Techno−Economic Analysis and Multi‐Objective Optimization of Cross‐Flow Wind Turbines for Smart Building Energy Systems

Abstract This work reports a technical, economic, and environmental investigation of the possibility of using a recently developed smallscale crossflow wind turbine (CFWT) to supply the energy demand of buildings for different integration scenarios. For this purpose, three CFWT‐assisted building energy system configurations with heat pumps, with and without batteries, and two‐way interaction with the local grid in two residential building models in Iran and Germany are investigated. Triobjective optimization with a Nondominated Sorting Genetic Algorithm (NSGA‐II) is performed for finding the optimal configuration of the energy system in different configurations. For economic assessment, the Capital Budgeting Analysis method is used with four indicators, namely, payback period (PP), net present value (NPV), internal rate of return (IRR), and profitability index (PI). The results show that due to different energy market regulations and prices, different integration scenarios and system configurations can outperform others in Germany and Iran. Overall, due to the exchange rate instability and low energy tariff in Iran, in order for the project to be feasible, either the CFWT cost must fall to below 30% of its current cost or the local electricity price should increase significantly to get a Levelized cost of energy of as low as 0.6 $ kWh−1.


Introduction
The limited resources of fossil fuels and their environmental effects have seriously necessitated taking a pivot in the energy sources to be sustainable and pollution-free. [1] In this regard, of a 2-kW WT, 4-kW PV, 4-kW inverter, and six battery strings. This scenario has NPC of 20527 $, a capacity shortage of 1.1%, and an unmet electricity load of 0.9%. Akbari et al. [13] simulated the same hybrid system without batteries in Khvaf town, Razavi Khorasan Province. They concluded that the optimal design providing 93% of the electricity includes 13-kW PV, 10-kW inverter, and four horizontal axis wind turbines (HAWT) of 100 kW in ongrid mode with an initial investment of 440 000 $ has a 4-year payback period. For a residential complex in Tehran, Ashrafi Gudarzi et al. [14] studied a WT-PV-battery hybrid system; additionally, Abbaszadeh et al. [15] researched a WT-PV-gas generator system. Each separately indicated that hybrid systems instead of diesel systems could significantly reduce GHG emissions while achieving cost savings. Farahi and Fazelpour [16] analyzed the WT-PV-battery-diesel generator system for commercial, public, and residential buildings in Tehran, Kish, and Binaloud. Moreover, they did a sensitivity analysis on fuel cost, thereby announced Kish is the best region for all buildings. Residential buildings in Kish with a 60-kW PV, a 2.5-kW WT, 480-kW diesel generator, and 16 batteries have a total COE of 0.250 $ kWh −1 and NPC of 15511300 $. The common features of these researches are that they all used three-bladed HAWTs, and their renewable energy system was a hybrid system with batteries or, even in some cases, a gas generator. The disadvantages of these researches are the pollution caused by generators, the high initial investment cost, and the large space required for installing PV panels, WTs, batteries, and other equipment regardless of the maximum available space in residential buildings in Iran.
SSWTs have become increasingly popular in windy areas; however, few studies have focused on them in buildings worldwide. Rodriguez-Hernandez et al. [17] evaluated 28 SSWT models at 18 locations in the Valley of Mexico Metropolitan area to estimate annual energy production. Borunda et al. [18] studied 5-kW and 10-kW SSWT to supply the residential demand in different Mexican cities with Intelligent Bayesian decision-making. Ibrik [19] analyzed 10, 100, and 220 kW SSWTs at eight sites in West Bank cities. Using a small-scale HAWT, geothermal heat pump, and preheating with solar energy, Ozgener [20] experimentally heated and cooled a greenhouse in Izmir, Turkey. He concluded that the WT alone can provide 3.1% of the total annual electricity consumption of the greenhouse (3568 kWh) or 12.5% of the total annual electricity consumption of the secondary water pumping, brine pumping, and fan coil (892 kWh). Mahmuddin et al. [21] employed the Capital Budgeting Analysis method for a 300-watt HAWT in the South Sulawesi coastal area, Indonesia, and determined the project is not feasible for wind speed below 7 m/s. Abdelrahim et al. [22] examined and optimized wind-driven charging stations for electric vehicles in high-rise buildings in Malacca, Malaysia. The findings show that a 5-kW RC-5K-A SSWT can generate 214 272 kWh yr −1 at the cost of 0.081 $ kWh −1 , confirming the future feasibility of WT as a source of energy-infrastructure support. Rahimi and Thaghafi [23] studied the 5.5-kW HAWT in the on-grid mode in Borujerd, Iran and acknowledged that the PP is 14 years, although Shahbazi et al. [24] declared the project with 5-and 10-kW Hummer at the height of 10 m in a greenhouse in Robat Karim, Tehran (the same country as the Ref. [23#23]), requires government support to become feasible. All the SSWTs considered in the above investigations globally were small-scale HAWTs. There are also two studies addressing the technoeconomic feasibility of small-scale vertical axis wind turbines (VAWTs) in Iran, one analyzing a 2-kW Savonius in Isfahan [25] and another a hybrid Savonius-Darrieus in Marvast, Yazd. [26] One of the emerging types of SSWTs for building applications is CFWTs, on which there are not that many pieces of research. Indeed, although the scientific literature on smart energy buildings is exceptionally rich in terms of CFWTs, it suffers from several important gaps. First, after so many extensive studies focused on the performance and efficiency enhancement of CFWT, there is no research on its techno-economic feasibility in the world. Second, no complete and comprehensive assessment has been conducted on small-scale household WTs with/without batteries in on-grid/off-grid mode. The first and the second gaps can also be concluded from Table 1. Three, the high investment costs of renewable-based building energy systems, a substantial part of which is related to batteries, have discouraged building owners from implementing and installing them in their buildings; which could be addressed by studies in this class, the possibility of two-way interaction of buildings with the electricity grid, and the possibility of multi-generating by the WT employing heat pump units in the building. This work aims to provide a feasible and reliable path by employing all the above-mentioned gaps toward developing CFWT-integrated smart building energy systems. By considering this purpose, this study proposes a smart novel configuration integrated with the CFWT and air source heat pump having a two-way interaction with the power grid. It helps the building owners to supply a part of their energy demands and compensate for building energy bills by selling their extra production electricity to the networks. Apart from this, two other CFWT-based smart building energy systems equipped with batteries and heat pumps in on-grid mode are presented. To get a better insight into the superiority and advantages of each configuration in different climates and countries, they are compared from technical, environmental, and economic aspects of different regions of Iran and also Germany as the case studies. Multiobjective optimization based on the Nondominated Sorting Genetic Algorithm (NSGA-II) approach is applied to the configurations to determine optimal operating conditions, and various economic indices are employed for the economic analyses. Figure 1 demonstrates the schematic diagram of each analyzed smart building system. Systems consist of CFWT, wind charge controller, inverter, and battery. They differ based on grid interaction and battery presence or absence. A detailed description of each configuration is provided below: a) Configuration 1 -with CFWTs, wind charge controller, and inverter without connection to the building, as shown in Figure 1a. This configuration is popular for on-grid areas in Iran. According to Table 2, the electricity tariff for the household sector is much cheaper than the wind feed-in tariff. [27,28] Therefore, buying electricity from the grid and selling generated CFWT power to the grid is more economical. Unlike the electricity tariff in the household sector, which varies by hour, consumption level, and climate zone, the wind feed-in tariff (FiT) is always fixed during a year. b) Configuration 2 -with CFWTs, wind charge controller, and inverter, as illustrated in Figure 1b. Depending on the building's electricity demand assessment, the smart controller decides whether to sell electricity to the grid or supply the building's demand. Otherwise, when there is not any electricity demand, the extra generated electricity sells to the grid. c) Configuration 3 -with CFWTs, wind charge controller, inverter, and battery, as presented in Figure 1c. In this configuration, the priority is to supply building electricity through the grid (a better decision based on the tariffs in Iran). Additionally, the battery is applied to supply the building's required electricity when the grid power is cut off, or the building demand is high during peak tariff hours. In addition, it has a two-way interaction with the electricity grid to sell extra electricity and purchase from the grid.

Experimental Section
The demand energy profile of buildings for one year is calculated by Carrier HAP 4.90 software. Afterward, to perform economic and environmental assessment, the modeling of the proposed smart building systems is analyzed via MATLAB software with a time resolution of 1 hour. Finally, the tri-objective optimization using the NSGA-II algorithm approach is applied to the scenarios to find the best operating condition from technical, economic, and environmental facets.

