Technical and Economic Evaluation and Optimization of an Off-Grid Wind/Hydropower Hybrid System

ABSTRACT


Introduction
The lack of electricity networks in remote areas and the cost of connection of these areas to national networks attributable to the undesirable geographical state of the region has led to the utilization of alternative energy sources, independent of the network [1,2]. In recent years, diesel generators have been considered to supply the power required in these areas [3,4]. However, the use of this generator has problems such as pollution caused by diesel fuel consumption [5], which elicits lung and respiratory diseases. Diesel fuel is considered a finite and non-renewable source, and in recent years, attention has been focused on reducing the consumption of fossil fuels and replacing it with clean and renewable energy [6][7][8][9]. The use of renewable energy, in turn, has problems such as the high cost of its construction and the dependence of its energy production on climate change [10, Ren et al. [30] analyzed the performance of a hybrid hydropower-photovoltaic-wind system and implemented multi-objective optimization for reducing the variance and maximum power generation through an owl search algorithm. Three climatic conditions were employed to analyze the efficiency of the hybrid system. In changing climates, the system continued to generate power. The Pareto front results demonstrated that the developed algorithm is the algorithm that comes closest to the Pareto front. As a result, improving the entire multi-objective integrated system was the most precise.
Lu et al. [31] applied an optimization model for the combination of a hydro-wind hybrid system. They asserted that coordinated operation augments the power output stability and generation profits by 4.66%. Tan et al. [32] evaluated the benefit and risks of the operation of the high-sized hydropowerwind-photovoltaic system and considered forecast uncertainty. They applied a case study in the Yalong River, China. They reported that the total power generation, power generation profit, and energy transmission line use efficiency all grew dramatically. They added that the detrimental impact of photovoltaic power and wind generation variability on the power grid can be mitigated.
Sterl et al. [33] conducted a study on decarbonizing the electricity mix via a hydropower plant integrated with wind power in Suriname. The results indicated that thermal power can be effectively supplanted by hydro-supported wind power systems. They reported that Suriname can achieve 20%-30% penetration of hydro-supported wind power. An innovative universal method to resolve the integrated systems module size was given by Zhang et al. [34] for the optimal arrangement of a gridconnected hydropower-wind-photovoltaic system. It was determined that the optimum capacity proportion of wind and PV for the integrated system was 0.256 and 0.744, individually. Zhao et al. [35] evaluated a PV-wind-hydro multi-energy system, relevant to Yalong River basin specifications. Their results demonstrated that the PV and wind power capacity rate is best gauged at 0.57:0.43.
In retrospect, in this study, a small hydropower system with a wind turbine is used as a hybrid system to generate and supply the electricity in a remote village in Ardabil, a northwestern province of Iran. Information about the river discharge located in this village and also the intensity of the wind in the region is obtained from the Iran power plant investment company (SANA) and included in the modeling of the combined system. In addition, an approximate reliability model is used in this paper. The purpose of this work is to investigate the responsiveness of the proposed system to provide reliable power required by rural consumers during the year, and fills a literature gap within this scope.

Combined System Structure
A small hydro and wind power generation system is comprised of a combination of hydroelectric generators, wind turbines, battery banks, inverters, controllers, and other necessary tools and cables. A schematic diagram of an energy hybrid system is shown in Fig. 1 . In this system, small hydropower and wind turbines operate together for meeting the load demand . The generating power after meeting the load demand feeds the battery until the charge is complete. On the other hand, when energy is scarce, the battery will discharge its stored energy to help with small hydropower and wind turbines to meet load demand until it runs out of power. In other words, for predicting the performance of hybrid systems, first, it is necessary to model each of the components of the system (small hydropower, wind, and battery) and then their combination can be assessed to meet the load demand. Small hydropower, wind turbine, battery bank, and dump load are based on the following priority : the supply of power in the first place, then the battery bank, and finally the dump load are in demand. Electricity generated by small hydropower and wind turbines is regulated by a voltage regulator and when there is a shortage of load electricity generated by an additional hybrid system is stored by battery banks. But as known, the amount of electricity generated by the small hydropower turbine and wind energy in a hybrid system depends on the flow of river water and wind speed. One source of energy is at a lower level of production while the other is usually at a higher level of production. For example, in winter, the wind speed is generally at a high level while the river flow is at a low level. It is should be addressed that the simultaneous utilization of multi-generation sources generally augments the reliability of load demand [36].

