In recent decades, energy has become essential for human advancement, leading to increased global energy demand. While fossil fuels have historically powered industrialization and economic growth, their environmental impact has spurred a transition to renewable energy sources, which are more sustainable and better suited for addressing climate change. Renewable Energy Technologies (RETs) such as solar, wind, biomass, hydro, and geothermal offer key benefits, including reduced greenhouse gas emissions and improved energy access (Pfeiffer and Mulder, 2013).
A developing country like Bangladesh faces challenges in ensuring electricity supply due to limited infrastructure, growing demand, and reliance on fossil fuels. Bangladesh also struggles to supply electricity to some remote locations, such as Manpura Island, which is home to approximately 90,000 people and has promising ecotourism potential.
Manpura lacks consistent power access because it is 18 km from the mainland and 96 km from the national grid. The island currently relies on limited diesel-based electricity. The West Zone Power Distribution Company (WZPDCL) powers Manpura Island. In 1997, WZPDCL established a substation with four diesel power generators near the center of Manpura Island (2 No. Hathazari Union). This substation can supply only 890 customers for six hours each evening. Currently, one generator is malfunctioning, so the remaining three supply electricity. The island’s other three unions (1 No. Manpura Union, 3 No. North Sakuchia Union, and 4 No. South Sakuchia Union) are powered by mini solar grids established by local people or private companies, which serve 2,500 to 2,600 consumers. However, this electricity is costly, at approximately BDT 30 per unit.
In 2021, the Bangladeshi government, through WZPDCL, made an agreement with Western Manpura Solar Power Ltd. (WMSPL) to supply electricity to the entire island. WMSPL proposed a hybrid system with a 3.0 MW capacity, consisting of Photovoltaic (PV), a Battery Energy Storage System (BESS), and a Diesel generator (DG). WMSPL was to sell electricity to the government at a rate of 0.2485 $/kWh (29.82 tk/kWh). However, due to political instability, the project was cancelled, and no progress has been made. The proposed Cost of Energy (COE) in this agreement was considered somewhat high for Bangladesh.
Despite Manpura being an Upazilla of Bangladesh, its residents suffer significantly from a lack of electrification. The island’s economy primarily depends on agriculture and sea fishing. However, electricity shortages disrupt food production, and much of the harvested fish spoils. Additionally, the main market and Upazilla city receive some electricity, but its quality is very poor compared to standard requirements. This means the provided voltage in this small area experiences significant disturbances, which is a concerning issue for consumers.
The objective of this research is to propose a comprehensive solution for electrifying Manpura Island with minimum energy cost and reliable power quality. A case study on an Indonesian island demonstrates that an off-grid system can be a feasible solution for remote islands like Manpura (Josia, 2023). This study focused on the reliability of the proposed power system. Another study on a different island incorporated PV, BESS, and DG to create a grid that reduced fuel consumption by 79.3% (Azman and Sulistiawan, 2024), indicating that renewable energy integration leads to less fossil fuel consumption. A hybrid system simulated for Manpura Island in Bangladesh achieved a COE of $0.262 USD (31.44 Tk/kWh) [1 USD = 120 Tk] and an NPC of $16.3 million USD. This study highlighted the economic and environmental benefits of integrating renewable resources into isolated regions (Arefin et al., 2017). However, it had limitations in addressing the technical aspects of the proposed system, and the energy cost was slightly higher than existing micro-grid solutions.
This research aims to design, optimize, and analyse an off-grid power system for Manpura Island using MATLAB/Simulink and HOMER Pro. The research objectives are divided into two parts: economic analysis and technical analysis. The economic analysis focuses on designing and simulating an off-grid power system comprising diesel generators, PV panels, and a Battery Energy Storage System (BESS) to achieve minimum energy cost. The technical analysis investigates how variations in weather conditions influence the system’s output and the effects of harmonics on performance. Additionally, it seeks to evaluate the economic feasibility of the proposed system while assessing its environmental impact, ensuring a sustainable and cost-effective energy solution for the island.
Given that only one prior study has been conducted on Manpura Island, further research is needed to adequately address the challenges of electrifying the location. The proposed off-grid system provides empirical insights and practical guidance for local stakeholders and may offer a reliable solution for researchers and commercial initiators.
The paper is organized as follows, section 1 discusses the motivation, aims, and objectives of this work. Section 2 reviews related research. Section 3 describes the methodology of the entire work. Section 4 presents the obtained results and various analyses. Finally, Section 5 provides the conclusion and recommendations.
