The power supply mode of island microgrid with a variety of complementary energy resources is one of the most effective ways to solve the problem of future island power supply. Based on the characteristics of seawater desalination system and water demand of island residents, a power allocation strategy for seawater desalination load, storage batteries, and diesel generators is proposed with the overall consideration of the economic and environmental benefits of system operation. Furthermore, a multiobjective optimal operation model for the island microgrid with wind/photovoltaic/diesel/storage and seawater desalination load is also proposed. It first establishes the objective functions which include the life loss of storage batteries and the fuel cost of diesel generators. Finally, the model is solved by the nondominated sorting genetic algorithm (NSGA-II). The island microgrid in a certain district is taken as an example to verify the effectiveness of the proposed optimal method. The results provide the theoretical and technical basis for the optimal operation of island microgrid.

There has been a long coastline and a large number of islands in China. Safe and reliable power supply is the basic guarantee for the development and construction of the islands and the daily life of the residents. In the past, the islands used to adopt submarine cables or diesel generators for power generation, which leads to frequent breakdowns and serious environmental pollution. Generally, it has abundant renewable energy such as solar energy and wind energy on islands and surrounding regions. With the development of distributed generation and microgrid technologies, the island microgrid could be one of the effective ways to meet the demand of island power supply with a variety of complementary energy resources in future [

As to independent microgrids, in [

As to independent island microgrid with wind/photovoltaic/diesel/storage and seawater desalination load, in [

In the optimization process of island microgrid in the future, the renewable energy generation, controllable power sources, and seawater desalination load would be the typical combination form of island microgrid if the production construction, the need of the residents, environmental, and economic benefits were taken into consideration. Considering the current research situation, even though some research results have been worked out, there are still some problems in two aspects.

Firstly, for the multiobjective process mode of optimal operation, the current research results show that there are multiple objectives of microgrid optimal operation, but most processes are to convert the multiobjective optimization problems to single objective optimization problem [

Secondly, for the regulation functions of the seawater desalination load, seawater desalination system is used to satisfy the need of water demand. If the unit capacity and storage capacity are suitable, it would be stable and complementary for energy output of intermittent power such as wind/photovoltaic energy.

Therefore, this paper shows the typical structure of island microgrid which consists of wind/photovoltaic/diesel/storage and analyzes the function of each part and then proposes the optimal operation strategy of island microgrid. On this basis, source optimal operation model of island microgrid has been established and solves the problem with the nondominated sorting genetic algorithm (NSGA-II). The operation optimization of independent island microgrid is achieved.

The island microgrid system discussed in this paper mainly consists of photovoltaic generation, wind power generation, diesel generators, an energy storage system, desalination load, and conventional load, as shown in Figure

Structure of the island microgrid system.

In an island microgrid system, the photovoltaic and wind power generation systems are uncontrollable micropower source, and they should be scheduled to generate as much power as possible during generation scheduling optimization. As typical controllable load, the desalination system could be used to smooth the volatility of photovoltaic and wind power on the basis of meeting water demand of island residents. The energy storage system can also be used to smooth the volatility of photovoltaic and wind power generation. Moreover, it can cooperate with the diesel generator to ensure the stability of system voltage and frequency. This could better enhance the stability of the microgrid system.

The seawater desalination load is considered as controllable load, which could help to adjust system power output. According to the water demand of island residents, reservoir volume, and the number of desalination units, the upper and lower bounds of desalination load per hour can be calculated. Let

The lower limit of desalination load is determined by the minimum water volume of the reservoir, current water demand, and current water volume of the reservoir.

The current water volume is sufficient:

The current water volume is insufficient:

The upper limit of desalination load is determined by the maximum water volume of the reservoir, current water demand, and current water volume of the reservoir.

The current water volume is low. When satisfying the current water demand and turning on all the units, the upper limit of water reservoir still cannot be reached:

The current water volume is low. When satisfying the current water demand and turning on all the units, the upper limit of water reservoir will be reached:

Moreover, the desalination load upper and lower limit can be determined according to the maximum and minimum number of desalination units in each time period:

The power regulation function of desalination system is as follows: when solar and wind energy is sufficient, the desalination units should be turned on as many as possible to utilize more renewable energy. When solar and wind energy is insufficient, the desalination units should be turned off as many as possible to provide relief for diesel generators and energy storage system.

If the system power balance cannot be achieved by adjusting desalination system, the battery charging/discharging power should be adjusted.

According to state of charge and the rated power limit of battery charging and discharging machine, the upper and lower limit of the battery output during each time period can be calculated to determine the regulating range of the battery. Let

The power regulation function of the battery is as follows: when the power output of renewable energy is high, it can be used to charge the battery and make use of the excessive energy that cannot be fully used by desalination system; when the power output of renewable energy is low, the battery can be set to discharge to meet the load demand of microgrid.

