Effects of Power Optimizer Application in a Building-Integrated Photovoltaic System According to Shade Conditions

Abstract

A building-integrated photovoltaic (BIPV) system produces power using photovoltaic (PV) modules as building exterior materials, whose architectural performance serves the same functions as those of existing building materials. Most relevant studies targeted general PV modules used in the building-mounted and -attached types. This study aims to integrate the building elevation-type BIPV system and exterior materials to secure both exterior material performance and PV electrical performance by embedding a power optimization device in an integrated system of BIPV modules and exterior materials. Thus, the advantages of economy, safety, aesthetics, and ease of maintenance can be achieved. In this study, experiments were conducted on elevation-type BIPV modules with and without a power optimizer, that is, a DC/DC converter, under various shade conditions, and the power loss rate of the BIPV system was analyzed. The power optimizer-equipped BIPV system was experimentally observed to have a PV power-loss rate approximately 2–3 times lower than that of the BIPV system without a power optimizer when the shade ratio of one module was approximately 10–75%. This exterior material-integrated BIPV-specific power optimization device reduces dependence on fossil fuels for power production and improves energy sustainability, contributing to the spread of zero-energy buildings and carbon neutrality.

1. Introduction

To achieve carbon neutrality, the International Energy Agency presented a roadmap based on new and renewable energy sources rather than fossil fuels to transition the global energy sector to a net-zero energy system by 2050 [1]. Major countries, especially those in Europe, are actively supporting the expansion of building-integrated photovoltaic (BIPV) systems through subsidy policies as part of the mandatory zero-energy buildings [2]. Because BIPV systems are not subject to separate site restrictions, their expansion and support are expected to further increase with the expansion of the construction of zero-energy buildings.

Photovoltaic (PV) systems are classified into the building-mounted, -attached, and -integrated types depending on their installation [3,4]. Building-mounted installations are the most common and require a separate site and external structure because they attach to a general PV module while protruding from the building roof. Building-attached PVs refer to the close attachment of general PV modules to buildings without the need for building materials. In comparison, BIPVs function as a building material and are externally installed on various building parts during the construction stage, such as windows, exterior walls, and roofs, without the need for additional external structures. A BIPV system can produce electrical power while exhibiting the same architectural performance and functions as other, existing building materials. Furthermore, BIPV modules are structurally different from regular PV modules in terms of the cell materials and glass-bonding methods [5,6], which facilitate their application in urban buildings. Domestic and international research has focused on four major areas of BIPV systems: their economics, safety, aesthetics, and maintenance.

This work started with studying the integration of the building elevation-type BIPV system and exterior materials. In addition to the development of the BIPV module and curtain-wall integration system, power optimization methods to prevent BIPV power loss were reviewed. BIPV-specific modules installed as elevation-type BIPV systems are generally manufactured as the glass-to-glass (G2G) type, and their size is designed to match the building elevation. Accordingly, their capacity is determined; therefore, the capacity and wiring configurations are diverse. In addition, because it is a building-elevation type BIPV, some output loss occurs owing to the influence of shading from surrounding buildings. Finally, this study aims to secure both exterior material performance and PV electrical performance by embedding a power optimization device in an integrated system of BIPV modules and exterior materials. First, to study PV power optimization among BIPV and exterior material integration methods, this study examined the possibility of using a power optimizer. In future research, we plan to develop and demonstrate a BIPV-specific power optimizer by developing a power optimization algorithm dedicated to BIPV modules.

In a PV system, a DC/DC converter, which is a module-level power electronic (MLPE) device, serves as a power optimization device. Micro-inverters and power optimizers are representative MLPE devices that optimize the power of individual PV modules to minimize the impact of module-level power loss on the entire system in existing centralized or string inverters. The micro-inverter includes a DC/DC converter and DC/AC inverter, and the power optimizer is a standalone DC/DC converter. Although the efficiency and lifespan of both devices are similar and high, reportedly approximately 10–25 years, significant differences exist in their in cost and maintenance. The cost of a micro-inverter is approximately 10–20% higher per W than that of a power optimizer [7]. Because the micro-inverter is a dual circuit of a converter and an inverter, it is expected to require more maintenance and higher maintenance costs than a power optimizer composed of a single circuit. The power optimizer adjusts the DC voltage to minimize the impact of the mismatch caused by the shading or failure of cells and modules, thereby minimizing the loss of PV power in the module, string or entire array. BIPV modules applied to buildings are manufactured differently in terms of their size, capacity, and color depending on the building. When using a power optimizer, the flexible wiring between modules can be of different specifications, making design and construction easy. Additionally, a power optimizer can cut off power in the event of a short circuit and has a real-time performance monitoring function for individual modules, enabling safe operation and maintenance. Currently, many companies are developing and selling power optimizers; however, they are difficult to directly use because they are manufactured only for general PV modules and do not fit BIPV modules in terms of the input voltage, maximum power point tracking (MPPT) voltage range, voltage adjustment method, and general inverter connection.

