Portal de Periódicos da CAPES

Improved power management control strategy for renewable energy‐based DC micro‐grid with energy storage integration

2018; Institution of Engineering and Technology; Volume: 13; Issue: 6 Linguagem: Inglês

10.1049/iet-gtd.2018.5019

1751-8695

Manoj Kumar Senapati, Chittaranjan Pradhan, Subhransu Ranjan Samantaray, Paresh Kumar Nayak,

Frequency Control in Power Systems

IET Generation, Transmission & DistributionVolume 13, Issue 6 p. 838-849 ArticleFree Access Improved power management control strategy for renewable energy-based DC micro-grid with energy storage integration Manoj Kumar Senapati, Manoj Kumar Senapati Department of Electrical Engineering, Indian Institute of Technology Dhanbad, Dhanbad, IndiaSearch for more papers by this authorChittaranjan Pradhan, Corresponding Author Chittaranjan Pradhan cp11@iitbbs.ac.in School of Electrical Sciences, Indian Institute of Technology Bhubaneswar, Bhubaneswar, IndiaSearch for more papers by this authorSubhransu Ranjan Samantaray, Subhransu Ranjan Samantaray School of Electrical Sciences, Indian Institute of Technology Bhubaneswar, Bhubaneswar, IndiaSearch for more papers by this authorParesh K. Nayak, Paresh K. Nayak Department of Electrical Engineering, Indian Institute of Technology Dhanbad, Dhanbad, IndiaSearch for more papers by this author Manoj Kumar Senapati, Manoj Kumar Senapati Department of Electrical Engineering, Indian Institute of Technology Dhanbad, Dhanbad, IndiaSearch for more papers by this authorChittaranjan Pradhan, Corresponding Author Chittaranjan Pradhan cp11@iitbbs.ac.in School of Electrical Sciences, Indian Institute of Technology Bhubaneswar, Bhubaneswar, IndiaSearch for more papers by this authorSubhransu Ranjan Samantaray, Subhransu Ranjan Samantaray School of Electrical Sciences, Indian Institute of Technology Bhubaneswar, Bhubaneswar, IndiaSearch for more papers by this authorParesh K. Nayak, Paresh K. Nayak Department of Electrical Engineering, Indian Institute of Technology Dhanbad, Dhanbad, IndiaSearch for more papers by this author First published: 10 September 2018 https://doi.org/10.1049/iet-gtd.2018.5019Citations: 18AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract This study presents an improved power management control strategy of a hybrid direct current (DC) micro-grid (MG) system consisting of photovoltaic cell, wind turbine generator, battery energy storage (BES), fuel cell (FC), and electrolyser. Based on the voltage and state of charge of BES, FC, and electrolyser, the proposed control scheme improved the dynamics of the DC-link voltage and contributes a better power management between each generation/source and load. A gain control technique is implemented in the grid-side inverter controller to regulate the modulation index and improving the voltage stability of the DC-link. Furthermore, the PI-controller gains of BES are tuned dynamically based on the deviation in voltage and its derivative using Takagi–Sugeno-fuzzy control to enhance the transient response of the voltage. For a reliable operation of the DC MG under standalone or prolonged islanding mode of operation, a priority-based load shedding algorithm is proposed for maintaining proper power coordination between different energy sources and storage devices. Owing to smoother and faster voltage response, the proposed control schemes can comply with the grid code requirements of the changing configuration of the modern renewable energy integrated DC MG. The effectiveness of the proposed control strategy is tested by comparing the existing scheme through MATLAB/Simulink®. Nomenclature G ideality factor of photovoltaic cell q electron charge of photovoltaic cell k Boltzmann constant ns number of solar panels connected in series np number of solar panels connected in parallel Ipv output current of the solar panels Irs reverse saturation current in photovoltaic cell Vpv output voltage of the solar panels Tp surface temperature of the solar panels R radius of wind turbine blade Cp power coefficient of wind turbine Idc_opt optimum DC-link current of wind system Vdc_opt optimum DC-link voltage of wind system Vw wind speed β wind turbine pitch angle ρ air density ω turbine angular velocity λ tip speed ratio λopt optimum tip speed ratio of wind power ωopt optimum rotor speed of wind turbine nf number of fuel cells connected in series Tfc absolute temperature of fuel cell Eif ideal standard potential of fuel cell ne number of cells of electrolyser Te temperature of electrolyser Vrev, e reversible cell potential of electrolyser Ae electrode area of electrolyser