The Case Studies
The cities of Ardabil, Zahedan, and Babolsar, located in Iran based on Figure 2, were considered case studies of this work. According to Köppen-Geiger climate classification, they are in hemi boreal, arid, and humid subtropical climates, respectively. [36] Also, based on the climate and wind characteristics similarity, Oldenburg city in Germany was chosen for comparison and illustrated in Figure 2. Two residential buildings are contemplated: a 160 m 2 house on a field of 14 m × 11.5 m and a 4-story apartment with two 65 m 2 units. The house is simulated in low-consumption and high-consumption patterns. Residents of the low-consumption house have proper timing and saving in using electrical energy. The prerequisite data for calculating the cooling, heating, and electrical profiles with Carrier HAP software, contained the characteristic of the buildings, local weather information, comfort standards, and hot water demand profiles. Cooling and heating systems were selected based on the city climate and available equipment.

Economic Analysis
The total capital expenditure during the lifetime (20 years) was calculated and compared with the grid electricity costs, a conventional method in Iran, to evaluate the proposed Global Challenges 2023, 7, 2200203 Table 1. Literature review (Bat = Battery, Gen = Generator, GHP = Geothermal heat pump).
The net present value over the lifetime is defined as below where x is the discount rate and tot, OM C i is the total operational and maintenance costs for the ith year. Moreover, the IRR describes an investment return rate as the percentage at which the NPV Global Challenges 2023, 7, 2200203  Average electricity tariff for the household in Iran [27] 0.0024 $ kWh −1 + 0.00012 $ kWh −1 annually (in the last 11 years)
value is zero. AC wind is the annual cost of wind-based and AC SP is the annual cost of separation production systems, which are written as [40] wind Bought Sold electricity c electricity is the electricity price based on Table 2 and Demand  E is the total electricity demand for the whole year. Another ranking investment index where the value of 1 indicates the break-even point is PI In the levelized cost of energy (LCoE), the cost of producing 1 kWh of electricity is calculated by dividing the present value of all the costs by the CFWT electricity production in the project lifetime [41] LCoE (1 ) is the CFWT electricity production in the ith year. Another important economic indicator is the PP, defined as the number of years necessary to compensate for the initial costs Furthermore, dependency on the grid (DoG) indicates the percentage of the annual energy requirement provided by purchasing from the grid, as follows

Environmental Analysis
Due to the increase in global environmental pollution, especially the worrying emission of GHG and its dangerous effect on rising temperature and changing wildlife and human life worldwide, environmental analysis has become more significant than ever. Here, the carbon dioxide emission reduction rate (CDERR) is calculated to compare the proposed integrated building CFWT configurations and demonstrate their superiority over the conventional method based on Equation (10) CDERR 100 where CDE wind is the carbon dioxide emissions of windbased and CDE SP is the separation production systems in kg. The carbon dioxide emission coefficient (δ) based on the energy balance sheet published by the Ministry of Energy is 0.645 kgCo 2 kWh −1 . As a result of electricity generation Global Challenges 2023, 7, 2200203 from CFWT, the non-emission of Co 2 can be calculated as follows [42] Co 1000

Multiobjective Optimization
The purpose of multiobjective optimization is to improve several conflicting performance indexes simultaneously according to the limitations and constraints of the problem. In many engineering fields, there are conflicts between objective functions, such as cost, time, and performance. Thus, finding a balance point between them is essential to maintaining system performance while reducing cost and time.
The Genetic Algorithm (GA) is an efficient method to solve problems that are usually intractable or time-consuming to solve by other methods. [43,44] Compared to other methods, it has the fastest computational speed and the highest number of best objective functions. [45] Further, GA was not getting stuck in local optimal points. In the present study, tri-objective optimization using the Nondominated Sorting Genetic Algorithm (NSGA-II) approach [46] is implemented to find the best decision variables for each configuration. MATLAB software was applied to maximize NPV (Equation (3)) and CDERR (Equation (10)) as fitness functions while simultaneously minimizing dependency on the grid (Equation (9)) as a cost function. Variable parameters for optimization were battery capacity and number of CFWT with a domain of 0 kWh < CAP Bat < 10 kWh and 0 < N WT < 10, respectively.
After the algorithm termination, the Pareto front is obtained for each generation. Choosing an answer among a set of answers (Pareto front) is difficult for most decision-makers. However, researchers have introduced different criteria; for example, the CS point is based on the nearest distance to the ideal point, and the ES point is the intersection of the line passing through Threat point c with an angle of 45° with the Pareto front. [47]

Results and Discussion
The technical, economic, and environmental assessments of the proposed smart energy home systems are surveyed by using MATLAB software. A prerequisite information required to perform the simulation is the house energy demand profile, calculated with Carrier HAP, a powerful and comprehensive tool for designing HVAC systems and modeling annual energy performance and energy costs. Time-dependent, hourly and monthly charts are first extracted to detect the best configuration. Then, a comparative parametric analysis is conducted to evaluate the effect of the main parameters on the technical, economic, and environmental indicators. Finally, considering NPV, CDERR, and dependency on the grid as objective functions, multiobjective optimization based on the NSGA-II approach is applied for the best city considering all aspects.