Hydropower System Model
For the desired location, two main parameters of waterfall height and water flow were measured at 5 m and 187 L/s (0.187 m 3 /s), respectively. The available potential of electric power from water sources is equal to 7.4375 kW, which is obtained from the following equation [37]: where , , and ℎ are turbine efficiency, generator efficiency, and hydraulic power, respectively. The output power of the theory according to the location of the turbine is obtained by the following equation [38]: In the above relation, is the density of water, is the electric discharge coefficient, is the acceleration of gravity and ℎ is the height of the waterfall. Kaplan-type turbines (small size and higher speed) can be used due to the topology and hydrological properties of the area [38]. According to the type of turbine and its technical specifications, the turbine efficiency is considered to be 90%. Owing to its variable rotor speed characteristic compared to a synchronous generator, an induction generator is employed to produce variable speed [39,40]. Generator efficiency is considered to be above 90%. The main specifications of the generator are tabulated in Table 1.

Wind Turbine Model
Three main factors can govern the output of a wind energy conversion system (WECS), which are the output power curve (obtained by aerodynamic efficiency, mechanical transmission, and electrical conversion efficiency). From a selected wind turbine, wind speed distribution from a selected installation site, and wind tower heights are extracted. It should be noted that choosing a suitable model to simulate the output power is very important. The most simple model for simulating the output power of a turbine can be described by [41]: where is the relative electrical power, is the low cut-off wind speeds, is the relative wind speed, and is the high cut-off wind speed. In small-scale wind turbines, low cut-off wind speeds are comparatively small, and wind turbines can function straightforwardly even when wind speeds are not too high. The important point is that changes in wind speed with altitude, affect both the evaluation of wind benefits and the wind turbine design. In general, two mathematical models are utilized for modeling vertical wind speed profiles in similar areas. The first is the log law, which is based on the boundary layer flow in fluid mechanics. The second is the law of power, broadly employed by researchers. Generally, the power law characterizes a simple model for the speed profile of vertical wind. The fundamental form of this method is given as: where is the wind speed at the height of the hub Z (m/s) (height of the turbine above ground) and is wind speed measured at reference height. The changes increase with parameters such as day, season, nature of the earth, temperature, wind speed, and assorted combined mechanical and thermal parameters. The value 1.7 is commonly used whenever no specific information from the site is available.

Battery Model
The lead-acid type is set for the battery bank, which is employed for storing surplus electrical energy for regulating the voltage of the system and for supplying power to load at lower wind speeds. These batteries operate under very precise conditions. Therefore, it is problematic for predicting the time's energy will be supplied to or extracted from the battery. Parameters such as charge current, charge efficiency, inductor discharge, and battery capacity affect battery behavior. Most battery models focus on three different parameters, such as battery state of charge (SOC), floating voltage, and battery life.

Battery SOC
As already known, energy is stored in the battery when the wind turbine generated power and hydropower is more than the load. Energy is extracted from the battery when the generating power cannot meet the load needs, and when the power output by both the wind turbine and the hydropower turbine is inadequate and the storage is exhausted, the load will be cut off. The battery capacity, alike all chemical processes, relies on temperature. Generally, the battery capacity changes are defined by addressing the temperature coefficient : In this equation, ′ is the operating battery capacity at a battery temperature of [K], is the nominal or relative battery capacity (given by the manufacturer), = 6% is the temperature coefficient in degrees, or else it is given by the manufacturer. To find the true battery SOC, having the initial SOC, current discharging, or charging time is vital. Nevertheless, approximately many storage systems aren't ideal and losses take place during charge and discharge and throughout periods of storage. Battery SOC at t+1 can be calculated through the: where is the amount of self-discharge that relies on the battery health and total charge and the suggested value is 0.2%/day. It is problematic for measuring separate discharge and charge efficiencies; thus, manufacturers typically define a specific efficiency limit. Regularly, the battery charge efficiency is equivalent to the adjusted efficiency and the discharge efficiency is set as one. Battery current in the hybrid system at time t can be described by: The efficiency is found based on the load profile and inverter characteristics are considered to be 92%. For simplicity, we assume that the wind turbine has a DC output, so it is not necessary to use a rectifier .Nonetheless, if a wind turbine is intended for being connected to an AC network, rectifier losses must be taken into account for a portion of the wind energy that goes from AC to DC.

Battery Floating Charge Voltage
The battery floating charge voltage response is modeled under both the charge and discharge conditions via the equation-fit technique, whereby models the battery as a black box and defines the floating charge voltage through a polynomial in periods of battery SOC and current. The battery floating charge voltage of the ( ′ ) can be calculated by: In other words, the temperature coefficient is used to calculate the effects of temperature on battery voltage predictions: where is the calibrated battery voltage, thereby the temperature effects are taken into account. For the temperature range considered by the battery, the temperature coefficient is considered constant at -4mV/C in cell V2. In the previous equation, the parameters A, B, C, and D are battery current functions that can be obtained from the quadratic polynomial equations as follows [42]: The parameters can be calculated via the least-squares method through equations expressed with battery performance data (given by the manufacturer). The parameters used for the battery are defined as: Here, I > 0 indicates the battery charge process and I < 0 indicates the battery discharge process.

