402

IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 4, NO. 2, APRIL 2013

A Novel Operation and Control Strategy for a Standalone Hybrid Renewable Power System
A. M. Osman Haruni, Student Member, IEEE, Michael Negnevitsky, Senior Member, IEEE, Md. Enamul Haque, Senior Member, IEEE, and Ameen Gargoom, Member, IEEE

AbstractThis paper proposes a novel operation and control strategy for a renewable hybrid power system for a standalone operation. The proposed hybrid system consists of a wind turbine, a fuel cell, an electrolyzer, a battery storage unit, and a set of loads. The overall control strategy is based on a two-level structure. The top level is the energy management and power regulation system. Depending on wind and load conditions, this system generates reference dynamic operating points to low level individual subsystems. The energy management and power regulation system also controls the load scheduling operation during unfavorable wind conditions under inadequate energy storage in order to avoid a system blackout. Based on the reference dynamic operating points of the individual subsystems, the local controllers control the wind turbine, fuel cell, electrolyzer, and battery storage units. The proposed control system is implemented in MATLAB Simpower software and tested for various wind and load conditions. Results are presented and discussed. Index TermsBattery storage, electrolyzer, energy management and power regulation system, fuel cell, load management, standalone hybrid power system, wind energy conversion system.

I. INTRODUCTION N remote and isolated areas, diesel generators are commonly used to provide electricity because grid connections are often neither available nor economically viable. Diesel generators are popular in remote area power system applications for their reliability, low installation cost, ease of starting, compact power density, and portability [1], [2]. However, diesel generators are becoming expensive to run; they also need frequent maintenance. Most importantly, they pollute the environment. Hybrid power generation systems that combine different renewable energy sources and energy storage systems offer an environmentally friendly alternative for standalone operations. However, there are several challenges for the hybrid power system. Appropriate control and coordination strategies among various elements of the hybrid system are required so it can deliver required power. Renewable-energy-based hybrid systems must also be also reliable and cost-effective.
Manuscript received September 23, 2011; revised August 05, 2012; accepted October 06, 2012. Date of publication December 21, 2012; date of current version March 18, 2013. The authors are with the Centre for Renewable Energy and Power Systems (CREPS), University of Tasmania, Hobart, Tasmania 7001, Australia (e-mail:amharuni@utas.edu.au; michael.negnevitsky@utas.edu.au; mehaque@utas.edu.au; ameen.gargoom@utas.edu.au). Color versions of one or more of the gures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identier 10.1109/TSTE.2012.2225455

Renewable energy sources such as solar, wind, bio-mass, geothermal, and tidal energy can play an important role as a primary power source for the hybrid power generation system. However, favorable conditions for renewable energy are necessary as the cost of electricity is heavily dependent on local weather patterns. As an example, a case study conducted in Tunisia [3] demonstrated that a hybrid system based on the solar power and battery storage can offer a cost-effective solution compared with the diesel-generator-based standalone power system. In favorable wind conditions, wind-turbine-based hybrid power systems can offer a cheaper option to a solar-energy-based hybrid renewable power supply [4]. Due to the intermittent nature of the wind, the instantaneous power extracted from the wind turbines often does not match the instantaneous load demand. As a result, energy storage systems are essential for continuous and reliable operation [5][11], [13], [14]. Energy storage devices provide transient stability during sudden wind and load variations [5][7]. Moreover, they are very useful for load-leveling applications [8], [9]. Different types of energy storage systems such as pumped hydro, compressed air energy storage, ywheel energy storage, thermal, hydrogen, batteries, superconducting magnetic energy storage, and super-capacitors are used in different applications for different purposes. Pumped hydro and compressed air energy storages are low cost options [8]. However, they have lower efciencies. Moreover, pumped hydro is mostly dependent on the geographical location. Flywheel energy storage, batteries, superconducting magnetic energy storage, and super-capacitors have a higher energy density and a very fast time response [8], [9]. As a result, they can support a sudden change in power demand and provide better transient stability. On the other hand, fuel cells and electrolyzers have higher power density with slower time responses [9][11], [13], [14]. Therefore, they are more suitable for long-term load leveling applications. Considering the application of energy storage systems in the wind-turbine-based hybrid power system, a combination of the fuel cell, electrolyzer, and battery can represent the most suitable option. First, excess power from wind can either be stored in the battery storage system or used to generate hydrogen by the electrolyzer. Second, batteries respond very quickly, which ensures better stability of the hybrid system during transient periods caused by sudden changes of wind and load. Third, this combination can improve the efciency of the system by sharing power so as to allow the operation of a fuel cell in a high efciency region. Recent studies on hybrid power systems focus on various issues such as the size and cost optimization [15][17], power

1949-3029/$31.00 2012 IEEE

HARUNI et al.: NOVEL OPERATION AND CONTROL STRATEGY

403

Fig. 1. Structure of the proposed hybrid power generation system.

