1. Introduction
With the consumption of a large amount of fossil fuel, various environmental problems have become increasingly severe [
1,
2]. Many countries, like China, need to increase their share of renewable energy in total electricity generation and optimize the existing energy structure [
3,
4]. Moreover, there is still 22.5% of the rural population without access to electricity in 2016 globally [
5]. Since the wind/solar/battery hybrid power system (WSB-HPS) can make the most use of renewable energy [
6,
7] and the market price of small-capacity renewable power generation equipment has dropped significantly, it is highly possible to establish a 100% renewable energy system for residential households [
8]. So, it is worth to study the techno-economic feasibility for this system.
From the aspect of system optimization, electricity economic benefit, power supply reliability, and environmental benefit are usually considered. Petrillo et al. [
9] proposed a model to minimize the cost of the system throughout the whole lifetime, taking into account purchase costs, installation costs, and equipment replacement and maintenance costs. Rodolfo et al. [
10] additionally considered the capital recovery factor and used the total cost and electrical power to calculate the unit cost of energy. Li Chong et al. [
11] adopted the total net present value and the levelised cost of energy to analyze the techno-economic feasibility.
The reliability of the system power supply is mainly calculated based on the amount of load exceeding the amount of power supply or its ratio of total power consumption [
12,
13,
14]. In terms of the environmental benefit, many studies have focused on reducing greenhouse gases such as carbon dioxide emitted from diesel generators in hybrid power generation systems [
14,
15,
16,
17]. There are mainly two kinds of methods to calculate capacity configuration. Mathematical programming is more suitable for single-objective optimization problems [
18,
19,
20]. The heuristic algorithms, such as particle swarm algorithm and genetic algorithm, are ideal for solving nonlinear problems and multi-objective optimization [
21,
22,
23,
24]. Besides, some researchers use mixed integer linear programming to analyze [
25].
In terms of system structure and economic analysis, Li Chong et al. [
11] performed research on the off-grid hybrid power system in Urumqi, China. Based on HOMER software, the effects of ambient temperature, resident load, and the tilt angles of the photovoltaic (PV) component are considered. Based on a remote village in India, Sen et al. [
26] used four kinds of renewable energy sources for a hybrid power system. According to different load types such as residential, agricultural, commercial, and industrial, the optimal off-grid capacity configuration is derived. Celik et al. [
27] performed a new techno-economic analysis for small autonomous hybrid power system under optimization purposes. Tao et al. [
28] presented a detailed feasibility study and economic analysis of a hybrid solar-wind-battery power system for an island.
To improve energy efficiency, some researchers designed a combined cooling, heating, and power (CCHP) system [
23,
29] or a combined heating and power (CHP) system [
30,
31]. However, for many households, as in China, the mainly thermal equipment, such as a water heater, is usually powered by solar thermal radiation or electricity. Moreover, the use of gas turbine will cause CO
2 emission and impose high costs of operation and maintenance [
32].
Most of the research focuses on a single aspect, such as optimization target design and algorithm optimization, and lacks comprehensive study on the whole system. For example, when the system is under continuous cloudy or no wind weather conditions, the insignificance of the battery capacity will seriously affect the power supply reliability. Also, the different types of equipment need to be selected according to the load and renewable source. Different from analyzing traditional large capacity microgrid, the capacity of equipment required by a household is usually fixed and discrete. In the actual application process, some uncertainties on the energy structure need further sensitivity analysis.
To solve the above problems, this paper uses actual annual weather data to fully consider the impact of dynamics and random weather on the energy structure. The total net present value (NPV) model of the system is established by taking into account all factors as much as possible. Taking a certain region in Hangzhou, China as an example, the best 100% renewable energy structure for the household is obtained by comparing the NPV and the levelised cost of energy (LCOE). Sensitivity analysis of the NPV was carried out, and the changes of renewable energy structure under different renewable resource richness are studied, which provides a reference for the actual planning and operation.
