Economic Feasibility Comparison between Building-Integrated Photovoltaics and Green Systems in Northeast Texas

Abstract

Various types of photovoltaic (PV) modules have been developed to exploit solar energy, a major renewable energy resource. One of the popular types of PV modules is building-integrated photovoltaics (BIPV), which are PV modules used as building materials. The goal of this study is to conduct an economic feasibility analysis of BIPV on the rooftop of the Keith D. McFarland Science Building at Texas A&M University, Commerce, Texas. To this end, a polynomial regression (PR) model is developed to estimate the electricity generation quantity of solar energy in the subject area, where the maximum temperature is 104 °F (40 °C) in summer. The developed PR models are used to estimate the potential profit of BIPV on the rooftop of the subject building, and the results are compared with the profit of a green roof system at the subject building. The economic feasibility analyses show that the levelized cost of electricity (LCOE) of the green roof system is approximately 39.77% higher than that of the BIPV system at a discount rate of 5%. Thus, the BIPV system is more profitable than the green roof system; consequently, this research will contribute to the implementation of BIPV on building rooftops and the expansion of renewable energy use in preference to fossil fuel.

1. Introduction

Renewable energy use to reduce greenhouse gas (GHG) emissions has been receiving worldwide attention [1]. Similar to other areas, the implementation of green building concepts has become popular in the field of building management. To this end, a traditional PV module is used as a building material, which is known as a building-integrated photovoltaic (BIPV) module [2]. Photovoltaic (PV) modules can be installed as a building skin (e.g., an outer wall) to enable the generation of renewable energy, in addition to performing the basic functions of building materials [3]. Yang and Athienitis [4] found that buildings consume 30% of global energy, so utilization of BIPV has a high potential to mitigate GHG emissions and enhance energy use efficiency. Moreover, it may be appropriate to use BIPV instead of a regular window, because this can significantly reduce solar radiation through windows into buildings, where solar radiation results in increased air conditioner use in summer [5].

BIPVs are of various types and can be classified as a roof, a window, a facade, or other shading device of a building [6]. They can be illustrated as a novel energy production system that is a combination of solar PV modules (or panels) and existing building materials. In particular, they are a highly reliable long-term system, because most parts involving PV modules have a warranty duration of 20–25 years [7]. The major advantages of BIPV can be summarized as follow: (1) innovative design, (2) sunscreen and power generation, (3) reduction in the carbon footprint of the building, (4) thermal insulation, (5) acoustic insulation for increased comfort, (6) increased value of the building, (7) environment-friendly nature, and (8) sustainability [6]. Notwithstanding the many advantages, the most critical factor to consider in BIPV installation is associated with environment-friendly building construction (or the construction of net zero energy buildings (NZEBs)) to meet the sustainable goal of worldwide agencies (e.g., the net zero emissions by 2050 scenario) [8]. For this reason, the BIPV market is expected to be one of the fastest-growing areas in the PV industry. The estimated growth from 2011 to 2017 was approximately 50% or more [6]. In 2022, the global BIPV size was valued at USD 19.82 billion, and the compound annual growth rate from 2023 to 2030 is expected to be 21% [9]. However, 80% of BIPV products are still based on rooftop PV modules (e.g., roof-integrated PVs, skylights, atria, and traditional rooftop PVs) even though there are multiple types of BIPVs [2].

Although the outlook for the BIPV market is positive, this does not guarantee economic success in the management of all buildings. The performance of a BIPV system in terms of solar energy productivity and insulation capability can vary according to its installation environment. As BIPV plays multiple functional roles (e.g., solar energy generation and building operational cost reduction), multiple factors, such as the photovoltaic module temperature, shading ratio, installation angle, orientation, and climate issues, should be simultaneously considered [10]. This implies that the performance of an individual BIPV case should be accurately estimated, and its economic feasibility should be investigated by using an appropriate assessment tool.

In fact, Gholami and Røstvik [11] showed that a BIPV system is a feasible alternative as an envelope material for a building, after considering cases in 30 countries. Lie et al. [12] claimed that a BIPV system should be installed in high solar irradiance regions to achieve economic feasibility in building management. Sun et al. [13] conducted a feasibility assessment of a BIPV system for a commercial building in a high-density urban district. Similar to Lie et al. [12], solar radiation is identified as the most critical factor to determine the electricity generation performance of a BIPV system. Kamel and Fung [14] addressed energy saving from their feasibility analysis of a BIPV system combined with an air-source heat pump. However, because the existing studies have not considered multiple factors (solar radiation, maximum temperature, daily minimum temperature, precipitation, and wind speed) for their energy generation quantity estimations of BIPV systems, they may have conducted the economic feasibility analysis inaccurately. Note that the performance of a traditional PV system is heavily influenced by the microclimate conditions [15]. This implies that an electricity estimation model under various microclimate conditions should be adopted for an accurate feasibility analysis of a BIPV system.