management [14], [19][21], power quality [21], and reliability [22]. However, little attention has been paid to the load management under inadequate energy reserves during low wind conditions, which may lead to a system blackout. In fact, load management strategies are essential for providing the required power supply during these conditions. In this paper, a standalone hybrid power supply scheme suitable for remote and isolated communities is presented. The system consists of an interior permanent magnet (IPM) synchronous generator-based wind turbine, battery storage, fuel cell, and electrolyzer unit as shown in Fig. 1. This paper proposes a novel energy management and power regulation system (EMPRS). This ensures coordinated control and exible energy management of the hybrid power system. The EMPRS generates the dynamic operating point for each subsystem of the hybrid power system which ensures: 1) optimum utilization of wind resources, 2) proper management of energy storage resources, and 3) continuous operation of the hybrid power system under unfavorable wind and insufcient energy storage conditions. Based on the dynamic operating points of each subsystem generated by EMPRS, the wind energy conversion system is controlled by a rectier in order to: 1) extract optimum power by regulating the rotor speed, and 2) ensure efcient operation of the IPM synchronous generator by regulating the - and -axes components of the stator current. The fuel cell is controlled by the hydrogen ow regulator and boost converter. The electrolyzer is controlled by a buck converter. The battery storage is controlled by a bidirectional dcdc converter. The performance of the proposed control system is tested under different wind and load conditions, and results validate the effectiveness of the system. II. SYSTEM COMPONENTS AND CONTROLLERS MODELING A. Wind Energy Conversion System (WECS) Modeling 1) Wind Turbine Model for Optimum Energy Extraction: The can be expressed as aerodynamic rotor power from wind [23] (1) where is the air density, is the rotor swept area, is the wind speed, and are the cut-in and cut-off wind speed, respectively, and is the power coefcient which is a function of the speed ratio and the pitch angle .

The speed ratio of the wind turbine can be dened as (2) is the rotor speed, and is the radius of rotor. where From (1) and (2), for a particular wind speed, the output power is proportional to the rotor speed and can be expressed as (3) . where From (3), optimum aerodynamic rotor power from the wind turbine can be extracted by controlling the rotor speed . For a particular wind speed, the optimum power is given as [23] (4) . where Fig. 2 demonstrates the power generated by a turbine as a function of the rotor speed for different wind speeds. The optimum power extraction from the wind refers to extracting the necessary power under varying wind speed conditions. As an example, for a particular wind speed , the optimum power is generated by keeping the rotor speed equal to either or . However, as is higher than the base rotor speed, the control system has to choose the rotor speed . If the wind speed drops from to , the control system sets the rotor speed to to extract the required power. 2) IPM Synchronous Generator Model: From Fig. 3, the voltage equations of the IPM synchronous generator in the and -axes are expressed as follows [24]: (5) (6) where and are the - and -axes components of the stator voltage, respectively; is the stator resistance; and are the - and -axes components of the stator current, respectively; is the frequency; and is the ux linkage. The torque equation of the IPM synchronous generator can be expressed as follows [24]: (7)

404

IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 4, NO. 2, APRIL 2013

Fig. 4. The -axis current versus output electric power. Fig. 2. Wind turbine characteristic curve.

Fig. 3. (a) The - and (b) -axes circuits of the IPM synchronous generator.

where is the number of pole pairs, and is the generated torque of the IPM synchronous generator. From (7), the -axis stator current component for constant torque can be expressed as a function of the -axis stator current component (8) The maximum efciency of the IPM synchronous generator can be achieved by minimizing the copper and core losses. From Fig. 3, the copper and the core losses for the IPM synchronous generator can be determined as follows [25]: (9) (10) is the core loss component. where The output power from the generator can be given as

Fig. 5. Machine side converter controller.

torque. In the second stage, two controllers are used to regulate the - and -axes currents under specic torque or power conditions in order to minimize losses. B. Fuel Cell Model and Control The model used in the paper is based on the dynamic proton exchange membrane fuel cell model (PEMFC) discussed in [10] and [12]. This model is based on a relationship between the Nernst voltage and the average magnitude of the fuel cell stack voltage [10] (12) where is the fuel cell voltage, is the number of fuel cells connected in series, is the Nernst voltage, and is the irreversible voltage losses. The Nernst voltage developed in the fuel cell is dened as follows [10]:

(11) The optimum value of can be determined from the output power versus the -axis stator current curve based on (5)(11), as shown in Fig. 4. From Fig. 4, the optimum value of the - axis current component is chosen such that the output power of the IPM synchronous generator is maximized. The corresponding value of can be obtained from (8). 3) Machine Side Converter Controller Design: The machine side converter shown in Fig. 5 consists of three controllers working on the principle based on (5)(11). In the rst stage, a controller is used to regulate the speed by controlling the

(13) where is the voltage associated with the reaction free energy, is the universal gas constant, is the Faradays constant, is the fuel cell absolute temperature, the partial pressure of hydrogen in the anode, and are the partial pressure of oxygen and water available in the cathode, respectively. ,

HARUNI et al.: NOVEL OPERATION AND CONTROL STRATEGY

405

Fig. 7. Fuel cell controller. Fig. 6. (a) and (b) powercurrent characteristics of a fuel cell.