2. Literature review
An economic and technical analysis was conducted to provide electricity for an ice-cream factory in Japan using PV and fuel cell technology (Habib et al., 2024). To ensure 100% reliable electricity, especially for islands, prioritizing renewable energy over fossil fuel-based sources is essential (Barreto, 2018), as standalone diesel generators are often insufficient. Manpura Island, abundant in sunlight and already equipped with small PV plants, presents an ideal location for a hybrid power system integrating PV panels with diesel generators. In Bangladesh, however, PV system installations frequently underperform due to poor design and management. Despite a government investment of $1.5 billion since 2014 in 42 renewable energy projects with a total capacity of 2500 MW, actual power generation from these plants has only reached 38 MW.
Research has also explored integrating renewable sources into the Dhaka metro rail system, demonstrating significant energy savings through PV integration (Islam et al., 2025). While most efforts have focused on household energy needs, few studies have addressed the electrical load requirements of other sectors, such as healthcare.
A comprehensive study highlighted that the Bangladesh government’s limited funds restrict grid extension in remote areas. The study recommended island micro-grids with solar PV as cost-effective and sustainable alternatives. Using HOMER Pro, three case studies determined optimal configurations based on daytime irrigation needs, revealing a cost of energy (COE) between $0.217 and $0.249 USD/kWh and net present costs (NPC) ranging from $0.847 to $0.991 million USD (Shoeb and Shafiullah, 2018). Another researcher investigated various hybrid system configurations, including solar-diesel, wind-diesel, and biomass-diesel combinations. They advocated for the use of renewable and non-renewable resources alongside advanced power electronics and energy storage to enhance system flexibility and availability, particularly in off-grid regions (Paska et al., 2009).
The primary challenges with grid-connected PV systems include transient responsiveness, steady-state inaccuracy, and a lack of control over voltage overshoot. These issues lead to inadequate power quality and potential harm to the power system. Researchers have designed DC to AC PI inverters using Particle Swarm Optimization (PSO) algorithms to address these problems. The proposed approach demonstrated superior performance compared to previous documented research, achieving a Total Harmonic Distortion (THD) with a transient reaction time of 0.1853s, grid voltage of 2.72%, and current of 0.29%. The PI controller with the PSO algorithm reduced voltage overshoot by 11.1% and settling time by 32.6% compared to conventional PI controllers. Consequently, this approach successfully improves the power quality of the utility grid, facilitating smoother integration of PV systems. However, it is also observed that PV integration can decrease system stability and the availability of system inertia (Roslan et al., 2020).
A study utilized RETScreen software to evaluate potential sites in Egypt for a 10 MW PV-grid plant, focusing on economic, GHG, and energy output metrics. The analysis identified Wahat Kharga as the optimal location due to its superior economic performance, energy output, and GHG emission reduction potential. Conversely, Safaga exhibited the lowest energy output and profitability. Based on these findings, the construction of large-scale PV power plant projects is recommended (El-Shimy, 2009).
The integration of renewable energy technologies (RETs) can introduce significant harmonics into a system. To mitigate these disturbances in grid-connected systems, researchers have developed a method employing LC filters for harmonic attenuation. They also investigated passive damping strategies to reduce resonance effects, ensuring compliance with harmonic standards. The effectiveness of these methods was validated through simulations (Ahmed et al., 2007).
Given the long distance from the mainland, integrating Manpura into the national grid is not feasible. As a developing country, Bangladesh faces limitations in extending its grid to remote islands like Manpura. Therefore, this research proposes an off-grid system for the area.
A preliminary study conducted in Chhattisgarh, India, in 2013 compared the economic aspects of off-grid systems versus national grid expansion (Sen and Bhattacharyya, 2014). This research demonstrated that establishing an off-grid system can be a more viable solution for certain remote areas, offering economic advantages, environmental sustainability, and reliability. Another study proposed and analyzed an off-grid hybrid power plant model combining solar and wind energy with a polymer electrolyte fuel cell for energy management. This design aims to minimize fossil fuel usage and CO2 emissions, with MATLAB/Simulink modeling and HOMER Pro optimization proving its efficacy in providing energy to remote areas (Tribioli et al., 2016).
