination [6]. The data are utilized to validate

the mathematical components of the large-scale PV plant to determine the interrelated value of the

optimization approach. The LCOE considered in the system is calculated as the ratio of the current value

of the plant's cumulative life cycle to the total electricity generated over the life period [7]. This ratio is

subjected to the design constraints that should be met where X = [Ns, Np, Nr, Fy, β, DIM1] represents the

vector of the design non-constants [8]. The flowchart below represents the optimization algorithm.
 Figure 4. Classification of simulation software based on its application

The LCOE is used as the fair operation in the optimization technique since it determines the

most profitable construction of the large PV system. The system constraints, defined in terms of

inequalities, determine both the upper and lower permissible values of the optimization process.

The fiscal value of electric power is assumed to stay constant during the model's operational life

period [9]. When the electric energy is constant, the optimization technique is separated from

variables that are unrelated to the PV plant system. Iteratively, the optimization process generates

fresh sets of design variable values. The process is repeated up to a point where an optimum

solution is obtained which makes the LCOE value minimum. In this design procedure, genetic

algorithms are proposed to select an optimal value of the variables found in vector X [10].

Therefore, the process minimizes the PV plant LCOE due to its ability to generate a global

optimum solution with computational simplicity.

3.3 Optimization results

The created solar PV optimization is applied to a fictitious solar PV system as well as a genuine and

existent solar PV system. The outcomes are examined for efficiency, consistency, and computation time.

Many of the grid-tied flat roof solar PV system test cases are modeled and the tilt angle optimized for

different geographical regions in Cambodia using the hypothetical configuration. The developed

optimization is employed in designing the yearly mean for energy generation to ensure model

consistency.

Table 1. Results power Output monthly (12 months) and permeant ratio (%) of 90MW solar PV.
 GlobHor DiffHor T_Amb ℃ GlobInc GlobEff EArray E_Grid PR ratio

kWh/m2 kWh/m2 kWh/m2 kWh/m2 MWh MWh

January 158.7 60.12 26.54 182.2 179.1 14195 13902 0.825

February 144.3 74.75 28.36 156.5 153.4 12129 11880 0.821

March 170.9 81.83 29.69 176.7 173.1 13546 13264 0.811

April 161.4 84.84 29.85 158.3 154.5 12193 11936 0.815

May 184.8 79.43 29.30 173.1 168.4 13301 13023 0.813

June 171.7 82.01 28.47 157.9 153.2 12271 12012 0.823

July 179.3 78.32 28.52 165.7 161.0 12823 12552 0.819

August 159.2 87.72 28.27 153.9 149.9 12001 11742 0.825

September 133.2 68.55 27.61 133.5 130.1 10383 10149 0.822

October 139.8 84.40 27.88 147.0 143.6 11534 11286 0.830

November 129.9 70.27 26.60 143.2 140.2 11279 11038 0.834

December 146.4 59.61 26.19 169.5 166.4 13293 13017 0.830

Year 1879.6 911.85 28.10 1917.3 1872.9 14898 145801 0.822

Figure 5. Average lifetime cost and energy
 Figure 6. central receiver solar thermal power analysis.

4 Power Control Method

The huge infiltration of renewable energy sources is significantly expanding as a result of the increased

awareness of environmental issues and the deregulation of electric power. A photovoltaic (PV) system is

one of the renewable energy sources that provide a long-term vision for power distribution [11].

However, the connection of a large number of scattered energy sources to the electrical power grid

produces instability when there are electrical disturbances in the grid. One of the disturbances is voltage

sags, which are caused by short connections between phases and ground, overloads, and lightning. Low

voltage ride-through regulations have been set at the grid connection to minimize these consequences.

These regulations require the distributed energy sources to inject reactive current for supporting grid

balance amid faults. The figure below shows the power control flowchart scheme.
 Figure 7. Power control flowchart scheme

Grid voltage sags are thought to make the task of the grid-connected inverter more difficult.

The degeneration of power quality caused by power ripples, as well as the increase of current

harmonic deception, are among the deteriorated performances [11]. To avoid these disadvantages,

a 90MW grid-connected PV system with a reactive power control approach is proposed [12]. It

has positive and negative current injection arrangements that offer active power and reactive

current that does not exceed the inverter's current limit. During voltage sag, the current injection

system supports the voltage. The inverter’s power control of the PV system ensures that there is

power balance during grid fault.

5 Simulation Analysis

The methodology above is used in MATLAB software and GA functions available in the Global

Optimization Toolbox library. The accuracy of the computations achieved by the design tool enables the

optimization of the 90MW PV plant installed in Cambodia [2]. The technical and economic parameters in

Cambodia are used to obtain the analysis and the system design. The input in the mathematical models

used in the optimization algorithm.

 The design optimization results used where the minimization or maximization of the life

period energy generation of the 90MW PV plant are taken as the objective functions in the

optimization procedure [4]. For different configurations of the PV plant are copied in all cases.

