SimPower-based analysis and design of a hybrid wind–diesel-superconducting magnetic energy storage system for simultaneous frequency and voltage control

Abstract

Power quality control in a stand-alone power system is a demanding task. For satisfactory operation, such systems are being augmented with fast-acting energy storage devices. In this article, a stand-alone wind–diesel system augmented with a small-rating superconducting magnetic energy storage system is considered for both reactive and real power balance. Suitable controllers are proposed which force the superconducting magnetic energy storage system to exchange both reactive and real power with the system under various perturbations. A simulation platform is developed in SimPower to virtually validate the system model and control design aspects. Superconducting magnetic energy storage system and its power electronic interface are represented by average value models and the various controller parameters are tuned using Genetic Algorithm. Simulation results show that both voltage and frequency of the system are improved.

Keywords

Genetic Algorithm voltage source inverter average model superconducting magnetic energy storage system proportional–integral–derivative controller asynchronous generator chopper duty cycle floating mode

Introduction

Isolated systems are gaining popularity in supplying energy to remote locations. Such sources when coupled with renewable resources such as wind energy provide a suitable alternative to present energy deficit situation. Wind energy (Hunter and Elliot, 1994; McGowan and Connors, 2000; Nacfaire, 1989) integration is beneficial in regions not connected to grid power system, such as remote locations and islands, and this is advantageous on account of fuel savings and so on. With fast progress in power electronic technology and the developments in multi-level inverters, this resource is finding more and more penetration into power generation. Earlier, several studies had been carried out on modelling such wind-powered systems (Jeffries et al., 1996), with no consideration to the use of energy storage in these systems.

The wind is a highly variable source as it may not be available all the time. To add to it, the wind speed fluctuates quite frequently and since these wind turbines (WTs) are generally coupled with induction generators, such variations translate into reactive and real-power disturbances in the network, thus worsening the quality of supply (Manwell et al., 2010). There are several methods suggested in the literature aiming to reduce these voltage and frequency oscillations. The use of flexible AC transmission system (FACTS); (Heetun et al., 2016) devices like SVCs (static VAR compensators) and STATCOM is one such method which can effectively reduce voltage oscillations in these wind-driven systems. This scheme, however, does not tackle the problem of frequency oscillations in such systems. For successful operation of wind-penetrated stand-alone power systems, energy storage devices are increasingly being used (Gao, 2015; Mubarak, 2017; Sumper et al., 2016). Several studies show their importance in such systems (Kouba et al., 2016; Lone and Mufti, 2004). Initially, these took the form of battery/inverter (Sebastián, 2016), but the batteries have a major limitation on account of its charging and discharging rate. Installation of flywheel (Faraji et al., 2017; Vidyanandan and Senroy, 2016) as energy storage gained popularity recently but their operation demands a lot of space and because of its high inertia starting the unit coupled with the flywheel requires special provisions.

Superconducting magnetic energy storage systems (SMES) can absorb/generate real and reactive powers with faster response to system needs and therefore finds a potential use in FACTS applications (Ise et al., 1987; Molina and Mercado, 2011). The SMES has a far superior dynamic performance in contrast to other storage technologies (Hayashi et al., 2006). Also, its life and operation are unaffected by the count of charging or discharging cycles unlike the typical batteries. Typically, the estimated life of an SMES unit surpasses 20 years and its cost, with technological improvements, is anticipated to reduce to approximately 25% of its current cost. Furthermore, they involve no toxic chemicals and give out no inflammable gases. Improvements in SMES on technical grounds and greater knowledge regarding handling of cryogenic systems will help in faster penetration of SMES units into the existing system. In the above context, SMES, in the near future, is expected to cater to the needs of a diverse electrical power system. SMES is, therefore, markedly suited for simultaneous frequency and voltage control of these wind-driven systems which are greatly affected by wind variations.

Recently, several studies have been undertaken to study the performance of SMES in such wind-driven systems (Zargar et al., 2017b). However, no proper control strategy is considered for SMES and the controller parameters have not been optimized for better performance. Moreover, converter dynamics have not been considered here. Several other studies (Zargar et al., 2017a) have used advanced control strategies. However, such studies have used active power modulation for frequency control (controlling only active power) with no regard for reactive power control which results in deterioration in system voltage. Furthermore, DC link voltage is assumed constant and thus no DC link dynamics is taken into consideration.

