Designing circular economy strategies in distributed generation for small- andmedium-sized enterprises using Monte Carlo simulation

DOI: 10.1111/jiec.13536 RESEARCH ARTICLE Designing circular economy strategies in distributed generation for smalland medium-sized enterprises using Monte Carlo simulation Jaime González-Domínguez1Gonzalo Sánchez-Barroso1 Francisco Zamora-Polo2Justo García-Sanz-Calcedo1 1Departamento de Expresión Gráfica, Universidad de Extremadura, Badajoz, España 2Departamento de Ingeniería del Diseño, Escuela Politécnica Superior, Universidad de Sevilla, Sevilla, España Correspondence Justo García-Sanz-Calcedo, Departamento de Expresión Gráfica, Universidad de Extremadura, Avenida de Elvas, Badajoz, España. Email: [email protected] Editor Managing Review: Jooyoung Park Funding information Interreg VA España-Portugal Program, Grant/Award Number: 0475_LOCALCIR_4_E; Junta de Extremadura (cofounded by European Regional Development Fund), Grant/Award Number: GR21098 Abstract The circular economy (CE) strategies in energy communities enable firms to efficiently manage the excess of photovoltaic energy they produce, and thereby enhance their sustainability. Thus, the present research aims to compare the economic and financial profitability and greenhouse gas (GHG) emissions of shared photovoltaic self-consumption versus individual self-consumption in the region of Extremadura (Spain). Six firms with complementary energy profiles were selected, analyzing their hourly energy consumption. In addition, the Monte Carlo method was used to generate 30,000 simulations, reducing the uncertainty caused by the variability of the firms’ energy consumption. The results show that collective generation covers the energy needs more efficiently, reducing the cost of energy consumed by 14.38% and generating better cost–benefit ratio. They also show that the CE strategy of the energy community allows firms to obtain a considerable reduction of GHG emissions associated with the photovoltaic energy consumed. KEYWORDS circular economy, energy community, industrial ecology, Monte Carlo, photovoltaic energy, SMEs 1INTRODUCTION The circular economy (CE) represents a new model of production and consumption, leaving a linear economy of production, use, and disposal (López Ruiz et al., 2020) with the aim of reducing resource consumption, waste, and minimizing environmental impact (Momete, 2020). Its inclusion in energy communities significantly affects the way energy is managed in cascades, transitioning toward a collaborative economy (Gomes et al., 2022). Energy communities provide support for this energy transition. The energy transition and the circular economy share a similar focus on such concepts as sustainable growth while respecting the environment and the economy (Chen & Kim, 2019). Implementing circular economy strategies in energy community projects could facilitate the energy transition and enhance the overall circularity of these initiatives (Mishra et al., 2022). In this context, some authors have carried out research on energy-based industrial symbiosis (Fraccascia et al., 2021). This strategy is based on reducing the amount of energy coming from external industrial systems, allowing concomitant reductions in environmental impact (Aissani et al., 2019) and energy dependence (Liu et al., 2017). More specifically, the aim is to reuse the waste materials and energies obtained during different This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. © 2024 The Author(s). Journal of Industrial Ecology published by Wiley Periodicals LLC on behalf of International Society for Industrial Ecology. Journal of Industrial Ecology 2024;1–14. wileyonlinelibrary.com/journal/jiec 1 2GONZÁLEZ-DOMÍNGUEZ ET AL. production processes, allowing energy costs to be reduced through a circular approach (Shah et al., 2020). In this context, there appears the concept of energy cascade models which allow the residual energy of one firm to be used by another (Kikuchi et al., 2016). Tsvetkova et al. (2015) designed a biogas-based distributed energy system to be replicated in six locations in order to achieve economy of replication. Li et al. (2015) developed indicators to evaluate the efficiency of industrial symbiosis and found that these strategies improve the firms’ energy efficiency and financial costs. However, most research in this area has focused on the use of waste and heat to generate energy for other firms rather than on the use of surplus distributed energy. Distributed solar power generation (Roberts et al., 2019) increases sustainability and contributes to compliance with new environmental measures. The production of electricity close to the company reduces the losses associated with its distribution (Ruiz-Romero et al., 2013). One option for the use of surplus photovoltaic (PV) energy generated is to sell it to the electricity market (State Government, 2018, 2019). However, this is a less than optimal option economically and environmentally. In this sense, another option is to generate electricity collectively in what are known as energy communities which can maximize the benefits of PV installation. In this way, collective distributed power generation has the