Parametric Study of the Configurations
The desired location for installing the CFWT must have suitable wind speed and direction conditions. The wind direction should not have sudden fluctuations, the average wind speed should be high, and the highest prevailing wind speed should be in one direction. The best way to display and analyze these features is through Wind rose and wind speed frequency diagrams, as shown in Figures 3 and 4. The dispersion of the wind direction in Babolsar from the southeast to the northwest makes the best installation option to be VAWT. The dominant wind direction is south to the southwest, with an average wind angle of 200° in Ardabil. If the turbine blades are placed at this angle, they can absorb maximum southwest wind potential from the front and northeast from behind the turbine. This angle is 120° for Zahedan and 265° for Oldenburg. In Figure 4, the average annual wind speed data at each city is classified by the Sturges' rule. The wind speed in Ardabil is 3.87 ± 2.46 m s −1 (3.87 is mean and 2.46 is variance), Zahedan is 3.48 ± 2.05 m s −1 , Oldenburg is 3.31 ± 2.19 m s −1 , and Babolsar is 1.139 ± 1.2 m s −1 . Thus, the occurrence probability of winds exceeding 1.2 m s −1 is higher in Ardabil, Zahedan, and Oldenburg than in Babolsar. Another weather information, the hourly variation of ambient temperature, is shown in Figure 5, where a noteworthy point in this figure is a significant difference, about 16.7 C°, between the day and night temperatures of Zahedan because of the dry weather.
The average hourly hot water consumption profile for a person living in a family of four on working days and holidays is illustrated in Figure 6. [48] The power consumption characteristics of electrical equipment are presented in Table 4. These data have been used to simulate the demand profile of heating, cooling, and electricity in buildings, as presented in Figure 7. The cooling device is an evaporative cooler in hot and dry cities or a heat pump in other cities. Also, a Wall-hung gas boiler is used for heating and hot water supply to the building. The electrical energy used by the cooling and heating equipment is added to the electricity demand profile. Figures 8 and 9 show the hourly and monthly electricity demand in the low-consumption house for four cities. It can be seen that Zahedan, which has a hot and dry summer climate, has the largest electricity demand, followed by Babolsar.
According to the turbine dimensions in Table 5 and the apartment and house rooftop area, installing a maximum of ten 0.5-kW CFWTs is possible. Figure 10 shows the monthly variation of the produced CFWT power, electricity demand, and electricity bought and sold in three building types in Babolsar with configuration 2. According to Figures 10a,b, monthly electricity costs are saved, and owners also receive a long-lasting income from selling CFWT electricity. In the case of Figure 10c, the 8-unit apartment consumed all the produced CFWT power, and there is nothing left to sell to the grid. Configuration 2 reduces energy costs by 39% to 69% in a low-consumption house, 25-69% in a high-consumption house, and 9% to 50% in each apartment unit, depending on the season. Figure 11 shows the PP for each configuration and building in Babolsar. Due to the PP of the apartment being longer than the project lifetime, it cannot be displayed in this figure. Moreover, not only does configuration 1 has a shorter PP, but also it has a higher IRR and NPV that induce more cost-effectiveness. When the power consumption pattern of a house is observed, it reduces home electricity costs further and results in a significantly different PP between low-and high-consumption houses. The usage of CFWT in the apartment is problematic because of the following reasons: • Most buildings have unused rooftop space under high wind speed conditions, making WTs a suitable option. Nonetheless, Turbulent and eddy wind flows caused by adjacent trees and high-rise buildings are critical subjects that must be studied before installing the WTs.
• Apartment residents face the challenges of operating the WT system, such as obtaining all owners' consent before the integration, participating in the financial profits and losses, determining a person responsible for the WT system monitoring and maintenance, and understanding financial contracts and regulations.
The effect of city and climate is investigated by considering building type, configuration, and turbine price as fixed parameters. For this purpose, CFWTs integrated with the house at the current price are considered for simulation. Referring to Figure 12, the CFWT electricity production is high in fall and winter in Ardabil (average of 629 kWh at six months), which leads to the purchase of a small portion of electricity demand from the grid. The CFWTs in Zahedan produce an average of 300 kWh of electricity per month. However, due to the high electricity demand from April to October, a small amount of CFWT electricity is for sale, and even a part of the electricity demand must be purchased from the grid. The CFWTs in Oldenburg have uniform productions throughout the year, and electricity demand is less than the turbine production. All this has led to less buying