, and can be expressed using the following differential equations [10]:

(14)

Fig. 8.

characteristics of an electrolyzer.

where and are the molar follow of hydrogen and oxygen, respectively; , , and are constants; is a constant, which can be dened by the relationship between the rate of reactant hydrogen and is the fuel cell current [10] (15) The output voltage of a fuel cell at normal operating conditions is determined by the irreversible voltage loss , which can be classied into three types: the activation voltage loss , ohmic voltage loss , and concentration voltage loss [12]. The and powercurrent characteristics of a fuel cell can be seen in Fig. 6. The output power of a fuel cell is determined as follows: (16) In order to design a control strategy for the fuel cell, the hydrogen ow has to be regulated to achieve the output power based on (12)(16). Moreover, as the fuel cell voltage varies according to the dynamic operating point, as shown in Fig. 6, a controlled boost converter is used to interface the fuel cell with the dc link of the system. The fuel cell controller is shown in Fig. 7. C. Electrolyzer Model and Control The electrolyzer consumes the electric power to produce hydrogen. The alkaline electrolyzer model is used in the application. The voltage drop across each electrolyzer cell is given by [8], [33]

are parameters of ohmic resistance, is the electrolyzer current, and is the area of electrode. The total voltage drop across the electrolyzer is dened as [8] (18) is the number of cells. where The total power consumption of the electrolyzer is given as (19) The electrical characteristics of the electrolyzer depend on the voltage, current, and temperature. The nonlinear relationship of the electrolyzer cell voltage and current at a given temperature is shown in Fig. 8. The hydrogen production rate can be expressed as a function of applied current as follows [8], [14]: (20) where is the Faradays constant, is the current density, and is a function of the current density and temperature. In normal operating conditions, the hydrogen outlet rate should be equal to the hydrogen production rate so that the pressure and stored hydrogen quantity in the cathode can be maintained as constant. Based on the ideal gas law, the resultant pressure of hydrogen can be written as [8], [14] (21) is the cathode volume, is the partial preswhere sure of hydrogen in the cathode, is the molar hydrogen outow rate to hydrogen tank, and is the ideal gas constant. In order to control the power ow in the electrolyzer, the input current has to be controlled. A buck converter is used to regulate the power ow in the electrolyzer by regulating the electrolyzer current based on (17)(21) as shown in Fig. 9.

(17) where is the voltage drop across the electrolyzer, is the thermodynamic cell voltage, is the electrolyzer temperature, and are parameters for the electrolyzer over voltage,

406

IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 4, NO. 2, APRIL 2013

Fig. 10. Battery charger/discharger controller. Fig. 9. Electrolyzer controller.

D. Compressor and Tank Model The relationship between the molar ow rate from the electrolyzer and the compressor power is given as follows according to the polytrophic model [8], [14]:

where (22) where is the compression efciency, and is the polytropic work, is the polytrophic coefcient, and is the pressure of storage tank. Hydrogen storage is the difference between the hydrogen produced by the electrolyzer and the hydrogen used by the fuel cell as (23) The pressure of stored hydrogen can be derived as in the hydrogen tank (24) where is the hydrogen storage tank volume, is the ideal gas constant, and is the temperature of the tank. E. Battery Storage System Modeling and Control Among different battery technologies, Liion batteries represent a suitable option for fuel-cell-based hybrid energy storage systems due to their high energy density and efciency, light weight, and good life cycle [26]. The generic Liion battery model is used [27]. The battery state of charge (SOC) is an indication of the energy reserve and is expressed as follows [27]: (25) where is the battery current, and is the battery capacity. The battery controller is a bidirectional dcdc converter that stabilizes the dc link voltage during sudden wind and load changes. The controller is shown in Fig. 10. F. Output Voltage and Frequency Controller In a standalone power system, various loads (linear, nonlinear, balance, and unbalanced three phase loads) can be conFig. 11. Load side inverter controller.

nected. As a result, in the proposed system, we use a three phase four wire inverter with split capacitor as shown in Fig. 11. The controller of the inverter compares the -, -, and -axes components of the output voltage with their reference values . Based on the error signal, a set of PI controllers generates appropriate signals to the PWM signal generator. III. ENERGY MANAGEMENT AND POWER REGULATION SYSTEM (EMPRS) The EMPRS ensures a continuous operation of the hybrid system via the coordination of the wind turbine, energy storage system, and loads. The EMPRS works in three stages. In the rst stage, the EMPRS predicts the wind and load prole for a specied period of time. In the second stage, based on the wind and load prole and the status of energy reserve, it schedules the maximum load that can be supplied by the system. In the third stage, it determines the operating condition of each subsystem. A. Wind and Load Prediction An accurate wind and load prediction is a key factor to ensuring a robust performance of the EMPRS. In several studies conducted earlier, it was demonstrated that an accurate forecasting system can be developed for the short-term (up to 15 min) forecasting of wind and load conditions [28][32]. An integration of wind and load forecasting in the EMPRS will allow the implementation of the load curtailment in advance, thus avoiding system blackouts, as will be demonstrated in the following. B. Load Scheduling Based on the wind and load prediction, the power balance equation of the hybrid system can be expressed as follows: (26)

HARUNI et al.: NOVEL OPERATION AND CONTROL STRATEGY

407

where and are the load demand and battery power, respectively. From (26), during high wind conditions, the excess power is consumed by the electrolyzer and battery storage as follows: (27) During low wind conditions, the power decit from the wind can be supplied by the fuel cell and battery storage as follows: (28) The energy balance equation can be obtained by integrating (26)
Fig. 12. Load management algorithm.

