Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.
Advertisement
Scientific Reports volume 15, Article number: 39103 (2025) Cite this article
1761 Accesses
Metrics details
The adoption of photovoltaic systems is growing as part of the global shift to renewable energy. Integrating photovoltaic systems into buildings enhances sustainability by enabling on-site generation, reducing energy costs, and promoting environmental preservation. Ensuring long-term reliability requires a comprehensive analysis. This study analyzes a grid-connected photovoltaic system, operated and maintained by the Power Electronics and Renewable Energy Laboratory (PEARL) for research. The system consists of Poly-Crystalline (Array 1) and Mono-Crystalline silicon (Array 2) panels with a total capacity of 3.575 kWp, monitored over 36 months from January 2020 to December 2022. Eleven performance parameters were analyzed following the IEC 61724 guidelines. Additionally, Modified Akima cubic Hermite (MAKIMA) methods were utilized to forecast the degradation rate of each photovoltaic technologies degradation rates in 2023. The results indicate that Array 1 and Array 2 produced an average AC energy output of 3881.67 kWh and 1120.48 kWh, with 86.74% and 56.30% performance ratios, respectively. The forecasting analysis predicted degradation rates of 10.58% for Array 1 and 11.99% for Array 2. These findings provide valuable information on the durability and efficiency of PV materials in tropical climates, contributing to optimized system performance, longevity, and material selection in high-temperature, high-humidity environments.
The renewable energy (RE) sector has experienced remarkable growth in recent decades, emerging as a cornerstone in the global transition toward sustainable energy systems. Among the various RE sources, solar energy (SE) stands out as one of the most reliable and accessible options due to its scalability and consistent availability, particularly in regions with high solar irradiance^1,2. Its adaptability, from residential rooftops to large-scale installations, further underscores its growing prominence in modern energy landscapes^3,4. According to the International Energy Agency (IEA), global solar photovoltaic (PV) generation capacity reached 672.5 GW in 2022 and rose to 875.46 GW in 2023, reflecting an annual growth rate of approximately 23.18%⁵, with projections indicating sustained growth in the coming years.
In Southeast Asia, countries like Malaysia are positioning themselves as emerging contributors to the RE agenda. Malaysia’s Ministry of Natural Resources, Environment, and Climate Change (NRECC) has set a target of achieving a 31% RE share in the total installed capacity by 2025. The Malaysia Renewable Energy Roadmap (MyRER) outlines a strategic vision to reduce dependence on fossil fuels and accelerate the adoption of clean energy, primarily driven by the expansion of solar PV⁶. Looking ahead, Malaysia aims to achieve significant emissions reductions by 2030 and 2035 in alignment with its Nationally Determined Contributions (NDC) under the Paris Agreement, specifically a 45% reduction in greenhouse gas (GHG) emissions intensity by 2030 and a 60% reduction by 2035^7,8.
Extensive research has been conducted to assess the performance and economic feasibility of PV systems across diverse geographic locations characterized by tropical climatic conditions⁹, as well as a range of installation configurations^10,11. These studies have encompassed various PV system types, including on-grid, off-grid, residential, and large-scale configurations, each evaluated for its potential to optimize energy generation and cost-effectiveness^12,13,14. More recently, the integration of PV systems into building structures, commonly referred to as Building-Integrated Photovoltaics (BIPV), has emerged as a promising solution for urban energy generation. Given the spatial limitations commonly encountered in densely populated urban environments, BIPV systems offer a strategic alternative to traditional ground-mounted installations, thereby enhancing the feasibility of solar adoption in space-constrained settings¹⁵. Ghazali et al.¹⁶ investigated the integration of PV systems into the vertical facades of high-rise buildings in Malaysia. Utilizing the System Advisor Model (SAM) developed by the National Renewable Energy Laboratory (NREL), the study assessed five different design scenarios using Heterojunction Intrinsic Thin-film (HIT-Si) modules with a nominal efficiency of 15.6%. Results indicated that vertical PV systems could generate between 400 and 700 MWh annually, significantly surpassing conventional roof-mounted systems, which produce approximately 240 MWh/year. The economic analysis showed a payback period of around 12 years for vertical systems, compared to 6 years for roof systems. Saleheen et al.¹⁷ conducted a performance analysis of a 232.5 kWp grid-connected PV system installed on the rooftop of Monash University, Malaysia, using 5-minute interval technical and meteorological data collected in 2019. The system generated approximately 301.5 MWh, closely matching the target of 305.0 MWh. Key performance indicators included a performance ratio (PR) of 85.4%, a capacity utilization factor (CUF) of 14.85%, and a levelized cost of energy (LCOE) of 0.396 MYR/kWh, along with a carbon emission reduction of 177 metric tons of carbon dioxide (CO₂). Akhter et al.¹⁸ evaluated the performance of a grid-connected PV system in Kuala Lumpur, comparing three PV technologies: Poly-crystalline (p-Si), Mono-crystalline (m-Si), and Amorphous silicon (a-Si). Using data from January 2016 to December 2019, the study demonstrated the system’s potential to mitigate over 28,000 kg of CO₂ emissions.
Anang et al.¹⁹ assessed a 7.8 kWp grid-connected rooftop PV system in Kuala Terengganu, Malaysia, operating under the feed-in-tariff scheme from 2018 to 2019. The study using PVsyst software uncovered that an optimal tilt angle of 5° could enhance annual energy production by 4.8%. The economic analysis of the system demonstrated that it generates an estimated annual net profit of RM 3,000, with a payback period ranging from 5 to 7 years, alongside an estimated CO₂ reduction of 7.45 tons per year. De Lima et al.²⁰ analyzed a 2.2 kWp PV system installed at the State University of Ceará, Brazil, operating from June 2013 to May 2014. Under tropical savanna conditions, the system achieved an annual energy yield of 1685.5 kWh/kWp, an average daily final yield of 4.6 kWh/kWp, a PR of 82.9%, and a capacity factor (CF) of 19.2%. Singh et al.²¹ presented a study in the moderate climatic region of Imphal, Manipur, India. Simulation using PVsyst and comparison with measured data yielded a PR of 74.4%, a CUF of 14.31%, and an average PV system efficiency of 14.4%.
Saxena et al.²²investigated the feasibility of grid-tied rooftop PV systems across seven Indian cities. A 100-kWp system was simulated under local climatic conditions, revealing a PR between 70% and 80%, a CUF between 19% and 21%, an estimated annual output of approximately 170 MWh, and a payback period of 5 to 6 years. Gopi et al²³. analyzed the operational performance of a 2 MWp photovoltaic plant in Kuzhalmannam, Kerala, over two years (2018–2019). The tropical monsoon climate caused energy generation to decrease by about 35% during the monsoon season. The plant’s overall PR averaged 73.39%, with a CUF of 15.41%. A summary of additional previous research studies situated in tropical climates is provided in Table 1. These studies offer broader insights into system performance, environmental impacts, and degradation analysis across diverse tropical regions.
Despite these valuable insights, a critical gap persists in long-term, field-based assessments of PV performance in tropical climates like Malaysia’s. Many existing studies, such as those conducted in Malaysia, Nigeria, and other regions, emphasize short-term data or focus solely on economic returns. These studies often lack comprehensive evaluations of system degradation and operational efficiency over time, as seen in cases where degradation analysis was not conducted. Furthermore, few studies integrate forecasting models to predict performance decline, limiting their utility in long-term planning. To bridge this gap, the present study conducts a rigorous 36-month field evaluation of a grid-connected PV installation in Malaysia, comparing the performance of Mono-crystalline and Poly-crystalline technologies under identical environmental conditions. This study provides novel insights into the performance and degradation of grid-connected PV systems in tropical climates, particularly in Malaysia. The following key contributions of this study are outlined as follows:
A direct comparative analysis is conducted between two widely utilized PV technologies (Mono-crystalline and Poly-crystalline systems) by evaluating their performance under identical environmental conditions. By eliminating external variables, the benchmarking approach ensures an objective and unbiased assessment, thereby providing a comprehensive understanding of the relative performance of both technologies.