from the grid and more selling. On the other hand, the turbine production in Babolsar is deficient because of the maximum of 329.56 kWh in May; therefore, more electricity is purchased from the grid, and it does not seem cost-effective compared to other cities. Figure 13 delineates the monthly price of bought/sold electricity from/to the grid in each configuration, assuming that the number, price, and capacity of CFWTs in the house are constant. Based on this figure, in configuration 2, Ardabil and Zahedan pay a small amount for the electricity bills and earn profits from selling surplus turbine electricity. In Oldenburg, Global Challenges 2023, 7, 2200203  the electricity bill has decreased by an average of 83.6% per month; however, there is a negligible profit from selling surplus turbine electricity, a max of 1.54 $. Using configurations 2 and 3 has led to a drastic electricity bill reduction in Babolsar (87.31% and 92.21%) and Zahedan (94.43% and 97.17%). Likewise, the annual energy cost in Oldenburg decreased significantly by 83.6% in configuration 2 and 98.65% in configuration 3. A remarkable aspect of Oldenburg is that selling the electricity stored in batteries to the grid during periods of low demand can earn a high profit.
The PP in all configurations is depicted in Figure 14, where the graphs of Ardabil, Zahedan, and Babolsar are logarithmic curves, and the graph of Oldenburg is linear due to the constant annual FiT. As mentioned, Ardabil experiences higher wind speeds and frequency than other cities. Subsequently, the CFWT power generation will be higher, and the PP will be smaller in all configurations. If a 5-year PP in configuration 1 is Global Challenges 2023, 7, 2200203   Table 4. The list of appliances with their nominal power, standby power, daily frequency, and operational cycle duration. [49,50]  desired, the CFWT price should reach 6.7%, 6%, and 5% of the current price in Ardabil, Zahedan, and Oldenburg, respectively. The graphs of configuration 3 for Ardabil and Zahedan slightly deviated from their normal behavior. This deviation is the PP time jump that occurred due to battery replacement. According to Figure 15, in the 12th year, the battery replacement costs will  result in a negative net cash flow and a sudden bad jump in economic metrics, including the PP. According to Table 6, for the cities of Iran, the PP in configuration 1 is the smallest; however, in Oldenburg, configuration 3 has the smallest PP. In other words, configuration 3 with the CFWT current price has a PP of less than eight years, and it is the best configuration in Oldenburg. Meanwhile, configurations 1 and 2 will have a PP less than the system lifetime if the CFWT price decreases. The purchased expensive household electricity and sold cheap wind electricity are the reasons for the economic attractiveness of configuration 3 compared with the first one. On the other hand, the results in Iran are entirely different from Germany due to the cheap and available household electricity. For instance, only configuration 1 in Ardabil at the current price will have a PP of 19.6 years, while other indexes show a nonfeasible project (with NPV, IRR, and PI of −22 900, 0.32, and 0.3, respectively). The PP of other cities of Iran at the CFWT current price is more than the system lifetime, but its reduction depends on the devaluation of the CFWT. That means by reducing the CFWT price to 10% of the current price in all cities and configurations, the PP will be less than 14 years, except for configuration 3 in Babolsar. In short, it can be acknowledged that configuration 1 in Iran will be more economically profitable if the following two conditions are met: 1. Choosing cities with an average speed of more than 3 m s −1 ; and 2. A devaluation of CFWT from 29% to 10% of its current price. Figure 15 shows the economic analysis criteria for Ardabil, including cost, income, cumulative cash flow, and cumulative NPV. 5-kW total capacity CFWTs with a price reduction of up to 6.7% are used to plot this figure. What stands out from Figure 15 is that in the zeroth year, the year of the construction, the project is only costly and has no income. The cumulative cash flow is negative in the 8th year and positive in the 9th; thereby, the project reaches net profitability in the 9th year.
The reduction rate of carbon dioxide emissions is shown in Figure 16. CDERR in the presence of the CFWT will be higher in any city with less electricity demand, less purchased from the grid, and more sold to the grid. As a result, the CDERR parameter increases in Babolsar, Zahedan, Ardabil, and Oldenburg, respectively. It should be mentioned that the pattern of consumption and type of electrical equipment in all cities is assumed to be the same to investigate the turbine performance.
Another proposed correction factor is the LCoE, calculated using the CFWT current price. Using configuration 1, the most economical configuration in Iran, the LCoE for the 5, 10, and 20 years PP is calculated in Table 7. The number of 4 and 7 CFWTs are considered from the multiobjective optimization results, as presented in Table 8. If the 10-year PP is desired, the local electricity price should increase significantly to get the LCoE of 0.6 $.