If the system power balance still cannot be achieved by adjusting desalination system and battery output, the diesel generator output should be adjusted.

To ensure the operating temperature and prolong the service life of the generator, the minimum power output of the diesel generator should be 30% of its rated power. Let the rated power of the diesel generator be

The power regulation function of the diesel generator is as follows: when the renewable energy is extremely insufficient, the load demand in the microgrid system can be met by turning on diesel generators to balance the power in the system.

Diesel generators, batteries, and desalination load are controllable to some extent. Considering the economic efficiency and utilization of renewable energy, in the real operation, the desalination load is first scheduled and then the battery charging/discharging power is adjusted. Finally, the diesel generator output adjustment is done to maintain system power balance.

Let

Comparing net power

The net power is less than or equal to the lower limit of desalination load:

The basic water demand should be met:

The net load is insufficient for the desalination units, so the batteries and diesel generators should be adopted to meet load demand. Specifically, the maximum discharging power is first determined by the rated capacity of the battery, state of charge (SOC), and discharging depth. The vacancy should be supplied by diesel generators.

Let

If the vacancy is less than the maximum discharging power of the battery,

Discharging power output of battery during period

If the vacancy is more than the maximum discharging power of the battery,

The battery operates in the highest discharging rate. The diesel generator needs to be started and its power output is determined by the remaining power vacancy.

If the remaining power vacancy is more than the minimum output limit of diesel generators,

The output of diesel generator

If the remaining power vacancy is less than the minimum output limit of diesel generator,

The output of the diesel generator is its minimum power output limit, and in this paper it is 30% of its rated power, that is,

The net power is more than the lower limit of desalination load and less than the upper limit of desalination load:

The number of desalination units

Correspondingly, the output of desalination units is

The net load still has some vacancy, which can be absorbed by battery:

If the battery cannot meet the charging condition, these vacancies should be abandoned. The diesel generators do not need to be started.

The net power is more than the upper limit of desalination load:

To utilize as much renewable energy as possible, the desalination units operate at full power rate:

The net load still has some vacancies, which can be absorbed by the battery:

If the battery cannot meet the charging condition, these vacancies should be abandoned. The diesel generators do not need to be started.

The aim of such strategy is to reduce the operating time of diesel generators, and the detailed operating flowchart is shown in Figure

The flowchart of energy exchange strategy in island microgrid.

In the dispatching model of microgrid which consists of wind, solar, diesel generator, and desalination load proposed in this paper, the objective is to minimize the life loss of batteries and minimize the fuel cost of diesel generators.

The relationship between the accumulated service life of lead-acid battery and its discharging depth can be seen from the relationship between its effective weighted factor (EWF) and its SOC. It is shown in Figure

The relationship between EWF and SOC of lead-acid battery.

When the SOC value is 0.5, that is, the discharging depth of the battery is 0.5, the EWF is 1.3, which means the battery service life will add 1.3 Ah when it discharges 1 Ah [

Suppose the service life loss factor of the battery is

When the initial discharging depth

When the final discharging depth

When the final discharging depth

When the final discharging depth

When the initial discharging depth

When the final discharging depth

When the final discharging depth

When the initial discharging depth

The objective function is

where

From another perspective, to minimize the fuel cost of diesel generators is another way to improve the utilization of renewable energy. This is because that in the microgrid system with photovoltaic generation, wind power generation, diesel generators, and an energy storage system, the main power source is the wind, solar renewable power generation system, and the diesel generators. When the fuel cost of diesel generators decreases, the output of diesel generator decreases as well. This will increase the proportion of renewable energy in the whole system and improve the utilization of renewable energy.

The power output constraints of diesel generators are as follows:

System power balance constraints are as follows:

Battery charging/discharging constraints are as follows:

Battery capacity constraint is as follows:

Desalination units number constraints are as follows:

According to the model of optimal operation, the problem can be expressed as follows:

The solution procedure of multiobjective optimization model proposed in this paper is essentially a nonlinear optimization problem which contains multiple decision variables, and the model is solved by the nondominated sorting genetic algorithm (NSGA-II). NSGA-II was put forward by Deb and so forth and is modified based on the algorithm of NSGA in 2002 [

The fast nondominated sorting method based on classification is employed, and its computational complexity is

The notion of crowding distance is proposed to show the fitness value of different elements in the same level after fast nondominated sorting procedure to ensure that the individuals in Pareto front can expand the scope of Pareto frontier evenly;

The mechanism of the elite reservation is introduced. The new generation is created by the competition between the offspring individuals and the parent individuals. This mechanism can improve the overall level of evolutionary population. Specific process is shown in Figure

The process of the power allocation strategy by NSGA-II.

The system studied in this paper is an independent island microgrid system with wind/ photovoltaic/diesel/storage and seawater desalination load on an island in China. The parameters of each component in the system are given in Table

Parameters of each component.

