1. Introduction
Electricity generated by fossil fuel-based resources such as oil, natural gas, and coal are widely employed in the world currently [
1,
2]. This results in ozone layer depletion, global warming, and environmental pollution [
3,
4,
5]. It is reported that the share of renewable energy in electrical energy generation has risen from 34.7% in 2019 to 36.9% in 2021 [
6]. In parallel, the World Green Building Council (WGBC) has issued a new sight to decrease 40% carbon dioxide emissions by 2030 for realizing 100% net zero emission buildings by 2050 [
7,
8]. Consequently, the search for affordable green energy for building applications becomes more urgently for these targets. Solar energy is one of the most favourable renewable energy sources, conducive to solving the challenges related to the greenhouse effect and environmental pollution.
In light of existing solar energy techniques, photovoltaic (PV) technology is the most broadly deployed to produce electricity for building applications [
9,
10,
11]. In the UK, the PV system installations are facilitated by the government incentive measures involving the export tariff (ET) and feed-in tariff (FiT) [
12,
13]. Presently, numerous researches have been performed to optimize the PV systems for various building applications. For example, Simola et al. [
14] assessed grid-connected PV systems for building applications in Finland and demonstrated that the optimized sizes are 89 kWp for grocery store, 28 kWp for dairy farm, and 5.2 kWp for domestic building, respectively. Mateus et al. [
15] explored electrical energy output of a PV unit in a detached family house in Portugal and revealed that the PV array could meet over 60% of the house energy consumption. Nacer et al. [
16] evaluated the feasibility of PV systems in Algerian dairy farms and found that 4.8 kWp PV array is suitable for coastal field farm, and 4.32 kWp system is proper for highland farm. Osma-Pinto and Ordóñez-Plata [
17] analysed the effects of various parameters on the PV electrical energy output including roof category, height of installation, and ambient velocity. They discovered when the installation height of the PV module is in the range of 50 cm to 75 cm on a green-vegetated roof, the PV module could produce 1.3 ± 0.4% more electrical energy in comparison with one installed on a concrete roof. Saheli et al. [
18] performed a feasibility assessment of PV coupled to an oscillating water column (OWC) system in Iran and found that approximately 95% of the energy is provided by the PV module in three ports, namely, 11,885.6 kWh/year for Noshahr, 11,607.4 kWh/year for Anzali, and 13,419.2 kWh for Torkaman, respectively.
Economic evaluations of the PV system have been explored broadly in the domains of buildings and industries. A number of researches have been implemented to forecast the system payback period (PBP) and net present value (NPV) based on the life-cycle cost (LCC) and levelized cost of electricity (LCOE) approaches in different countries. Saheli et al. [
18] implemented a financial analysis of the OWC system with PV module in three different ports, Iran, and revealed that the LCOEs are 5.46
$/kWh in Noshahr, 5.59
$/kWh in Anzali, and 4.83
$/kWh in Torkaman, respectively, indicating that the Torkaman is the best place for the PV module installation. Zimmerman et al. [
19] conducted an economic analysis of vertically mounted PV panels with the third-generation material in the USA, and demonstrated that the LCOE is within the range from 18.3
$/kWh to 23.1
$/kWh when the vertical PV panels are mounted facing towards south; by comparison, the LCOE falls in the range between 11.8
$/kWh and 14.2
$/kWh when the vertical PV panels are installed facing towards east or west. Ozcan et al. [
20] completed a techno-economic investigation of a PV system with 30 years’ life expectancy and found that the system could produce approximately 3913.84 kWh~4323.94 kWh of electrical energy and the PBP is less than 7 years. Ahmed et al. [
21] discovered that the PV plant has the lowest cost of energy, 0.026 USD/kWh, resulting in the lowest PBP of 3.2 years in Pakistan. McKenna et al. [
22] built a financial methodology to control the PV self-consumption for a domestic building in the UK and confirmed that approximately 45% of the power requirement could be covered by the mounted system, leading to a power cost saving of approximately £138 annually. Koppelaar [
23] established numerical model of solar PV for investigating the net energy ratio (NER) and energy payback time (EPT) value, and found that an average of NER for mono- and polysilicon solar-PV could reach 8.6 and 9.2 times whereas the average EPT is 3.9 and 2.9 years, respectively. Mirzania et al. [
24] investigated the income loss induced by the cancellation of the FiT scheme and indicated that the return on investment (ROI) and internal rate of return (IRR) are the vital factors of the profitability in the model; the project NPV is also improved when the ROI and IRR are enhanced, especially for larger solar PV arrays.