In 2018, an off-grid hybrid system was designed and analyzed for rural communities in Karnataka, India (Suresh et al., 2020). This research sought to minimize the Net Present Cost (NPC) and Cost of Energy (COE) for the target area. The system incorporated PV, Battery Energy Storage Systems (BESS), biomass, biogas, fuel cells, and a wind farm, all utilized during optimization due to their local availability. The optimized COE was 0.163 $/kWh. The research employed HOMER software and a Genetic Algorithm (GA). A metaheuristic optimization technique was also established to minimize the NPC for providing electricity to a transmitter station in India, integrating PV, BESS, biogas, and a pumped hydro energy system (Das et al., 2019). A comprehensive analysis comparing on-grid and off-grid systems for electrifying a domestic community revealed that the on-grid system yielded a lower COE due to its grid sell-back capabilities (Hassan, 2021).
While most researchers focus on optimizing systems with respect to NPC and COE, there is a notable lack of technical studies analyzing the proposed systems. Specifically, comprehensive economic and technical research for Manpura Island is absent. There is significant potential in designing an off-grid system that minimizes NPC and COE, with the added benefit of a thorough technical evaluation. This represents a clear research gap for extensive analysis on Manpura Island
3. Methodology
An off-grid system consists of several components, including photovoltaic panels, wind turbines, hydro-electric turbines, diesel generators, biogas and biomass plants, DC-AC converters, and various types of loads (Akinsipe et al., 2020). The availability of renewable sources is also location-dependent (Mikhail et al., 2020). Our research followed several general steps. First, we estimated the island’s load demand. Next, we investigated available resources to meet this demand. Based on these resources, a system was proposed and then optimized for economic efficiency. Finally, the optimized system was technically evaluated from an electrical perspective. This sequential process is illustrated in Figure 1.
Figure 1. Methodology of analysis of off-grid system for Manpura Island.
3.1 Load demand estimation
To design the proposed off-grid system, we first need to estimate the island’s load demand. To do this, we must comprehensively assess the community’s energy consumption patterns. A survey of the island’s residents and establishments—including schools, colleges, markets, and banks—was conducted to estimate the load demand. Data was also collected from a field visit to Manpura Island.
Now, individual load groups reach their respective peak demands at different times of the day. For example, institutional loads run from 9:00 AM to 5:00 PM. Domestic loads run throughout the day but have a peak demand between 7:00 PM and 10:00 PM. Therefore, engineers introduce a factor called the group diversity factor. This factor combines the demand of all groups to create a total load.
From Table 1, the summation of all groups’ maximum loads is 4.245 MW. The diversity factor will be greater than one. As Manpura is an island, the diversity factor will vary between 1.4 and 1.6 (Sargent et al., 1994). In this research, we will use a value of 1.5 to determine the system’s peak demand. According to equation (1), the system’s maximum demand will be 2.83 MW.
Table 1. Estimation of maximum load.
| Load group | Category | No. of units | Maximum load of each unit (Watt) | Maximum load of all unit
(kW) |
Maximum load of group (kW) |
| Domestic load | Family | 17080 | 200 | 3416 | 3576.5 |
| Mujib 100 years gift house | 535 | 300 | 160.5 | ||
| General institutional load | Primary school | 43 | 500 | 21.5 | 60.9 |
| High school | 9 | 800 | 7.2 | ||
| Madrasa | 39 | 500 | 19.5 | ||
| College | 3 | 1500 | 4.5 | ||
| Court | 1 | 1000 | 1 | ||
| Bank | 2 | 1600 | 3.2 | ||
| Guest house | 1 | 2000 | 2 | ||
| Weather office | 1 | 1000 | 1 | ||
| WZPDCL | 1 | 1000 | 1 | ||
| Emergency institutional load | Police station | 2 | 1200 | 2.4 | 13.9 |
| Fire service | 1 | 1500 | 1.5 | ||
| Hospital | 1 | 5000 | 5 | ||
| Community clinic | 5 | 1000 | 5 | ||
| Commercial load | Market | 8 | 59000 | 472 | 502 |
| Workshop | 10 | 1500 | 15 | ||
| Small industries | 2 | 7500 | 15 | ||
| Agricultural load | Irrigation pump | 5 | 7000 | 35 | 35 |
| Religious load | Mosque | 115 | 400 | 46 | 48 |
| Temple | 5 | 400 | 2 | ||
| Others | Orphanage | 3 | 600 | 1.8 | 9.6 |
| Officer’s club | 2 | 1500 | 3 | ||
| Other clubs | 4 | 1200 | 4.8 |
Electrical demand is increasing daily. During the design of the hybrid system, a safety factor of 25% was incorporated. Given that the island’s peak demand is 2.83 MW, the system will be designed to provide an additional 25% of this value. Therefore, the proposed system capacity will be 3.54 MW, rounded up to 3.6 MW.