PV-module mismatching and shading are the main causes of performance degradation in PV systems. Owing to the need for the optimal power management of individual modules, research has been conducted to develop and verify algorithms that perform MPPT of individual modules through DC/DC converters [8,9]. However, most power optimization-related studies focused on experiments and simulations targeting general PV modules used in building-mounted and -attached systems, with a focus on ground and rooftop PV systems. The National Renewable Energy Laboratory compared the power performance under normal and shade conditions with commercial micro-inverters and power optimizers for large-scale PV systems [10]. The average power loss recovery rate due to shade was 35% for micro-inverters and 25–35% for power optimizers. Power optimizers are also less expensive than micro-inverters. Kim et al. [11] developed a power optimizer for general PV modules, installed it in a large-scale PV system, and compared and analyzed the DC voltage, current, and power of the module before and after installation. Lee et al. [12] used an MLPE device to evaluate the feasibility of the north-facing installation of a PV system based on an analysis of the annual operation data of a domestic rooftop PV system comprising general PV modules. Sinapis [13,14] conducted DC and AC performance analyses by constructing a string inverter, power optimizer, and micro-inverter system for general PV modules, and the author found that the MPLE improved power performance by up to 35% under partial shade conditions. In addition, Sarwar et al. [15] evaluated the converter of a PV system based on MLPE devices through simulations and showed that 25–35% more power was generated when an MLPE device was used in shaded conditions. Ravits et al. [16] predicted the system reliability and summarized the converter requirements based on the connection of string inverters, micro-inverters, serial power optimizers, and parallel power optimizers in relation to the application of power optimizers to BIPV systems. Additionally, Ravits et al. [17] studied the power performance of each PV-module type (c-Si, c-Si half-cells, and CIGS) and the applicability of commercial module-level converters for one prototype unit of a curtain-wall BIPV system. They reviewed and highlighted the need to develop a module-level converter suitable for the electrical requirements of BIPV modules. Lastly, Spiliotis et al. [18] designed a DC/DC converter built into a BIPV system and verified it through simulations.

Given that previous studies focused on general ground or rooftop PV modules rather than BIPV ones, and that they were limited to concepts, further research considering the characteristics of BIPV systems is required. Therefore, this study aimed to utilize a power optimizer stored in a curtain-wall frame, thus integrating the exterior materials of a BIPV module [19]. First, we conducted a review of the use and storage of electric devices such as power optimizers, reflecting the characteristics of BIPV exterior material integration. Then, we conducted experiments on BIPV system modules with and without a power optimizer under shade conditions, analyzing their power-loss rates. The results confirmed the effect of power-optimizer application to a BIPV module in shady conditions.

2. Experimental Methods

2.1. Configuration

The experiments were conducted in an office building in Goyang-si, Gyeonggi-do, South Korea. Pictures and architectural drawings of the building are presented in

Table 1. Images of the building where the experiments were conducted.

Category	Front Side	Left Side	Roof Floor
Picture
Drawing

. The weather information for the region as of 2022 provided by the Korea Meteorological Administration [20] was an average daily temperature of 13.2 °C, average minimum temperature of 9.2 °C, average maximum temperature of 18.0 °C, average relative humidity of 64.4%, average wind speed of 2.4 m/s, total annual solar radiation of 1479.4 kWh/m², and precipitation of 1775.3 mm. Two BIPV systems, each having a capacity of 342 W, were installed on the roof of the building, facing south at an inclination angle of 90° (vertical). Figure 1 shows the experimental BIPV systems installed on the rooftop. The BIPV systems are herein referred to as A (left 2-panel system in the front picture; A-a and A-b) and B (right 2-panel system; B-a and B-b).