Optimization Results
As discussed, the conflict between objectives in configurations 2 and 3 reveals the importance of multiobjective optimization. Using a tri-objective optimization approach based on the NSGA-II, objective functions of NPV, CDERR,   20 20 and grid dependency are optimized for Zahedan city by the devaluation of CFWT to 29% of the current price. Accordingly, the Pareto front diagram for all generations is shown in Figures 17 and 18. The 2D diagrams are illustrated in Figures 17b and 18b to better comprehend the Pareto front diagram. CS and ES methods are used to calculate the best points shown along with the ideal point in these figures. The optimization results, including the optimal value of the decision parameters and the obtained objective functions, are listed in Table 8. According to the table, if maximizing NPV is the only goal, configuration 2, with an NPV of 2588.86 $, is the desired condition. However, the lowest grid dependency and the highest CDECRR occur in configuration 3 with 28.38% and 115%, respectively. In view of the fact that NPV is more important than the other two objective functions for assessing wind-based building energy systems, configuration 2 seems to be more economic. As seen in Table 8, the optimum battery capacity value is about 1 kWh, and the number of turbines in configurations 2 and 3 is almost the same. The scatter distributions of variable parameters are illustrated in Figure 19 for a better image of the optimum range. From Figure 19a, it can be concluded that the optimal number of turbines is more than one. Referring to Figure 19b, since all the number of turbine optimal points have been distributed in the entire range from 0 to 10 with a concentration at the upper bound, it is not a sensitive parameter. However, the battery capacity is a sensitive parameter, and keeping it between 1 to 6 kWh leads to optimal technical and economic conditions simultaneously.

