Abstract
This article investigates the effect of coordinated control of variable frequency transformer and fuzzy based flywheel energy storage system for frequency stabilisation of wind penetrated two area power system. Wind turbine generator and fuzzy based flywheel energy storage system are incorporated in two different areas. Variable frequency transformer, set up in series with the tie-line near wind-embedded area, is modelled as a tie-line power flow controller between interconnected areas. Variable frequency transformer introduces the phase angle shift between interconnected areas in accordance with the power demand on account of load and wind speed variations in wind-embedded area, while actual power is supplied from fast acting flywheel energy storage system incorporated in other area. Fuzzy controller for flywheel energy storage system is designed in such way that frequency deviations are mitigated continuously during disturbances and nominal storage capacity of flywheel energy storage system is restored on incessant basis, after every load and wind disturbance is tackled.
Keywords
Introduction
The rapid increase in energy demand and limited fossil fuel availability has led to substantial growth in harnessing the energy from non-conventional sources in the last few decades. Among the many non-conventional energy sources available, wind energy is proved to be the most promising one. Firstly, because of its substantial potential to fulfil the gap between generation and demand and secondly, because of the progressive growth in power electronic technologies from the last few decades. High wind energy penetration unquestionably improves the system reliability but its very incorporation poses a great threat to the power quality because of the inherent fluctuating nature of wind. For this, fast acting energy storage systems (ESS) can be used which can effectively smoothen the effect of wind speed fluctuations (Ahsan and Mufti, 2019; Vidyanandan and Senroy, 2016).
Load frequency control study plays a vital role in determining the reliability and stability of an interconnected power system. The two maintain objectives of LFC are to sustain the area frequency and tie power at predetermined values during sudden area load variations. However, the conventional LFC structure has limit in reducing the large frequency deviations (during sudden load variations) due to the sluggish response of governor. Thus, it calls for a fast acting energy storage systems which can adequately diminish the effect of sudden area load variations. For a quite longer time, many advanced fast acting energy storage systems such as battery energy storage system (BESS), super_magneting energy storage system (SMES), flywheel energy storage system (FESS) and super_capacitor energy storage system (SCESS) have been in use for mitigating the sudden frequency variations (Mufti et al., 2009, 2015). Higher cost of SMES on account of expensive helium requirement, slow charging and discharging rate of BESS, lesser storage capacity of SCESS are some of the limitations that these very energy storage systems possess. FESS on the other hand is advantageous over these energy storage systems in a way that they possess efficiency as high as 90%, longer life up to 20 years, cycling life up to 100,000 cycles, energy and power density up to 424 kWh/m3 and 2000 kW/m3 respectively (Eckroad and Gyuk, 2003; Lazarewicz and Rojas, 2004).
For a quite longer time, application of FESS has been limited because of their susceptibility to mechanical failures. But now significant improvements have been made in its structure. To state a few, nowadays rotary parts of flywheel are held by magnetic bearings to avoid friction, wind shear is avoided by placing the structure in vacuum, power exchange is done through power electronic interfaces etc. (Gonzalez et al., 2016). Moreover, energy storage capacity of FESS has increased significantly and is nowadays up to 500 MJ with a much faster response time of 0.25 cycle, because of the recent advancements in power electronic interfaces (Hebner et al., 2002).
Integration of FESS for frequency control of wind-based systems has been reported in studies by Suvire and Mercado (2012), Achour et al. (2017) and Yao et al. (2016). Incorporation of FESS for frequency regulation of micro-grid has been analysed in studies by Vidyanandan and Senroy (2016), Jin et al. (2017) and Nair and Senroy (2016). Fuzzy controlled FESS has been investigated in study by Manzoor and Mufti (2020). Application of PID controlled FESS for LFC of an interconnected power system has been reported in study by Lee and Kim (2013) in which FESS is placed in each area of an interconnected power system.
With the ever increasing number of areas in an interconnected power system, it may not be economically reasonable to introduce ESS in every area of an interconnected power system. It becomes profitable if a good capacity ESS unit positioned in an area is able to stabilise the frequency oscillations of other interconnected areas. For this, researchers have proposed the use of phase shifters (flexible AC transmission system (FACTS) device) with ESS for LFC. Application of solid state phase shifter in coordination with SMES for LFC has been reported in study by Ngamroo et al. (1999). LFC by combined control of SMES and Thyristor controlled phase shifter (TCPS) has been analysed in study by Bhatt et al. (2010). Incorporation of interline power flow controller (IPFC) and redox flow battery for LFC has been explored in study by Chidambaram and Paramasivam (2012). The literature survey clearly indicates that power electronic based phase shifters are in common use with ESS for frequency regulation of an interconnected power system (Dash et al., 2015).
Conversely, these power electronic based phase shifters pose a great threat to the power quality in terms of harmonic pollution, voltage surge susceptibility and resonance. Moreover, apart from providing inadequate inertia and restricted overload support, they are unable to connect asynchronous power system areas and this is one of the major limitations of using these phase shifters. At this, a novel power transmission technology called variable frequency transformer (VFT) can be incorporated in place of these power electronic based phase shifters. VFT is a rotating phase shifting transformer which apart from linking asynchronous power system areas can overcome the other limitations of power electronic based phase shifters such as harmonics, overloading proficiency and resonance. Moreover, VFT can provide adequate inertia due its large rotating mass which can significantly improve the frequency control during sudden load variations (Ambati et al., 2015; Ambati and Khadkikar, 2016; Bakhsh and Khatod, 2014; Merkhouf et al., 2008; Piwko et al., 2005; Pratico et al., 2010).
Keeping above discussion in view, the aim of this article is to analyse the effect of coordinated control of VFT and fuzzy based FESS for frequency stabilisation of wind penetrated two area power system. Wind turbine generator is installed in area #1 and fuzzy controlled FESS is installed in area #2 as shown in Figure 1. VFT is modelled as a tie line power flow controller between area #1 and area #2, whose task is to assist the control of FESS locally in area #1. To ensure continuous control, fuzzy controller for FESS is appointed with two tasks, one is to minimise the frequency deviations during disturbances and other is to restore the nominal storage capacity of FESS on incessant basis, after every load and wind disturbance is tackled. The developed model is investigated for different area load and wind speed variations.