A thorough evaluation is performed using eleven core performance metrics, as defined by the IEC 61,724 standardized methodology. This methodology ensures the robustness, international comparability, and reliability of the results, thereby offering a valuable reference for future global investigations into PV system performance.
The Modified Akima Cubic Hermite Interpolation (MAKIMA) algorithm is utilized to model the future degradation of both PV systems. This methodology introduces an innovative approach to degradation profiling, specifically tailored to PV systems operating under tropical climatic conditions.
The reliability of the degradation forecasts is rigorously assessed through quantitative validation techniques. This validation process strengthens the credibility of the predictive model, positioning it as a practical tool for asset management and strategic decision-making in the renewable energy sector.
This paper is structured as follows: Sect. 2 outlines the PEARL PV system and details the eleven performance evaluation criteria. Section 3 presents and analyzes the results, drawing comparisons with existing literature. Finally, Sect. 4 concludes with a summary of key findings and their broader implications.
This section describes the technical data and specifications of the PEARL grid-connected PV system and the list of performance analysis parameters.
The PEARL PV system is a grid-connected photovoltaic system that was installed in October 2015 and has been operating continuously since its installation. It is situated on the rooftop of the Engineering Tower Building at the Universiti Malaya, with geographical coordinates of 3.07° north latitude and 101.39° east longitude. The PV system is operated and maintained by the Power Electronics and Renewable Energy Laboratory (PEARL) for research purposes and comprises two distinct technologies: Array 1, consisting of Poly-Crystalline silicon (2 kWp), and Array 2, consisting of Mono-Crystalline silicon (1.875 kWp). The combined total output power of the system is 3.875 kWp. Both arrays are fixed in a south-facing azimuth orientation, with optimal inclination angles of 10° for Array 1 and 10° for Array 2, respectively. A description of the PEARL PV system is presented in Table 2, while the schematic overview of its connection layout is depicted in Fig. 1.
Overview of the connection scheme of the grid-connected PEARL PV system.
Array 1 consists of sixteen poly-crystalline modules (serial number PV-AE125MF5N), each with a rating of 125 Wp at Standard Test Conditions (STC), resulting in a total capacity of 2 kWp. Array 2 comprises twenty-five Mono-crystalline modules (serial number SQ75), each with a rating of 75 Wp at STC, yielding a total capacity of 1.875 kWp. Figure 2 illustrates the installed Array 1 and Array 2 at the PEARL system site, representing the Poly-Crystalline and Mono-Crystalline array systems, respectively.
Layout PEARL’s PV system on site: (a) Array 1 and (b) Array 2.
The PEARL PV system is connected with two inverters via string connection. The two inverters with a rated power of 1600 W are connected to Array 1 and Array 2. Table 3 shows a description of PEARL’s photovoltaic system’s inverters.
The PV system is integrated with the SMA Sunny Sensor Box, which continuously monitors system performance and collects meteorological data, including solar irradiance, ambient temperature, module temperature, and wind speed. The system is powered by the SMA Power Injector, enabling seamless communication between the sensor box and the SMA Sunny WebBox via a communication bus. Data is retrieved through the Sunny Portal, with selectable time intervals of 5, 15, or 30 min.
To ensure the reliability and accuracy of the analysis, a systematic quality check was performed on the collected dataset prior to further processing. First, the raw power data was extracted and converted into energy values based on the associated timestamps to ensure proper time alignment and integration. All monitoring data used in this study were collected at 15-min intervals throughout the entire assessment period. Although the original system was configured to record at 5-min resolution, data inconsistencies in the 2020 dataset, specifically its availability only at 15-min intervals, necessitated a standardized approach. To ensure methodological consistency and enable valid year-to-year comparisons, all data were retrieved and analyzed at 15-min intervals.
Additionally, the analysis was limited to data within the time window of 07:00 to 19:00, which corresponds to the typical range of daylight hours in Malaysia. Sunrise in the region typically occurs between 06:50 and 07:30, while sunset ranges from 18:50 to 19:30. Selecting this range ensures the inclusion of meaningful daylight generation data while avoiding periods with zero irradiance and power values due to darkness.
To further preserve the integrity of the dataset, no data imputation, filtering, or interpolation was performed. This approach maintains the authenticity of the original measurements and enhances the scientific validity of the results. By using the raw data in its original form, the findings of this study provide a reliable and transparent reference point for future research, particularly for comparative assessments across different climatic conditions.
Both systems, Array 1 and Array 2, were evaluated using standard performance indicators such as yield, performance ratio, and capacity factor. These metrics were calculated based on definitions provided in IEC 61724-1:2017³⁴, which serves as a reference for PV system performance analysis. While the methodology follows IEC-defined formulas, the study does not claim full compliance with the standard due to limitations in sensor uncertainty data and system classification. The performance assessment also draws on analytical approaches from Akhter et al.¹⁸ and de Lima et al.²⁰, which are widely used in similar research.
Energy output for alternating current (E_AC) is defined as the cumulative energy output of a system over a specified period. In conjunction, the E_DC is the ratio between the energy output of AC and the inverter efficiency, while the energy output of DC represents the actual performance of the PV module/array. Both energy outputs can be computed using (1) and (2), respectively.
Where n represents the total number of 15-min interval over the study period, and (:varDelta:t) is the duration of each interval.
The yield assessment comprises three distinct yields, each defined by its specific purpose. The reference yield (Y_R) is determined by dividing the total in-plane irradiation (H, Wh/m²) by the reference irradiance of the PV system (G, W/m²). Here, H represents the solar irradiation on the module plane, while G_STC denotes the irradiance under standard test conditions (1,000 W/m²). (3) outlines the method for calculating Y_R.
The array yield (Y_A) is defined as the ratio of the DC energy generated by the module to the nominal power of the installed PV system. It is representing the equivalent number of hours per day that the module operates at its maximum power. The installed PV system capacity (P_PV, kW_p) is used in this calculation, with (4).
The final yield (Y_F) is defined as the ratio of the AC energy output to the rated capacity of the installed PV system. This metric is critical for assessing the system’s performance from a utility perspective, with using (5).
The PR is a critical metric for evaluating the efficiency of the PV system, independent of geographical location. It quantifies the relationship between the actual energy output of the PV plant and the theoretical energy output that would be generated if the system operated continuously at its nominal efficiency under STC. The formula for PR is provided in (6).
The CF is a parameter that quantifies the operational duration of a power plant over a specified period. It is calculated by dividing the actual energy output by the total possible energy output during the period, which is based on the rated AC capacity (P_AC) multiplied by the total number of operation hours, in this case, approximately 36 months. A capacity factor of 100% signifies that the plant operates continuously at its rated capacity. The formula for CF is presented in (7).
where M is the number of days in the measuring year.
Determining the efficiency of PV modules is essential as it relates to the quality of modules. The same applies to inverter performance and overall system efficiency. Typically, efficiency is used as a general measurement for comparison between PV plants. (8), (9), and (10) are the formulas to determine array, system, and inverter efficiency, respectively.
where A_m represents the surface area (m²) of modules for both Array 1 and Array 2 systems.
The array losses (L_A) represent the losses due to array operation that highlights the inability of an array to fully utilize the available irradiance. While the system losses (L_s) are due to losses in converting the DC power output from DC to AC power by the inverter²⁴. L_A and L_s are calculated using (11) and (12).
Modified AKIMA Interpolation (MAKIMA) is an enhanced version of the original AKIMA interpolation method introduced by Hiroshi Akima in 1970³⁵. The original method was designed to construct cubic splines that closely mimic the appearance of natural, hand-drawn curves. MAKIMA extends this approach by employing cubic Hermite polynomials to interpolate between non-equidistant data points and introduces a geometric slope adjustment to address edge cases where both the numerator and denominator may be zero, thereby enhancing numerical stability³⁶. MAKIMA is particularly effective for handling oscillatory or irregularly spaced data, outperforming conventional interpolation methods that often struggle to maintain a flat or natural response under such conditions. This makes it highly suitable for forecasting degradation in PV systems, where efficiency data often exhibits short-term fluctuations and nonlinear trends. In this study, MAKIMA was applied to forecast PV degradation patterns observed over the period 2020 to 2022, using 2019 as the baseline year. The overall flow of the MAKIMA-based degradation estimation and forecasting process is illustrated in Fig. 3.