The BIPV systems consisted of modules, inverters, and two BIPV modules each, with the same capacity connected in series. The BIPV modules were separately manufactured by KOES at our request; each was a colored BIPV module made of single-crystal silicon, glass-to-glass, and was dark gray with a capacity of 171 W, consisting of 36 cells. The B-b module was 30 cm longer than the other three, and the power efficiency calculated using the module area differed; however, the electrical specifications were identical to those of the other modules owing to the arrangement of the same cell configuration. The inverter was a 350 W micro-inverter (MI-350, Hoymiles; efficiency ≥ 94.79%) installed for each array. In this experiment, the space available at the experimental site was insufficient to install multiple modules; therefore, two modules were connected. Because this array has a small capacity, it did not match the capacity of a general small-capacity inverter; therefore, a micro-inverter was inevitably used as a DC/AC inverter. To examine the BIPV power-loss rate with and without a power optimizer, a power optimizer was installed in each module of array A. The power optimizer (NE-GC-02, Nanoom Energy; capacity 600 W) was manufactured for general PV modules, which is why the internal algorithm was modified separately to allow operation in the experimental environment with the BIPV module. Unlike general PV modules, the capacity and operating-voltage range of the BIPV modules were so small that there was a difference in the perception of low power. We lowered the minimum value of the low-power recognition section so that it could operate even when connected to a BIPV module. The power optimizer adjusts the output by adjusting the voltage using, for example, a buck, buck-boost, or another converter. The operation method can differ due to the circuit used by each manufacturing company. The details of the optimization technology are not generally disclosed because it is confidential to each manufacturing company. The power optimizer used in this experiment was a buck converter with full power processing (FPP) operation. This converter lowers the output voltage relative to the input voltage. It is also called a step-down converter and comprises a switch, inductor, diode, and condenser. FPP is commonly used in PV applications to achieve MPPT operation by handling the entire amount of power generated by the PV module. The inverters of the BIPV systems were connected to the distribution board on the roof, and PV power was supplied to the building.

Table 2. Specifications of the BIPV systems.

Category	Parameter	Value
PV * System	Capacity [W]	342
	Direction/Slope	South/90°
PV Module	Rated power [W]	171
	Maximum power voltage [V]	20.77
	Maximum power current [A]	8.25
	Open circuit voltage [V]	24.2
	Short circuit current [A]	8.68
	Efficiency [%]	10.1, (B-b) 8.4
	Cell material	Monocrystalline
	Cell configuration	6 × 6
	Dimensions (L × W × H) [mm]	1462 × 1156 × 12, (B-b) 1762 × 1156 × 12
	Weight [kg]	47
Inverter	Rated power [W]	350
	Input voltage range [V]	16–60
	Efficiency [%]	94.79
	Dimensions (L × W × H) [mm]	178 × 153 × 28
	Weight [kg]	1.98
Power Optimizer	Rated input DC power [W]	600
	Maximum input voltage [V]	60
	MPPT ** operating range [V]	8–60
	Dimensions (L × W × H) [mm]	150 × 100 × 42
	Weight [kg]	0.625

* PV: photovoltaic, ** MPPT: Maximum power point tracking.

summarizes the specifications of the devices used to construct the experimental BIPV systems.

2.2. Analysis

The power performance experiments were conducted for approximately two months, from 7 September 2023 to 23 October 2023. To analyze the operating characteristics of the BIPV systems, a watt meter, solar-irradiance meter, thermocouple, and a data logger were installed. The collected data included the AC power, slope solar irradiance, and air and module temperature. The data were measured at 1 min increments. A power meter (ADpower HPM-100A) was installed on the output side of the inverter to measure and store AC power. The solar irradiance meter (EKO MS-60) was installed at 90°, the same inclination angle as that of the modules, and the solar irradiance on the incline surface was measured. A thermocouple (TC-T) was installed at the top and back of module B to measure the air and module temperatures. The solar irradiance meter and thermocouple were connected to a data logger (GRAPHTEC midi LOGGER GL840) to save data. The measurement error for each device was observed to be in the range of ±0.5% for the power meter, ±1% for the solar-irradiance meter, and ±0.5% for the data logger.