This article simulates the behaviour of SMES in hybrid wind system and discusses its use for frequency and system voltage control employing Genetic Algorithm (GA) technique to optimize the controllers.

GA – a brief description

The GA TOOLBOX in MATLAB, with integral square error (ISE) as a fitness function, is employed to optimize the different controller gains defined in the following chromosome

c h r o m o s o m e = [k p_{v b} k i_{v b} k d_{v b} \dots k p_{v} k i_{f} k d_{f}]

where $k p_{v b}, k i_{v b}, \dots$ are the propotional, integral gains of voltage controller. The procedure is depicted in the block diagram (Figure 1).

Figure 1.

Block diagram showing GA tuning.

The mathematical formulation for calculating ISE is as follows

ISE = \int_{0}^{\infty} {[e (t)]}^{2} d t

(1)

Here, $e (t)$ represents the error signal and may represent any of the system deviations like frequency and voltage.

SMES: a brief overview

Energy is stored in an SMES system as magnetic field produced by the flow of direct current in the coil. The conductor carrying the current is in a superconductive state as it is operated at cryogenic temperatures and therefore effectively has no resistive losses (Buckles and Hassenzahl, 2000). The concept of SMES was concieved with the idea of very large plants mainly for the purpose of load levelling.

For its higher efficiency and better cycling capability within short time intervals, it is best suited for high power but shorter duration requirements. Also, the negative effects of such renewable sources concerning power quality issues (mainly due to presence of power converters in such sytems) can be mitigated with this energy storage.

At present, the major hindrance associated with SMES is its very high capital costs (specifically the cooling units using either liquid helium at 4.2 K or superfluid helium at 1.8 K).

SMES configuration

Figure 2 shows the interface between SMES and AC grid. A power electronic interface called the Power Conditioning System (PCS) connects the SMES unit to the grid. The PCS includes a voltage-sourced converter (VSC) together with DC-DC chopper, a DC link capacitor, a coupling transformer and associated control.

Figure 2.

Interface connecting SMES to AC system.

The SMES output is controlled through a DC-DC 2-Q chopper. A bi-directional power converter (AC/DC) connects the unit to the AC grid. This converter is generally a three-phase bridge made of semiconductor switches functioning as a voltage source inverter (VSI); (Sebastián and Alzola, 2012). The converter used is either a conventional 6-pulse type or a 12-pulse type. The control of power flow from the SMES unit to the AC system and vice versa is achieved by issuance of firing pulses based on controllable parameter deviations. The DC bus voltage builds up to its nominal value once suitable gating signals are issued to the semiconductor switches (usually insulated-gate bipolar transistors (IGBTs) are used) and this voltage remains constant thereafter. Adjusting the duty cycle of the 2-Q chopper changes the polarity and magnitude of the voltage applied to the SMES coil. This in turn enables charging and discharging of the SMES.

During transient load, when there is a sudden increase in power demand, there is an instantaneous discharge of stored energy by the SMES through PCS. The SMES coil begins to store back its reference energy once the control process starts its operation to take the power system to the newly established equilibrium condition. Similarly, during an abrupt fall in power demand, the SMES responds almost instantaneously charging to its maximum value, thereby taking in some of the surplus energy from the system, and it releases this excess energy once the system goes back to its steady state and the SMES-stored energy again returns to its reference value (Iqbal et al., 2009).

Control strategies

The model employs proportional–integral–derivative (PID) controllers for frequency and voltage control. The frequency controller generates reference power based on frequency deviation and this reference signal is fed to SMES. Active and reactive power balance is achieved using $I_{d} and I_{q}$ as reference currents. These currents are based on respective frequency and voltage deviations. A proportional controller is also used for SMES current with low gain to bring it back to its reference after a disturbance. A lower gain is essential as taking a higher gain deteriorates system frequency.

The system considered consists of a wind-driven induction generator. As discussed earlier, such systems are prone to fluctuations in frequency owing to wind variations and power changes. On the other hand, the presence of induction generators causes active and reactive power disturbances which translate into voltage variations in the system. So, for maintaining power quality of the system, simultaneous modulation of active and reactive power by energy storage is required.