potential to minimize electricity consumption and environmental impact through the optimal sharing of surplus PV energy. As an instrument for collective energy generation, energy communities are defined in Directive (EU) 2018/2001 as a legal entity open to voluntary participation and controlled by stakeholders, whose participants are natural persons, small and medium-sized enterprises (SMEs), and local authorities (European Parliament and of the Council, 2018). These initiatives seek to provide the participants with economic, social, and environmental benefits. Previous research has addressed the relationship between the circular economy and energy communities. Thus, this strategy allows for the appropriate use and optimization of surplus energy and contributes to the principles of circular energy (Mishra et al., 2022). Therefore, collective distributed generation contributes to economic growth, social development, and environmental responsibility (Palafox-Alcantar et al., 2020). The potential of distributed energy generation has been analyzed for in different types of buildings. Yan et al. (2021) established the most appropriate technology combinations for distributed energy generation in three types of commercial buildings and different climate zones, using multidisciplinary design optimization (MDO). López Prol and Steininger (2020) determined the economic profitability of photovoltaic installations in residential, commercial, and industrial buildings in Spain, establishing the opportunities for self-consumption or shared storage. Fina et al. (2019) evaluated the profitability of energy communities made up of mostly residential buildings in comparison with the solar energy generation of individual buildings. Li and Ma (2020) determined the benefits and limitations of peer-to-peer (P2P) electricity trading in residential communities, analyzing the factors that influence this new form of inter-community energy trading. There are precedents on collective PV power generation in residential buildings. However, few of them are applied to distributed PV generation in a cluster of firms. Analyzingscientificcontributions,thisresearchaimstofillthescientificgapintheevaluationofdistributed power generationstrategyinSMEsin an area with high solar radiation. In this way, it will analyze and quantify the profitability and the reduction of the environmental impact (greenhouse gas[GHG]emissions)whenSMEsdecide to produce energy collectively. The Monte Carlo method is implemented in order to reduce the uncertainty associated with the variability of the firms’ energy demand, increasing the applicability of the results and allowing useful information to be obtained for the scientific community. 2METHODS 2.1 General overview To develop the energy community, six SMEs from different economic sectors were selected. They were located close to each other in the Extremadura region (Spain). Extremadura is in the west of Spain and borders with Portugal, at longitude −6.15◦and latitude 39.2◦. It has an area of 41,635 km2and a total of 1,054,245 inhabitants. Figure 1shows the map of the region of Extremadura. Two scenarios of distributed photovoltaic energy generation were considered for this research. Each of the steps of the methodology, applications, and scenario considered will be detailed below. Figure 2shows a flow chart for each scenario in this research. The Monte Carlo method allows the inherent uncertainty due to the variability of the firms’ energy demand to be reduced (Monie et al., 2021) since it generates a multitude of simulations of their energy consumption for different load situations: low, medium, and maximum (Díaz López, 2018). In this way, the results will not be affected by changes in the firms’ energy demand. 2.2 Firm selection and facility design The firms selected have to be located close enough together to form part of an energy community. While the number of firms that constitute that community can vary, it must be sufficient to guarantee the complementarity of the energy flow during the hours of photovoltaic production. First, the firms’ energy profile was drawn up based on the hourly energy consumption data supplied directly by them. This was followed by the selection of firms with complementary patterns of energy consumption to ensure collective photovoltaic self-consumption. For this purpose, the energy 15309290, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jiec.13536 by Universidad De Sevilla, Wiley Online Library on [26/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License GONZÁLEZ-DOMÍNGUEZ ET AL.3 FIGURE 1 Map of the region of Extremadura (Google, 2005). consumption was observed during the hourly period studied for this research which corresponded to the daily hours in which photovoltaic energy can be generated. This period depends on the season of the year and includes a part of the total firms’ daily consumption. The case analyzed in this research had six firms to ensure an efficient energy flow within the energy community. The selected firms were SMEs in accordance with the European Union criteria as they had a size of 0 to 250 employees and a turnover of less than 50 million euros (The Commission of the European Communities, 