The experimental groups for the installation or non-installation of the power optimizer under shade conditions were organized into six cases, as listed in

Table 3. Classification of the experimental cases.

Case	Condition	I–V and V–P Curves in Simulation
1	No installation of the power optimizer
2	Installation of the power optimizer in module A	Maximum power efficiency: 100%
3	Shading 10% of modules A-b and B-b along with the conditions of Case 2	Maximum power efficiency: 95.8%
4	Shading 25% of modules A-b and B-b along with the conditions of Case 2	Maximum power efficiency: 87.6%
5	Shading 50% of modules A-b and B-b along with the conditions of Case 2	Maximum power efficiency: 71.2%
6	Shading 75% of modules A-b and B-b along with the conditions of Case 2	Maximum power efficiency: 53.3%

. In Case 1, the power optimizer was not installed; in Case 2, the power optimizer was installed in each A module; and in Case 3, a shade ratio of 10% was applied to modules A-b and B-b, along with the conditions of Case 2. In Cases 4, 5, and 6, the shade ratio was increased to 25%, 50%, and 75%, respectively. Shade was directly provided using blackout sheets based on the cell-unit area of the BIPV module. The same shading conditions were created by shading the module located on the right side of each array based on the front. The I–V and V–P curves in Table 3 were modeled and simulated in this experimental environment using Solar Pro, a PV-system simulation program. The simulations presented in Table 3 show that only the I–V and V–P curves of module B changed under each shading condition, the module that did not have a power optimizer.

3. Results and Discussion

3.1. Daily BIPV Operation Characteristics

Figure 2 summarizes the slope solar irradiance, PV power of the BIPV systems, and air and module temperature data measured on September 9, 2023, collected at 1 min increments. As shown in Figure 2a, the patterns of the measured solar irradiance and PV power were similar. Figure 2c shows the power of module A as the actual measured value and the value calculated using Equation (1), which used the solar irradiance and PV specifications. The difference between the calculated and actual measured values was less than 5%. In the equation, $P_{AC}$ [W] is the AC power of the PV system, $G_m$ [W/m²] is the actual solar radiation, $A$ [m²] is the area of the PV array, ${\eta }_{Mod}$ [%] is the efficiency of the PV module, and ${\eta }_{Inv}$ [%] is the inverter efficiency.

$$P_{AC}=G_m\times A\times \frac{{\eta }_{Mod}}{100}\times \frac{{\eta }_{Inv}}{100}$$

(1)

The daily BIPV operational characteristics were analyzed using measured data and meteorological information provided by the Korea Meteorological Administration.

Table 4. Daily operating characteristics of the PV systems by case.

Case	Date	Weather (Mean Cloud Amount/d, Rainfall/d, Mean Temperature/d)	Solar Irradiance (kWh/m2/d)	Maximum PV Power (W)		PV Energy (kWh/d)
				Installation PO *	Non-Installation PO	Installation PO	Non-Installation PO
1	9 September 2023	3.3, 0 mm, 25.8 °C	4.20	196.1 (57.3%)	198.8 (58.1%)	1.29 (95.1%)	1.31 (93.6%)
2	22 September 2023	2.0, 0 mm, 20.5 °C	4.66	213.5 (62.4%)	225.8 (66.0%)	1.38 (91.4%)	1.49 (96.3%)
3	1 October 2023	0.1, 0 mm, 18.4 °C	5.04	225.2 (65.8%)	229.3 (67.0%)	1.45 (89.2%)	1.53 (91.3%)
4	5 October 2023	2.3, 0 mm, 14.2 °C	5.39	224.7 (65.7%)	201.0 (58.8%)	1.49 (85.2%)	1.39 (77.2%)
5	11 October 2023	2.3, 0 mm, 17.6 °C	4.67	170.0 (49.7%)	141.5 (41.4%)	0.95 (62.6%)	0.88 (56.5%)
6	17 October 2023	0.0, 0 mm, 13.8 °C	5.52	96.5 (28.2%)	83.4 (24.4%)	0.56 (31.4%)	0.57 (31.1%)

* PO, power optimizer.

lists the weather information, maximum PV power, and daily PV energy on specific days in each case. The parenthetical value of the maximum PV power indicates the maximum PV power efficiency compared with the rated output, and the parenthetical value of the daily PV energy is the daily PV energy compared with the daily PV power calculated using Equation (1). The specific day for each case was selected as a day with clear weather (0 ≤ average daily cloud amount < 5.5) during the experiment, according to the weather classification standards of the Korea Meteorological Administration, which resulted in an environment similar to the standard test condition (STC) of the rated output. The maximum PV power efficiency was less than 70%, and the power loss was attributed to the vertical installation of the BIPV modules [21]. The maximum PV power for each case varied because of the influence of the experimental conditions, real-time weather changes, and minute-by-minute measurements.