A exponential voltage of battery B exponential capacity of battery H polarisation voltage of battery Q battery capacity ibat battery current Rbat internal resistance of battery PAC AC-load power, kW Pb battery power, kW Pe electrolyser power, kW Pfc fuel cell power, kW Pg combined power generated by PV and wind PG utility grid power, kW PL DC- load power, kW Ppv PV power, kW Pw wind power, kW Vdc DC-link voltage, V Vdc ∗ rated value of the DC-link voltage, V VL line voltage on the AC-side m PWM modulation index of the grid-side inverter 1 Introduction Nowadays, renewable energy sources like solar, wind, tidal, biomass, or small-scale hydro-based distributed generations (DGs) are gaining popularity as clean sources of energy [1]. DGs are limited to a few kilowatts to megawatts and are interconnected at the distribution substation, distribution feeder, or to the customer load. As the penetration of DGs increases, to manage the power, coordination between generation and demand becomes more challenging in the distribution system due to their highly intermittent nature. Therefore, the concept of micro-grid (MG) has been developed as an effective way to support a reliable and better power system operation of DGs and storage devices via distribution feeders to smaller loads. MG platforms provide several advantages such as uninterrupted service to customers, improved reliability, power quality, and operational optimality over conventional distribution systems for both grid-connected and islanded mode of operation [2, 3]. An extensive research works have been carried out on various control aspects of AC MG systems such as grid integration [4], grid-connected and islanded mode of operation [5], power-sharing among multiple energy sources [6, 7], protection [8, 9] etc. On the other hand, in the recent years, direct current (DC) MG system has attracted increasing popularity due to the requirement of single-stage converter cost when compared with AC MG and easy controls the phase and reactive power control [2]. Apart from the above advantages, DC MGs can provide a better power quality performance during voltage sags or blackout scenarios in the utility grids [2]. A typical DC MG consists of mainly four power point terminals, namely, generation, load, energy storage devices, and grid interface [2, 3]. For a stable and economic power system operation with maintaining the DC-bus voltage constant, the optimum power flow among these terminals should be necessary during power system contingencies. Few works have been reported on hybrid energy sources-based DC MG for voltage stability and controlling the uninterruptible power management in residential complexes, telecommunication buildings, Internet data centres etc. [10–12]. In [13], a power flow management and voltage control scheme is proposed in an isolated DC MG consisting of photovoltaic (PV) and battery storage device. However, in this scheme, the grid-connected and islanded mode of operation with consideration of other renewable sources and energy storage systems are not considered. The operation and power management strategy of a DC MG consisting of only PV and battery storage system is proposed in [14]. In this scheme, the maximum renewable energy utilisation is achieved during different operating modes of the MG by considering the DC voltage control, battery power management, state of charge (SoC) of the battery, and DC load. However, other renewable sources and energy storage systems are not included in this study. A reliable power control scheme in DC distribution system for balance and line fault conditions are highlighted [1]. However, the scheme is limited to a DC distribution system consisting of a wind power generation system, battery, and DC loads. In [15], a voltage variation-based control strategy and frequency band-based power-sharing scheme are proposed for variable wind generation and multiple slack terminals of DC MG system. In [3], the authors have proposed a three-level autonomous control based on a DC voltage variation scheme for the same DC MG system as presented in [15]. The performance of the control scheme is evaluated for different operating conditions such as load shedding, generation curtailment, and generation fluctuation. Another voltage control scheme based on fuzzy control with gain scheduling technique is proposed in [16] for accomplishing both power-sharing and energy management task in a DC distribution system. However, the control strategies for the fuel cell (FC) and dump load (i.e. electrolyser) are not considered in [3, 15, 16]. In [17], the voltage droop control