The goal of this study is to conduct an economic feasibility analysis of BIPV on the rooftop of the Keith D. McFarland Science Building at Texas A&M University, Commerce, Texas at 33.2410° N, 95.9104° W. The rooftop PV, which is the most popular BIPV type, is considered due to its feasibility in terms of implementation cost and material availability. In addition to the feasibility analysis of BIPV installation, the study considered green roof installation, which is another popular building management option for green buildings (referring to energy-efficient buildings) as well as vegetated buildings [16]. In particular, a green roof (a vegetation roof), which is one type of vegetated building, is known as a thermal- and energy-efficient building because it protects the building from excessive solar radiation in summer [16,17,18,19,20]. Moreover, He et al. [21] showed the efficient hydrological performance of a green roof in Singapore, which has a tropical climate. Due to these advantages, the green roof market size was valued at USD 1.41 billion in 2022, and its compound annual growth rate from 2023 to 2030 is expected to be 16.2% [22]. In other words, the market size is expected to reach USD 1.41 billion by 2023. According to the planting medium depth [23], there are three types of green roof: intensive (for trees and shrubs), semi-intensive (for shrubs and sedums), and extensive (for grasses and sedums). Although an intensive green roof can provide a thicker roof to protect a building’s surface from direct solar radiance [24], this study considers the extensive green medium (less than 6 inches) that is widely adopted in practice [25]. The intensive green roof has a heavy weight of soil, so thoughtful inspection of the structural safety of a building is required [24]. On the other hand, the extensive green medium can also mitigate the surface temperature of a building, even though its depth is less than 6 inches [25]. However, similar to BIPV, a green roof requires high investment and management costs, so this study conducts an economic feasibility analysis on both systems.

To conduct the feasibility analysis, first, by using the historical climate data of the rooftop of the subject building, the potential solar energy generation quantity (i.e., electricity quantity) is estimated. To this end, a polynomial regression (PR) model is developed to capture the non-linear variability between multiple climate factors (e.g., temperature, humidity, solar radiation) and the electricity generation quantity (i.e., a dependent variable) [26]. In addition, the electricity use of the subject building involving a heating, ventilation, and air conditioning (HVAC) system is estimated, so that the economic feasibility of BIPV on the rooftop of the subject building can be analyzed. The result is also compared with the economic feasibility of a green roof system. The economic feasibility analysis shows that the levelized cost of electricity (LCOE) of the green roof system is approximately 39.77% higher than that of the BIPV system at a discount rate of 5%. Thus, we can conclude that the BIPV system is more profitable than the green roof system. As a result, this research will contribute to the implementation of BIPV on building rooftops and the expansion of renewable energy use.

The remainder of the paper is organized as follows. Section 2 illustrates the historical climate data and field study data of a green system at the Keith D. McFarland Science Building at Texas A&M University, Commerce, Texas. Section 3 introduces a new BIPV system at the subject building and a solar energy estimation model based on polynomial regression. Section 4 shows the economic feasibility comparison results between a BIPV system and a green roof system on the rooftop at the subject building. Section 5 concludes the study and findings.

2. Data Collection

As mentioned in Section 1, this study collected data from the rooftop at the Keith D. McFarland Science Building at Texas A&M University, Commerce, where the average maximum temperature in summer is 104 °F (40 °C). The floor space of the subject building is 10,169.45 m², and the size of the rooftop available for the green roof installation is 2378.97 m². Figure 1 illustrates the subject building and its rooftop [27]. Currently, the rooftop consists of an outdoor unit of a HAVA system and a green roof system comprising 86 grids (0.18581 m² per grid) with four different plant types, namely purple trailing lantana (Lantana montevidensis), new gold lantana (Lantana × hybrida ‘New Gold’), white trailing lantana (Lantana montevidensis ‘White’), and hardy ice plant (Delosperma cooperi) [28]. The green roof is four inches (10.16 cm) in depth with an extensive green medium. The planting period is from 30 May 2017 to October 2017.