Figure 3 shows the BIPV operating characteristics on a day with clear weather recorded for each case. The power patterns of A and B, the BIPV systems with and without the power optimizer, were generally similar; however, from 16:00 onwards, shade thrown by the parapet was observed on the A-a module; therefore, arrays A and B differed. Specifically, in Case 1, before installing the power optimizer, the PV powers of A and B were similar. In Case 2, however, the PV power of B was higher than that of A, as the power optimizer was installed in A. This change is believed to be caused by the power consumption of the power optimizer installed in A. Case 3, the condition in which 10% shade was thrown onto each array module along with the conditions of Case 2, exhibited a more narrow PV power gap between A and B compared to that observed in Case 2, likely owing to the operation of the power optimizer. In Cases 4 to 6, where the shade ratio increased, the PV power of A was higher than that of B due to the power optimizer.

3.2. BIPV Power Loss Rate Analysis with and without a Power Optimizer

To analyze the BIPV power loss rate according to whether the system included a power optimizer, a linear regression analysis of the solar irradiance and PV power was performed for a specific day in each case (Figure 4). Equation (2) is a linear regression equation and Equation (3) is used for calculating the coefficient of determination. $x$ is the solar irradiance, and $y$ the PV power. In Equation (2), the ${\widehat{\beta }}_0$ and ${\widehat{\beta }}_1$ are the coefficients of the linear-regression equation, and the coefficient of determination in Equation (3) is expressed as $R^2$. The results of the linear-regression analysis demonstrated a linear relationship for all the coefficients of determination greater than 0.96. Module A was estimated to have low PV power (<300 W/m²) due to the shade thrown by the parapet after 16:00. In array B, the BIPV system without a power optimizer, the PV power was divided into two modules to determine the maximum power point.

$$\begin{gathered} {\widehat{y}}_i={\widehat{\beta }}_0+{\widehat{\beta }}_1{x}_i \\ {\widehat{\beta }}_0=\frac{{\sum }_{i=1}^n{y}_i}{n}-{\widehat{\beta }}_1\frac{{\sum }_{i=1}^n{x}_i}{n}=\overline{y}-{\widehat{\beta }}_1\overline{x} \\ {\widehat{\beta }}_1=\frac{n{\sum }_{i=1}^n\left(x_i{y}_i\right)-\left({\sum }_{i=1}^n{x}_i\right)\left({\sum }_{i=1}^n{y}_i\right)}{n{\sum }_{i=1}^n\left(x_i^2\right)-{\left({\sum }_{i=1}^n{x}_i\right)}^2}=\frac{\frac{1}{n}{\sum }_{i=1}^n{\{(x}_i-\overline{x})(y_i-\overline{y})\}}{\frac{1}{n}{{\sum }_{i=1}^n{(x}_i-\overline{x})}^2} \end{gathered}$$

(2)

$$R^2=1-\frac{{\sum }_{i=1}^n{\left(y_i-{\widehat{y}}_i\right)}^2}{{\sum }_{i=1}^n{\left(y_i-\overline{y}\right)}^2}$$

(3)

Figure 5 displays the PV power in each case according to the solar irradiance calculated using the linear regression equation. For reference, by comparing the calculated values of Cases 1 and 2, we found that the PV power efficiency loss increased to approximately 3% owing to the installation of the power optimizer. At low solar irradiance levels, the PV power was unstable and the PV power-loss rate was slightly higher; however, overall, the loss rates were similar and depended on the solar irradiance level. Regardless of whether the BIPV system included a power optimizer, the PV power-loss rate of each system increased with the shading ratio. This can be confirmed by Figure 5 as the slope of the linear regression equation gradually decreases as the shading ratio increases. Additionally, owing to the operation of the power optimizer, the PV power loss of B, the system without the power optimizer, was greater than that of A, which included a power optimizer.