strategy is proposed to achieve a stable load sharing among the load and sources in the DC MG using a quadratic and linearly constrained optimisation problem. However, the authors have taken a linear model system and not considered the voltage performance during the fault scenario of the MG in the grid integration mode of operation. In [18], the authors have proposed an optimisation technique for the renewable DG using ant-lion optimisation technique. Mainly, the authors have focused on the optimal location and size of DG in a radial distribution system. In this study, the voltage performance of the PV and wind-based MG using this optimisation technique is evaluated under variable load condition. However, the authors have not highlighted the voltage performance of the MG by considering the energy storage elements (i.e. FC, battery), dump load, and grid-connected system. In [19], the authors have proposed an energy management scheme of DC MG, which consists of PV, diesel generator, and supercapacitor-based energy storage. However, wind and energy storage systems are not considered in this study. Using a sliding mode scheme for DC bus voltage control for a standalone system consisting of PV and battery is presented in [20]. However, in this study, wind power, FC with grid-connected scenarios are not taking into consideration. The authors are presented in the literature for the hybrid AC–DC MG focusing on each of the broad aspects of control, namely modelling, power management, coordinated control, voltage stability analysis, power quality, and protection strategies [21]. In [22], the authors have focused on fast transient response, low fluctuations of DC-link voltage, good rejection of disturbance due to the sudden change in active power for the grid-connected mode. However, the energy storages, dump load, islanding mode, as well as fault conditions are not considered in this study. The above study clearly shows that the dynamic power management performances of the DC MGs consisting of multiple renewable sources and energy storage devices have not evaluated for grid-connected, islanded mode of operation, and line fault scenarios. For an effective power management and to cope with the increasing power demand profile, it is necessary to propose an improved power coordination strategy in renewable energy-based multi-energy storage DC MG system. In this paper, wind and solar energy-based DC MG system with three energy storages, namely battery, FC, and electrolyser are considered as shown in Fig. 1. Out of the three energy storages, the battery can act either as a source or as a sink. As a result, it can charge (discharge) within its specified limits with respect to the surplus (deficit) of power generation. The FC is only acting as a source when the availability of hydrogen is sufficient and the electrolyser acts as a sink for the system. In [23], the authors have highlighted that the voltage-based control scheme as presented in [2] has the limitation in achieving proper power coordination among battery, FC, and electrolyser. Fig. 1Open in figure viewerPowerPoint Schematic diagram of the studied DC MG system In this paper, both voltage and SoC sensitive-based control scheme of battery, FC, and electrolyser are proposed to coordinate the power management in an effective manner. It results in improving the dynamic performance of the DC-link voltage during system contingencies. Since battery storage device acts as an important energy storage device to enhance the DC-link voltage response of the DC MG when compared with FC and electrolyser to provide the surplus power for balance between generation and demand. Hence, in this paper, the gains of the proportional plus integral (PI) controller are adaptively tuning by Takagi–Sugeno (TS)-fuzzy logic control with respect to system events to improving the DC-link voltage dynamic performance. It is well known that the TS-fuzzy-based control is more flexible due to its linguistic rule-based consequent which can produce an infinite number of gain variation characteristics and contributes a better solution for non-linear control problems [24, 25]. Further, a proportional gain control strategy of the grid-side inverter is implemented to keep the modulation index within a reasonably practical limit and contribute a proper real power management with maintaining the DC bus voltage constant. The performance of the proposed control strategy is investigated for both grid-connected and islanded mode of operation as well as a line fault condition and found to be superior to the existing methods. In [26], a load