The simultaneous control of frequency and voltage is achieved through the combined action of the chopper’s duty cycle and the VSC switching. Frequency is measured through a three-phase phase-locked loop (PLL) using alternator voltage as its input. Comparing the actual frequency obtained from PLL with the reference (1 pu), an error signal is generated which is fed to the frequency controller. The controller then generates a reference power signal $(P_{s m}^{*})$ which is then used to compute the average chopper current $(I_{i s})$ according to equation (4), and then using equation (3), the duty cycle of the chopper is calculated. Alongside this action, a voltage measurement block measures the load bus voltage (in pu) and capacitor voltage (in V) and compares these with their nominal values (i.e. 0.7071 pu and 800 V, respectively). The error signal so generated is fed to two PI controllers which generate two reference signals $I_{d}$ and $I_{q}$ . These currents are then used as basis to produce a set of three-phase voltages based on DC bus and load bus voltage deviations. This voltage is used as a control signal to be fed to the average model-based VSC.

The cumulative action described above results in simultaneous control of both the system parameters, that is, frequency and voltage.

Mathematical modelling

In Figure 3, Q1 and Q2 represent the switches (IGBT) with similar duty cycle. D1 and D2 are the diodes.

Figure 3.

Chopper connection with SMES.

Based on the chopper duty cycle D, ‘three’ operating areas can be observed for the chopper configuration as follows (Iqbal et al., 2009)

• D < 0.5 • D = 0.5 • D > 0.5

The chopper current $I_{i s}$ (average value) is

\begin{array}{l} I_{i s} = \frac{1}{T} [\int_{0}^{D T - T / 2} i_{s m e s} d t + \int_{T / 2}^{D T} i_{s m e s} d t] \\ = \frac{i_{s m e s} (D - 0.5)}{0.5} \end{array}

(2)

where $i_{s m e s}$ represents the instantaneous value of SMES current. Rearranging equation (2), the duty cycle D of the chopper (Figure 4) can be expressed as follows

D = 0.5 + \frac{I_{i s}}{2 i_{s m e s}}

(3)

Let the desired input power to the SMES be represented by $P_{s m}^{*}$ ; then, the required average chopper current, $I_{i s}$ , is attained from equation (4)

P_{s m}^{*} = V_{d c} I_{i s}

(4)

The stored energy within the SMES coil is given as follows

E_{s m} = \frac{1}{2} L i_{s m e s}^{2}

(5)

Figure 4.

Control model for SMES.

System description

The sytem comprises a synchronous machine (SM) in conjunction with wind turbine generator (WTG), consumer load and SMES energy storage coupled with the grid through PCS. The WTG supplies uncontrolled real power, whereas the consumer load consumes uncontrolled active power. The SMES may consume/generate controlled active power based on frequency deviations. A 75 kVAR capacitor is provided for reactive power requiremnents of the asynchronous machine. The SMES power reference $P_{r e f}$ can be controlled based on frequency deviation. Based on such deviations, the SMES can supply or consume active power. The SMES operating mode is reflected in the DC bus capacitor voltage and the chopper’s duty cycle.

For $D = 0.5$ , the SMES operates in the floating/standby mode. The floating mode is the steady-state operating point of the system. In this mode, no energy transfer occurs between SMES unit and grid. The coil current remains fixed at its steady-state value and the voltage across it is zero. With no energy transfer taking place, all the energy remains stored in the coil (in our case, this energy is 56.25 kJ). Furthermore, the DC bus (capacitor) voltage is maintained at the reference value of 800 V. This mode is representative of the normal operation of the system prior to and after the occurrence of the fault.

For $D < 0.5$ , the SMES operates in the discharging mode. This condition occurs when there is an increase in the power consumption such as an increase in load demand or a reduction in wind speed. The SMES coil is put to use to bridge this deficit between supply and demand. Consequently, there is a decrease in coil current implying a negative rate (di/dt). The decreasing current results in negative voltage appearing across the coil. The duty cycle D controls the magnitude of both the voltage appearing across the coil and across the capacitor. Furthermore, the rate of change in coil current is dictated by D.