The flowchart of the proposed MAKIMA model for forecasting degradation rate.
In this context, the relative degradation rate of the PV module is evaluated based on the change in system efficiency over time. It is calculated with respect to the baseline year (2019) using (13).
Where (:{{eta:}_{sys}}_{y}) represents the system efficiency in year and y indicates the assessed year.
The application of MAKIMA interpolation was conducted through MATLAB’s built-in function, which internally uses the Modified Akima method³⁷. Although our manuscript did not derive this interpolation from first principles, MATLAB documentation clearly describes this mathematical foundation. Given that PV efficiency may fluctuate non-linearly, MAKIMA provides an appropriate framework for modeling such behavior. It begins with a dataset of points that represents time and denotes the corresponding PV system efficiency. To maintain smooth interpolation, MAKIMA calculates local slopes ((m_{i})) using a weighted average of neighboring differences in (14).
Where (:{m}_{i}) denoted the slope between adjacent efficiency values, (:w) is the weight factor determined by the absolute difference in adjacent slopes.
In the case where (:left| {m_{{i + 1}} – m_{i} } right|) is approximately zero, a geometric correction strategy is applied to avoid numerical instability. This may involve using one-sided differences or an averaged fallback slope. If the correction does not produce satisfactory results, the geometric correction process is repeated with adjusted weights or alternative estimates until convergence is achieved. This formulation prevents instability, particularly division by zero, by applying a geometric correction. Once the slopes are computed, MAKIMA constructs a cubic Hermite polynomial for interpolation over the interval [(x_{i} ,x_{i} + 1)] using (15).
The Hermite basis functions are defined in (18) to (21), ensuring smooth transitions and continuity in both the function values and their first derivations across each subinterval. These properties make the interpolation particularly robust for modeling nonlinear or oscillatory degradation behaviors.
Accurate degradation forecasting is essential for evaluating the long-term performance of PV systems at specific locations. This includes identifying defects or material-related factors that contribute to reduced energy output over time. MAKIMA’s precision and adaptability make it a reliable tool for supporting such performance assessments³¹.
To evaluate the forecasting performance of the MAKIMA model, metrics such as root mean square (RMSE), mean absolute error (MAE), mean square error (MSE) and mean absolute percentage error (MAPE) are employed. These metrics assess the error between the real and predicted value and determine the accuracy of MAKIMA model in calculating the degradation rates of both PV arrays in 2023. The formula for each metric is defined as follows using (22), (23) and (24):
where (S) and (:mathop Slimits^{ – }) represent the real and predicted value of the degradation rate, m represents the number of sample and (:i) is a subscript indicating individual data points being used in the calculation of the error.
The subsequent section provides a comprehensive analysis and evaluation of eleven performance metrics and the degradation rate of the proposed forecasting method known as MAKIMA on both Array 1 and Array 2, respectively.
The monthly average output for 2020, 2021, and 2022 are shown in Fig. 4a, Fig. 4b, and Fig. 4c, respectively, while the average of module temperature are depicted in Fig. 4d, Fig. 4e, and Fig. 4f. The annual average (:{E}_{AC}) and the module temperature across the assessment period are illustrated in Fig. 5a and Fig. 5b, respectively.
For Array 1, the highest outputs consistently occurred during the first half of the year (January to May) and October across all three years, while the lowest outputs were observed in June, July, August, September, November, and December. Specifically, in 2020, January saw the peak output (222.36 kWh), and June recorded the lowest output (176.15 kWh). In 2021, February reached the highest output (225.09 kWh), with August as the minimum (162.36 kWh). In 2022, the highest output occurred in January (202.36 kWh), with August showing the lowest (102.13 kWh). The annual average outputs were 200.88 kWh in 2020, 189.66 kWh in 2021, and 183.38 kWh in 2022.
For Array 2, the highest outputs were recorded in January, February, March, May, October, and December, while the lowest outputs were in April, June, July, August, September, and November. In 2020, January had the highest output (113.91 kWh), and June had the lowest (95.80 kWh). In 2021, October recorded the peak output (109.00 kWh), with September showing the lowest (82.99 kWh). In 2022, the highest output occurred in December (102.92 kWh), while the lowest was in August (61.50 kWh). The annual averages were 109.93 kWh in 2020, 97.64 kWh in 2021, and 93.42 kWh in 2022.
Comparison of AC energy production (a) to (c) and monthly average module temperature (d) to (f) for 2020, 2021, and 2022.
The monthly average module temperatures over the period from 2020 to 2022 displayed in Fig. 4d to Fig. 4f shows a clear correlation between monthly solar irradiation and module temperature across the three-year assessment period. In general, months with higher solar irradiation corresponded with elevated module temperatures, reflecting the direct influence of solar energy absorption on the thermal behavior of PV modules. In 2020, March recorded one of the highest irradiation values (126.97 kWh/m²), coinciding with the peak module temperature of 40.33 °C in April, likely due to thermal inertia and sustained irradiance over the preceding weeks. Similarly, in 2021, February showed the highest irradiation level (130.25 kWh/m²), aligning with a notable module temperature of 39.66 °C. A comparable pattern was observed in 2022, where March received 122.08 kWh/m² of irradiation and exhibited a module temperature of 39.23 °C. Conversely, months with lower irradiation, such as August and September in 2022, when irradiation dropped to 60.47 kWh/m² and 71.29 kWh/m², corresponded with the lowest module temperatures (35.98 °C and 34.97 °C). These findings reinforce the well-established relationship between irradiance and module temperature, whereby higher solar input intensifies thermal loading on the module surface. However, interannual variability, particularly in 2021, also suggests that other meteorological factors such as ambient temperature, wind speed, and cloud cover, can modulate this relationship, occasionally leading to deviations from a strictly linear trend.
Annual assessment of (a) average output of AC energy and (b) average module temperature.
Figure 5a illustrated the overall average AC energy output from Array 1 was 2.24 MWh, while Array 2 generated 1.18 MWh, resulting in a combined total of 3.42 MWh during the assessment period. Both arrays consistently achieved their highest outputs between January and March, with the lowest recorded in August and September, particularly in 2021 and 2022. This drop aligns with Malaysia’s monsoon season, marked by increased cloud cover and rainfall, which diminishes solar radiation. A notable trend is the decline in energy output over the three years. Array 1’s output fell from 200.88 kWh in 2020 to 183.38 kWh in 2022, while Array 2’s output decreased from 109.93 kWh in 2020 to 93.42 kWh in 2022. This reduction reflects growing climatic variability, potentially influenced by cloud cover, rainfall, and other atmospheric factors.
Figure 5b presents the monthly average module temperatures for the evaluation year ranged from 35.58 °C to 39.18 °C, consistent with the thermal behavior expected in PV systems. Throughout the year, higher levels of solar irradiation generally coincided with elevated module temperatures, indicating that solar energy input is a primary driver of thermal loading on PV modules. March recorded one of the highest irradiation levels (125.65 kWh/m²), with the module temperature peaking at 39.18 °C. Similarly, January and February exhibited high irradiation values of 124.29 kWh/m² and 121.35 kWh/m², respectively, which corresponded to elevated module temperatures of 37.06 °C and 38.48 °C. Conversely, the lowest module temperatures were observed in December (35.58 °C) and November (36.26 °C), aligning with moderate irradiation levels of 106.42 kWh/m² and 105.08 kWh/m². Overall, the thermal profile reflects typical seasonal variation, with elevated temperatures during the first half of the year and a gradual decline toward the end. These trends are particularly relevant for PV system performance, as sustained high module temperatures can reduce energy conversion efficiency due to the negative temperature coefficient of power output.
The solar irradiation production throughout the assessment period.