Conclusion
This work focused on smart building systems integrated with CFWTs to reach maximum energy performance, reliability, and cost-effectiveness for residential buildings. The proposed smart systems have been simulated with/without batteries in on-grid mode and two-way interaction with the electricity grid  to provide electricity and cooling for two typical residential buildings in selected cities of Iran and Germany. Comparisons and analyses have been made considering the weather conditions, tariffs, and energy regulations of both countries. After calculating the building demand energy profile, MATLAB software is used to simulate and compare all configurations from technoeconomic-environmental aspects. Multiobjective optimization based on the NSGA-II approach is applied to study  the influence of variable parameters containing the number of turbines and battery capacity on the performance of configurations 2 and 3. These parameters are examined by evaluating their impact on the NPV, CDERR, and grid dependency. In summary, the substantial conclusions are drawn as follows: • A low-consumption house is the best building model for installing a CFWT in Iran. Configurations 1, 2, and 3 will be efficient, respectively. • Although Zahedan and Babolsar have the highest electricity demand because of their hot and dry summer, configurations 2 and 3 have reduced Zahedan's electricity bill by 94.43% and 97.17%, and Babolsar's by 87.31% and 92.21%.
• Configuration 1 is the most cost-effective and the best in Iran, which will increase the indicators if two conditions are met: 1. Utilization in windy, moderate climate cities with an average wind speed of over 3 m s −1 ; and 2. Devaluation of CFWT to 29% of the current price by allocating a subsidy or producing a similar model at a lower price. • The best option from the economic aspect in Oldenburg is configuration 3. Due to its expensive household electricity tariff and cheap wind FiT, it varies from the best configuration in Iran. • The PP less than the system lifetime in Iran is achieved by reducing CFWT price or changing LCoE. Thus, in Ardabil, Zahedan, and Babolsar, if the 10-year PP for configuration 1 is taken into consideration, either the CFWT price should decrease to 21%, 20%, and 6% of the current price, or the local electricity price should increase significantly to get LCoE of 0.54 $, 0.59 $, and 1.63 $, respectively. • Configurations 2 and 3 are currently not attractive in Iran, since the wind FiT is 17 times higher than the household electricity tariff. • The tri-objective optimization in Zahedan with the devaluation of CFWT to 29% of the current price indicates the NPV, grid dependence, and CDERR of seven turbines have values of 2330 $, 38.42%, and 90%, respectively, by choosing configuration 2. Furthermore, selecting configuration 3 changes objectives to 1253 $, 28.38%, and 115%, as well as variable parameters to eight turbines and 1 kWh battery capacity.