Table 1. Historical climate data in Commerce, Texas (edited from [29]).

Month	Max. Temperature (°C)	Avg. Temperature (°C)	Min. Temperature (°C)	Precipitation (mm)	Wind (m/s)
January	20.32	11.02	−4.44	2.03	4.11
February	23.72	16.07	7.92	2.03	4.81
March	24.60	18.42	9.17	0.51	5.18
April	25.18	20.78	14.65	2.54	5.44
May	29.38	24.04	18.08	1.02	5.04
June	31.23	27.37	21.91	4.57	4.73
July	33.38	30.26	24.34	1.78	3.99
August	32.04	28.61	24.59	5.08	3.78
September	30.69	26.81	20.99	0.51	3.91
October	28.02	20.98	7.99	1.78	4.16
November	26.27	16.37	8.56	0.76	3.80
December	21.70	9.07	−5.00	3.30	3.67

describes the monthly climate data in Commerce, Texas, U.S. in 2017 [29]. Commerce has 2.16 mm of rainfall and 4.39 m/s of wind monthly, on average. In particular, 44.12% of the total precipitation of 25.91 mm is in June, July, and August. The annual average temperature is 20.82 °C, so the location is classified as a humid subtropical climate (i.e., Cfa) under the Köppen climate classification [30]. The subject location has a mild winter and hot summer.

The electricity consumption of the subject building was collected over 2017, and

Table 2. Energy consumption data of the subject building in Northeast Texas.

Month	Electricity Consumption (kWh)	Electricity Cost (USD)	HVAC (KWh)	Max. Surface Temperature (°C) 1	Max. Green Roof Temperature (°C) 2
January	338,347.93	57,519.15	140,865.43	44.56	22.89
February	248,121.81	42,180.71	50,639.32	47.91	26.24
March	236,843.55	40,263.40	39,361.06	48.77	27.10
April	197,482.49	33,572.02	0.00	49.34	27.67
May	200,753.10	34,128.03	3270.61	53.47	31.80
June	225,565.28	38,346.10	28,082.79	55.28	33.61
July	248,121.81	42,180.71	50,639.32	57.40	35.73
August	257,144.42	43,714.55	59,661.93	56.08	34.41
September	250,377.47	42,564.17	52,894.97	54.75	33.08
October	230,076.59	39,113.02	32,594.10	52.13	30.46
November	214,287.02	36,428.79	16,804.53	50.41	28.74
December	270,678.34	46,015.32	73,195.85	45.92	24.25

¹ The maximum building surface temperature; ² the maximum building surface temperature under the green roof.

shows the collected data [31]. The building has an annual energy consumption of 2,917,799.83 kWh/year, so its annual electricity bill is USD 496,025.97. Note that the unit electricity price is USD 0.17/kWh. April has the lowest monthly electricity use of 197,482.49 kWh, due to the low utilization of HVAC. Similarly, May and November have low costs of electricity. On the other hand, January has the highest consumption of electricity due to the use of the heating function of HVAC in winter, while August has the highest consumption of electricity to cool the subject building in summer. Given the size of the building, electricity of 23.91 kWh/m²/month is consumed.

In addition, the maximum green roof temperature is lower than the maximum surface temperature of the subject building, so it is a useful system to reduce the use of HVAC and the ensuing electricity cost.

Table 3. Electricity use (kWh) of the subject building with a green roof system in Northeast Texas.

Category	January	February	March	April	May	June	July	August	September	October	November	December
Non-green roof	232,377	207,204	204,653	203,863	219,270	237,836	268,465	248,233	231,672	210,216	204,293	219,226
Green roof	201,536	198,659	198,324	198,196	199,457	201,231	204,231	202,243	200,636	198,628	198,154	200,064

describes the electricity use of the subject building under two cases: (1) a non-green roof case, and (2) a green roof case. Approximately 11% of the total electricity use can be reduced by the use of a green roof system. However, additional economic feasibility analysis should be conducted, because the green roof system requires additional installation, operation, and maintenance costs (see Section 4.1).