The annual global horizontal irradiation (GHI), depicted in Fig. 6, decreased over the same period from 1357.37 kWh/m² in 2020 to 1233.83 kWh/m² in 2022, showing a clear negative correlation between solar irradiation and energy generation, especially in June through September. The sharp decline in GHI in August 2022 (60.47 kWh/m²) and September (71.29 kWh/m²) is particularly significant. The impact of high rainfall and cloud cover during these months demonstrates the challenge posed by Malaysia’s monsoon season. The reduced GHI levels resulted in a substantial drop in energy output, with Array 1 generating just 102.13 kWh and Array 2 producing only 61.50 kWh in August 2022. This suggests that seasonal weather patterns play a critical role in solar power generation, where periods of high cloud cover, especially during the monsoon, can severely limit energy capture.
The data shows that the lowest energy outputs typically occurred during the Southwest Monsoon period (from late May to September), which is associated with increased cloud cover and heavy rainfall, reducing solar radiation. In contrast, the Northeast Monsoon (from November to March) and inter-monsoon periods (from April to May and October to November) are characterized by clearer skies and less rainfall, resulting in higher solar radiation and consequently higher energy outputs. This seasonal and transitional weather pattern highlights the significant impact of local climatic conditions on solar energy generation, particularly in tropical regions like Malaysia. Therefore, when designing solar PV systems in such environments, it is essential to consider these seasonal variations. Additionally, integrating alternative energy sources or enhancing energy storage systems to manage the effects of cloud cover and rainfall during the monsoon and inter-monsoon periods can help ensure a stable and reliable energy supply.
Figure 7 presents the monthly average array yield over the assessment period, revealing significant seasonal variations in energy production. The highest average array yields for both arrays were observed in January, February, March, and October, while August recorded the lowest values. Specifically, the average array yield for Array 1 ranged from 87.43 h to 118.55 h, while Array 2 exhibited yields between 50.13 h and 65.52 h. The cumulative array yields for 2020, 2021, and 2022 showed a progressive decline for Array 1, from 1295.56 h to 1133.14 h, and for Array 2, from 736.89 h to 668.61 h.
Monthly average array yield during the period of assessment.
This image is for Fig. 2aFigure 8 further illustrates the monthly average final yield, with Array 1 exhibiting its highest average final yield in January (105.08 h), February (104.34 h), and March (107.36 h), and its lowest in August (79.37 h) and September (81.09 h). The monthly average final yield for Array 1 ranged from 79.37 h to 107.36 h, with total yields of 1201.51 h in 2020, 1123.95 h in 2021, and 1035.98 h in 2022. In contrast, for Array 2, the lowest monthly average yields were recorded in February (56.27 h) and May (56.99 h), with the lowest yield of 44.25 h occurring in September. The monthly average array yield for Array 2 ranged from 44.25 h to 56.99 h, and its total yield was 674.51 h in 2020, 625.75 h in 2021, and 582.83 h in 2022.
Monthly average final yield during period of assessment.
The reference yield values obtained during the assessment period highlight significant seasonal variability. Peak yields were observed in January (124.29 kWh), February (121.35 kWh), and March (125.65 kWh), which can be attributed to the optimal conditions for energy generation during the Northeast Monsoon. This period is characterized by clearer skies and reduced rainfall, resulting in a more favorable alignment of actual solar irradiance with the system’s theoretical capacity. In contrast, the lowest reference yield was recorded in August (90.32 kWh), reflecting the impact of the Southwest Monsoon, during which increased cloud cover and reduced sunlight significantly limited the PV system’s potential output.
This analysis underscores the profound influence of climatic and seasonal factors on the performance of solar PV systems. Higher yields in the first quarter of the year demonstrate the advantage of clearer skies and less precipitation, while the significant drop in mid-year energy output highlights the constraints imposed by the monsoon seasons. These findings are crucial for the design and optimization of solar energy systems in tropical climates, such as Malaysia, where integration of complementary energy sources or advanced energy storage systems may be necessary to ensure a consistent and reliable energy supply throughout the year.
The monthly CF for both Array 1 and Array 2, illustrated in Fig. 9, offers critical insights into the operational efficiency of the PV systems. The highest CF values were recorded between January and May, with values ranging from 10.22% to 11.81%, indicative of favorable weather conditions that facilitated optimal energy generation. In contrast, the lowest CF was observed in August and September, recorded at 8.39% and 8.70%, respectively. Notably, a marked improvement of 1.64% in capacity factor was observed from September to October, highlighting the transition to the inter-monsoon phase, which brings clearer skies and more consistent sunlight. This shift underscores the importance of understanding how seasonal transitions impact energy output, as changes in cloud cover and precipitation directly affect solar radiation availability. February emerged as the month with the highest capacity factors for both arrays, with Array 1 recording 15.34% and Array 2 8.27%. These values reflect optimal energy generation conditions, attributed to stable, clear weather. Conversely, August showed the lowest capacity factors, with Array 1 at 10.67% and Array 2 at 6.11%.
Annually, the overall average monthly CF for Array 1 was 12.79%, compared to 7.16% for Array 2. This discrepancy highlights the impact of system design, particularly the material and efficiency of the panels, in determining overall performance. Array 1’s higher CF suggests that Polycrystalline panels may offer superior performance in fluctuating weather conditions compared to Array 2’s potential use of Monocrystalline panels. This comparison emphasizes the importance of carefully selecting panel types and system configurations to optimize performance in regions subject to seasonal variations and unpredictable weather patterns. The findings reinforce the critical influence of local weather patterns on solar PV performance, particularly in tropical regions where shifts between rainy and dry seasons significantly impact energy output.
Monthly average of capacity factor generated for two arrays.
Figure 10 illustrates the monthly averages of the PR for both arrays. For Array 1, the highest PR was recorded in June at 93.51%, suggesting optimal system performance during this month. In contrast, Array 2 achieved its highest PR in December at 62.45%, which can be attributed to relatively stable weather conditions conducive to efficient energy conversion. On the other hand, the lowest PR for Array 1 was observed in November (81.13%) and December (81.06%), while Array 2 exhibited its lowest PR values in March (51.87%) and April (51.03%). These lower PR percentages indicate a potential deterioration in system efficiency, likely caused by less favorable weather conditions, including increased cloud cover, reduced sunlight, and possible fluctuations in temperature. The annual average PR for Array 1 and Array 2 were 87.81% and 45.55%, respectively, reflecting a notable difference in their overall performance. The higher PR of Array 1 highlights the influence of system design and possibly panel quality, which contribute to its ability to perform better under a wider range of conditions. The overall PR of the system, at 82.9%, serves as a benchmark for evaluating the overall performance of the PV system.
When comparing the performance of the PEARL PV system with other similar studies, it is clear that Array 1 demonstrates superior performance with a PR of 87.81%. This figure surpasses the PR typically reported in similar settings, reflecting the optimized design of the system, which likely includes high-quality panels, advanced inverters, and a strategically chosen installation site. This comparison further emphasizes the importance of not only the geographical location but also the climatic factors and system maintenance protocols when designing and evaluating PV systems.
Monthly average on PR for two arrays during the period of assessment.
Figure 11 presents the monthly average efficiencies for the array, system, and inverter components of both Array 1 and Array 2, as depicted in Fig. 11a, Fig. 11b, and Fig. 11c, respectively. These efficiency parameters are crucial indicators of the overall performance and operational effectiveness of the PV system throughout the assessment period.
Monthly average efficiency of two arrays for the (a) array, (b) the system, and (c) inverter.
Figure 11a illustrates the monthly average array efficiency, revealing that the overall efficiency reached its peak between May and October, with values ranging from 9.06% to 9.71%. The lowest efficiencies occurred in January (8.52%) and December (8.50%). Array 1 demonstrated its highest efficiency during the summer months, specifically from May to September, with a peak of 12.54% in June, while its lowest efficiency was recorded in December at 10.66%. Array 2 exhibited its highest efficiency during the same period, although its lowest efficiency was observed in April (5.70%). The positive trend in array efficiency from April to June is indicative of the system’s optimal performance under conditions of increased sunlight exposure, minimal cloud cover, and moderate ambient temperatures. Notably, Array 1 consistently outperformed Array 2, highlighting its superior ability to convert solar energy into electrical power under varying environmental conditions.