shedding algorithm is implemented on the basis of SoC of the battery. However, the authors have not considered the voltage level of DC MG while deciding the percentage of load shedding. Because the DC-link voltage is an important system element in DC micro-grid and its performance depend on the every system events such as the fluctuations in load, the variety of power generation/storage capacity and the change in the system parameters, etc. Hence, in this paper, both SoC and DC bus voltage level are taken into account while deciding the load shedding strategy for maintaining the DC-link voltage within its specified limits, for a reliable and stable power system operation. The main contributions of this paper are summarised as below: To achieve an effective real power management scheme among the variable generations (wind and solar) and energy storage devices such as the battery, FC, and electrolyser, for mitigating the DC-link voltage fluctuations during system contingencies. TS-fuzzy-based controller is proposed to tune the PI gains of the battery system for achieving a smoother DC output voltage during different operating scenarios. Propose a proportional gain-based control scheme for regulating the modulation index of the DC MG inverter for improving the dynamics performance of the output voltage. To take an appropriate decision for load shedding, the SoC of battery and DC bus voltage-based algorithm is implemented. The rest of the paper is organised as follows. The control strategies employed for different components of the proposed hybrid DC MG are illustrated in Section 2. In Section 3, the outlines of different control methods of the proposed control schemes are employed during different operating conditions of the hybrid DC MG. The simulation results are presented in Section 4. Finally, the conclusion of the proposed work is provided in Section 5. 2 Control strategies employed for different components of the proposed hybrid DC MG The schematic of the hybrid DC MG system studied in the present work is shown in Fig. 1. The main components are: (i) variable renewable energy sources, i.e. PV and permanent magnet synchronous generator (PMSG)-based wind turbine system, (ii) FC and dump load, (iii) variable load scenario, (iv) battery energy storage (BES), (v) DC–DC converters, and (vi) voltage source converters (VSC). In the present study, the mathematical modelling and characteristics of each of the components used are simulated by using MATLAB/Simulink and are well documented in [14–16, 23–27]. The brief modelling of individual components of the DC MG and the proposed control strategies of the present study are discussed in this section. The detailed simulation data of the DC MG are provided in Tables 1–3. (Figs. 2 and 3) Table 1. Modelling/simulation data of the DC micro-grid DC micro-grid system/parameters Values photovoltaic cell [24] short-circuit current 8.21 A open circuit voltage 33 V maximum power voltage () 26.3 V maximum current () 7.61 A no. of cells per module 54 no. of cells connected in series (ns) 10 no. of cells in parallel (np) 3 Boltzmann constant (k) 1.38 × 10−23 J/K electron charge (q) 1.6 × 10−9 C rated power 6 kW PMSG [28] flux linkage 0.433 Vs number of poles 10 rated speed 152.89 rad/s armature resistance (Rs) 0.39 Ω stator inductance (Ls) 0.0082 H rated torque 40 N m rated power 6.5 kW inertia 0.01189 kg/m2 fuel cell [25] absolute temperature (Tfc) 1273 K faraday's constant (F) 96,487 C/kmol universal gas constant (R0) 8314 J/(kmol K) ideal standard potential (Eif) 1.18 V number of cells series (nf) 215 constant (Kr) 0.842 × 10−6 hydrogen molar constant (KH2) 8.43 × 10−4 molar constant of water (KH2O) 2.81 × 10−4 molar constant of oxygen (KO2) 2.5 × 10−3 response time of hydrogen (τH2) 26.1 s response time of water flow (τH2O) 78.3 s response time of oxygen flow (τO2) 2.91 sec ohmic loss/cell (r) 32,813 × 10−8 Ω rated power of fuel cell 8 kW electrolyser [25] r1 0.0015Ωm2 r2 −6.019 × 10−6m2C−1 s1 2.427 V s2 −0.0307 V C−1 s3 3.9 × 10−4 V C−2 t1 0.214 A−1m2 t2 −9.870 A−1m2 C t3 119.1 A−1m2C2 ne 64 Vrev 1.1647 V rated power of electrolyser 10 kW battery [25] number of batteries in series 15 rated voltage of each battery 12 V rated capacity of each battery 55 Ah DC micro-grid system/parameters Values utility grid and distribution line PCC voltage (line voltage) 400 V grid voltage (line voltage) 11k V base voltage 11 kV frequency 50 Hz distribution line inductance 1.05 × 10−3 H/km distribution line resistance 0.1153 Ω/km distribution line length 14 km X/R ratio of distribution line 2.86 X/R ratio of utility grid 0.7 