(29) where is the total energy produced by the wind energy conversion system, is the total energy consumed by the load, is the total energy supplied by the battery, is the total energy consumed by the electrolyzer, and is the total energy supplied by the fuel cell. However, in real hybrid system operation conditions, (29) is only valid for the following conditions: 1) If , the excess energy is stored in the hybrid system. 2) If , the energy decit from wind is balanced by the fuel cell and battery storage. In this condition, the fuel cell and battery storage can produce required power provided the hydrogen storage and SOC of the battery are available. The system may experience blackout conditions if the energy reserves are not sufcient to meet the load demand. The robustness of EMPRS depends on its prediction accuracy. Although it has been demonstrated that the short-term prediction error can be as low as 1% for normal conditions, this can increase under sudden wind guests or sudden changes of large industrial loads. As a result, sufcient reserves must be allocated to offset the prediction error of up to 5%. Moreover, the unlikely event of no wind conditions may occur and continue for a relatively long period. During this period, the hybrid system will totally rely solely on stored energy. Thus, energy reserves that can serve high priority loads for a sufcient period have to be preserved. As a result, in this study a portion of energy from the battery storage system is dedicated as a reserve. For normal operation conditions, the SOC range of the battery is assumed to be between 75% and 95%. For emergency operating conditions, the SOC is allowed to be as low as 40%. However, it has to be noted that these ranges are case specic. A number of constrains such as historic wind and load conditions, economic sizing of the storage system, security, and reliability issues of the system can affect the optimum range of SOC for both normal and emergency operating conditions. In this study, the selected ranges are used for demonstration only. A wider allowable SOC range can be considered for specic study.

Considering practical operational aspects during low wind conditions, management of energy reserves of the hybrid system is vital. In order to ensure the system operation, the load curtailment is adopted. The load management algorithm is shown in Fig. 12. It is described as follows: 1) Calculate the total energy difference between the wind energy and the load demand (30) , check SOC of the battery and the status of 2) If the hydrogen storage. If , no load curtailment is required. If and extra energy from wind is not sufcient to bring the SOC to 75%, the load curtailment is executed. In this condition, the EMPRS allows the SOC of the battery to be 75%. 3) If , check SOC of the battery. If and the fuel cell and the battery have enough reserves to supply the energy decit, no load curtailment is needed. For other conditions, load curtailment is implemented. In this condition, the EMPRS allows the SOC of the battery to be 75%. To implement the load curtailment, loads are divided according to their priority. The loads such as hospitals, police stations, etc. can be considered as high priority or emergency loads. The hybrid system has to full the power demand of these loads at any condition. On the other hand, some lighting loads, washing machine loads, etc. can be considered as low priority loads and can be switched off when required. C. Operation Point of Each Subsystem The EMPRS generates the operation point based on the current wind and load conditions and actual limitations of each subsystem. Limitations include the maximum and minimum power of the fuel cell for operation in the ohmic region, the maximum charging or discharging power , and the maximum and minimum state of charge of the battery storage system. In this paper, the allowable range of SOC is assumed to be from 40% to 95%. For normal operating conditions, the allowable SOC of the

408

IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 4, NO. 2, APRIL 2013

TABLE I DYNAMIC OPERATING POINT

Mode 4: Low wind condition , , and . In this condition, the wind turbine extracts the maximum available power. The decit power is provided by the battery and fuel cell during transient conditions. Once the system reaches a steady-state condition, decit power is supplied by the fuel cell only. Mode 5: Low wind condition , , and . In this condition, the wind turbine extracts the maximum power. The fuel cell provides its maximum power and the battery storage provides the remainder. Mode 6: Low wind condition , , and . In this condition, the wind turbine extracts the maximum power and the battery storage provides the necessary power as the power decit is lower than the minimum power limit of the fuel cell. Mode 7: Low wind condition , low {\rm SOC} , and . In this condition, the wind turbine extracts the maximum power. The battery cannot provide power as the SOC is close to 75%, which is the minimum limit for normal operation conditions. The fuel cell operates in its ohmic zone. Since power produced by the fuel cell is higher than , the extra power is used to charge the battery. When the SOC of the battery storage system reaches about , the operation of the system moves to Mode 6. Mode 8: No wind condition. In this condition, the wind turbine is unable to produce any power. The fuel cell and the battery provide the required power. Depending on the severity of this condition, the EMPRS allows the SOC of the battery storage to drop as low as 40%. IV. SIMULATION, RESULTS, AND DISCUSSION Simulation studies are conducted to evaluate the performance of the proposed system under varying wind and load conditions. A. Performance of the Local Controllers Under Different Wind and Loading Conditions In this section, the performance of the local controllers is evaluated under varying wind and load conditions. The parameters of the wind turbine, IPM synchronous generator, and energy storage system are shown in Table II. 1) Dynamic Operating Points of Subsystems Under Different Wind and Loading Conditions: Figs. 13(a) and (b) show hypothetical wind and scheduled load proles, respectively. The EMPRS determines the operating mode according to current wind and load conditions, and available energy reserves as discussed in Section III. Fig. 14(a) shows the power generation from the wind energy conversion system. Fig. 14(b) shows the electrolyzer, fuel cell, and battery power. Figs. 14(c) and (d) show the status of the hydrogen storage and the SOC of the battery, respectively. Fig. 14(e) shows the operation mode of the hybrid power system. From Figs. 13(b) and 14(a), the initial load demand is 0.5 kVA (0.48 kW, 0.05 kVAR), and the maximum available power after considering conversion losses is about 0.9 kW. In this condi-