Limitations of Battery Simulation
Battery SOC is used as a design variable to control overcharge and discharge protection. Overcharging may take place when the ballast power is generated by the hybrid system or when there is a low load demand. Herein, when the battery SOC reaches the maximum value = 1, the control system interferes and halts the charging process. In other words, if the charging mode is reduced to a minimum level = 1 − , the control system shuts off the load.

ACS Economic Model
In the economic method, grounded on the concept of the annual cost of the system (ACS), the annual cost of the system is combined with the annual marginal cost , the annual replacement cost , and the annual maintenance cost of . For this model, five important parts are considered: hydropower, wind turbine, battery, and other parts are equipment that is not included in the design variables. ACS can be expressed by:

Annual Investment Cost
The annual investment cost of each combination of the hybrid system is known as the installation cost. The investment cost can be calculated as follows: where is the life time combined (year), CRF is the capital recovery factor that is a rate for calculating the present value of one year (a series of equivalent annual cash flows). The CRF is calculated using the following equation [43,44]: The annual real interest rate i is given by the following equation that is pertinent to the nominal interest rate ′, and the annual inflation rate :

Annual Replacement Cost
The annual replacement cost is the annual amount of all replacement costs incurred during the life of the project. In the proposed hybrid system, only the battery necessities of being periodically supplanted during the project [45]: where is the cost of replacing the battery, is the lifespan of the battery (years), and is the meeting factor, equivalent to a rate for calculating the future value of a series of annual cash flows:

Maintenance Costs
The maintenance cost of the system is calculated by considering the given inflation rate: In this equation, ( ) is the n th -year maintenance cost. The lowest ACS is considered one of the optimal combinations that can ensure the required power supply reliability.

LCE Economic Model
An optimal integration of the hybrid wind and hydropower turbine energy system must be found, from both reliability and economic standpoints [46]. Based on the studied wind and hydropower hybrid power generation system, the energy leveling cost is expressed as the total cost of the hybrid system over the energy supplied. The economic model can be divided into three parts (wind turbine, hydropower turbine, and battery bank): That LCE is the Levelized cost of energy. In this equation, , and are the total investment cost and maintenance cost in the lifespan of hydropower, battery, and wind power generation systems, respectively, is the total investment cost and maintenance cost in the lifespan of the battery bank, ℎ and are the lifespans of a hydropower and wind system, respectively, is the life of the battery bank and ( , , ℎ) is the annual energy that is supplied by the wind-hydropower hybrid system. The combination with the lowest LCE is taken as the optimal combination that ensures the required reliability of the power supply.

Ant Colony Optimization (ACO)
A promising source of repetition is the collective behavior of ants in nature. To find food, they randomly search the space around their nest, and when they find food, according to the type and quality of food, they bring some of it to their nest, and when they return, they secrete a pheromone. The ants come to the food and as the number of ants increases, the amount of pheromone secreted increases, and the shortest possible path is identified. Using the ant colony algorithm (ACO), the longest and shortest paths from each node to another node within the network can be optimized, as well as the effect of activity dependence on resource movement. Since the ACO algorithm attempts to mimic the behavior of real ants using a simulation process, a probability function is used to describe the path. The probability of accessing point from point (transfer probability) is calculated according to the following formula: In this formula, is the amount of pheromone chemical in points ( , ), which is also called the ant sense. is the ratio of points of view ( , ) which is called the ant eye and is a heuristic of the problem and is defined as = 1⁄ , where is the distance between two nodes and . and are compatible parameters, introducing the importance of the path against the aspect ratio when choosing the next point to move. Finally, tabuK is the list of nodes visited in the current path, called ant memory. Fig. 2 shows the flowchart of the ant colony algorithm.

Objective Functions
In the proposed method, the output of the optimal measurement algorithm is the number of wind turbines, batteries, and hydropower turbines [47]. These values must be optimized so that the 20-year cost of the system is ultimately equal to the minimum possible value. The total cost of the system ( )($) is equal to the total investment costs ( )($) and maintenance costs ( )($): In which: vector includes the amount of power that can be extracted from a hydroelectric generator, wind turbine, batteries, inverter power, the number of charger batteries, and the optimal installation height of wind turbines. The and values are the payback ratio and the present value of the payment, respectively, which are calculated by the following equations: In the above equations, is the discount rate and is the useful life of the system.