In Bangladesh, power demand varies seasonally (Sarker et al., 2017). Agricultural load operates at full capacity during the summer, which is why the summer load profile is presented here. Winter load profiles are inherently lower than summer profiles, eliminating concerns for other seasonal loads. Figure 2 illustrates that the load profiles for Manpura Island during the summer (July) season approach 3.6 MW, with an average load of approximately 2.755 MW. The system’s load factor is 0.787. The peak demand typically occurs between 11:00 AM and 6:00 PM.
Figure 2. Estimated load profile of Manpura Island during summer season.
3.2 Global irradiation data
Figure 3 displays the monthly average solar radiation for Manpura Island, with data sourced from NASA’s Prediction of Worldwide Energy Resource (POWER) database. This data serves as input for the HOMER Pro software. Daily radiation typically ranges from 3.5 to 5.5 kWh/m2. Solar power generation is more efficient from October to May due to increased sunlight. Conversely, the period from June to September experiences higher rainfall and cloud cover, leading to reduced sunlight.
Figure 3. Mean solar global horizontal irradiation data of a month.
3.3 Temperature data
Temperature is also a critical factor in determining how much power a PV system will produce. In Manpura, the monthly average temperature varies from 20° to 30 °C throughout the year, as shown in Figure 4. These temperature fluctuations are important to consider when designing a solar power system on the island.
Figure 4. Monthly average temperature data.
3.4 Area calculation
To determine the area required for a PV plant, we must consider the shadow effect caused by the tilt of the panels. This calculation is crucial for estimating the necessary distance between panels to prevent shadows from being cast on nearby panels, which would lower power production. Therefore, to calculate the area of one PV panel, we must first determine the shadow length to mitigate the shadow effect. Figure 5 illustrates a PV panel tilted at an angle β. Considering the site’s latitude (∅) and the inclination angle (δ) at a critical condition, a shadow of length LS is generated. The shadow length of a PV panel can be calculated as follows:
where ∅ is the site’s latitude, δ is the inclination angle in the worst case, and β is the PV panel’s tilt angle, which is typically equal to the latitude of the location.
Considering β = 23.8041°, ∅ = 23.8041°, δ = 23.50° (worst case scenario), and L=1.65m, sin (23.8041°) ≈ 0.4037, δ + ∅ = 47.3041, tan (47.3041°) ≈ 1.0866
So, the shadow length of the PV panel is = 1.65 × 0.4037 × 1.0866 = 0.7238 m. And the total horizontal length is Lcos β + LS = 2.234 m.
Figure 5. Shadow length measurement of PV panel (PV panel tiled with angle β).
The PV array for the proposed design requires 3.6 MW. For the simulation, the LONGi Solar LR6-60PB 305 W monocrystalline solar panel module, which has 60 cells, was used. Each 305 W panel requires an area of approximately 2.22 m². Therefore, a 3.6 MW PV system using these panels would require a total area of approximately 27,855 square meters, or 0.027855 square kilometers.
3.5 On-site wind profile evaluation
Manpura Island, located in the Bay of Bengal, exhibits an inconsistent and uncertain wind profile. Wind speeds on the island range from a minimum of 1.63 m/s to a maximum of 5.81 m/s, according to data from the Bangladesh meteorological website. This significant variation in wind speed suggests that a wind farm at this location would not be economically feasible. Further supporting this conclusion, a study by Nadi et al. (2019) on offshore wind energy estimation, using real-time data, also indicated high variability in available wind energy, rendering a wind farm in this area economically and technically unreliable. Given Manpura’s proximity to the Chattogram region, an analysis of the wind potential in Patenga, Chattogram (Azad et al., 2026) can be considered when evaluating this location. That study found that the availability of wind potential follows a Weibull distribution pattern, confirming the inherent uncertainties in consistently generating power from wind turbines throughout the year.
3.6 Proposed system for Manpura Island
Given the uncertain availability of wind power, PV plays a crucial role in this system design. Figure 6 illustrates a system comprising PV with BESS and a diesel-based generator, with all components clearly marked. The PV panels generate power during the daytime, serving the electrical load, and any excess electrical energy from the PV is stored in the BESS. At night, the stored energy from the BESS and the diesel generator support the load.
Figure 6. Proposed hybrid power system configuration for Manpura Island.