Figure 11b depicts the monthly average system efficiency, which reflects the overall capacity of the PV system to convert solar energy into usable electricity. Both Array 1 and Array 2 achieved their highest system efficiency in June at 9.03%. For Array 1, system efficiency ranged from 10.02% to 11.61%, while for Array 2, it fluctuated between 5.26% and 6.45%. A notable efficiency increases of 9.5% from April to May underscores the significance of enhanced sunlight exposure and favorable environmental conditions. Despite fluctuations in solar radiation, both arrays maintained relatively stable system efficiencies, demonstrating the robustness of the system to compensate for seasonal and environmental variations. These results emphasize the importance of optimal sunlight exposure, moderate ambient temperatures, and low humidity levels in maintaining high system efficiency throughout the operational period.
Figure 11c illustrates the monthly average inverter efficiency for both arrays. Array 1 consistently exhibited higher inverter efficiency than Array 2, with a peak efficiency recorded in December (94.41%) and the lowest recorded in March (89.22%). Efficiency reductions were observed during the periods from January to March and from September to October, with decreases of −3.21% and − 1.12%, respectively. Conversely, efficiency improvements were noted from April to May and from November to December, with increases of 3.13% and 5.22%, respectively. The annual average inverter efficiency for Array 1 was 92.09%, while for Array 2, it was slightly lower at 91.08%. The lowest inverter efficiencies for the year were observed in October (89.84%) for Array 1 and in March (86.17%) for Array 2, further suggesting the seasonal influence on inverter performance. Interestingly, both arrays exhibited similar efficiency trends in April, May, August, and September, with December emerging as the month with the highest inverter efficiency, likely driven by favorable weather conditions that optimize solar energy capture.
The findings from the study by Ketjoy et al.³⁸ corroborate the observed seasonal variations in inverter efficiency. The research, which evaluated inverter performance over three years, reported minimal degradation, with only a −1% reduction in efficiency due to the inverter being housed in an air-conditioned environment. Furthermore, their study highlighted March as a period of optimal inverter efficiency, aligning with the highest solar irradiation and longest sunlight duration. The efficiency analysis presented in this study demonstrates that both Array 1 and Array 2 exhibit performance patterns that are strongly influenced by environmental conditions, with Array 1 consistently outperforming Array 2 across all efficiency metrics. The results further emphasize the importance of considering seasonal variations, solar radiation, temperature fluctuations, and system design when evaluating the performance of PV systems.
Figure 12 shows the monthly average array losses for two arrays. Both Array 1 and Array 2 experienced their highest losses in December, recording 15.03 h and 69.96 h, respectively. The lowest losses for both arrays were observed in September, with values of 0.20 h for Array 1 and 44.16 h for Array 2. On average, system losses for Array 1 and Array 2 were 6.48 h and 55.52 h, respectively. A significant discrepancy is evident, with Array 2 consistently incurring higher losses than Array 1. While Array 1’s losses ranged from 0.20 h to 15.03 h, Array 2 exhibited more pronounced fluctuations and higher peaks. The lowest value for Array 1 in September is attributed to the seasonal effects of the monsoon, where high humidity and rainfall lower ambient temperatures. This reduction in temperature has been proven to enhance PV module performance by improving efficiency and reducing array losses. Studies by Akhter et al.¹⁸ and Adaramola et al.³⁹ show that solar radiation levels are typically lower during periods of high humidity, which often coincide with heavy rainfall. Additionally, the accumulation of water droplets on the PV modules from rainfall and the humid environment aids in cooling the modules by facilitating heat transfer through evaporation, further improving array performance during the rainy months.
Monthly average array losses for two arrays.
The observed high losses in December for both arrays, particularly for Array 2, warrant a closer examination of the environmental conditions during this period. December typically corresponds with the tail end of the monsoon season, suggesting that while the weather may stabilize, other factors such as dust accumulation or micro-climatic effects might contribute to the increased losses observed. Additionally, while Array 1 shows relatively consistent performance throughout the year, the higher losses in Array 2 suggest possible manufacturing mismatch issues, which have been documented in similar studies conducted by Amin et al.⁴⁰ and Chekal Affari et al.⁴¹. These discrepancies are often linked to manufacturing defects and aging, as noted by the authors, but this study identifies aging as a primary cause of high array losses. The impact of such aging, especially without regular maintenance and cleaning, could lead to a degradation in the module’s ability to convert sunlight into energy, compounding overall system losses.
Monthly average system losses for two arrays.
The overall average system losses for Array 1 (7.98 h) and Array 2 (8.00 h) in Fig. 13 highlight the performance gap between the two arrays. Both arrays experienced the highest losses in September (10.01 h) for Array 1 and in March (12.15 h) for Array 2. These high losses are likely influenced by particulate matter or reduced solar radiation during seasonal transitions. This suggests that the system’s sensitivity to environmental shifts warrants further monitoring, particularly during such periods, to better capture these effects. The lowest losses occurred in December, with 5.16 h for Array 1 and 6.01 h for Array 2, possibly due to the monsoon’s cooling effect. However, this also suggests that factors like inverter performance and installation quality, such as cabling, could be influencing system efficiency. The discrepancies between arrays may not be entirely environmental but could also reflect differences in maintenance and system design.
Figure 14 presents the yearly efficiencies for Array 1 and Array 2 from 2019 to 2022, demonstrating a clear decline in performance over time. In Fig. 14a, the yearly average efficiency of Array 1 decreases almost linearly from 2019 to 2021, followed by a sharp drop of −3.35% between 2021 and 2022. Similarly, as depicted in Fig. 14b, Array 2 shows a gradual decrease in efficiency from 2019 to 2022, with a notable decline of −3.68% between 2021 and 2022. The yearly efficiency values for both arrays over the four-year assessment period are provided in Table 4. The primary cause of the efficiency decrease is system aging, but other contributing factors, such as maintenance practices, UV exposure, and environmental conditions⁴². In this study, system aging emerged as the predominant factor contributing to the observed degradation. However, the role of maintenance, particularly regular cleaning and inspection, must not be overlooked. Inadequate maintenance can accelerate the degradation process, as dirt and debris accumulation on panels can reduce light absorption, further lowering system performance. Thus, the study highlights the importance of developing more effective maintenance strategies, especially in regions prone to high dust accumulation and humidity.
Figure 15a and Fig. 15b illustrate the relative degradation rates for both arrays, defined as the deviation in system efficiency from its 2019 baseline. Over the four-year assessment period, Array 1 exhibited an average annual degradation rate of 2.22%, while Array 2 showed a slightly higher rate of 2.54%. The more significant degradation in Array 2 can be attributed to material aging, which reduces the overall efficiency of the PV system. Over time, components such as encapsulants, glass, cells, and junction boxes experience a decline in their optical and mechanical properties⁴². The exposure of the arrays to environmental factors, such as fluctuating temperatures, irradiance levels, humidity, and physical stress, further accelerates degradation^43,44.
Efficiency of both arrays during the four years of assessment.
(a) Relative degradation and forecasted data in Array 1. (b) Relative degradation and forecasted data in Array 2.
An additional evaluation was performed to forecast the relative degradation rate for 2023 using the MAKIMA method. The forecast predicted a degradation rate of 10.58% for Array 1 and approximately 12% for Array 2. These predicted degradation rates suggest that the efficiencies of both PV systems could decrease by 9.94% for Array 1 and 5.31% for Array 2. The MAKIMA method was also assessed for forecasting accuracy using performance errors such as RMSE, MSE, MAE, and MAPE. The RMSE for Array 1 was 0.011, the MAE was 0.016, and the MAPE was 0.78%, while for Array 2, the RMSE was 0.013, the MAE was 0.01, and the MAPE was 0.35%. These results demonstrate that the MAKIMA method is highly effective for short-term forecasting, with predicted values closely aligning with actual data.