battery storage system is assumed at 75% to 95%. However, for emergency operation, the SOC is allowed to be as low as 40%. The limitations also include the maximum and minimum power consumption of the electrolyzer , and the maximum pressure of the hydrogen cylinder . The summery dynamic operating points are outlined in Table I. Details of the dynamic operating points are described as follows: Mode 1: High wind condition , and . In this mode, the wind turbine extracts optimum power that is equal to the load demand. The electrolyzer and the battery cannot consume any power because the hydrogen tank pressure and the SOC of the battery have reached their maximum limits. Mode 2: High wind condition and . In this mode, the wind turbine extracts optimum power that is equal to the load demand plus the capacity of the battery and electrolyzer. Mode 3: High wind condition and . In this condition, the wind turbine extracts the maximum power. The excessive power from the wind is consumed by the electrolyzer and battery.

HARUNI et al.: NOVEL OPERATION AND CONTROL STRATEGY

409

TABLE II SIMULATION PARAMETERS

Fig. 14. Power balance and operation mode sequence of hybrid system. (a) Power from wind energy conversion system. (b) Dynamic interaction of energy storage system. (c) Hydrogen pressure (pu). (d) SOC (%). (e) Modes of operation.

Fig. 13. Wind and load proles. (a) Wind speed; (b) real power and reactive power.

tion, the EMPRS operates in Mode 3 as the excess power from the wind can be stored in the battery and can be consumed by the electrolyzer to produce hydrogen. At the time of 20 s, the SOC of the battery reaches its maximum limit (about 95%). As a result, the battery cannot consume any more power. In this

condition, the EMPRS goes to Mode 2, which allows the wind energy conversion system to extract optimum power (0.85 kW) by controlling the rotor speed. In this condition, the excess wind power is consumed by the electrolyzer. From Fig. 13(b), at the time of 40 s, the load increases from 0.5 to 0.65 kVA (0.6 kW). At this time, the pressure of the hydrogen storage system reaches its maximum value. As a result, the electrolyzer cannot consume any extra power from the wind energy conversion system. In this condition, the EMPRS goes to Mode 1, which allows the wind energy conversion system to extract optimum power (0.65 kW) by controlling the rotor speed. From Fig. 13(b), at the time of 50 s, the load increases from 0.65 to 1.5 kVA (1.4 kW). In this situation, as the maximum available energy from the wind turbine is less than the load demand and the power decit is within the fuel cell power generation limit, the EMPRS operates the system in Mode 4. However, at the time of 60 s, the wind speed drops from 12 to 8 m/s. As the decit power is about 1.2 kW, which is within the combined limit of the fuel cell and the battery, the EMPRS operates the system in Mode 5.

410

IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 4, NO. 2, APRIL 2013

At the time of 70 s, the load decreases from 1.5 kVA (1.4 kW) to 1.0 kVA (0.97 kW). In this condition, the EMPRS operates the system in Mode 4 as the decit power is within the limit of the fuel cell. At the time of 80 s, the wind speed increases from 8 to 11.7 m/s. In this condition, as the decit power (0.170 KW) is less than the minimum power limit of the fuel cell, the EMPRS operates in Mode 6. At the time of 100 s, a 0.1-kW single-phase load is connected to the system. At this time, as the SOC of the battery approaches its minimum limit, the EMPRS runs the system in Mode 7. In this condition, the fuel cell can produce sufcient power, which is used to balance the decit power and to increase the SOC of the battery. At the time of 110 s, the wind speed drops and the wind turbine cannot produce any power. The load power increased from 1.1 to 1.3 kVAr (1.25 kW). In this condition, the fuel cell and battery storage system supply the power. As this condition is considered as an emergency condition, the EMPRS allows the SOC of battery storage system as low as 40%. 2) Performance of the Wind Energy Control System: Fig. 15(a) shows power extracted from the wind turbine and power losses. The efciency of the wind energy conversion system is shown in Fig. 15(b). From Fig. 15(c), it can be seen that the wind energy control system extracts the optimum power by regulating the rotor speed of the IPM synchronous generator. From Figs. 15(d) and (e), we can see that the wind energy control system maintains an efcient operation of the IPM synchronous generator by controlling the - and -axes components of the stator current. 3) Performance of the Electrolyzer Controller: The buck converter is used to control the electrolyzer power ow. The electrolyzer voltage and current are shown in Figs. 16(a) and (b), respectively. 4) Performance of the Fuel Cell Controller: The boost converter is used to control the dc-link voltage and power ow of the fuel cell. The fuel cell voltage and current are shown in Figs. 17(a) and (b), respectively. 5) Performance of the Bidirectional Battery Controller: As shown in Fig. 14(b), the bidirectional dcdc converter controls the battery charging/discharging power in order to ensure the power balance and transient stability of the system. 6) Dynamic Interaction of the Energy Storage System: The dynamic interaction of each energy storage device is shown in Fig. 14(b). It can be seen that under any change of wind speed or load demand, the battery storage system provides or consumes the transient power owing the faster dynamics of its fuel cell counterpart. 7) Performance of the Inverter Controller: Fig. 18 shows the output voltage and frequency is regulated by the load side inverter controller under varying wind and load conditions. Fig. 19 shows the dynamic performance of the inverter at 70 s, when the load decreases from 1.5 to 1.0 kVA. Fig. 20 demonstrates the dynamic performance of the inverter at 100 s, when a load of 0.1 kW is connected to phase B. From Fig. 20, it is revealed that the current in phase B increases while the phase voltages remain constant. Fig. 21 demonstrates the active and reactive power demand response of the inverter.