3. Building-Integrated Photovoltaics on a Rooftop

This study considers a BIPV system on the rooftop of the Keith D. McFarland Science Building mentioned in Section 2. Figure 2 shows the BIPV system on the subject building. It consists of PV modules, a surge protector, a junction box, a DC/AC inverter, and a power meter. The generated electricity can be sent to either an external grid system or an internal grid system. Similar to the green roof system mentioned in Section 2, the subject BIPV system has the same rooftop space of 2378.97 m². Therefore, approximately 1189 mono-facial solar modules (i.e., LG405N2W-V5 with a size of 2 m × 1 m) with frames can be installed. In other words, the subject BIPV system can have a capacity of 481.55 kWp.

3.1. Electricity Generation Modeling

Table 4. Solar radiation and solar energy generation in Dallas, Texas.

Month	Solar Radiation (kWh/m2/day) 1	Electricity Generation Quantities (GWh/day) 2
		2017	2018	2019	2020	2021	2022
Jan.	4.85	2.84	6.58	7.77	11.42	20.90	42.32
Feb.	5.01	4.04	6.96	8.54	15.00	23.11	50.39
Mar.	5.31	5.65	8.19	9.32	14.29	32.32	54.52
Apr.	5.59	6.63	8.33	14.27	19.07	31.83	58.80
May	5.44	7.74	9.48	12.84	27.16	39.77	67.94
Jun.	5.68	7.70	12.67	15.90	30.53	47.47	80.20
Jul.	6.18	7.81	11.77	15.84	37.61	49.87	85.87
Aug.	6.26	7.03	11.35	14.97	34.68	52.10	71.84
Sep.	6.11	6.07	8.93	13.20	28.10	54.60	77.97
Oct.	5.5	6.48	7.00	11.61	23.45	43.90	61.10
Nov.	5.12	5.07	7.77	9.07	19.70	35.63	41.47
Dec.	4.53	4.74	6.26	10.06	19.00	32.23	35.81

¹ Data are adopted from [32]; ² data are adopted from [33].

describes the solar radiation and generation quantities of solar energy in Dallas, TX [32,33]. Since the subject area utilizes the weather station in Dallas Love Field, TX, similar solar radiation intensity and electricity generation patterns can be applied to the BIPV system on the rooftop of the subject building.

Figure 3 compares the solar radiation and solar energy generation data illustrated in Table 2. The data are normalized according to Equation (1):

$$X^{\prime }=\left(X-min\right)/\left(max-min\right)$$

(1)

By using the min–max normalization, Figure 3 shows that the solar energy generation quantities are heavily dependent on the solar radiation quantities. The R² values of years 2017, 2018, 2019, 2020, 2021, and 2022 are 55.04, 61.01, 67.08, 68.28, 71.98, and 82.18%, respectively.

$$R^2=1-RSS/TSS$$

(2)

In Equation (2), RSS is the sum of squared residuals, and TSS is the total sum of squares. Note that R² is the coefficient of determination, which is a well-known statistical tool to measure the estimation accuracy of a dependent variable from the independent variables [34].

In addition, the climate data and solar energy generation data are used to develop an electricity generation quantity estimation model based on polynomial regression (PR), which is one of the popular machine learning (ML) approaches. Note that its coefficients can be utilized to address the impact of independent variables on a dependent variable (i.e., electricity generation quantity) [1,26]. Moreover, unlike the simple regression model, PR enables the nonlinear relationship between variables to be captured [35]. The JAVA-based PR library proposed by [34] is used with $L=5$ in Equation (3). The modeling is conducted under the computing environment of Intel Core i5-8250U CPU @1.60 GHz. Equations (3) and (4) represent the PR model [1].

$$Y=g\left(X_1,\dots ,X_n\right)={\beta }_0+f_1\left(X_1\right)+\dots +f_n\left(X_n\right)+\epsilon , \epsilon ~N\left(0,{\sum }_{j=1}^n{\sigma }_j^2\right)$$

(3)

$$f_j\left(X_j\right)={\beta }_{j1}\left(X_j\right)+{\beta }_{j2}\left(X_j^2\right)+\dots +{\beta }_{jL}\left(X_j^L\right), j = 1, 2, \dots , n$$

(4)

where $f_j\left(X_j\right)$ is a polynomial function of $X_j$; ${\beta }_0={\sum }_{j=1}^n{\beta }_{j0}\left(X_j^0\right)$; $X_j^0=1$; ${\sum }_{j=1}^n{\beta }_{j0}\left(X_j^0\right)={\sum }_{j=1}^8{\beta }_{j0}$; ${\beta }_j$ is a coefficient of $X_j$; and ${\beta }_0$ is a constant. Supposing that the daily solar radiation (X₁), daily maximum temperature (X₂), daily minimum temperature (X₃), precipitation (X₄), and wind speed (X₅) are independent variables to estimate electricity generation quantity (Y) of solar energy, then

$$Y=0.00655+0.064X_1+1.251X_2-0.807X_3+0.429X_3^3+0.071X_4+0.322X_5^2$$

(5)