Proposed scheme.
Variable frequency transformer: An overview
Variable frequency transformer (VFT) is a rotating phase shifting transformer which functions at a regulating phase angle. It has three main components; rotating transformer with three-phase windings on both rotor and stator, dc drive motor system coupled to the rotor, and three phase collector system to conduct current between rotor windings and its fixed bus conduit. VFT is incorporated in line by connecting its stator and rotor sides to two separate power system areas as shown in Figure 2 and that power system areas can be synchronous as well as asynchronous. Power transfer through VFT is proportional to direction and magnitude of torque applied to rotor by dc drive motor system. In order to transfer power from one area to other, rotor is turned out of its original position with respect to stator by applying torque. When VFT is used to connect synchronous power system areas, rotational speed of rotor is zero and dc drive motor system is proposed to produce torque at zero speed. However, when power system areas are out of synchronism, rotor rotates at a speed proportional to the difference in frequencies between areas, and dc drive motor system is proposed to produce torque at rotational speed of rotor, with no effect on the power flow. Thus, VFT can flexibly connect both synchronous as well as asynchronous power system areas without affecting the power flow and that power flow is independently controlled by applied torque (Ambati et al., 2015; Ambati and Khadkikar, 2016; Merkhouf et al., 2008).

Variable frequency transformer.
Analysis of power flow in VFT
Considering VFT as an ideal lossless machine, the power balance equation can be expressed as (Merkhouf et al., 2008):
where
Since behaviour of VFT is similar to transformer, therefore magneto motive force of stator and rotor should balance each other as
where
Stator and rotor windings of VFT not only differ by number of turns but by frequencies as well, so
and
or
where
Since rotor speed is proportional to difference in frequencies between stator and rotor, therefore
and
where
f mec = mechanical speed of rotor in Hz;
w mec = mechanical speed of rotor in rpm.
The dc drive motor power can be expressed, using (1)–(6), as
The dc drive torque Tt can be expressed, using (1)–(7), as
The above equation confirms that dc drive motor torque Tt is independent of the rotational speed of rotor and power flow through VFT is directly proportional to applied torque Tt (Merkhouf et al., 2008).
Flywheel energy storage system
Flywheels are mechanical energy storage systems which stores energy in the form of kinetic energy, in cylindrical rotating mass which is coupled to the reversible generator motor. The generator/motor can be brushless DC machine, asynchronous machine, permanent magnet synchronous machine and so on (Gurumurthy et al., 2013; Lee et al., 2009; Vidyanandan and Senroy, 2016). FESS is charged by motoring action which raises its speed, and discharged by generating action which decreases its speed. Since FESS works in a variable frequency mode due to its changing speed, it necessarily needs power electronic converters for interfacing. The basic configuration of FESS is presented in Figure 3(a). The mathematical equations leading the dynamics of FESS are as follows