4. Results and Discussion
4.1 Economic analysis
The economic feasibility of the proposed 3.6 MW Diesel-PV hybrid system was analyzed using HOMER Pro software. HOMER Pro optimizes the sizes of system components to meet the maximum demand at the simulation’s case study location (Khalil et al., 2021). The system is designed for a lifetime of 25 years. Table 2 lists the system cost values used as input for the HOMER Pro software.
Table 2. Input values of costs in HOMER.
| Component | Size | Purchased cost ($) | Replacement ($) |
| PV | 305 W | 150 | 150 |
| Generator | 1 MW | 100,000.00 | 100,000.00 |
| 500 kW | 40,000.00 | 40,000.00 | |
| Inverter | 1 kw | 250 | 200 |
| Battery | 1 kWh | 100 | 80 |
4.1.1 Electrical summary
Figure 7 indicates the energy production of each sub-element of the system throughout the year, as achieved from the HOMER simulation result.
The system produces 15,087,243 kWh of total energy per year. Most of this energy, 4,984,029 kWh/year (almost 33% of the system’s total), is produced by the 1 MW fixed generator-1, as shown in the figure. The photovoltaic (PV) system is the next largest contributor, at 4,767,068 kWh/year. Table 3 compares the COE of the system’s individual components. The PV system, with a generation cost of $0.0324/kWh, is the most economic option, being 6 to 7 times cheaper than the diesel gensets (Babajide and Brito, 2021). This comparison clearly underscores the economic advantage of the PV system, offering substantial cost savings compared to the diesel generators.
Figure 7. Monthly electricity production of different components.
Table 3. Energy production cost of every component.
| Component | Generation cost ($/kWh) |
| PV | 0.0324 |
| 1MW Fixed Capacity Genset | 0.212 |
| 500 kW Genset | 0.209 |
4.1.2 Fuel summary
The fuel summary of HOMER simulation results, shown in Table 4, indicates an average fuel consumption of 24 hours per year. The total fuel consumption over the project’s lifetime is 2,757,792 L/year, averaging 7,555 L/day. Based on a diesel price of $0.87/liter, the total daily fuel cost is $6,572.87, leading to annual fuel expenses of $2,399,200.74. The data also reveals that the system consumes very small amounts of fuel between 7 am and 6 pm. However, from 6 pm to 7 am, fuel consumption significantly increases, reaching up to 1000 liters per hour.
Table 4. Fuel consumption summary.
| Quantity | Value | Units |
| Total fuel consumed | 2,757,792 | L/year |
| Avg. fuel per day | 7,555 | L/day |
| Avg. fuel per hour | 315 | L/hour |
Fuel consumption results in the production of harmful gases. Table 5 shows the annual amounts of different gases produced by the proposed system’s generators. Carbon dioxide is the dominant pollutant, with an alarming emission of 7,245,889 kg/year, underscoring its substantial contribution to greenhouse gas effects and climate change. In contrast, sulfur dioxide and nitrogen oxides are emitted at 17,784 kg/year and 18,032 kg/year, respectively. While still pollutants, their overall environmental impact is less severe. Other pollutants include unburned hydrocarbons and carbon monoxide. If only a diesel generator were used to supply the energy demand, the amount of emissions could increase at a larger rate.
Table 5. Quantity of produced pollutant gases.
| Quantity | Amount (Kg/year) |
| Carbon dioxide | 7,245,889 |
| Carbon monoxide | 29,156 |
| Unburned hydrocarbons | 1,489 |
| Particulate matter | 280 |
| Sulfur dioxide | 17,784 |
| Nitrogen oxides | 18,032 |
4.1.3 Renewable penetration
Renewable penetration is a key performance metric for renewable energy systems. It represents the proportion of a power system’s electricity generated from renewable sources relative to its total energy demand. Higher renewable penetration signifies a greater contribution of renewables to the overall energy mix, thereby reducing dependence on fossil fuels (Alizadeh et al., 2016). Table 6 illustrates the system’s overall renewable penetration. The installed, or nominal, renewable capacity accounts for 50.5% of the total installed capacity. However, the usable renewable capacity is 45% due to the derating factor of PV. Furthermore, the figure indicates that the share of total energy produced by renewable sources over a given period, or the renewable penetration, is approximately 34.4%. The remaining demand is met by conventional generators. While high renewable penetration offers benefits such as reduced fuel consumption, operational costs, and emissions, it may necessitate energy storage or backup generation to maintain system reliability.