2003). Thus, the participants of the energy community comply with the definition of the European Directive (European Parliament and of the Council, 2018). Firms 1 and 2 were restaurants, which fall within the industrial sector of hotel and catering activities. Firms 3 and 4 were car repair firms, whose industrial sector is the sale and repair of motor vehicles and motorbikes. Firm 5 dealt with the sale of food products and is included in the industrial sector of retail trade of food, drinks, and tobacco products in specialized establishments. Finally, firm 6 was financial brokerage and its industry sector is financial and insurance activities. Once the complementarity of the energy profiles of the selected firms was ensured; the mean monthly energy demand over 1 year was quantified. To obtain this value, the energy demanded by each firm over 3 years was used. In addition, the monthly consumptions were used to see the economic and financial viability of each of the scenarios. The photovoltaic system was sized to cover the energy demand during the period studied. The parameters and data used in this research to calculate the mean monthly production of the photovoltaic installation are given in Supporting Information S2 (Appendix 1). 2.3 Scenario description The first scenario consisted of an individual installation for each firm so that the excess produced is sold on the electricity market through the trading corporation, in accordance with the regulatory norm (State Government, 2019). In order to size the installation, each firm’s energy profile was studied. In this first scenario, the installation was sized to cover at least 50% of the energy demand during the least favorable months in the period studied. In this way, most of the energy demand was covered during the most favorable months, generating few surpluses, and obtaining a higher profitability of the PV installation. This is due to the fact that the sale of surplus energy in the electricity market does not allow the profit of the PV installation to be optimized. Inthesecond scenario,to optimizethephotovoltaicenergyflow inthe energycommunity,thefirmsweregrouped togetherfollowing twocriteria. First, the firms were grouped according to the complementarity of their energy profile and the seasonality of their energy demand. Thus, the hourly energyprofile andthe influenceofthe seasonsof theyearonenergydemandwere analyzedinorder togroup energycomplementaryfirmstogether. The second criterion was based on energy demand values. In particular, the excess of PV energy was established for it to be used by the company 15309290, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jiec.13536 by Universidad De Sevilla, Wiley Online Library on [26/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 4GONZÁLEZ-DOMÍNGUEZ ET AL. FIGURE 2 Flow chart of the two scenarios. with the highest level of energy consumption. Due to the proximity of the firms, no proximity criteria were considered for clustering. Regarding the photovoltaicinstallation in Scenario 2, itwassizedto try to coverasmuch of the energy demandofthe hourly period studied as possible,considering the limitations of the space available for the photovoltaic installation. In contrast to Scenario 1, in Scenario 2 the surpluses can be used to cover the energy demand of the firms in the cluster, resulting in a very cost-effective alternative. The photovoltaic energy flow between the firms of Scenario 2 depends on their energy consumption during the hourly period studied. Thus, the Monte Carlo method was applied to consider the randomness of energy consumption, which is essential to analyze the collective generation (Scenario 2) properly. The individual generation (Scenario 1) does not have this problem, as there is no photovoltaic energy flow between the firms. 2.4 Economic and financial analysis Aneconomic andfinancialstudy wascarriedout todeterminethe economicadvantagesand disadvantagesofthe twoscenarios proposed (Almaktar et al., 2021). On the one hand, the economic study made it possible to calculate an annual balance after implementing the photovoltaic installation, to determine the economic savings over 1 year. In this way, the benefits and revenues generated through excess energy from individual generation (Scenario 1) and the electricity bill savings generated by the energy sharing in collective generation (Scenario 2) were quantified. On the other hand, the financial viability study allowed the advantages and disadvantages of the investment to be analyzed under the two scenarios. In Scenario 1, the investment made by each firm depends on the cost of each firm’s PV installation. However, in Scenario 2 the cost of the distributed generation installation is shared among all the firms, with their share being proportional to their energy demand, so that they share energy at no cost. For the economic and financial analysis, all costs related to the photovoltaic installation were considered, including material and human resources costs.A service life of the installations of 25 years was considered (Ganesan & Valderrama, 2022). Table2lists the main parameters and costs considered in 15309290, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jiec.13536 