The main objective of this study is to assess energy and economic performance of the PV systems in three typical buildings in the East Midlands area of the UK, including an office building, a domestic building, and a poultry shed. At first, in the technology aspect, the PV systems are simulated to evaluate their monthly and hourly electrical energy production. The profiles of building electrical energy consumption are obtained by automatic meter reading (AMR). Then, an innovative economic model is developed according to the Monte Carlo method, and the 25 years’ cumulative NPV and PBP are explored on the basis of the FiT scheme through the @Risk software. Afterwards, the economic sensitivity analyses including NPV and PBP distributions, NPV variation with the uncertain input parameters, and comparative distribution of probability between NPV and PBP are achieved as well. Finally, the influences of government subsidies involving the FiT and new Smart Export Guarantee (SEG) schemes are investigated for the annual savings and PBP variation. The output of this study contributes to promoting the PV application in various building categories.
4. Economic Models
One of the main contributions of this paper is employing a financial approach to assess the PV system economic characteristics by taking full account of volatile economic alteration, uncertain influence elements concerning interest and discount rates, life expectancy of assets, expected maintenance budget, and payback period based on the @RISK 8.0 software. Moreover, the yearly income and saving are required to make a composition between the SEG and FiT schemes [
40,
41]. Specifically, the expenses of the entire PV system consist of the panel, support assembly, solar inverter, electrical and mechanical fixing, as well as structure anchor and cables protections costs. As stated in the studies [
21,
22,
23], all components of PV system would have 25 years of life expectancy excluding the inverter. In order to obtain an optimistic rate of return (ROR) from the PV system, the replacement period of the inverter is anticipated for 10 years on the basis of the product specification [
28]. If the inverter has been employed beyond 10 years, some issues will occur such as ultrasonic vibrations, over-voltage, over-current, mechanical and electrical wear of capacitors, etc. These issues not only have a direct bearing on the efficiency of the PV system and thermal loading of the inverter, but also result in the extra expense in light of maintenance over the operating period.
4.1. Net Present Value and Payback Period
To achieve the optimality and efficiency design, it is necessary to assess life-cycle cost (LCC) for the feasibility of investment:
where O
IE is the components of the PV unit outlay (£); O
SEEE is the electrical energy expense of the PV unit (£); O
MP is the mortgage outlay (£); O
M&I is the maintenance and insurance outlay (£); O
ITS is the savings and income outlay of the PV unit (£); and O
PE is the periodic outlay of PV unit (£).
The NPV represents the overall cash flow of a project, in which a positive value indicates that the project is going to be profitable while a negative value leads to a net loss. The NPV is expressed as:
where O
N is the net cash inflow over the operating phase N (£); and γ is the discount rate (%).
The PBP is employed to determine the time required to recoup the fund expended within an investment:
where X is the last period with a negative discounted cumulative cash flow (£); Y is the discounted cumulative cash flow (£); and Z is the discounted cash flow (£).