Table 6. Renewable penetration of the system.
| Parameter | Value | Units |
| Nominal renewable capacity divided by total renewable capacity | 50.5 | % |
| Usable renewable capacity divided by total capacity | 45.0 | % |
| Total renewable production divided by load | 34.4 | % |
| Total renewable production divided by generation | 31.6 | % |
4.1.4 Cost summary
This section primarily analyzes the cost summary to estimate the feasibility of meeting the island’s load demand at the lowest cost, thereby aiding effective decision-making. We will consider two cases based on the HOMER optimized results, first, a system with only PV and generators, and second, a system that includes an Energy Storage System (ESS). Figure 8 presents two bar charts displaying the cost summary results from the HOMER simulated data. The first chart shows that the initial capital cost is highest for PV at $1,770,491.80 and lowest for the 500 kW generators at $60,000. Fuel consumption is highest for the 1 MW generator-1, at approximately $14,786,408.88, and gradually decreases for the other generators. PV and the converter have zero fuel consumption. Replacement costs are low for all components, as they are not expected to be replaced within their operational lifetime. The second chart illustrates that when an ESS is incorporated, the initial capital cost for storage becomes high, reaching $4,000,000.00, due to the need for a large battery bank to support the load. The replacement cost for the ESS is also substantial, at approximately $8,236,521.20, because its maximum lifetime is five years. This necessitates replacing the batteries four times within the system’s 25-year lifespan. Furthermore, maintaining the ESS requires a significant amount of $5,171,006.62.
Figure 8. Cost summary of proposed system for (a) excluding and (b) including ESS.
The economic analysis of the hybrid power plant in Table 7 reveals that incorporating an ESS significantly increases both operating costs and total NPC, with values of $3,366,862.00 and $50,675,650.00, respectively. In contrast, the system without an ESS has lower operating costs of $2,550,668.00 and a total NPC of $36,124,300.00. The average per unit cost of energy during its lifetime (LCOE) is $0.201/kWh or 24.12 Tk/kWh (considering 1 USD = 120 Tk) for the system without an ESS. This is remarkably lower compared to existing small PV plants in Manpura, where the LCOE was more than 30 Tk/kWh. In contrast, the LCOE for the ESS-integrated system is $0.283/kWh or 33.96 Tk/kWh. This analysis indicates that while the ESS enhances reliability and energy availability, it does so at a greater economic cost, making the non-ESS option more cost-effective from a financial perspective. Therefore, the acceptable system for this research is the system without an ESS.
Table 7. Economic analysis of the system.
| System | Operating cost ($) | Total NPC ($) | LCOE ($/kWh) | |
| WITH ESS | 3,366,862.00 | 50,675,650.00 | 0.283 | |
| Without ESS | 2,550,668.00 | 36,124,300.00 | 0.201 | |
The tariffs for different fuel power plants in Bangladesh, as shown in Table 8, vary based on fuel costs, efficiency, and maintenance. Natural gas plants typically have tariffs ranging from 3 to 3.5 BDT/kWh, making them the cheapest option for electricity production in Bangladesh (Das et al., 2020). Due to the increasing price of coal, the cost for coal-based power plants has risen to approximately 12 BDT. For example, the Payra Power Plant’s tariff will be 12.73 BDT per kilowatt-hour in January 2025 and 13.80 BDT in February 2025, while the Rampal Power Plant’s tariff will be 11.47 BDT per kilowatt-hour in January 2025 and 13.02 BDT in February 2025. Furnace oil (heavy fuel oil) based power plants have an average tariff of 17 BDT per unit, accounting for approximately 23% of the country’s total electricity generation. Diesel-based power plants have the highest electricity production cost in Bangladesh at 37 BDT per unit, though they only account for about 0.92% of the country’s total electricity generation. Bangladesh has been importing electricity from the Adani group since 2023, initially at a cost of 14.02 BDT, which is projected to increase to 15.75 BDT by February 2025.
Table 8. Tariff based on different fuel type plants.
| Generation fuel type | Tariff (Tk/kWh) |
| Gas turbine power plant | 3-3.5 |
| Coal power plant | 12-13.8 |
| Heavy fuel oil power plant | 17 |
| Imported power from India | 15.75 |
| Diesel based Power plant | 37 |
| Govt. agreement with WMSPL | 29.82 |
| Proposed system (PV-diesel) | 24.12 |
Gas turbines, coal, and heavy fuel power plants are not feasible for Manpura Island due to capacity and transportation issues. The proposed renewable-based hybrid system has a lifetime cost of 24.12 BDT (0.201 $/kWh), which is higher than some existing power generation plants in Bangladesh. This is attributed to the use of diesel-based engines, higher initial capital costs, and the installation of advanced technology. Additionally, reliance on diesel generators can lead to significant fuel cost fluctuations, driving up the overall tariff. A substantial rise in diesel prices could increase the Cost of Energy (COE), while a decrease in solar irradiance could heighten the system’s dependence on diesel, thereby increasing operational costs. Smaller project scales may also result in less favorable economies of scale, leading to higher per-unit electricity costs.