by Universidad De Sevilla, Wiley Online Library on [26/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License GONZÁLEZ-DOMÍNGUEZ ET AL.5 TABLE 1 Characteristics of the firms’ roofs. Firm Type RA (m2) OA (m2) Inclination (◦) Azimuth (◦) 1 Gabled 150 0 12 24 2Flat 609 86.8 0 9 3 Gabled 920 0 18 9 4 Gabled 945 018 7 5 Gabled 2463 267 15 −7 6 Gabled 717 121 13 7 Abbreviations: OA, obstacle area; RA, roof area.. TABLE 2 Main parameters and cost for economic and financial analysis. Name Description Maintenance cost Annual cost of 3% of the total cost of the installation (Muñoz-Cerón et al., 2018) Degradation of PV production Decrease 0.8% each year (Domingos & Pereira, 2021) Discount rate (k) Discount rate of 3.4% Interest rate 4% interest rate on the total investment Sale rate 7% for sale the energy to the electricity grid Solar panels €128.04 (CYPE Prices Generator, 2021) Adjustable supports €70-144.70 (CYPE Prices Generator, 2021) Inverters €1985.34-4630.28 (CYPE Prices Generator, 2021) this research for the economic and financial analysis. To carry out these analyses, various financial ratios were calculated. The net present value (NPV) allows one to determine the benefits or losses of the investment for year n(Ponta et al., 2018), according to Equation (1). NPV =−I0+ n ∑ t=1 Ft (1+k)t(1) where I0is the initial investment, Ftis the cash flows in each period t,kis the discount rate, and nis the number of time periods over which the investment analysis is performed. The internal rate of return (IRR) is an indicator of the profitability of the investment. It is determined when Equation (1) is equal to 0. The benefit costratio(BCR) represents the benefit obtained foreacheurospent in the photovoltaicinstallation.Thisindicator is complementary to the NPVand provides useful information to investors about how much benefit can be obtained with the economic funds spent (Frej et al., 2021). The calculation of the BCR is shown in Equation (2) (Malka et al., 2022). BCR =∑benefits ∑cost = ∑n t=1 Ft (1+k)t I0 (2) where the initial investment, I0, includes the costs associated with the maintenance of the photovoltaic system. The last ratio used in financial analysis is the payback period (PP), which is the number of periods needed to recover the initial capital invested. Equation (3) shows the calculation of the payback period. PP =I0 AR (3) where PP is the payback period, I0is the initial investment made in the installation, and AR is the annual revenue obtained by the installation. This last parameter is obtained when carrying out the previous economic analysis. 15309290, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jiec.13536 by Universidad De Sevilla, Wiley Online Library on [26/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 6GONZÁLEZ-DOMÍNGUEZ ET AL. 2.5 Monte Carlo simulation method The Monte Carlo statistical method allows one to simulate a parameter that is random in nature, taking its behavior to be represented by that of a probability density function (PDF) of a random variable (Beltrán et al., 2020). It is thereby possible to construct a database that is large enough to mitigate the problems deriving from the uncertainty of the variables to which the study is sensitive. Following the Central Limit Theorem, the Monte Carlo method shows convergence at 1 √n,wherenis the number of cases analyzed (Ballio & Guadagnini, 2004). There are different variables that apply uncertainty to the results obtained in energy community research. However, energy demand is considered to be a variable whose randomness significantly affects the distribution of surplus photovoltaic energy. It affects the energy flow within the energy communities. Thus, the Monte Carlo method is used to generate a sequence of random numbers to simulate variations in energy demand. Random values of energy demand will be generated based on a PDF that describes the uncertainty of the variable (Wang et al., 2022). The firms’ energydemandsare assumedto berepresentableasarandomparameter,andaremodeled usingPDFs (Uniyal& Kumar, 2018).Ofthemanypossible PDFs (Weibull, normal, beta, etc.), it is found (Bordbari et al., 2018) that the uncertainty of electrical energy demand follows a normal distribution. A building’s energy consumption can be predicted by assigning a suitable probability distribution function (Kang & Wang, 2018). It is also possible to model this parameter with a beta distribution since its corresponding PDF has shape and range parameters that allow it to represent practically any other PDF (Yu et al., 2018). The mean of the normal distribution establishes the location of the probability mass function, and the standard deviation allows one to limit the range of values of the variable. Equation (4) gives the normal PDF form used in the Monte Carlo simulations. f(x, m, 𝜎)(4) where xis the sequence of numbers generated in the simulations, and mis the mean and σthe standard deviation of the energy consumption (expressed in kWh). Stochastic Monte Carlo simulations are a suitable alternative for estimating expressions that do not have analytical solutions (López-Agüí, 2008). Equation (5) lacks an analytically expressible primitive, so it cannot be solved with analytical methods. b ∫ a f(x)dx (5) where xis a continuous random variable distributed over the interval [a,b]andf(x) is the normal distribution expressed in Equation (4). The solution to this problem can be estimated from random values by determining which are below the function f(x). Thus, Equation (5) can be approximated by the expression in Equation (6). b ∫ a f(x)dx ≈N D(b−a)M(6) where Nis the number of random values below the function f(x),Dis the number of total random values, (b–a) is the interval of values of the variable x, M is the height of a rectangle containing all the points of the function f(x) in the interval (b,a). For a normal distribution, the value of Mcorresponds to the mean of the PDF, as this is the highest value of the function f(x). 