Array B exhibited an array power loss of approximately 20% when the shade ratio of one module was 25%, and more than 60% when the shade ratio of one module was 50%. In Case 6, because the shade ratio of one module was high (75%), A and B were exhibited similar PV power loss rates.

Figure 6 summarizes the PV power-loss rates listed in Figure 5 as average values. The average PV power-loss rate of A, which included the power optimizer, was 2.1%, 6.7%, 30.7%, and 65.5% in Cases 3, 4, 5, and 6, respectively, compared to Case 2. The PV power loss rate of B, wherein no power optimizer was installed, was 4.9%, 19.8%, 63.2%, and 67.8% in Cases 3, 4, 5, and 6, respectively. Therefore, the difference in the PV power-loss rates of the PV systems with and without a power optimizer was 2.8%, 13.1%, 32.4%, and 2.3% in Cases 3, 4, 5, and 6, respectively. Thus, the effect of the power optimizer application was significant when the shade ratio of one module was in the range of approximately 10–75%.

3.3. Discussion

By analyzing the experimental results, we confirmed that the output loss was reduced by applying the power optimizer to the BIPV module. However, while conducting the study, the need for several improvements was recognized. The power optimizer stored in the BIPV and exterior material integration system needs to lower the minimum power and voltage operation values, minimize the size, improve efficiency, and improve the power-optimization control algorithm by considering the characteristics of the BIPV module. Additionally, because the power optimizer manages DC power, it can be linked to a battery that charges and discharges. In future research, we plan to develop and demonstrate a power optimizer specifically for a BIPV system considering these issues.

In urban areas, a general PV system has the disadvantage of being difficult to secure in the installation area. However, a BIPV system can be installed on the exterior wall of a building and contribute to zero building energy by producing electricity. If a power optimizer is applied to the BIPV and exterior material integration system, the PV power efficiency can be increased by minimizing the loss due to shading, and the power and maintenance fees can be reduced while ensuring safety through performance monitoring and management of individual modules. In addition, the aesthetics of the building can be improved by configuring various modules, such as color and size. The increased energy efficiency and maintainability of the devices improve the environmental friendliness of the overall energy production. Therefore, we believe that this study can contribute to carbon neutrality by reducing power consumption and carbon emissions, promoting the spread of zero-energy buildings, and maintaining energy sustainability.

4. Conclusions

This study analyzed the power-loss rates of BIPV systems with and without a power optimizer and under shady conditions through an experiment targeting two BIPV systems, each having a capacity of 342 W. In these systems, two BIPV modules were connected in series, and the experimental period was approximately two months, from September to October 2023. The power-loss rates were analyzed using the slope solar irradiance and the PV power measured at 1 min increments on specific days with clear weather.

Six experimental cases were considered as follows. In Case 1, no power optimizer was installed, while in Case 2, a power optimizer was installed in module A. In Case 3, a shade ratio of 10% was applied to one of the two modules of each array, along with the conditions of Case 2, and in Cases 4, 5, and 6, the shade ratio was increased to 25%, 50%, and 75%, respectively.

The maximum PV power efficiency was less than 70% owing to the vertical installation of the BIPV modules. By comparing the results before and after the power optimizer installation (Cases 1 and 2), we found that the PV power efficiency loss was approximately 3% owing to the power consumption of the power optimizer. Moreover, the PV power-loss rate increased with the shading ratio. Notably, while the BIPV system with the power optimizer showed a PV power-loss rate approximately twice that of the system without the power optimizer owing to its operation, this effect was not significant when the shade ratio was too low or high.

The results demonstrated that the application of a power optimizer to an elevation-type BIPV system improves its power efficiency, thereby reducing dependence on fossil fuels for power production and improving energy sustainability. Therefore, we believe that this study can contribute to the spread of zero-energy buildings and carbon neutrality.

In this study, owing to the limitations of the experimental space and the absence of a power optimizer suitable for BIPV modules, the BIPV system was implemented on a small scale and a power optimizer for general PV modules was used. In the future, based on the present research results, a BIPV-specific power optimizer that can be stored in a curtain wall should be developed considering the characteristics of the BIPV module, thus allowing for assessment of its applicability.