Despite these challenges, the proposed system offers a lower COE compared to standalone diesel-based power plants and the power plant proposed by WMSPL. WMSPL’s integration of PV with BESS and a diesel engine results in a slightly higher COE than the proposed system. Given that a standalone diesel engine system in Bangladesh typically has an energy cost of around 35-40 BDT, the proposed COE of 0.201 $/kWh (24.12 Tk/kWh) can be considered an effective solution for electrifying this remote coastal region.
4.2 Technical analysis
To perform the technical analysis, we designed and simulated the system in MATLAB/Simulink. These simulations provide important information regarding the operational dynamics of the hybrid power system. This simulation environment allowed us to accurately model the hybrid power system, integrating components such as the Diesel generator, Solar PV panels, DC-DC boost converters, and inverters.
Figure 9. Actual schematic diagram of the proposed system.
The PV panels were integrated with MPPT algorithms to optimize their performance (Beriber and Talha, 2013). Figure 9 shows a schematic diagram of the proposed structure, which was designed in Simulink. The current-voltage (I-V) and power-voltage (P-V) curve of the PV array for different irradiances is displayed in Figure 10.
Figure 10. I-V and P-V characteristics curve of the PV array output.
4.2.1 PV output power
The simulation is performed at an irradiance of 1000 W/m² and a temperature of 25 °C. Since radiation levels change hourly or monthly, the system’s output will also vary (Madhukumar et al., 2020). Figure 11 illustrates the change in PV output with varying irradiance, simulated over a 1-minute period. During this simulation, the irradiance was gradually ramped up from 200 to 1000 W/m² and then ramped down to 300 W/m² every 0.1 seconds. This demonstrates that the output increases as irradiance increases. The system, designed for 1000 W/m², achieves its desired output of 3.6 MW at this irradiance, as observed between the 0.4 and 0.5-second marks. The fluctuations visible in the waveform are attributed to the sudden changes in solar irradiance.
Figure 11. PV array output power with variable solar irradiance.
The MPPT’s duty cycle continuously adjusts due to variations in irradiance and temperature. Figure 12 illustrates the duty cycle produced by the MPPT algorithm during a simulation at 1000 W/m2 and 25 °C. The plotted data demonstrates the MPPT’s responsiveness, ensuring the solar panels operate at their maximum power point despite fluctuations in solar radiation and temperature. The MPPT algorithm continuously adjusts the duty cycle to optimize power output.
Figure 12. Duty cycle generated by the MPPT to control the boost circuit.
The duty cycle generated by the MPPT is fed to the DC-DC boost converter as a gate pulse to the IGBT. The boost converter increases the voltage level achieved from the PV array. Figure 13 shows the boosted DC voltage. The voltage boosts to a higher value, nearly 1600V, for a few milliseconds and then gradually decreases. Due to the gate pulse generated by the MPPT and applied to the IGBT’s gate, the voltage waveform has some ripple. The magnitude of the converted voltage gradually decreases and reaches a desired steady-state value of nearly 600V, as displayed in the figure.
Figure 13. The output DC voltage of the boost converter.
The boosted DC voltage feeds into the next section, a DC-to-AC inverter. The output must be converted into AC to drive the island’s AC loads. Figure 14 displays one phase of the three-phase inverter’s output voltage. The inverter utilizes a PI-controlled PWM generator to trigger the gate of the switch. The inverter’s voltage waveform, generated using PWM gate pulses, consists of a series of high-frequency rectangular pulses that approximate a sinusoidal waveform.
Figure 14. The output voltage (single phase shown) of inverter.
These pulses switch between high and low states, with their width (duty cycle) modulated to match the desired amplitude of the sine wave at each moment. As the duty cycle increases, the average voltage during each pulse cycle increases, building the overall sine wave shape. Although the raw output appears as sharp rectangular pulses, the fundamental frequency and shape of the sine wave are embedded in the pattern. After passing through a filter, the sharp edges are smoothed out, producing a nearly sinusoidal waveform with reduced harmonic distortion, as shown in Figure 15.
Figure 15. Three phase output voltage after the LC filter.