2.6 Greenhouse gas emissions The market’s electricity consumption generates an environmental impact that depends on the nature of the region’s energy mix (Wilting et al., 2021). When the energy is renewable, the environmental impact is zero (Guillén-Lambea et al., 2023). The distributed generation of Scenarios 1 and 2 reduces the GHG emissions of the firms since they stop consuming energy from the market power grid. To calculate the reduction of GHG emissions in each scenario, the total energy compensated with the photovoltaic installation is obtained and its environmental impact is quantified based on the energy mix of the region. Equation (7) is the expression applied in this research to obtain the GHG emissions reduction. GH Greduction =EnergyPV ×Femission (7) where GHGreduction (expressed in tCO2eq/year) are the metric tons of CO2equivalent that are reduced each year due to distributed photovoltaic generation; EnergyPV (expressed in kWh) is the energy compensated annually with photovoltaic production; Femission (expressed in tCO2eq/kWh) is a factor that quantifies the amount of CO2equivalent that is emitted into the atmosphere when consuming 1 kWh in the region of Extremadura. The tCO2eq is the volume of GHG emissions equivalent to 1 metric ton of CO2. Analyzing the most important energy suppliers in the Extremadura region, 41.2% of the energy mix comes from renewable energy, 1.93% from high efficiency cogeneration, 26.33% from natural gas combined cycle, 15309290, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jiec.13536 by Universidad De Sevilla, Wiley Online Library on [26/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License GONZÁLEZ-DOMÍNGUEZ ET AL.7 FIGURE 3 Mean electrical energy consumption per firm. Underlying data are available in Table S1 of Supporting Information S1. 2.93% from coal, 1.13% from fuel oil/gas, 21.20% from nuclear, and 5.27% from other non-renewable sources. The value of Femission is 165 ×10−6 tCO2/kWh according to the National Commission for Markets and Competition (Comisión Nacional de los Mercados y la Competencia, 2022). 3RESULTS 3.1 Scenario 1: Individual production Supporting Information S2 (Appendix 2) shows Figure 3, which presents the mean monthly consumption of the six firms. Also, a description of Figure 3is given in that appendix. The data used to produce Figure 3are shown in Table S1 of Supporting Information S1. The mean monthly production of the photovoltaic installations was calculated for each company. Considering the data in Table 1,Firm1had an installation of 30 panels, Firm 2 of 43 panels, Firm 3 of 61 panels, Firm 4 of 49 panels, Firm 5 of 246 panels, and Firm 6 of 46 panels. This corresponds to installations of 9.9, 15.3, 20.13, 16.17, 81.8, and 15.18 kW, respectively. With these values, it is possible to determine the amount of energy covered by each firm’s installation (Figure 4). The data used to produce Figure 4are shown in Table S2 of Supporting Information S1. From Figure 4, it can be inferred that the energy production of the installed panels covers approximately 50% of the energy consumed during the period studied in the least favorable months in all the companies except Firm 1. This is because this firm’s roof area does not allow more panels to be installed, and it therefore cannot cover even half of the winter months’ energy demanded. This solution corresponds to a scenario in which the photovoltaic installation is done individually for each firm. 3.2 Scenario 2: Collective production The grouping of firms in Figure 5has been carried out following the criteria set out in Section 2. Thus, three groups have been formed, each with a photovoltaic installation designed according to its energy profile. These groupings were made based on the complementarity of their energy profile, the values of the energy consumption, and the seasonality of their energy demand. In this way, an adequate flow of surplus energy is achieved. In fact, Firms 3, 4, and 5 had a high energy demand during the summer season, while Firms 1, 2, and 6 did not present this behavior during these months. Thus, the surpluses generated by Firms 1, 2, and 6 can be used to cover the energy demand of Firms 3, 4, and 5, as these firms had a complementary energy profile. By analyzing the energy demand values, three groups were obtained. Group 1 comprised Firms 2 and 3; Group 2 comprised Firms 1 and 4; and Group 3 comprised Firms 5 and 6. Finally, it was observed that Firm 5 of Group 