Fig. 15. Performance of the wind energy conversion system controller. and power conversion loss . (a) Power from wind turbine . (d) -axis current . (e) -axis (b) Efciency (%). (c) Rotor speed . current

Fig. 16. Electrolyzer voltage and current. (a) Electrolyzer voltage. (b) Electrolyzer current.

B. Load Management of the System Under Low Wind Conditions The EMPRS performance is evaluated under realistic wind and load scenarios. Let us assume the proposed hybrid shown in Fig. 1 operates in an isolated area. The average load demand is assumed to be about 0.6 kW and peak load demand is 1 kW. A case study is performed under low wind conditions when the

HARUNI et al.: NOVEL OPERATION AND CONTROL STRATEGY

411

Fig. 17. Fuel cell voltage and current. (a) Fuel cell voltage. (b) Fuel cell current.

Fig. 20. Inverter response at 100 s when a load is connected to a single phase: (a) voltage and (b) current responses.

Fig. 21. Real power and reactive power responses.

Fig. 18. System voltage and frequency. (a) System voltage. (b) System frequency.

Fig. 22. Wind speed prole and generated wind power. (a) Wind speed. (b) Power from wind turbine.

Fig. 19. Inverter response during load change at 70 s when the load decreased from 1.5 to 1.0 kVA: (a) voltage and (b) current responses.

load peak occurs. Let us assume the energy reserve is low and cannot support the system without load curtailment. The wind speed prole is shown in Fig. 22(a). A hypothetical wind prediction is assumed with an error of 5%. The corresponding power extracted from the wind turbine is shown in Fig. 22(b). The load prole is shown in Fig. 23(a). All loads are divided into four categories. Type is the load with the highest priority that constitutes about 25%30% of the total load. Type is a high priority load that constitute about 25%30% of the total load. Type is the load with a medium priority that constitutes about 25%30% of the total load. Type

has the lowest priority. The load curtailment operation is shown in Fig. 23(b). Figs. 24(a) and (b) show the operation of the electrolyzer/fuel cell and associated hydrogen storage, respectively. Figs. 25(a) and (b) show the operation of the battery storage and associated SOC, respectively. From Figs. 2225, it can be seen that the EMPRS can operate the system without any load curtailment up to 2:00 hours. However, from 2:00 to 3:24 hours, the EMPRS curtails the load with the lowest priority as the stored energy is not sufcient to provide the required decit. At 3:24 hours, the hydrogen storage runs out. Because the energy stored in the battery is not sufcient to run the system in the normal operating condition, the EMPRS goes to the emergency mode (Mode 8). As a result, from 3:24 to 5:00 hours, the EMPRS curtails loads in order to prevent the system blackout. During this period, the EMPRS uses the emergency reserves of the battery to supply

412

IEEE TRANSACTIONS ON SUSTAINABLE ENERGY, VOL. 4, NO. 2, APRIL 2013

V. CONCLUSION A novel operation and control strategy for a hybrid power system with energy storage for a standalone operation is proposed. The performance of the proposed control strategy is evaluated under different wind and load conditions. From the simulation studies, it is revealed that the machine side converter is able to extract the optimum power. It is also able to operate the IPM synchronous generator with maximum efciency. The battery storage system is successfully controlled by a bidirectional converter. The fuel cell and electrolyzer are controlled using boost and buck converters, respectively. The overall coordination of the wind turbine, fuel cell, battery storage system, electrolyzer, and loads is done by the developed EMPRS. The obvious advantage of the EMPRS is that it can prevent the system from blackouts in the event of low wind conditions or inadequate energy reserves.
Fig. 23. Load condition. (a) Load prole. (b) Load management.