Figure 16 displays the three-phase output current when connected to a 3.6 MW load. Initially, the graph shows fluctuations lasting approximately 0.01 seconds as the system stabilizes during the transient phase. After this period, the current reaches its steady-state value, demonstrating system stabilization under the load. The current magnitude is approximately 7000 A, indicating the system’s ability to efficiently handle heavy loads. This smooth transition from fluctuation to stability highlights the system’s reliability and performance under large-scale load conditions.
Figure 16. Three phase output load current (transient behavior shown within 0.01s).
4.2.4 FFT analysis
It is important to maintain the harmonic distortion in the current and voltage waveforms below the standard threshold (Mishan et al., 2025). This ensures the efficient operation of electrical equipment and minimizes losses due to heat and interference. A Fast Fourier Transform (FFT) analysis is performed on the output voltage and current waveforms, shown in Figures 17 and 18, to measure the harmonic distortion. The harmonic analysis of the output voltage shows that the total harmonic distortion (THD) is approximately 7.89%.
Figure 17. Analysis of FFT of three phase output voltage (THD of 7.89%).
The harmonic analysis of the output current, shown in Figure 18, reveals a total harmonic distortion (THD) of approximately 1.63%.
Figure 18. Analysis of FFT of three phase output current (THD of 1.63%).
4.2.5 Output power analysis
To ensure reliability in the energy supply, consistent power output is essential, particularly in off-grid applications like Manpura Island. The hybrid system’s performance requires continuous monitoring and optimization to address fluctuations in load demand and generation capacity. Figure 19 illustrates the hybrid system’s output power under fixed load conditions, simulated at 25°C with an irradiation of 1000 W/m². The graph shows that the output power is lower for the initial 0.06 seconds before stabilizing at 3.17 MW. Although this is slightly below the desired output, it remains sufficient to meet the island’s demand.
Figure 19. Output power of the proposed hybrid system (calculated at load terminal).
After the power is measured at constant conditions, two cases are considered. Case 1: The temperature is kept constant, and the radiation is variable. Starting at 1000 W/m², the radiation is ramped down to 200 W/m² and then ramped back up to 1000 W/m². The temperature remains at 25 °C until the simulation concludes. Figure 20 shows that the output power varies with irradiance, decreasing as irradiance reduces and increasing as it increments.
Figure 20. Variation of output power with solar irradiance (variation of PV generation).
In case 2, the irradiance is kept constant while the temperature is varied. The temperature is ramped down from 25 °C to 10 °C, then ramped up to 50 °C, while the irradiance is maintained at 1000 W/m². Figure 21 illustrates that as the temperature drops, the power output rises, and vice versa. This demonstrates a non-linear relationship between temperature and output power.
Figure 21. Variation of output Power with Temperature (power generation fluctuation).
The proposed hybrid diesel generator and PV system experiences several losses that reduce its overall efficiency. The boost converter incurs various losses, including conduction, switching, core, gate drive, and capacitor losses. Similarly, the inverter also suffers from comparable losses, further diminishing the system’s output. Figure 22 illustrates the system’s input and output power. Despite an input of 3.6 MW from the PV, the output is 3.17 MW, resulting in a system efficiency of 88.05%. This indicates that approximately 88% of the input power is converted into usable output power, while about 12% is lost due to inefficiencies in the conversion process.
Figure 22. (a) DC input power from PV panel and (b) AC output power of the hybrid system.
5. Conclusions
The study focused on designing, simulating, and analyzing a 3.6 MW hybrid power plant (comprising PV panels and a diesel generator) for Manpura Island, showcasing its potential to meet local energy demands sustainably. Using MATLAB/Simulink and HOMER Pro, the analysis emphasized key techno-economic aspects such as energy production, renewable penetration, and cost-effectiveness. Results showed the system generates 15,087,243 kWh annually, with PV contributing 31.6% and a 1 MW diesel generator producing 33%. The PV generation cost was significantly lower ($0.0324/kWh) compared to diesel, which ranged from $0.209 to $0.212/kWh. The boost converter maintains the output voltage level at 600V. The results also showed that the output current waveform is almost distortion-less, with a THD of 1.63%, and the output voltage THD is 7.89%. Ultimately, the simulated system has an overall efficiency of 88.05%. This research provides empirical insights and design considerations for researchers and commercial stakeholders involved in designing hybrid power plants for remote islands. It offers actionable insights for decision-makers on implementing a hybrid power plant for Manpura Island, demonstrating the system’s adaptability to varying environmental conditions while ensuring a stable power supply.