For $D > 0.5$ , the SMES operates in the charging mode. This condition occurs when there is load rejection or an increase in wind speed; the SMES coil now absorbs the surplus power. Consequently, there is an increase in the coil current implying a positive rate (di/dt). The increasing current results in positive voltage appearing across the coil.

Simulink model

The Simulink model of the system is shown in Figure 5. Various components mentioned in the following section such as the induction generator, the SM, the consumer load, the circuit breaker and the coupling transformer have been taken from SimPower systems library of Simulink.

Figure 5.

Simulink model.

WTG induction generator

The synchronous machine block taken from SimPower systems library is used to model a three-phase induction generator (squirrel cage) which is connected to the WT. A negative mechanical torque input indicates that the machine is operating in the generating mode. A fourth-order state–space is used to model the electrical part of the machine, while the mechanical part is modelled by a second-order system. The electrical parameters are referred to the stator. The two-axis reference frame (d-q frame) is the reference for all rotor and stator quantities.

A 275-kW induction machine is modelled in this system with an inertia constant of 2 s. The induction generator (IG) output is connected to the grid, while its mechanical input is connected to the WT block. Inside the WT block seen in Figure 5 are the WT characteristics. These characteristics define the mechanical torque to be applied to the IG as a function of the generator’s shaft speed and the wind speed. There is no control over the WTG active power since the WT used has no pitch control. This is the worst-case scenario as the wind variability directly impacts the WTG active power.

Diesel system

The SM is coupled with diesel engine which is equipped with speed governor. The speed governor in the study is considered to have a speed regulation R and includes an integral gain Ki. The SM of this unit is simulated by a standard d-q axis model. The d-q frame of reference is one which is aligned to generator rotor direct and quadrature axis and rotates at the same speed as the rotor.

The diesel generator has a power rating of 300 kVA with rated voltage of 480 V. (taken from SimPower library).

Average converter model

The three-phase AC/DC voltage source converter is modelled by an average value model. The input current is obtained from the generated output power and DC bus voltage. Figure 6 shows the modelling of the three-phase converter.

Figure 6.

DC/AC converter model.

SMES model

The SMES coil is modelled by an inductive coil of 5 H which is coupled with the DC bus through 2-Q class B DC-DC chopper (average model). The coil carries a current of 150 A. Thus, the energy content of the coil is $(1 / 2) L I^{2}$ which gives 56.25 kJ. The SMES model is shown in Figure 7. The duty cycle of the chopper is taken according to equation (3).

Figure 7.

Chopper average model coupled with SMES coil.

Other auxiliaries

The VSC is coupled with the grid through a 480/220 V Y/Δ transformer. The coupling transformer serves the purpose of isolation and stepping-down high voltage. The three-phase breaker block simulates the operation of a three-phase circuit breaker and its closing and opening times are taken to initiate a load step or load rejection, respectively. The consumer load in the steady state, taken as 200 kW, is modelled by a three-phase resistive (R) load.

Simulation results

Different graphs shown here depict the performance of the system parameters, that is, system frequency and voltage with and without using SMES in the system. Here, the frequency is compared for both load step (Figure 8(b)) and wind speed step (Figure 9(b)). The system voltage, plotted in pu, for load step and wind speed step is shown in Figures 8(d) and 9(d), respectively.

Figure 8.

Load step: (a) power, (b) frequency, (c) SMES current and (d) load voltage.

Figure 9.

Wind speed step: (a) power. (b) system frequency, (c) SMES current and (d) load voltage.

The power supplied by SMES is considered positive/negative when it is either consumed or produced, respectively, following the passive sign convention as is depicted in the power plots in Figure 8(a). Figure 8(c) shows the SMES current. Starting from steady state at $t = 0 s$ , the power consumed by the the SMES and load is 0 and 200 kW, respectively. The steady-state speed of wind is 8 m/s and the power generated by the alternator and WTG is 155 and 90 kW, respectively; the SMES coil current is 150 A with the system in reference state.