The degradation of PV modules represents a significant challenge for the solar energy industry, influencing material selection, manufacturing processes, and equipment aging⁴⁵. The NREL reports an average degradation rate of less than 1% per year for PV panels. However, in warmer climates and rooftop installations, degradation rates can exceed − 9.55% to −0.4% per year⁴⁶. Some studies also report an average degradation rate of −0.3% per year across multiple modules⁴⁷. These increased degradation rates in tropical climates, such as Malaysia’s, can significantly reduce the power output and lifespan of PV modules. This poses a challenge for the long-term reliability and efficiency of PV systems, potentially hindering the widespread adoption of solar energy, particularly in regions with harsh climatic conditions. By focusing on these areas, the solar industry can work toward enhancing the reliability and performance of PV systems, reducing degradation rates, and improving their long-term sustainability.
A comparative analysis was conducted on the performance of PEARL’s system, which incorporates two distinct PV technologies: Poly-crystalline (Array 1) and Mono-crystalline (Array 2). The results indicate that Array 1 consistently outperformed Array 2 across all evaluation metrics. Solar irradiation and humidity were identified as significant factors influencing the performance of these systems. The primary findings are presented in Table 5, along with descriptions from previous literature for comparison with the present study underscores notable disparities in the operational outcomes of PV systems across different geographic conditions.
Systems studied by Ghazali et al.¹⁶, Saleheen et al.¹⁷, and Raghoebarsing et al.⁴⁸ demonstrated higher Y_F, ranging from 3.68 to 3.70, in contrast to the PEARL system’s Y_F of 1.87. Similarly, array and system efficiencies reported in prior studies, particularly Raghoebarsing et al.⁴⁸ (11.90% array efficiency, 97.20% system efficiency) and Ghazali et al.¹⁶ (12.52% array efficiency), significantly exceeded those achieved by the PEARL installation (3.07% and 1.72%, respectively). These differences can be attributed to multiple factors, including more stable solar irradiance profiles, larger system capacities, and potentially superior maintenance regimes. Furthermore, inverter efficiencies approaching or exceeding 96% were recorded in Raghoebarsing et al.⁴⁸ and Luna Carlosama et al.⁴⁹, whereas the PEARL system’s inverter efficiency was moderately lower at 91.08%, likely reflecting technological and operational constraints inherent to small-scale or older PV installations.
Despite the lower performance metrics in yield, efficiency, and CF (6.40% for the PEARL system), a notable strength lines in the PR. The PEARL system achieved a PR of 86.74%, outperforming many systems installed in other regions studied by de Lima et al.²⁰, Saxene et al.²² and Sekyere et al.²⁵, with theoretically Malaysia is more favorable in solar resources. This high PR indicates a remarkably efficient internal system operation relative to the available solar input, even under the challenging climatic conditions of Kuala Lumpur, characterized by high humidity, frequent rainfall, and heavy cloud cover during the monsoon period. The findings reveal that while absolute energy yields and efficiency indicators are important, they do not necessarily reflect the intrinsic quality or optimization of PV system performance in all environments. In this context, the PEARL system’s superior PR highlights its operational resilience, effective system integration, and minimal internal losses.
Therefore, although the PEARL system did not achieve outstanding results in conventional performance indicators, its exceptional PR firmly establishes its competitive advantage within the tropical PV sector. These results emphasize the critical need for performance evaluations to consider PR alongside absolute metrics, particularly when assessing PV system viability in regions subject to significant climatic variability. The present study thus contributes a nuanced understanding of system behavior, providing valuable insights for future deployments in similar tropical settings.
The main objective is to assess and analyze the performance of a 3.575 kW_p grid-connected PV system, installed on the rooftop of the Engineering tower building at the Universiti Malaya, Kuala Lumpur. A comprehensive performance study was carried out from January 2020 to December 2022 (36 months). The main findings of the present study can be outlined as follows:
The yearly mean output of AC energy for 36 months for Array 1 and Array 2 is 2240.96 kWh and 1176.93 kWh, respectively.
The maximum average AC energy output for Array 1 and Array 2 was 214.72 kWh in March and 106.86 kWh in May, while the minimum was recorded at 158.74 kWh in August and 82.98 kWh in September, respectively.
The total average GHI in 2020, 2021, and 2022 was 1357.37 kWhm^{– 2}, 1290.47 kWhm² and 1233.83 kWhm² respectively, with a total average (36 months) of 388.67 kWhm^{– 2}.
The yearly average of final yield for the two arrays are 3.07 h/d for Array 1 and 1.72 h/d for the Array 2 module, respectively.
Annual Average array losses (La) were recorded as 0.21 h per day for Array 1, 1.83 h per day for Array 2.
The annual average System Losses (L_s) were recorded for 0.26 h per day for Array 1 and 0.26 h per day for Array 2.
The system’s yearly performance ratios for Array 1 and Array 2 are 86.74% and 56.30%, respectively.
System capacity factors for Array 1 and Array 2 are 17.05% and 14.93%, respectively.
Annual array and system efficiency for both arrays are (11.70% and 10.8%) for Array 1 and (6.40% and 5.82%) for Array 2, respectively.
Forecasted Degradation rates for both Array 1 and Array 2 using the MAKIMA method are approximately 10.58% and 11.99% (~ 12%), respectively.
The MAKIMA model achieved an MAPE of 0.78% for Array 1 and 0.35% for Array 2, respectively.
Furthermore, the MAKIMA method was proposed to forecast the degradation rate in 2023 for both arrays, successfully capturing the exponential increase in degradation. The efficiency of both systems drops to 9.94% and 5.31%, respectively. For enhanced precision, a longer dataset is recommended for degradation calculation. A technical evaluation determined that Array 1 (Poly-crystalline) is optimal for the given location and environmental conditions, while Array 2 (Mono-crystalline) is less suitable. For future work, it is intended to analyze the techno-economic performance of both systems to identify the most cost-effective PV system for Malaysia and other similar tropical climates.
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.
Amorphous silicon
Alternating current
Building integrated photovoltaic
Cadmium telluride
Capacity factor
Copper Indium Gallium Selenide
Capital utilization factor
Direct current
Degradation rate
Alternating current energy
Direct current energy
Solar irradiation
Greenhouse gas
Reference irradiation
Heterojunction intrinsic thin-film
International energy agency
Array losses
Levelized cost of energy
System losses
Array efficiency
Inverter efficiency
System efficiency
Monocrystalline
Mean average error
Modified akima cubic hemite interpolation
Mean average percentage error
Mean square error
Malaysia renewable energy roadmap
Nationally determined contributions
Natural resources, environment and climate change
National renewable energy laboratory
Polycrystalline
Alternating current power
Power electronic and renewable energy laboratory
Photovoltaic rated power
Performance ratio
Photovoltaic
Renewable energy
Root means square error
Solar energy
Array yield
Final yield
Reference yield
Zang, H. et al. Day-ahead photovoltaic power forecasting approach based on deep convolutional neural networks and meta learning. Int. J. Electr. Power Energy Syst. 118, 105790. https://doi.org/10.1016/j.ijepes.2019.105790 (2020).
Article  ADS  Google Scholar 
Ayompe, L. M., Duffy, A., McCormack, S. J. & Conlon, M. Measured performance of a 1.72 kW rooftop grid connected photovoltaic system in Ireland. Energy. Conv. Manag. 52, 816–825. https://doi.org/10.1016/j.enconman.2010.08.007 (2011).
Article  ADS  Google Scholar 
Purohit, I. & Purohit, P. Performance assessment of grid-interactive solar photovoltaic projects under india’s National solar mission. Appl. Energy. 222, 25–41. https://doi.org/10.1016/j.apenergy.2018.03.135 (2018).
Article  ADS  Google Scholar 
Cubukcu, M. & Gumus, H. Performance analysis of a grid-connected photovoltaic plant in Eastern Turkey. Sustain. Energy Technol. Assess. 39, 100724. https://doi.org/10.1016/j.seta.2020.100724 (2020).