Case Study
The geographical specifications of the study area, Khalkhal, Ardabil, Iran, are given in Table 2. The optimization problem examines the ability or inability of the composition obtained for the system, as well as the degree of reliability of the system in meeting consumer demand, which is shown in Fig.  3. These constraints are tested in all simulation steps and if not implemented in one step, the resulting combination is unsuitable and is removed from the possible designs. According to Fig. 3, the peak electricity consumption of the site is related to the summer season. Therefore, the integrated system must be optimized to be able to supply the energy demand. According to studies on the region, on these days of the year, the region has suitable conditions for installing hydroelectric generators in the path of water transmission canals and also receiving suitable wind energy for installing wind turbines, and finally providing reliable energy required by the site during the year. Information about wind intensity in this region is obtained from the Renewable Energy and Energy Efficiency Organization (SATBA) [49] and applied in the simulation. This information is shown graphically in Fig. 4. The objective function of the problem is a hybrid model based on the cost of power plant construction and the reliability of the proposed system, which is calculated and presented by the ACO algorithm.
Information about water flow through the river located in the reference site and also values related to wind intensity are applied numerically at the beginning of the optimization problem. According to the figure 2, the output power of small wind and hydropower units is calculated in each simulation step and the load supply condition is checked at the same time. If the above condition is not satisfied in one iteration, the plan is removed from the existing plans, and the proposed algorithm searches for the optimal point of the problem with the aim of the lowest possible cost and also provides a reliable load under different weather conditions. The technical specifications and prices of the equipment used in this research are shown in Table 3 based on the new standards and the latest prices.

Results and Discussion
The proposed system is related to a remote village in Ardabil to supply a 50 kW load at peak consumption. To investigate the response or non-response of the proposed system to load demand during the year, an approximate reliability model has been used in this study. The total energy generated by the wind turbine and hydropower system is shown in Fig. 5. As seen in Fig. 5, the lowest average of total energy production by wind turbines and hydropower turbines is related to the spring and summer months, whereby the lowest amount is related to June with an average of 198 kW. After June, September and August are on average with 225 and 235 kW, respectively. However, the highest average production for the studied system is related to the cold months of the year, winter and autumn, with the highest average in February, producing 556 kW, and June producing 545 kW of electricity. Fig. 6 shows the amount of electricity generated by a hydropower turbine.
According to Fig. 6, the highest amounts of electricity generation by turbines are related to the wet months of the year, spring and summer, specifically the highest average is related to September with 14.3 kW, while for this year the lowest average is related to December with 5.6 kW. To find the optimal composition of the proposed power plant, a MATLAB programming environment has been used . The program was performed with an initial population of 60 people and 200 iterations for the proposed algorithm and the optimal combination presented in Table 4.  According to the results of the hybrid system simulation, the proposed algorithm is implemented with different populations and the results were obtained in 5 stages of the simulation. The best solution was obtained in 150 iterations of the hybrid system simulation. For this case, to supply the required electric charge, it will need 10 wind turbines and 23 kVA batteries, as well as a 14.6 kW hydropower turbine, which will cost $ 202,138. According to the results, the accuracy of the ACO is corroborated and has a lower cost for the construction and operation of a combined system in the study area. Fig. 7 and Fig. 8 present the convergence process of the ACO considering the hydropower turbine and the hydropower turbine, respectively.
According to the results, the COE in the wind-hydropower system is gauged at 0.261 $/kWh battery storage. The studied system is implemented to provide a reliable load at peak time (50 kW) and with an optimal combination targeted at achieving the minimum cost of construction of the power plant. The obtained value is significantly high for system reliability and is suitable for implementing the system in the study area. It should be added that the system has a lifespan of 20 years and all prices are real and based on modern standards.

Conclusion
The main purpose of integrating small hydropower turbines with other energy sources is the reliability of the supply of electricity to remote villages under different weather conditions and, on the other hand, the reduction of construction and operation costs of the system. In this thread, the study examined the costs of a grid-independent wind-hydropower plant over 20 years. With a case study of a village in Ardabil, Iran. The river water flow in this area was measured at 146 L/s. The best waterfall height in this area was identified as 5 meters, which is the best point for installing a hydroelectric generator . In addition to the high reliability of the proposed system during the simulation process, the construction cost of the power plant is $ 45656 lower than the case without the use of a hydropower turbine, due to the cheaper hydropower system compared to wind turbines. The optimal combination of this power plant was achieved using the ACO algorithm. According to the results, the COE in the wind-hydropower system, with battery storage was obtained at 0.261 $/kWh. Therefore, adding a hydropower system resulted in reducing the total cost of constructing the system, augmenting the reliability of the system, and reducing the cost of energy production. The results of the simulation could be implemented in the study area by implementing the integrated wind and hydropower plant project at this site, while providing the necessary electricity from renewable and permanent sources, reducing the environmental pollution caused by the use of end-of-life fossil fuels.