REFERENCES
[1] B. Singh and J. Solanki, Load compensation for diesel generator-based isolated generation system employing DSTATCOM, IEEE Trans. Ind. Appl., vol. 47, no. 1, pp. 238244, Jan./Feb. 2011. [2] R. J. Best, D. J. Morrow, D. J. McGowan, and P. A. Crossley, Synchronous islanded operation of a diesel generator, IEEE Trans. Power Syst., vol. 22, no. 4, pp. 21702176, Nov. 2007. [3] M. El Mnassri and M. A. S. Leger, Stand-alone photovoltaic solar power generation system: A case study for a remote location in Tunisia, in Proc. 2010 IEEE Power and Energy Society General Meeting, Minneapolis, Jul. 2529, 2010, pp. 14. [4] P. Hearps and D. McConnell, Renewable Energy Technology Cost Review, Melbourne Energy Institute, Mar. 2011. [5] M.-S. Lu, C.-L. Chang, W.-J. Lee, and L. Wang, Combining the wind power generation system with energy storage equipment, IEEE Trans. Ind. Appl., vol. 45, no. 6, pp. 21092115, Nov./Dec. 2009. [6] P. K. Goel, B. Singh, S. S. Murthy, and N. Kishore, Isolated windhydro hybrid system using cage generators and battery storage, IEEE Trans. Ind. Electron., vol. 58, no. 4, pp. 11411153, Apr. 2011. [7] T. Zhou and B. Franois, Energy management and power control of a hybrid active wind generator for distributed power generation and grid integration, IEEE Trans. Ind. Electron., vol. 58, no. 1, pp. 95104, Jan. 2011. [8] N. Gyawali and Y. Ohsawa, Integrating fuel cell/electrolyzer/ultracapacitor system into a stand-alone microhydro plant, IEEE Trans. Energy Convers., vol. 25, no. 4, pp. 10921104, Dec. 2010. [9] C. Wang and M. H. Nehrir, Power management of a stand-alone wind/ photovoltaic/fuel cell energy system, IEEE Trans. Energy Convers., vol. 23, no. 3, pp. 957967, Sep. 2008. [10] A. Hajizadeh, M. A. Golkar, and A. Feliachi, Voltage control and active power management of hybrid fuel-cell/energy-storage power conversion system under unbalanced voltage sag conditions, IEEE Trans. Energy Convers., vol. 25, no. 4, pp. 11951208, Dec. 2010. [11] S. D. G. Jayasinghe, D. M. Vilathgamuwa, and U. K. Madawala, Direct integration of battery energy storage systems in distributed power generation, IEEE Trans. Energy Convers., vol. 26, no. 2, pp. 977685, Jun. 2011. [12] C. Wang, M. Nehrir, and S. R. Shaw, Dynamic models and model validation for PEM fuel cells using electrical circuits, IEEE Trans. Energy Convers., vol. 20, no. 2, pp. 442451, Jun. 2005. [13] P. Garcia, L. M. Fernandez, C. A. Garcia, and F. Jurado, Energy management system of fuel-cell-battery hybrid tramway, IEEE Trans. Ind. Electron., vol. 57, no. 12, pp. 40134023, Dec. 2010. [14] T. Zhou, B. Francois, M. el H. Lebbal, and S. Lecoeuche, Real-time emulation of a hydrogen-production process for assessment of an active wind-energy conversion system, IEEE Trans. Ind. Electron., vol. 56, no. 3, pp. 737746, Apr. 2009. [15] T. K. A. Brekken, A. Yokochi, A. von Jouanne, Z. Z. Yen, H. M. Hapke, and D. A. Halamay, Optimal energy storage sizing and control for wind power applications, IEEE Trans. Sustain. Energy, vol. 2, no. 1, pp. 6977, Apr. 2011.

Fig. 24. Fuel cell, electrolyzer power, and hydrogen status. (a) Power from electrolyzer/fuel cell. (b) Hydrogen storage.

Fig. 25. Battery power and SOC. (a) Battery power. (b) SOC.

power only to the loads with the highest priority. The wind returns at 3:48 hour. However, the EMPRS still considers the situation as an emergency because the SOC of the battery is too low for the normal operation. During this period, if the wind power exceeds the emergency load, the excess power is used to charge the battery.