transformer rated KVA 24 rated KV 11/0.4 kV (Y–Y) frequency 50 Hz resistance 0.001 p.u. inductance 0.03 p.u. magnetisation resistance 500 p.u. magnetisation reactance 500 p.u. DC voltage level of different subsystems solar voltage 263 V fuel cell 203 V electrolyser 86 V line voltage on AC-side 400 V low level DC-link voltage 450 V high-level DC-link voltage 660 V LCL filter (AC-side) filter inductance: inverter-side (Li) 11 × 10−3 H filter inductances: grid-side (Lg) 0.8 × 10−3 H filter capacitance (Cf) 12 × 10−6 F damping resistance (Rf) 1.14 Ω switching frequency of PWM (f sw) 4000 Hz Table 2. Value L and C used in DC–DC converters (Fig. 1) DC/DC converter Inductance in Henry, mH Capacitance in Farad, µF boost converter (PV) 5.5 (Lpv) 2500 (Cpv) buck converter (wind) 10 (Lw) 1500 (Cw) boost converter (fuel cell) 0.45 (Lf) 7500 (Cf) buck–boost converter (battery) 12 (Lb) 1000 (Cb) buck converter (electrolyser) 20 (Le) 500 (Ce) centre boost converter 0.9 (Ld) 6000 (Cd) Table 3. Value of the PI-controller gains of the micro-grid Subsystems Proportional gain (Kp) Integral gain (Ki) boost converter (Fig. 1): PI-1 0.008 0.125 fuel cell (Fig. 2a) PI-2 0.008 0.125 PI-3 0.13 0.5 PI-4 0.0065 0.009 electrolyser (Fig. 2b) PI-5 80 550 PI-6 6 81 inverter controller (Fig. 3) PI-7 7 90 Fig. 2Open in figure viewerPowerPoint Control block diagram for coordinated operation(a) FC, (b) Electrolyser Fig. 3Open in figure viewerPowerPoint Proposed control scheme during grid-connected islanding mode of operation In Fig. 1, and are the output voltage and current of the solar panels, respectively. is the PV power, and are the inductor and capacitor of the PV converter, respectively. and are the power electronics switch (i.e. MOSFET, insulated-gate bipolar transistor) and pulse width modulation (PWM) duty cycle of PV converter, respectively. and are the output voltage and current of the wind power, respectively. is the wind power, and are the inductor and capacitor of the wind power converter, respectively. and are the power electronics switch and duty cycle of wind power converter, respectively. and are the output voltage and current of the FC, respectively. is the FC power, and and are the inductor and capacitor of the FC converter, respectively. and are the power electronics switch and duty cycle of FC converter, respectively. is the output current of the battery, is the battery power, and are the inductor and capacitor of the bi-directional battery converter, respectively. and are the power electronics switch of bi-directional battery converter, and are the duty cycle of bi-directional battery converter, is the electrolyser power, and are the power electronics switch of electrolyser converter, and are the duty cycle of electrolyser converter, and are the inductor and capacitor of the main DC/DC MG converter, respectively. and are the power electronics switch and duty cycle of the main DC/DC MG converter, is the load power, and is the grid power, and are low- and high-level DC-link voltage of the DC MG. 2.1 Control strategy employed for the PV cell The output current of the equivalent circuit of the PV cell is expressed as follows [24]: (1)Similarly, the maximum output power () of the PV cell is represented as follows [15]: (2)where and are the output voltage and current of the solar panels, respectively. and are the number of solar panels connected in series and parallel, respectively. k is the Boltzmann constant, q is the electron charge, G is the ideality factor, is the surface temperature of the solar panels, and is the reverse saturation current, and and are the maximum voltage and current of the solar panel, respectively. The detailed modelling of PV cell is described in [23]. The PV system is interfaced to the DC MG through a DC–DC converter as shown in Fig. 1. The widely used perturb and observe (P&O)-based method is employed in this study, for maximum power point tracking (MPPT) [24]. In this study, variable nature of irradiance and temperature are considered to validate the proposed control scheme. The detailed simulation data of the PV cells are provided in Tables 1–3. 