Equation (3) is used to estimate the electricity generation quantities of the six years, 2017–2022, and the R² values of the 2017, 2018, 2019, 2020, 2021, and 2022 data sets are 83.92, 84.59, 86.94, 91.29, 77.86, and 87.55%, respectively. Note that Equation (5) uses the normalized data according to Equation (1).

3.2. Levelized Cost of Energy

The levelized cost of electricity (LCOE) for a photovoltaic (PV) system can be represented as the cost of the BIPV system divided by the energy generation quantities over its lifetime [36]:

$$LCO{E}_1=\frac{{\sum }_{t=0}^T{C}_t}{{\sum }_{t=0}^T{Q}_t}$$

(6)

where $C_t$ is the cost (USD) of BIPV at time t; $Q_t$ is the energy generation quantity at time t; and T is the lifetime of the BIPV. In this case, the LCOE₁ can be represented in USD per kWh. By dividing the net present value (NPV) of total costs over the lifetime by the net present value (NPV) of electrical energy produced over the lifetime [37], Equation (6) can be rewritten as Equation (7):

$$LCO{E}_2=\frac{{\sum }_{t=0}^T{C}_t/{\left(1+r\right)}^t}{{\sum }_{t=0}^T{E}_t/{\left(1+r\right)}^t}$$

(7)

where $E_t$ is the energy value (USD) produced at time t; and r is the discount rate (%) [38]. Equation (7) can be expanded as Equation (8):

$$LCO{E}_2=\frac{{\sum }_{t=0}^T\left(I_t+O_t+M_t+F_t\right)/{\left(1+r\right)}^t}{{\sum }_{t=0}^T{E}_t/{\left(1+r\right)}^t}$$

(8)

where $I_t$ is the investment cost (USD) of BIPV at time t; $O_t$ is the operational cost (USD) of BIPV at time t; $M_t$ is the maintenance cost (USD) of BIPV at time t; and $F_t$ is the interest expenditure (USD) at time t [36].

4. Economic Feasibility Comparison

The proposed modeling techniques are used to conduct the economic feasibility comparison between a green roof system and a BIPV system. The electricity generation model is used to estimate the electricity quantities produced by both systems, and the LCOE values are computed for the economic feasibility comparison. Figure 4 represents the procedure of the economic feasibility comparison between green roof and BIPV systems. Once the subject systems are identified, performance measurements such as the LCOE, data collection period, discount rate, and pricing parameters (e.g., unit electricity price per kWh) are defined. In addition, data involving the investment cost, operational cost, maintenance cost, interest expenditure, and energy value are collected to compute the LCOE. After the calculation, an alternative comparison is conducted in terms of the LCOE, net present value (NPV), and break-even point (BEP), and the best alternative is selected. The detailed information is addressed in the following sections.

4.1. Economic Feasibility of a Green Roof System

As mentioned in Section 2, the subject green roof system enables a reduction in the electricity use of HVAC from 317,519 to 31,571 kWh/year, so the total electricity consumption can be reduced from 2,687,309 to 2,401,36 kWh/year. Assuming the unit electricity price of USD 0.17/kWh in Texas, U.S. [39], the energy-use reduction of HVAC can eventually mitigate the electricity cost of the subject building from USD 456,842.57 to 408,231.31/year. However, the operation and maintenance costs of the green roof system illustrated in

Table 5. Installation, operation, and maintenance costs of the green roof system.

Category	Item	Quantity	Unit Cost (USD)	Cost (USD)
Installation	Extensive grid (grid)	6394	5	31,970.00
	Planting medium (ton) 1	261.73	68	17,797.49
	Plants (plant) 2	25,576	9.99	255,504.24
	Irrigation system (unit)	1	3800	6000
Operation and maintenance	Water (L) 3	1,116,801.8	0.00185	2065.20
	Electricity (kWh) 4	6172.49	0.17	1049.32

¹ Composition (volume): 1/3 coarse horticultural vermiculite, 1/3 peat moss, and 1/3 blended compost; ² each extensive grid has four plant types (i.e., purple trailing lantana (Lantana montevidensis), new gold lantana (Lantana × hybrida ‘New Gold’), white trailing lantana (Lantana montevidensis ‘White’), and hardy ice plant (Delosperma cooperi)); ³ water rates are considered from [40]; ⁴ electricity is used for a water pump connected to the irrigation system [41].

should be considered to conduct an economic feasibility analysis of the green roof system.