Article  Google Scholar 
International Energy Agency. Renewable Energy Progress Tracker. International Energy Agency, Paris. (2024). Available at: https://www.iea.org/data-and-statistics/data-tools/renewable-energy-progress-tracker
Sustainable Energy Development Authority (SEDA). Malaysia. RE Generation Statistics & Monitoring. Sustainable Energy Development Authority Malaysia. Available at: https://www.seda.gov.my/statistics-monitoring/re-generation/
Ministry of Energy and Natural Resources Malaysia. Report on Peninsular Malaysia Generation Development Plan 2020 (2021–2039). KeTSA, Malaysia. (2021). Available at: https://www.st.gov.my/ms/contents/files/download/169/Report_on_Peninsular_Malaysia_Generation_Development_Plan_2020_(2021-2039)-FINAL.pdf
Economic Planning Unit. National Energy Policy 2022–2040. Economic Planning Unit (EPU), Malaysia. (2022). Available at: https://www.ekonomi.gov.my/sites/default/files/2022-09/National%20Energy%20Policy_2022_2040.pdf
Ahmad, M. & Zeeshan, M. Multi-objective optimization of concentrated solar power plants from an energy-water-environment nexus perspective under distinct Climatic conditions – Part A: Techno-economic analysis. J. Clean. Prod. 375, 134099. https://doi.org/10.1016/j.jclepro.2022.134099 (2022).
Article  Google Scholar 
Singh, R., Sharma, M. & Yadav, K. Degradation and reliability analysis of photovoltaic modules after operating for 12 years: A case study with comparisons. Renew. Energy. 196, 1170–1186. https://doi.org/10.1016/j.renene.2022.07.048 (2022).
Article  CAS  Google Scholar 
Prasad, M. & Prasad, R. Bifacial vs monofacial grid-connected solar photovoltaic for small islands: A case study of Fiji. Renew. Energy. 203, 686–702. https://doi.org/10.1016/j.renene.2022.12.068 (2023).
Article  Google Scholar 
Ramadhani, U. H., Lingfors, D., Munkhammar, J. & Widén, J. On the properties of residential rooftop azimuth and Tilt uncertainties for photovoltaic power generation modeling and hosting capacity analysis. Solar Energy Adv. 3, 100036. https://doi.org/10.1016/j.seja.2023.100036 (2023).
Article  Google Scholar 
Bhattacharya, S., Sadhu, P. K. & Sarkar, D. Performance evaluation of Building integrated photovoltaic system arrays (SP, TT, QT, and TCT) to improve maximum power with low mismatch loss under partial shading. Microsyst. Technol. 30, 583–597. https://doi.org/10.1007/s00542-023-05564-0 (2024).
Article  CAS  Google Scholar 
Shrivastava, A. et al. Solar energy capacity assessment and performance evaluation of a standalone PV system using PVSYST. Mat. Today Proc. 80, 3385–3392. https://doi.org/10.1016/j.matpr.2021.07.258 (2023).
Article  CAS  Google Scholar 
Sarkar, D., Sadhu, P. K. & and A new hybrid BIPV array for enhancing maximum power with reduced mismatch losses under extreme partial shading scenarios. Energy Sour. Part A Recover. Utilization Environ. Eff. 44, 5172–5198. https://doi.org/10.1080/15567036.2022.2083271 (2022).
Article  CAS  Google Scholar 
Ghazali, A., Salleh, E., Lim, C. H., Sohif, M. & Sopian, K. Performance and Financial Evaluation of Various Photovoltaic Vertical Facades on High-rise Building in Malaysia. Energy Build. https://doi.org/10.1016/j.enbuild.2016.11.003 (2016).
Article  Google Scholar 
Saleheen, M. Z., Salema, A. A., Islam, M., Sarimuthu, S. M., Hasan, M. Z. & C. R. & A target-oriented performance assessment and model development of a grid-connected solar PV (GCPV) system for a commercial Building in Malaysia. Renew. Energy. 171, 371–382. https://doi.org/10.1016/j.renene.2021.02.108 (2021).
Article  Google Scholar 
Akhter, M. N., Mekhilef, S., Mokhlis, H., Olatomiwa, L. & Muhammad, M. A. Performance assessment of three grid-connected photovoltaic systems with combined capacity of 6.575 kWp in Malaysia. J. Clean. Prod. 277, 123242. https://doi.org/10.1016/j.jclepro.2020.123242 (2020).
Article  Google Scholar 
Anang, N., Nur Azman, S., Muda, S. N. A., Dagang, W. M. W., Daud, M. Z. & A. N. & Performance analysis of a grid-connected rooftop solar PV system in Kuala Terengganu, Malaysia. Energy Build. 248, 111182. https://doi.org/10.1016/j.enbuild.2021.111182 (2021).
Article  Google Scholar 
de Lima, L. C., de Araújo Ferreira, L. & de Lima Morais, F. H. B. Performance analysis of a grid connected photovoltaic system in Northeastern Brazil. Energy. Sustain. Dev. 37, 79–85. https://doi.org/10.1016/j.esd.2017.01.004 (2017).
Article  Google Scholar 
Singh, T. S. D., Shimray, B. A., Singh, A. B. & Meitei, S. N. Performance measurement of 5 kWp rooftop grid-connected SPV system in moderate Climatic region of Imphal, Manipur, India. Energy. Sustain. Dev. 73, 292–302. https://doi.org/10.1016/j.esd.2023.02.009 (2023).
Article  Google Scholar 
Saxena, A. K., Saxena, S. & Sudhakar, K. Energy, economic and environmental performance assessment of a grid-tied rooftop system in different cities of India based on 3E analysis. Clean. Energy. 5, 288–301. https://doi.org/10.1093/ce/zkab008 (2021).
Article  Google Scholar 
gopi, A., Sudhakar, K., Ngui, W. K., Kirpichnikova, I. M. & cuce, E. Energy analysis of utility-scale PV plant in the rain-dominated tropical monsoon climates. Case Stud. Therm. Eng. 26, 101123. https://doi.org/10.1016/j.csite.2021.101123 (2021).
Article  Google Scholar 
Palm, F. et al. Performance evaluation of Burkina faso’s 33 MW largest Grid-Connected PV power plant. Energies 16, 6177. https://doi.org/10.3390/en16176177 (2023).
Article  Google Scholar 
Sekyere, C. K. K. et al. Performance evaluation of a 20 MW grid-coupled solar park located in the Southern oceanic environment of Ghana. Clean. Eng. Technol. 5, 100273. https://doi.org/10.1016/j.clet.2021.100273 (2021).
Article  Google Scholar 
Malvoni, M., Kumar, N. M., Chopra, S. S. & Hatziargyriou, N. Performance and degradation assessment of large-scale grid-connected solar photovoltaic power plant in tropical semi-arid environment of India. Sol. Energy. 203, 101–113. https://doi.org/10.1016/j.solener.2020.04.011 (2020).
Article  ADS  Google Scholar 
Mulcué-Nieto, L. F. et al. Energy performance assessment of monocrystalline and polycrystalline photovoltaic modules in the tropical mountain climate: the case for Manizales-Colombia. Energy Rep. 6, 2828–2835. https://doi.org/10.1016/j.egyr.2020.09.036 (2020).
Article  Google Scholar 
do Nascimento, L. R., Braga, M., Campos, R. A., Naspolini, H. F. & Rüther, R. Performance assessment of solar photovoltaic technologies under different Climatic conditions in Brazil. Renew. Energy. 146, 1070–1082. https://doi.org/10.1016/j.renene.2019.06.160 (2020).
Article  CAS  Google Scholar 
Junedi, M. M. et al. Environmental and economic performance assessment of integrated conventional solar photovoltaic and agrophotovoltaic systems. Renew. Sustain. Energy Rev. 168, 112799. https://doi.org/10.1016/j.rser.2022.112799 (2022).
Article  Google Scholar 
Ngure, S. M., Adaramola, M. S. & B., M. A., O., A. E. & and Techno-economic performance analysis of grid-tied solar PV systems under tropical savanna Climatic conditions in Kenya. Int. J. Ambient Energy. 44, 1–16. https://doi.org/10.1080/01430750.2023.2224329 (2023).