5. Case Study
Based on the real monitoring data, the load curve of a household is shown in
Figure 3. The total annual electricity consumption of the household is 5935 kWh. There are three peak load times in one day, which are located at 7 am, 12 noon and 8 pm. At other times, the load is used for the indoor lighting and the standby state of electrical appliances such as refrigerators and televisions. Since there is no heating pipe in the area, it is dependent on the air conditioner for heating in winter. The simulation time is set to 1 h, and the whole lifetime of the system is set to 25 years.
The system optimal search space is set, as shown in
Table 4. The capacity of PVs is from 0 to 20 kW, and the number of wind turbines is from 0 to 10. In total, there are 640,000 combinations for the household.
5.1. Techno-Economic Feasibility Analysis
This paper adopts the DC bus structure. The DC bus voltage is 300 V, and the AC bus voltage is 220 V. The results of different structures in the lifetime of the system are obtained with the help of the simulation software HOMER. Furthermore, the techno-economic feasibility is analyzed in depth.
As shown in
Figure 4 and
Table 5, the optimal structure W-P-B (1) is obtained by comparing the system NPV and LCOE. The PV capacity is 5 kW, the wind turbine capacity is 7 kW, and the energy storage battery capacity is 34.56 kWh. According to wind speed, different types of wind turbines can be used to improve the utilization of wind. The LCOE is calculated to be
$0.146 based on the total power generation, less than two times the current price of electricity in the rural area of Zhejiang Province (
$0.08). Compared with the P-B system and the W-B system, the cost of the W-P-B (1) is reduced by 44.2% and 18.8%, respectively.
Moreover, the electricity price of diesel generators is usually between $0.29~$0.44, approximately two or three times of the WSB-HPS. As a result, the WSB-HPS has shown significant economic advantages in areas with abundant renewable resources and proven the feasibility of 100% renewable energy supply for residential households. Besides, it is necessary to calculate the power generation of each piece of equipment in order to ensure that the regular load demand is met. Based on this, the corresponding analysis suggestions are given.
5.1.1. System Power Generation Analysis
As shown in
Table 6, the hybrid power system will waste a lot of excess electricity power (174% in this paper). If the household replaces traditional fossil energy-consuming equipment with the electrical equipment, such as electric vehicle and electric heating system, the economic benefit can be significantly improved.
Figure 5 presents the output power of PV is even lower in summer than in winter, although the irradiance in summer is higher than in winter. Because the efficiency of PV is sensitive to temperature, and in summer, the internal temperature of PV can reach more than 50 °C.
As shown in
Figure 6, the inconsistency between the peak load and the peak of renewable energy leads to the charge and discharge behavior of the energy storage battery. During the daytime, solar energy is concentrated and abundant, which increases the SOC of the energy storage battery. The distribution of wind energy is more dispersed and concentrates in the morning and evening. By making full use of the coordination effect of the energy storage battery, the reliability of the resident electricity is guaranteed.
According to the statistics of
Figure 7, the frequency that the state of charge (SOC) is above 60% is 94.5%. The frequency that the SOC is above 80% is 83.5%. In February and August, due to the heavy load and the maximum discharge depth of battery, there was a shortage of capacity. The maximum power shortage was 3.5 kW, and the total unmet load was 51.1 kWh, accounting for 0.86% of the total electricity consumption. Therefore, the system can use standby energy storage equipment in February and August temporarily, which will reduce battery life loss and improve power supply reliability significantly.
5.1.2. Sensitivity Analysis
Sensitivity analysis is used to verify whether the best result has robustness under the allowable fluctuation range of external factors, and provide a reference for practical projects in different renewable sources. The sensitivity analysis was carried out in five aspects, which is shown in
Figure 8.
Figure 8a shows that the NPV decreases exponentially with the increase of the allowable shortage capacity. When
Lloss increases from 0% to 1%, the NPV decreases by 15.6%.
Figure 8b,c show that the NPV decreases linearly with the increase of renewable resources richness, and the effect of wind speed is more significant than that of irradiance.