3 was the one with the highest energy consumption of all the firms. The surpluses of each group were used by the firms with the highest consumption. In sum, Firm 5 could use the surpluses of Firm 6 and the other two groups. The first two groups had a photovoltaic installation with a peak power of 86.79 kW, and the third had one of 125.07 kW peak power. The 86.79 kW installation consisted of 263 solar panels, and the 125.07 kW installation had 379 panels. While these 379 panels were installed on the roofs of Firms 5 and 6, the 86.79kW installation required not only the roofs of Firms 2 and 4 but an additional space. This additional space was an available plot of land next to the cluster and cleared of buildings, with a total of 2600 m2. As can be seen from Figure 3(Supporting Information S2), the mean monthly energy demand is different for each of the firms in the same group, allowing the energy produced in the installations to adequately cover the demand. A normal distribution was used to model the uncertainty of the 15309290, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jiec.13536 by Universidad De Sevilla, Wiley Online Library on [26/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License 8GONZÁLEZ-DOMÍNGUEZ ET AL. FIGURE 4 Monthly comparison of mean electricity production versus mean consumption. Underlying data are available in Table S2 of Supporting Information S1. energy demand of the firms. A standard deviation of 25% was considered from the analysis of the energyconsumption of the firms under study. The mean energy demands considered for the Monte Carlo simulations corresponded to the values given in Figure 3(Supporting Information S2). With this,30,000casesweresimulatedtoquantifytheuncertaintycorresponding to these firms’ monthly consumption. The 30,000 simulations allow the uncertainty generated by the variability of energy demand to be reduced by more than 150 times, so that the error is significantly less than 1%. For these cases, the photovoltaic installations cover on average 100% of the annual energy consumed in the hourly period studied by the first four firms (Groups 1 and 2). Firms 5 and 6 (Group 3) cover, with the PV system, 97.38% and 76.68%, respectively. By sharing the overages of the collective 15309290, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jiec.13536 by Universidad De Sevilla, Wiley Online Library on [26/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License GONZÁLEZ-DOMÍNGUEZ ET AL.9 FIGURE 5 Grouping considered in the second scenario. TABLE 3 Economic analysis for Scenarios 1 and 2. Scenario 1 Economics ratios Firm 1 Firm 2 Firm 3 Firm 4 Firm 5 Firm 6 Total Annual savings (%) 51.8 74.6 93.64 83.93 90.8 96.05 – Annual savings (€/year) 1965.3 2540.8 3995.5 3135.0 16,204.3 3022.6 30,863.5 Surplus energy revenue (€/year) –61.67 26.8 12.0 129.7 106.7 336.87 Scenario 2 Economics ratios Firm 1 Firm 2 Firm 3 Firm 4 Firm 5 Firm 6 Total Annual savings (%) 99.81 99.81 99.81 99.81 99.81 99.81 – Annual savings (€/year) 3795.15 3405.75 4266.51 3735.77 17,851.58 3146.71 36,201.47 Surplus energy revenue (€/year) 1132.10 1132.10 1132.10 1132.10 1132.10 1132.10 6792.6 generation, 99.81% of the energy demand of the six firms is covered. Thus, it can be observed that collective generation almost completely covers the mean energy demand of the hourly period studied for all the firms in the 30,000 Monte Carlo simulations. Indeed, the main advantage offered by the grouping of firms for the collective generation of solar energy is the use of such individual overages to cover the entire collective’s energy demands, thus avoiding over-sizing the installations in order to cover adequately the energy consumed throughout the year (González González et al., 2018). 3.3 Economic analysis As described in Section 2, the economic analysis considers energy consumption, photovoltaic energy production, surpluses generated, and the installation maintenance. Table 3lists the annual savings on each firm’s electricity bill for Scenarios 1 and 2. The percentage annual savings was very high except for Firm 1 which had no roof area on which to place a photovoltaic installation of greater power. The annual revenues from the overages of individual generation were low, and the percentage of energy covered by the six firms was 85% of their energy demand during the period studied. This is consistent with the sizing of the installation to cover 50% of the energy demand during the least favorable months, as 100% of the energy demand would be covered during the most favorable months. This covers a large part of the energy consumed by the firms without generating much surplus in other more favorable months. In this scenario, if the percentage of energy covered during the least favorable months were to increase, the profitability of the installation would not increase. The investment would be higher, and surpluses with a lower economic value would be produced since its sale to the electricity market does not generate a high profit. Collective generation 15309290, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/jiec.13536 by Universidad De Sevilla, Wiley Online Library on [26/08/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License