Load step

With the system in the above steady state, an additional load of 50 kW is connected to the consumer bus at $t = 1 s$ by the closure of the three-phase breaker. Figure 8(a) shows the SMES power corresponding to the load step. Figure 8(b) shows the comparison between the frequency of two identical systems, one without any storage and the other with SMES. Figure 8(c) shows the SMES current behaviour on load disturbance. It is seen that the SMES current falls down to $≃ 126 A$ which implies an energy content of 39.69 kJ indicating that the coil has released a maximum of 16.56 kJ of its stored energy content. Figure 8(d) shows the corresponding voltage (of a phase in rms) in the two systems. The instantaneous load step of 50 kW is supplied by the SMES as shown in Figure 8(a).

It is observed from the above plots that the proposed scheme results in better frequency and voltage control with a significant reduction in peak–peak variations. It is observed further that the transients after the occurrence of a disturbance has lesser amplitude with respect to reference for frequency plot (Figure 8(b)) and better stability (lesser fluctuations) for voltage plot (Figure 8(d)).

Wind step

Here, the effect of wind variance is observed in Figure 9. The wind speed increases from its steady-state speed of 8 to 9 m/s. The corresponding power from WT increases from 90 kW (in steady state) to 144 kW. The corresponding frequency plot and SMES current plot are shown in Figure 9(b) and (c), respectively. It can be observed from the plot in Figure 9(c) that the peak SMES current goes to $≃ 180 A$ implying an energy content of 81 kJ which is an addition of 24.75 kJ to its steady-state energy content.

It can be observed from the power plot (Figure 9(a)) that at the instant of wind speed step, the additional power is taken up by the SMES, and at steady state, the surplus power is reflected in the reduction in diesel generation.

Similar to the plots corresponding to load step, the system with SMES shows a much better performance than the one without it. It is observed further that the transients, for the system without SMES, after the wind disturbance have severe fluctuation for voltage in contrast to the system with SMES (Figure 9(d)).

Load rejection

A combined frequency plot is shown in Figure 10(b) where simultaneous effect of load step, wind step and load rejection is observed. It can be seen from Figure 10(a) that a load of 50 kW is initially added at $t = 1 s$ , a wind speed step is given at $t = 12 s$ and finally the load of 50 kW is taken down at $t = 22 s$ .

Figure 10.

Simultaneous disturbances: (a) power and (b) frequency.

It can be observed from Figure 10(a) that at $t = 1 s$ (load addition), the SMES instantly supplies 50 kW power; the wind, except some fluctuations, supplies the same amount of power as earlier; and the diesel system gradually begins to supply the deficit amount of power. At $t = 12 s$ (wind speed step), the wind power output increases; this surplus power in the system is absorbed by the SMES at that instant and gradually it releases this power and this is reflected in reduced power supply from the diesel system. A similar response (to the wind speed step) is observed for the load rejection at $t = 22 s$ with the diesel system reducing its power output in steady state.

Observation

As it is evident from the plots, the system performance has improved significantly with the proposed scheme. Tables 1 and 2 summarize the performance of the two systems considered with the fourth column (italicized) depicting the improvement in system parameters with the use of SMES.

Table 1.

Frequency and voltage comparison for load step.

Parameter	Without SMES	With SMES	% reduction with SMES
Peak–peak frequency deviation	0.0034	0.0011	309.09
Peak–peak voltage deviation	0.0048	0.0024	200

SMES: superconducting magnetic energy storage system.

Table 2.

Frequency and voltage comparison for wind step.

Parameter	Without SMES	With SMES	% reduction with SMES
Peak–peak frequency deviation	0.0036	0.001	360
Peak–peak voltage deviation	0.0044	0.0015	293.93

SMES: superconducting magnetic energy storage system.

Conclusion

This article proposes an effective way of incorporating SMES energy storage into an isolated wind–diesel system with faster real and reactive power exchange capabilities. An average value–based SMES model and its associated power conversion system is used and its performance is validated through SimPower systems. The dynamics of all the components involved in the system have been considered. It has been shown that the suggested scheme results in better quality of supply under wind and load disturbances. It is also observed that not only the maximum deviation in frequency is considerably reduced but also the scheme results in better transient response, thereby reducing the impact of such disturbances on the system. The system voltage is better regulated and controlled with this system.