Figure 8d shows that the energy storage batteries still account for a higher proportion of the total cost, compared with other equipment. The NPV of the system increases linearly with the increase in the price of the energy storage battery.
Figure 8e shows that the changes of NPV and LCOE are not apparent when the maximum allowable discharge depth is set to a different value. Since the capacity shortage occurs only in a few days throughout the year, the effect of the higher discharge depth on the battery life is not obvious.
Figure 9 and
Figure 10 analyze the impact of the richness of renewable resources on the optimal energy structure. The overall trend shows that the capacity of each part of the system decreases with the increase of renewable resource richness. However, as can be seen in conjunction with
Figure 6, the PV output power is mainly concentrated around noon, which can basically meet the load peak demand at noon. However, if the SOC of the battery is very high at this time, the excess solar energy will not be stored in time. This will cause much waste. The distribution of the output power of the wind turbine is more dispersed within one day, which is more similar to the residential load curve. It is beneficial to replenish the energy of the energy storage battery in time when the load is low. Therefore, the effective utilization rate of wind energy resources will be significantly larger than that of irradiance resources. Compared with the annual average irradiance, with the increase of the average annual wind speed, the capacity of each piece of power generation equipment is significantly reduced. When the average wind speed reaches 6 m/s, wind power completely replaces PV power.
5.2. Grid-connected System Analysis
According to the analysis of the above system, much electric power is wasted. If the WSB-HPS is connected to the grid in some conditions, the difference of the techno-economic feasibility between the systems needs to be analyzed. The grid-connected system structure is shown in
Figure 11.
According to
Table 7, the W-P system is still the best of different configurations for grid-connected. Compared with the 100% renewable energy system, residents can achieve
$8079 during the whole lifetime of the system. The price of the energy storage system is still higher than that of the power grid as a backup power source. As the penetration rate of renewable energy increases, the system economic benefits have increased significantly. As shown in
Table 8, the emissions of atmospheric pollutants and corresponding fines are also reduced.
If the grid has requirements for the volatility of purchasing power, it is necessary to consider the economic benefit of energy storage. By increasing the purchase price beyond the fluctuation range as a constraint, the simulation results are shown in
Figure 12. When the energy storage capacity is insufficient, the higher electricity price will increase the system NPV. When the energy storage capacity is excessive, the higher energy storage cost will increase the system NPV.
As shown in
Figure 13, when the hybrid power system produces more electricity than the load consumption, the excess electricity will be sold to the grid. When renewable resources are insufficient, the power grid acts as a backup power source to sell electricity to the household. According to
Table 9, only about 10% of electricity needs to be purchased from the grid.
6. Conclusions
Based on the lifecycle of the system, the techno-economic feasibility of 100% renewable energy for residential households is analyzed. For the windy areas in China, the renewable hybrid power system has a significant economic advantage compared to a diesel generator. Based on the case in this paper, the optimal energy structure contains 5 kW PV, 7 kW wind turbine, 5760 Ah battery, and a 6.2 kW convert. The levelised cost of energy of the renewable energy system is around 0.146 $/kW. The unmet load is only 0.86% of the load consumption, which is concentrated in February and August, and the excess electricity is 174%. For the grid-connected situation, the renewable energy system can bring $8079 in 25 years with the LCOE is −0.062 $/kW.
The analysis results have some reference value for the actual operation planning: (1) NPV decreases exponentially with the increase of allowable shortage capacity of the system; (2) Due to the feature of residential load, the effect of wind resource richness on the economy and capacity of the system is more significant than that of irradiance resource richness; (3) The high internal temperature of PV makes the conversion rate less than ideal in summer; (4) Changing the energy storage allowable depth of discharge within a specific range has little effect on the economy for residential household; 5) The temporary backup energy storage in high-load months can significantly improve the reliability and economy of the system.
This paper only calculates and analyses the load of a general residential household. The simulation analysis is not carried out for the case with special load requirements, such as electric vehicles. In further research, a large number of residential loads will be collected for the cluster analysis. The impact of different residential load types on the techno-economic feasibility of 100% renewable energy power system can be studied.