4.2. Monte Carlo Method
The Monte Carlo method has been employed within the context of the economic management of the renewable energy system because it takes into account the probability density functions and processes considerable data for random input variables. Traditional economic analysis approaches, such as the LCOE and LCC, usually use point values to estimate upper and lower boundaries. This means that they are restricted, owing to the lack of a sense of the likelihood of various results. In comparison, the Monte Carlo method is a comparatively straightforward and mature approach involving uncertainty or fuzzy design and forecast parameters in quantitative models. Furthermore, the calculation process of the Monte Carlo method needs to be carried out many times, and the outcomes are based on its own setting of input designated parameters randomly from pre-defined variable distributions. Accordingly, in order to analyse the economic sensitivity of the PV system with the Monte Carlo method, a proper scope and a distribution for each variable are required to be first defined, such as log normal, normal and triangular. To be more specific, very high capital investment is possible in terms of the PV module price, which would have a big range with a higher frequency in the lower value, hence a log-normal distribution is defined as the parameter input. Similarly, the electricity price is somewhere closer to the lower termination of the scope, so a triangular distribution is chosen for the calculation process.
Table 5 illustrates the values of A, B, and C as the representation of different input parameters for the three categories of distributions. For a log-normal distribution category, A represents the mode, B represents the µ parameter, and C represents the σ parameter. For a normal distribution category, A is the average value, and B is the standard deviation. For the triangular distribution category, A is the low termination of the distribution boundary, B is the peak value, and C is the upper termination of the distribution boundary. For constant parameters, the scope denotes the value utilized. For the stochastic analysis of the economic sensitivity presented in this study, 10,000-iteration repetitions of the Monte Carlo simulation are completed by using the @RISK software. The viability of each scenario can be obtained when each input parameter is changed in the range of assumed future market condition.
Figure 3 depicts the simulation program of the economic analyses for the three PV systems.
6. Conclusions
In this study, energy and stochastic economic rentability of the PV unit application are investigated through three example cases (including an office building, a domestic building, and a poultry shed) located in the East Midlands, UK. The energy model of the PV unit is developed to assess its monthly electricity generation, module efficiency, and surface temperature variation by the EES 8.4 software; meanwhile, the economic model is established to explore the NPV and PBP for three different systems based on the Monte Carlo method and resolved by @RISK software. A 25-year economic analysis is conducted in view of the UK discount rate, inverter replacement cost, maintenance cost, cumulative savings, and NPV. Meanwhile, as the sensitivity evaluations of the PBP and NPV are executed, the comparisons of annual savings and PBP with the FiT and SEG schemes are performed as well. Therefore, some vital outcomes can be summarized as follows:
The highest and lowest electricity production periods of PV systems occur in June and December, reaching 1241.79 kWh and 332.12 kWh for the office building, 494.45 kWh and 151.33 kWh for the domestic building, as well as 1890.32 kWh and 484.94 kWh for the poultry shed, respectively.
The annual electricity generation is less than the annual electricity consumption for the three cases. However, in summer, the PV systems could produce sufficient electricity from 10:00 to 16:00 to meet the building demands.
The maximum average surface temperatures of the PV systems are in the range of 30.1 °C to 32.01 °C in June, and the corresponding electric efficiencies are in the range of 15.17% to 15.48%. By comparison, the minimum average surface temperatures are in the range of 8.87 °C to 10.18 °C in December, associated with electric efficiencies varying from 10.11% to 10.89%.
Based on the 25-year lifetime, the NPVs of the office building, domestic building, and poultry shed are £9108.4, £1717.91, and £7275.86, with the corresponding PBPs of 6.15 years, 9.12 years, and 9.41 years, respectively.
The economic sensitive evaluation indicates that the capital investment and discount rate have significant effects on the NPV and PBP; however, the income tax rate has less influence on the NPV and PBP.
Yearly savings on the basis of the FiT and SEG schemes with flexible and fixed export tariffs are £513.69, £346.37, and £215.66 for the poultry shed, £300.88, £146.43, and £84.16 for the domestic building, as well as £389.15, £177.21, and £100.33 for the office building, respectively.
The PBPs on the basis of the FiT scheme are 9.41 years for the poultry shed, 9.34 years for the domestic building, and 6.15 years for the office building, which are far less than those under the SEG scheme with a fixed export tariff for 27.16 years, 26.93 years, and 22.13 years as well as the SEG scheme with flexible export tariff for 47.18 years, 46.85 years, and 41.92 years, respectively. The FiT scheme has the shortest payback period compared with the SEG scheme.