Components | Parameters |
---|---|

Photovoltaic battery capacity/kW | 285 |

Wind power capacity/kW | 350 |

Storage battery capacity/kWh | 300 |

The maximum charge/discharge power of storage battery/kW | 100 |

Diesel generator rated power/kW | 300 |

In addition, the maximum electricity load is 300 kW, and daily water demand is about 500 tons on the island. Eight seawater desalination machines compose the desalination system. The rated power of each machine is 25 kW and the rated daily water production is 100 tons. The maximum reservoir capacity of the system is 160 tons, and the minimum water requirement is 48 tons.

The data about wind, solar radiation, the conventional load in winter and summer, and the water demand is shown in Figures

The speed distribution of the wind.

The annual radiation data.

The conventional load in winter and summer.

The annual mean wind speed is about 10 m/s at the height of 90 meters on the island, which is influenced by the coastal air current. Higher wind velocity appears in winter and in summer the speed is relatively low.

The monthly variation range of total solar radiation is 260 MJ/m^{2}–663 MJ/m^{2}. The summit appears in April and May, and the highest value achieves 663 MJ/m^{2} in May. The trough value of radiation appears from November to January of the next year, and the lowest value achieves 260 MJ/m^{2} in December. The annual solar radiation is about 5695 MJ/m^{2}, which means the solar energy on the island has great development potential.

The total daily water consumption of residents is about 500 tons, the peak period is 9:00 a.m. and 7:00 p.m. The maximum of daily water demand achieves 22.8 tons at 7:00 p.m. and the minimum value is 16.3 tons.

The population quantity is 200. The maximum iterations are 5000 in optimization procedure. The crossover rate is 0.3. The mutation rate is 0.5. The optimization results and the extreme solution of Pareto are shown in Figure

The extreme solution of Pareto optimal concentration considering the effect of auxiliary power regulation.

Storage battery life loss | Diesel fuel consumption/RMB | |
---|---|---|

The minimum of storage battery life loss | 0 | 4021 |

The minimum of diesel generator fuel consumption | 0.0012 | 3230 |

The daily water demand condition.

From Figure

From Figure

In Figure

Life loss expenses are converted into loss expenses in order to reflect the economic efficiency of optimal solution in general, and the loss expense of battery is

Total loss expense of system operation consists of two parts: the expenses of storage batteries life loss and the fuel consumption of diesel generators:

Parameters comparison of seawater desalination load considering auxiliary power regulation and without auxiliary power regulation.

Parameters | Considering auxiliary power regulation | Not considering auxiliary power regulation |
---|---|---|

The minimum of storage battery life loss | 0 | 0 |

The maximum of diesel generator fuel consumption/RMB | 4021 | 4507 |

The maximum of storage battery life loss | 0.0012 | 0.0015 |

The minimum of diesel generator fuel consumption/RMB | 3230 | 3924 |

The minimum of total loss expenses/RMB | 3576 | 4357 |

Storage battery life loss in minimum total loss expenses | 0.0011 | 0.0014 |

Fuel consumption in minimum total loss expenses/RMB | 3256 | 3947 |

Considering auxiliary power regulation of seawater desalination load, the extreme solution is superior over the condition without auxiliary power regulation. From the overall economic point of view, both the value of life loss and fuel consumption become lower than the condition without auxiliary power regulation at the minimum of total loss expenses, and the total expenses are also lower. The reduction of fuel consumption enhances the renewable energy utilization. Auxiliary power regulation of seawater desalination load significantly increases the operating economic efficiency and renewable energy utilization of island microgrid.

A suitable solution from Pareto front is selected to get the output results of the various microsources in 24 hours at the minimum of total loss expenses. The output results are shown in Figure

The result of Pareto front considering the effect of auxiliary power regulation.

In Figure

Figure

The output results of the various microsources in 24 hours without auxiliary power regulation at the least total loss expenses conditions.

Figure

The output results of the various microsources in 24 hours without auxiliary power regulation at the least total loss expenses conditions.

SOC of storage battery at the least total loss expenses condition in 24 hours.

Water capacity of the seawater desalination system in 24 hours.

For the island microgrid with wind/photovoltaic/diesel/storage and seawater desalination load, a multiobjective optimization model and its solving method have been proposed. The controllable sources such as the seawater desalination load, storage batteries, and diesel generators have been taken into consideration and the power allocation strategy is proposed. The model is analyzed by solving a problem on a practical island, and the rationality of the proposed model and the power allocation strategy is verified.

During the process of practical island microgrid optimization, according to the project owner’s specific investment requirement and generation proportion of renewable energy, the optimal solution method can be applied in the island microgrid. The theory evidence and technical support of the island microgrid optimal operation can be provided through the method.

The authors declare that there is no conflict of interests regarding the publication of this paper.

This work is supported by the National Natural Science Foundation of China (no. 51277067) and the Project of Capital SCI & TEC Resources Platform (no. Z131110000613053).