HARUNI et al.: NOVEL OPERATION AND CONTROL STRATEGY

413

[16] T. K. Saha and D. Kastha, Design optimization and dynamic performance analysis of a stand-alone hybrid wind-diesel electrical power generation system, IEEE Trans. Energy Convers., vol. 25, no. 4, pp. 12091217, Dec. 2010. [17] L. Wang and C. Singh, Multicriteria design of hybrid power generation systems based on a modied particle swarm optimization algorithm, IEEE Trans. Energy Convers., vol. 24, no. 1, pp. 163172, Mar. 2009. [18] H. Dagdougui, R. Minciardi, A. Ouammi, M. Robba, and R. Sacile, A dynamic decision model for the real-time control of hybrid renewable energy production systems, IEEE Syst. J., vol. 4, no. 3, pp. 323333, Sep. 2010. [19] S.-K. Kim, J.-H. Jeon, C.-H. Cho, J.-B. Ahn, and S.-H. Kwon, Dynamic modeling and control of a grid-connected hybrid generation system with versatile power transfer, IEEE Trans. Ind. Electron., vol. 55, no. 4, pp. 16771688, Apr. 2008. [20] M. L. Rahman, S. Oka, and Y. Shirai, Hybrid power generation system using offshore-wind turbine and tidal turbine for power uctuation compensation (HOT-PC), IEEE Trans. Sustain. Energy, vol. 1, no. 2, pp. 9298, Jul. 2011. [21] E. Muljadi and H. Ed. McKenna, Power quality issues in a hybrid power system, IEEE Trans. Ind. Appl., vol. 38, no. 3, pp. 803809, May/Jun. 2002. [22] Y. Zhang, A. A. Chowdhury, and D. O. Koval, Probabilistic wind energy modeling in electric generation system reliability assessment, IEEE Trans. Ind. Appl., vol. 47, no. 3, pp. 15071514, May/Jun. 2011. [23] K. E. Johnson, L. Y. Pao, M. J. Balas, and L. J. Fingersh, Control of variable-speed wind turbines: Standard and adaptive techniques for maximizing energy capture, IEEE Control Syst. Mag., vol. 26, no. 3, pp. 7081, Jun. 2006. [24] P. C. Klause, O. Wesynczuk, and S. D. Sudhoff, Analysis of Electric Machinary and Drive System, 2nd ed. Hoboken, NJ: Wiley, Feb. 2002, 0-471-14326-X. [25] W. Qiao, L. Qu;, and R. G. Harley, Control of IMP synchronous generator for maximum wind power generation, IEEE Trans. Ind. Appl., vol. 45, no. 3, pp. 10951105, May/Jun. 2009. [26] A. Khaligh, A. M. Rahimi, Y. J. Lee, J. Cao, A. Emadi, S. D. Andrews, C. Robinson, and C. Finnerty, Digital control of an isolated active hybrid fuel cell/Li-ion battery power supply, IEEE Trans. Veh. Technol., vol. 56, no. 6, pt. 2, pp. 37093721, Nov. 2007. [27] Sim Power Systems, Reference, Hydro-Qubec, MathWorks, Inc.. Natick, MA, 2011. [28] A. Kusiak and Z. Zhang, Adaptive control of a wind turbine with data mining and swarm intelligence, IEEE Trans. Sustain. Energy, vol. 2, no. 1, pp. 2836, Jan. 2011. [29] A. Kusiak and Z. Zhang, Short-horizon prediction of wind power: A data-driven approach, IEEE Trans. Energy Convers., vol. 25, no. 4, pp. 11121122, Dec. 2010. [30] C. W. Potter and M. Negnevitsky, Very short-term wind forecasting for tasmanian power generation, IEEE Trans. Power Syst., vol. 21, no. 2, pp. 965972, May 2006. [31] N. Amjady, F. Keynia, and H. Zareipour, Short-term load forecast of microgrids by a new bilevel prediction strategy, IEEE Trans. Smart Grid, vol. 1, no. 3, pp. 286294, Dec. 2010. [32] A. M. O. Haruni and M. Negnevitsky, An articial intelligence approach to develop a time-series prediction model of the arc furnace resistance, J. Adv. Computat. Intell. Intelligent Informatics, vol. 14, no. 6, pp. 722729, Sep. 2010. [33] P. Artuso, R. Gammon, F. Orecchini, and S. J. Watson, Alkaline electrolysers: Model and real data analysis, Int. J. Hydrogen Energy, vol. 36, pp. 79567962, 2011.

A. M. Osman Haruni (S10) received the B.Sc. degree in electrical and electronic engineering from Bangladesh University of Engineering and Technology in 2003, and the M.Eng.Sc. degree from University of Tasmania in 2008. He is currently working toward the Ph.D. degree at the University of Tasmania, Australia. He worked as a engineer with Siemens Bangladesh Limited in the Power Transmission and Distribution (PTD-MV) Department from 2004 to 2006. His research interests include power electronics control for renewable energy technologies, load modeling and its application in power system operation and control, and use of articial intelligence in power systems.

Michael Negnevitsky (M95SM07) received the B.S.E.E. (Hons.) and Ph.D. degrees from the Byelorussian University of Technology, Minsk, Belarus, in 1978 and 1983, respectively. Currently, he is Chair Professor in Power Engineering and Computational Intelligence and Director of the Centre for Renewable Energy and Power Systems at the University of Tasmania, Hobart, Australia. From 1984 to 1991, he was a Senior Research Fellow and Senior Lecturer in the Department of Electrical Engineering, Byelorussian University of Technology. After arriving in Australia, he was with Monash University, Melbourne, Australia. His interests are power system analysis, power quality, and intelligent systems applications in power systems. Dr. Negnevitsky is a Chartered Professional Engineer, Fellow of the Institution of Engineers Australia, Member of CIGRE AP C4 (System Technical Performance), Member of CIGRE AP C6 (Distribution Systems and Dispersed Generation), Australian Technical Committee, and Member of CIGRE Working Group JWG C1/C2/C6.18 (Coping with Limits for Very High Penetrations of Renewable Energy), International Technical Committee.

Md. Enamul Haque (M97SM10) graduated in electrical and electronic engineering from Rajshahi University of Engineering Technology [formerly, Bangladesh Institute of Technology (BIT)], Rajshahi, Bangladesh, in 1995. He received the M.Engg. degree in electrical engineering from University Technology Malaysia in 1998, and the Ph.D. degree in electrical engineering from University of New South Wales, Sydney, Australia, in 2002. He has worked as an Assistant Professor for King Saud University, Saudi Arabia, and United Arab Emirates University for four years. He is currently working as a Lecturer in renewable energy and power system with the School of Engineering, University of Tasmania, Australia. His research interests include smart energy systems, control and grid integration of renewable energy sources and energy storage system, micro grid system with hybrid wind/solar/fuel cell systems, power electronics applications in smart-grid, micro-grid and power system applications.

Ameen Gargoom (M08) received the B.Sc., M.Sc. (with honours), and Ph.D. degrees in 1994, 2001, 2007, respectively, all in electrical power engineering. He worked as a consultant engineer with Al-Emara Co. for Engineering Consultants, Libya, for six years before joining the University of Garyounis in 2001 as an Associate Lecturer. In 2008, he joined Tasmania University as a Research Fellow. Currently he is working as a Lecturer with the School of Engineering, University of Tasmania, Australia. His research interests include power electronics control for renewable energy technologies and smart grid systems, new techniques for power quality monitoring and classication, and the application of signals processing techniques in power systems.