LCOE value obtained during the design optimization method is slightly low than the one obtained

by applying minimal-cost and maximal-energy objective functions [10]. Successful computation

of the optimal global solution is verified by performing the optimization process using an

exhaustive search algorithm. The simulation results are as shown below.

Table 2. Analysis of the simulation

Load Generated Injected

Bus Voltage Mag Angle MW Mvar MW Mvar Mvar

No. Degree

1 0.981 -2.387 85.000 76.500 0.000 0.000 0.000

2 1.000 0.000 90.000 81.000 411.135 58.603 0.000

3 0.949 -6.137 75.000 67.500 0.000 0.000 0.000

4 0.926 -7.892 45.000 40.500 0.000 0.000 0.000

5 0.924 -8.022 45.000 40.500 0.000 0.000 0.000

6 0.952 -6.128 60.000 54.000 0.000 0.000 0.000

Total 400.000 360.000 411.135 85.603 0.000

Figure 8. The system analysis results
 Figure 9. The system analysis results

6 Conclusion

Large-scale PV plants are reported to be capable of lowering the cost per watt of regular electricity

installed. A new technique for computing the best contour of large PV plants is given in this article. The

optimization procedure takes into account the system's impacts based on the quantity, kind, and

placement of the components on the installation site. Energy generation with high precision should be

used in computations performed by the proposed system. This system is a high-fidelity example, with

components such as solar panels, DC-DC converters, DC-AC inverters, and three-phase service grids

planned and built mathematically as a replication model [13]. The high-accuracy feature has been

approved using the empirical data of the 90MW PV plant in Cambodia.

The model's accuracy is determined by the model's hierarchical structure level, model

measurement accuracy, model identification, structure accuracy, measurement accuracy,

simulator algorithm metrics, and calibration parameter accuracy [14]. The paper is based on the
physical ranking of the PV power model structure, making the system a high-fidelity model. The

results suggest that using the optimization strategy can result in a lower cost of energy generated.

This allows for the most advantage to be produced over the PV module's operational lifetime.
References

[1] A. Ismail, R. Ramirez-Iniguez, M. Asif, A. Munir, and F. Muhammad-Sukki, Progress of solar

photovoltaic in ASEAN countries, vol. 48, Renewable and Sustainable Energy Reviews, 2015, pp.

399-412.

[2] Z. Zhongming, L. Linong, Z. Wangqiang, and L. Wei, Energy Sector Assessment, Strategy, and

Road Map., Cambodia, 2018.

[3] P. Giroux, G. Sybille, and H. Le-Huy, "Modeling and simulation of a distribution STATCOM

using Simulink's Power System Blockset.," In IECON'01. 27th Annual Conference of the IEEE

Industrial Electronics Society (Cat. No. 37243), vol. 2, no. IEEE, pp. 990-994, 2011, November.

[4] A. Luque and S. Hegedus, Handbook of Photovoltaic Science and Engineering., New York: Wiley,

2003.

[5] T. Yoshino, R. Inzunza, T. Ambo, E. Ikawa, T. Takahashi and N. Takahashi, MW-range PCS for

efficient operation of large PV plants, Proc. 3rd IEEE Int. Symp. Power Electron. Distrib. Gener.

Syst, 2012, pp. 249-253.

[6] W. Muneer, K. Bhattacharya and C. Canizares, Large-scale solar PV investment models, tools, and

analysis: The Ontario case, vol. 26, IEEE Trans. Power Syst., Nov 2011, pp. 2574-2555.

[7] L. Hassaine, E. Olias, J. Quintero and V. Salas, Overview of power inverter topologies and control

structures for grid-connected photovoltaic systems., vol. 30, Renewable and Sustainable Energy

Reviews, 2014, pp. 796-807.

[8] D. Weinstock and J. Applebaum, Optimal design of solar fields, Israel: n Proc. 22nd Convention

Electr. Electron. Eng, 2002, pp. 163-165.

[9] C. Deline, Partially shaded operation of a grid-tied PV system, Proc. 34th IEEE Photovoltaic

Specialists Conf, 2009.

[10] S. Rivera, B. Wu, S. Kouro, H. Wang and D. Zhang, Cascaded H-bridge multilevel converter

topology and three-phase balance control for large scale photovoltaic systems, 2012, pp. 690-697.

[11] A. Dhaneria, Grid-Connected PV System with Reactive Power Compensation for the Grid, IEEE

Power & Energy Society Innovative Smart Grid Technologies Conference (ISGT), 2020 February,

pp. 1-5.

[12] D. Kaudinya, P. Deepak, P. Balachandra and N. RAvindranath, Grid-connected versus stand-alone

energy systems for decentralized power—A review of literature, vol. 13, Renewable and

Sustainable Energy Reviews, 2009, pp. 2041-2050.

[13] E. Nfah, J. Ngundam, M. Vandenbergh and Schmid, Simulation of off-grid generation options for

remote villages in Cameroon., vol. 33, Renewable Energy, 2008, pp. 1064-1072.

[14] M. Ropp and S. Gonzalez, Development of a MATLAB/Simulink model of a single-phase grid-

connected photovoltaic system., IEEE transactions on energy conversion, 2009.