In Table 5, the investment cost of the green roof system is USD 321,271.73, with 6394 extensive green roof grid units in a rooftop space of 2378.97 m². The average installation cost of USD 130.42/m² is within the range of USD 107.64 to 215.28/m² suggested by [42]. In addition, considering the average water use for gardening in Texas [43], each grid (i.e., 0.61 m × 0.61 m × 0.11 m) requires 174.66 L/year. This implies that 1,116,801.8 L/year (USD 2065.20/year) of water would be consumed, together with electricity use of 6172.49 kWh/year (USD 1049.32/year). In other words, the annual expenditure is USD 3114.52/year, and the annual saving is USD 48,611.26. Figure 5 shows this result. The BEP of the subject green roof is approximately 8.4 years at an interest rate of 5%, and 7.5 years at 2%. Note that all the values in Figure 5 are net present values.

4.2. Economic Feasibility of a Building-Integrated Photovoltaics System

The economic feasibility of the subject BIPV mentioned in Section 3 is analyzed in a similar manner to the green roof system. According to [44], the PV module efficiency can vary over the year, and the conversion rate of solar irradiance into electric energy is 6–20%. In particular, in Texas, the minimum and maximum efficiency rates of PV modules are 15 and 20%, respectively [45]. This implies that the minimum and maximum electricity generation quantities in Texas can be 4.12 and 5.12 kWh/kWp/day, respectively. Figure 6 represents the relationship between electricity generation quantities and PV module efficiency. The electricity generation quantity data under different PV module efficiencies are collected from [46,47,48,49,50,51,52,53,54,55]. As the PV module efficiency increases, the electricity generation quantity increases.

Regarding the capacity of the subject BIPV system (i.e., 481.55 kWp), the minimum and maximum electricity generation quantities are 1984.43 and 2466.92 kWh/kWp/day, respectively. Figure 7 shows the monthly electricity generation quantities. Equation (5) is used to estimate the electricity generation quantities. Annually, an electricity quantity of 828.56 MWh/year can be produced by the subject BIPV, and its value is USD 140,855.86/year considering the unit electricity price of USD 0.17/kWh in Texas, U.S. [39].

Table 6. Installation, operation, and maintenance costs of the BIPV system.

Category	Item	Quantity	Unit Cost (USD)	Cost (USD)
Installation	PV module cost (module)	1189	141.75	168,540.75
	Structure cost (USD)	1	278,967.09	278,967.09
	Electric distribution system cost (USD)	1	132,870.07	132,870.07
	Other costs (USD) 1	1	572.84	572.84
Operation and maintenance	Operation and maintenance (USD) 2	1	11,619.02	11,619.02

¹ Other costs consist of a building permit fee and a connection cost to an external electric distribution system; ² these in total are assumed to be 2% of the total installation cost [56].

shows that the investment cost of the BIPV system is approximately USD 580,950.75, with 1189 PV modules in a rooftop space of 2378.97 m² and an average installation cost of USD 244.20/m². In addition, the operation and maintenance costs are assumed to be 2% of the investment cost of the BIPV system [56], so its value is USD 11,619.02/year. Based on the data illustrated in Table 6, LCOE₁ can be computed via Equation (6), and its result is USD 0.049/kWh, which is within the range of USD 0.0491–0.0605)/kWh suggested by [57]. Note that the values of the LCOE₁ of PV modules on a rooftop are USD 0.039–0.0680)/kWh [58] and USD 0.04–0.06)/kWh [59]. Considering the energy use reduction quantity of the total electricity consumption of 5,718,960 kWh for a 20-year lifetime, the LCOE₁ of the green roof system is USD 0.065/kWh. Thus, in terms of LCOE₁, the BIPV system is a more attractive system than the green roof system. However, since the BIPV system requires an 86% higher installation cost than the green roof system, it may be difficult to find potential users of BIPV systems.