2.2 Control strategy employed for the wind turbine system In this study, the PMSG-based wind turbine model used is described in detailed in [25, 27–30]. The aerodynamic mechanical power (Pw) and the maximum output power (Pmax) derived from wind by a wind turbine is evaluated as follows: (3)with (4)where is the air density, the wind speed, R the blade radius, Cp () the power coefficient, which is a non-linear function of the tip speed ratio () and turbine pitch angle , the optimum tip speed ratio at MPPT, the turbine angular velocity, the optimum rotor speed at MPPT, and a constant. The PMSG-based wind turbine is interfaced to the DC MG through an AC–DC stage consisting of a three-phase diode rectifier and a buck converter. Owing to the high cost and low reliability of the controlled rectifier, a simple diode-based rectifier is used in this study, for AC to DC conversion [28]. In this paper, the P&O-based control method is employed for extracting maximum power from the wind turbine [28]. The control block diagram of the PMSG set-up is shown in Fig. 4. Here, is used as a control parameter for perturbing. The mathematical expression of is defined as follows [28]: (5)where a is a constant and is used to scale the values of (rated DC voltage) and (rated DC current) of the wind system. and are the optimum DC voltage and current of the wind power system. Fig. 4Open in figure viewerPowerPoint Control block diagram of converter for the PMSG-based wind turbine system [28] The optimum value of , i.e. , is set close to 45° and initially, can be half or one-third of the or less. The optimum-relation-predicated (ORB) control of wind power system for calculating the reference DC-link current (Idc,w ∗) is shown in Fig. 4 [28]. Owing to quick current controllability, fast dynamics, and less hardware requirement, a hysteresis controller is used in this study, for controlling the duty ratio of the buck converter as shown in Fig. 4 [28]. The detailed simulation data of the PMSG-based wind turbine is given in Tables 1–3. 2.3 Control strategy employed for the FC In this study, a solid oxide fuel cell (SOFC) model is used. The detailed modelling of SOFC is presented in [24, 31, 32]. The output voltage () of the SOFC is expressed as follows: (6)where , , and are the partial pressures of hydrogen, oxygen, and water response, respectively. Eif is the ideal standard potential, R0 is the universal constant (J/(kmol K)); r is the ohmic losses of the FC stack (Ω), F is the Faraday's constant (C/kmol), nf is the number of fuel cells connected in series, Tfc is the absolute temperature (K), and I0 is the stack current (A). The SOFC is interfaced to the DC MG through a boost converter as shown in Fig. 1. During normal operating condition, FC operates in standby mode. At the inception of any abnormal condition, it supplies power to the DC MG within 2 s to ensure power balance and constant DC-link voltage [24]. In [24], the authors have used a PI-controller for regulating the power of the FC through the current control scheme. However, the authors have not considered the voltage level of the FC for regulating the power. In this controller, both voltage and SoC of the FC and current comparison of generation, load have been considered to improve the dynamic power response of the FC, as shown in Fig. 2a. In Fig. 2a, Ig is the combined generated current of PV and wind, ILd is the DC load current, and Ifc is the FC current in the DC MG. Whenever there is an increase in load, the extra load power is provided by the FC excluding the power supplied by the generation and the battery. The extra demanded power of the load is fulfilled by the FC when the SoC of the battery is <20% and the voltage level is 80%), the excess power is taken by the electrolyser. When SoC becomes >80%, the controller will increase the duty cycle as a function of overvoltage. However, whenever the power demand is very less than the generation, the electrolyser will receive the excess generated power for maintaining the power balance in the DC MG. In practice, the production of hydrogen is depend on the current of the electrolyser but not in voltage [25]. In order to stop the electrolyser to receive the excess power when the hydrogen tank is full, a control criterion is implemented which is shown in Fig. 2b. As the control signal is zero (i.e. hydrogen tank of the electrolyser is filled), the PWM pulse (De2) of the buck converter (Se2) is zero, as a result, the electrolyser disconnects from the DC MG which is shown in Figs. 1 and 2b. 2.5 Control strategy employed for the BES system In this work, a generic Lithium-ion battery model is taken [24, 25]. The model is implemented by using a controlled voltage source in series with a constant resistance. The controlled voltage source () and the battery voltage () are represented as (8) and (9), respectively (8) (9)where is the initial voltage of the battery, H the polarisation voltage, Q the battery capacity, A the exponential voltage of battery, B the exponential capacity, the internal resistance of battery, the battery current, and the charge dr