Article  Google Scholar 
Owolabi, A. B., Yakub, A. O., Luqman, R., Same, N. N. & Suh, D. Performance assessment of four grid-connected solar photovoltaic technologies under similar environmental conditions in Nigeria. Int. J. Energy Res. https://doi.org/10.1155/2023/9458440 (2023).
Article  Google Scholar 
Govindarajan, L., Batcha, M. F. B. M. & Abdullah, M. K. B. Performance assessment of large-scale rooftop solar PV system: a case study in a Malaysian public university. Discover Appl. Sci. 6, 328. https://doi.org/10.1007/s42452-024-06007-9 (2024).
Article  Google Scholar 
Mohamed, K., Shareef, H., Nizam, I., Esan, A. B. & Shareef, A. Operational performance assessment of rooftop PV systems in the Maldives. Energy Rep. 11, 2592–2607. https://doi.org/10.1016/j.egyr.2024.02.014 (2024).
Article  Google Scholar 
International Electrotechnical Commission. IEC TS 61724-1:2017. Photovoltaic System Performance – Part 1: Monitoring. International Electrotechnical Commission (IEC), Geneva, Switzerland. (2017).
Akima, H. A. New method of interpolation and smooth curve fitting based on local procedures. J. ACM (JACM). 17, 589–602. https://doi.org/10.1145/321607.321609 (1970).
Article  Google Scholar 
Ren, K., Moudud, A., Nielsen, P., Gussmann, M. & Rönnegård, L. Interpolation Methods to Improve Data Quality of Indoor Positioning Data for Dairy Cattle. Front. Anim. Sci. 3, https://doi.org/10.3389/fanim.2022.896666 (2022).
Moler, C. B. Numerical Computing with Matlab (Society for Industrial and Applied Mathematics, 2004).
Ketjoy, N., Chamsa-ard, W. & Mensin, P. Analysis of factors affecting efficiency of inverters: case study grid-connected PV systems in lower Northern region of Thailand. Energy Rep. 7, 3857–3868. https://doi.org/10.1016/j.egyr.2021.06.075 (2021).
Article  Google Scholar 
Adaramola, M. S. & Vågnes, E. E. T. Preliminary assessment of a small-scale rooftop PV-grid tied in Norwegian Climatic conditions. Energy. Conv. Manag. 90, 458–465. https://doi.org/10.1016/j.enconman.2014.11.028 (2015).
Article  ADS  Google Scholar 
Amin, N., Lung, C. W. & Sopian, K. A practical field study of various solar cells on their performance in Malaysia. Renew. Energy. 34, 1939–1946. https://doi.org/10.1016/j.renene.2008.12.005 (2009).
Article  CAS  Google Scholar 
Chekal Affari, B. et al. Power losses in PV arrays of field-aged modules. Microelectron. Reliab. 147, 115052. https://doi.org/10.1016/j.microrel.2023.115052 (2023).
Article  CAS  Google Scholar 
Bernardes, S., Lameirinhas, R. A. M., Torres, J. P. N. & Fernandes, C. A. F. Characterization and design of photovoltaic solar cells that absorb ultraviolet, visible and infrared light. Nanomaterials 11, 78 (2021).
Article  CAS  PubMed  PubMed Central  Google Scholar 
Sabo, M. L., Mariun, N., Hizam, H., Mohd Radzi, M. A. & Zakaria, A. Spatial energy predictions from large-scale photovoltaic power plants located in optimal sites and connected to a smart grid in Peninsular Malaysia. Renew. Sustain. Energy Rev. 66, 79–94. https://doi.org/10.1016/j.rser.2016.07.045 (2016).
Article  Google Scholar 
Irshad, A. S. et al. Optimization of grid-photovoltaic and battery hybrid system with most technically efficient PV technology after the performance analysis. Renew. Energy. 207, 714–730. https://doi.org/10.1016/j.renene.2023.03.062 (2023).
Article  Google Scholar 
Leitão, D., Torres, J. P. N. & Fernandes, J. F. P. Spectral irradiance influence on solar cells efficiency. Energies 13, 5017. https://doi.org/10.3390/en13195017 (2020).
Article  CAS  Google Scholar 
Mota, F., Neto Torres, J. P., Ferreira Fernandes, C. A. & Marques Lameirinhas, R. A. Influence of an aluminium concentrator corrosion on the output characteristic of a photovoltaic system. Sci. Rep. 10, 21865. https://doi.org/10.1038/s41598-020-78548-z (2020).
Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 
Rahman, T. et al. Investigation of degradation of solar photovoltaics: a review of aging factors, impacts, and future directions toward sustainable energy management. Energies 16, 3706 (2023).
Article  CAS  Google Scholar 
Raghoebarsing, A. & Kalpoe, A. Performance and economic analysis of a 27 kW grid-connected photovoltaic system in Suriname. IET Renew. Power Gener. 11, 1545–1554. https://doi.org/10.1049/iet-rpg.2017.0204 (2017).
Article  Google Scholar 
Luna Carlosama, C. F., Chuquen, M. & Mulcue Nieto, R. Jiménez García, F. N. Potential of photovoltaic generation in the Putumayo department of Colombia. Appl. Sci. 11, 5528 (2021). https://www.mdpi.com/2076-3417/11/12/5528
Article  Google Scholar 
Download references
Power Electronics and Renewable Energy Research Laboratory (PEARL), Department of Electrical Engineering, Faculty of Engineering, Universiti Malaya, Kuala Lumpur, 50603, Malaysia
Putri Nor Liyana Mohamad Radzi & Saad Mekhilef
School of Engineering, Swinburne University of Technology, Hawthorn, VIC, 3122, Australia
Saad Mekhilef
Department of Electrical Engineering, Faculty of Engineering, Universiti Malaya, Kuala Lumpur, 50603, Malaysia
Noraisyah Mohamed Shah
Department of Electrical Engineering, Rachna College of Engineering and Technology Gujranwala (A Constituent College of University of Engineering and Technology Lahore), Gujranwala, 52250, Pakistan
Muhammad Naveed Akhter
Centre de Développement des Energies Renouvelables, Unité de Recherche Appliquée En Energies Renouvelables, URAER, CDER, 47133, Ghardaia, Algeria
Samir Hassani
Department of Electrical Engineering, College of Engineering, University of Ha’il, Ha’il City, 81451, Saudi Arabia
Abdullah Albaker
Department of Electrical and Communication Engineering, United Arab Emirates University (UAE-U), Al Ain, UAE
Addy Wahyudie
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Search author on:PubMed Google Scholar
Putri Nor Liyana Mohamad Radzi: Conceptualization, Investigation, Methodologies, Formal analysis, Writing – original draft, Saad Mekhilef: Supervision, Writing – review & editing, Noraisyah Mohamed Shah: Supervision, Writing – review & editing, Muhammad Naveed Akhter: Supervision, Methodology, Writing – review & editing. Samir Hassani: Methodology, Writing – review & editing, Abdullah Albaker: Writing – review & editing, Funding, Addy Wahyudie: Writing – review & editing, Funding.
Correspondence to Putri Nor Liyana Mohamad Radzi, Saad Mekhilef, Noraisyah Mohamed Shah or Addy Wahyudie.
The authors declare no competing interests.
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://creativecommons.org/licenses/by-nc-nd/4.0/.
Reprints and permissions
Mohamad Radzi, P., Mekhilef, S., Mohamed Shah, N. et al. Performance evaluation and degradation analysis of grid connected photovoltaic systems for energy efficient buildings in tropical climates. Sci Rep 15, 39103 (2025). https://doi.org/10.1038/s41598-025-26765-9
Download citation
Received: 03 March 2025
Accepted: 30 October 2025
Published: 07 November 2025
Version of record: 07 November 2025
DOI: https://doi.org/10.1038/s41598-025-26765-9
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative
Advertisement
Scientific Reports (Sci Rep)
ISSN 2045-2322 (online)
© 2025 Springer Nature Limited
Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

source