It is also observed that with the suggested control, the SMES current returns to its steady-state value within short span $(≃ 15 to 16 s)$ and thus can be used for continuous control operation. The scheme is economical too as a small rating (5 H) coil is used.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Aaqib Ali Abass

References

Buckles

Hassenzahl

(2000) Superconducting magnetic energy storage. IEEE Power Engineering Review 20(5): 16–20.

Faraji

Majazi

Al-Haddad

, et al. (2017) A comprehensive review of flywheel energy storage system technology. Renewable and Sustainable Energy Reviews 67: 477–490.

Gao

(2015) Energy Storage for Sustainable Microgrid. Cambridge, MA: Academic Press.

Hayashi

Hatabe

Nagafuchi

, et al. (2006) Test results of power system control by experimental SMES. IEEE Transactions on Applied Superconductivity 16(2): 598–601.

Heetun

Abdel Aleem

Zobaa

(2016) Voltage stability analysis of grid-connected wind farms with facts: Static and dynamic analysis. Energy and Policy Research 3(1): 1–12.

Hunter

Elliot

(1994) Wind-Diesel Systems: A Guide to the Technology and Its Implementation. Cambridge: Cambridge University Press.

Iqbal

Mufti

Lone

, et al. (2009) Intelligently controlled superconducting magnetic energy storage for improved load frequency control. International Journal of Power & Energy Systems 29(4): 241.

Ise

Murakami

Tsuji

(1987) Charging and discharging characteristics of SMES with active filter in transmission system. IEEE Transactions on Magnetics 23(2): 545–548.

Jeffries

McGowan

Manwell

(1996) Development of a dynamic model for no storage wind/diesel systems. Wind Engineering 20: 27–38.

10.

Kouba

NEY

Menaa

Hasni

, et al. (2016) A novel optimal frequency control strategy for an isolated wind–diesel hybrid system with energy storage devices. Wind Engineering 40(6): 497–517.

11.

Lone

Mufti

MUD

(2004) Redox flow batteries: Modelled for power quality improvements in autonomous wind-diesel power systems. Wind Engineering 28(5): 577–586.

12.

McGowan

Connors

(2000) Windpower: A turn of the century review. Annual Review of Energy and the Environment 25(1): 147–197.

13.

Manwell

McGowan

Rogers

(2010) Wind Energy Explained: Theory, Design and Application. Hoboken, NJ: John Wiley & Sons.

14.

Molina

Mercado

(2011) Power flow stabilization and control of microgrid with wind generation by superconducting magnetic energy storage. IEEE Transactions on Power Electronics 26(3): 910–922.

15.

Mubarak

(2017) Designing of a Pv/Wind/Diesel Hybrid Energy System. Blurb, Inc., https://books.google.co.in/books?id=DH3NtAEACAAJ

16.

Nacfaire

(1989) Wind-Diesel and Wind Autonomous Energy Systems. Boca Raton, FL: CRC Press.

17.

Sebastián

(2016) Application of a battery energy storage for frequency regulation and peak shaving in a wind diesel power system. IET Generation, Transmission & Distribution 10(3): 764–770

18.

Sebastián

Alzola

(2012) Flywheel energy storage systems: Review and simulation for an isolated wind power system. Renewable and Sustainable Energy Reviews 16(9): 6803–6813.

19.

Sumper

Díaz-González

Gomis-Bellmunt

(2016) Energy Storage in Power Systems. Hoboken, NJ: John Wiley & Sons, https://books.google.co.in/books?id=wfpcCgAAQBAJ

20.

Vidyanandan

Senroy

(2016) Frequency regulation in a wind–diesel powered microgrid using flywheels and fuel cells. IET Generation, Transmission & Distribution 10(3): 780–788.

21.

Zargar

Mufti

MUD

Lone

(2017a) Adaptive predictive control of a small capacity SMES unit for improved frequency control of a wind-diesel power system. IET Renewable Power Generation 11(14): 1832–1840.

22.

Zargar

Mufti

MUD

Lone

(2017b) Modelling and performance assessment of a standalone hybrid wind-diesel-superconducting magnetic energy storage system using four-quadrant operation of superconducting magnetic energy storage. Wind Engineering 42: 496–509.