To identify the BEP of the subject BIPV, Figure 8 illustrates the NPVs of cost and revenue over the lifetime of the BIPV system. Similar to Section 4.1, the unit electricity price of USD 0.17/kWh in Texas, U.S. is considered [39]. The annual income is USD 140,855.86/year from the electricity production of 828.56 MWh/year, and the BEP of the subject BIPV is approximately 5.4 years at an interest rate of 5% and 4.7 years at 2%. Figure 9 illustrates the BEPs over different discount rates from 0.01 to 0.15. Because the BIPV system has a higher annual income than the green roof system, its BEPs are relatively stable under different discount rates.

Table 7. Levelized cost of electricity (LCOE) of the green roof and BIPV systems.

Discount Rate (%)	Green Roof System			BIPV System
	Discounted Cost (USD)	Discounted Energy Value (USD)	LCOE2	Discounted Cost (USD)	Discounted Energy Value (USD)	LCOE2
1.00	367,474.97	877,217.07	0.4189	790,622.30	2,541,821.88	0.3110
2.00	362,198.60	794,863.78	0.4557	770,938.30	2,303,195.21	0.3347
3.00	357,607.92	723,212.80	0.4945	753,812.35	2,095,579.52	0.3597
4.00	353,599.07	660,642.89	0.5352	738,856.96	1,914,277.11	0.3860
5.00	350,085.53	605,803.75	0.5779	725,749.36	1,755,375.36	0.4134
7.00	344,267.00	514,988.38	0.6685	704,042.76	1,492,228.99	0.4718
10.00	337,787.39	413,855.06	0.8162	679,869.97	1,199,185.34	0.5669
15.00	330,766.54	304,273.99	1.0871	653,678.02	881,663,52	0.7414

shows the values of the LCOE₂ of both the green roof and BIPV systems computed by Equation (8) under various interest rates between 1 and 15%. At the discount rate of 5%, the LCOE₂ of the green roof system is 39.77% higher than that of the BIPV system. Thus, we can conclude that the BIPV system is more profitable than the green roof system. Although the BIPV system requires higher installation costs than the green roof system, its BEP is 4.7 years shorter than that of the green roof system, due to its higher energy value (or annual revenue shown in Figure 8).

As a result, we conclude that the BIPV system is more profitable than the green roof system under the subject building. However, we should note that a green roof system is an appropriate option for a vegetated building, providing an air purification function and a habitat for birds and butterflies in addition to the energy use reduction [60]. Moreover, similar to other farming systems (e.g., agrophotovoltaic systems), the green roof system enables consideration of producing cash crops or vegetables to generate additional sales profits [15]. Thus, it is critical to determine an appropriate goal before the economic feasibility analysis is conducted.

5. Conclusions

This study conducted economic feasibility analyses of two popular rooftop renewable energy systems: a green roof system, and a BIPV system. In particular, the system installation on the rooftop of the Keith D. McFarland Science Building at Texas A&M University, Commerce, Texas (33.2410° N, 95.9104° W) is considered for an accurate estimation of the levelized cost of electricity (LCOE) of both systems. For the LCOE estimation of the green roof system, the historical climate and electricity use data of the subject building involving a heating, ventilation, and air conditioning (HVAC) system are collected. In addition, the installation, operation, and maintenance costs of a green roof system at the subject building are analyzed. To estimate the potential solar energy generation quantity (i.e., electricity quantity) by BIPV, a polynomial regression (PR) model is developed. Since PR enables the non-linear variability between multiple climate factors (e.g., temperature, humidity, solar radiation) and electricity generation quantity (i.e., a dependent variable) to be captured, this study developed the PR model with high prediction accuracy. The R² values of the PR model under the given test electricity generation data sets range (77.86–91.29%). The electricity generation quantities estimated by the PR model are used to predict the annual revenue of the BIPV system. The experiment results show that the levelized cost of electricity (LCOE) of the green roof system is approximately 39.77% higher than that of the BIPV system at a discount rate of 5%. In other words, the BIPV system is more profitable than the green roof system, even though it requires 86% higher installation cost than that of the green roof system. As a result, this research will contribute to the implementation of BIPV on the rooftops of buildings and the expansion of renewable energy use.

In future studies, other BIPVs such as a skylight type, an atria type, a shading type, a cladding type, and a curtain wall need to be considered for economic feasibility analysis. Although this study selects the rooftop BIPV due to its feasibility in terms of implementation cost and material availability on the subject building, more options should be considered to identify the most profitable solution for building management. Moreover, the study has to be extended to various cases for the generalization of the study result.