(a) Configuration of FESS and (b) simulink model of FESS.
At any instant, stored kinetic energy in rotor is as
where
At any instant, energy supplied from FESS is given as
The left behind kinetic energy in FESS is
where
The simulation model of FESS obtained from above equations is shown in Figure 3(b) (Vidyanandan and Senroy, 2016).
Fuzzy controlled flywheel energy storage system
In the present study, fuzzy controller is used to generate the power command for FESS. To ensure the continuous control, fuzzy controller is appointed with two tasks, one is to minimise the frequency deviations during disturbances, and other is to restore the nominal storage capacity of FESS on incessant basis, after every disturbance is tackled (Manzoor and Mufti, 2020). Whenever there is a sudden decrease in frequency on account of disturbance, power flows from the flywheel to the system to minimise the frequency deviations. As the governor control mechanism starts working to set the system to the new equilibrium condition, fuzzy controller ensures that power flows back from the system to the flywheel to restore its nominal speed. Similarly, whenever there is a sudden increase in frequency, power flows from the system to the flywheel which raises its speed. As the governor control mechanism starts working, again fuzzy controller ensures that power flows back from the flywheel to the system to restore its nominal speed. Therefore, the two controlling parameters for fuzzy controller are area #2 frequency/rotor speed deviations (ΔWm2) and FESS speed deviations (ΔWfess) as presented in Figure 4. Membership functions for fuzzy controller are given in Figure 5. Rule base for fuzzy controller is given in Table 1 (Manzoor and Mufti, 2018; Mufti et al., 2009) and in it one simple rule means if rotor speed deviation (ΔWm2) is NS (negative small) and FESS speed deviation (ΔWfess) is PM (positive medium), then power command for FESS is NM (negative medium).

Schematic diagram of fuzzy controller.

Membership functions of fuzzy controller: (a) FESS speed deviation (p.u), (b) rotor speed deviation in area #2 (p.u), and (c) power command for FESS (p.u).
Rule base for fuzzy controller.
Simulink model of proposed scheme
The power flow block diagram and simulink model of proposed scheme are given in Figures 6 and 7(a) (landscape orientation) respectively. Various components incorporated in the simulink model are as follows:

Block diagram showing the power flow for proposed scheme.

(a) Simulation model of VFT and fuzzy controlled FESS in two area power system and (b) governor and excitation model of area #2 integrated with flywheel model.
Wind turbine generator – Squirrel cage induction generator
Wind turbine generator is realised by squirrel cage induction generator (asynchronous machine) coupled to wind turbine (Abbas and Mufti, 2019). Mechanical input of squirrel cage induction generator (SCIG) is connected to wind turbine and its electrical output is connected to grid as shown in Figure 7(a). Wind turbine characteristics define the mechanical torque to be applied to squirrel cage induction generator as a function of generator’s rotor speed and wind speed. In the present study, there is no pitch control and this is the worst situation as wind disturbance directly affects the wind power output.
Wind turbine generator has a power rating of 10 MVA with a rated voltage and nominal frequency of 3 kV and 60 Hz respectively.
Diesel generator – Synchronous generator
The two areas (area #1/area #2) are realised by synchronous generators coupled to diesel governor and excitation system as shown in Figure 7(a) (Abbas and Mufti, 2019). The power input (Pmec) of synchronous generator is connected to diesel engine governor and field voltage input (Vf) is connected to excitation system. The excitation system uses the direct axis stator voltage (Vd) and quadrature axis stator voltage (Vq) to generate the field voltage (Vf) for the rotor. Diesel governor characteristics (speed regulation with integral gain) define the power command (Pmec) to be applied to synchronous generator as a function of generator’s rotor speed (Wm) and tie power (Ptie). Synchronous generator is modelled in dq reference frame. For area #2, there is one more input to diesel governor, that is flywheel power ‘Pfess’ as shown in Figure 7(b) (in consensus with Figure 6).
Diesel generator in each area has a power rating of 100 MVA with a rated voltage and nominal frequency of 3 kV and 60 Hz.
Variable frequency transformer – Wound rotor induction machine
Variable frequency transformer is realised by doubly fed wound rotor induction machine (WRIM) coupled to dc drive motor system (Bakhsh and Khatod, 2014; Merkhouf et al., 2008) given in Figure 7(a). Stator side of WRIM is connected to wind-embedded area (area #1) and rotor side is connected to FESS-embedded area (area #2) through a long transmission line of 350 km. Power transfer through VFT is a function of torque applied to rotor by dc drive motor system and in our study, dc motor system is realised by gain block. Tie power (Ptie) is assumed to be flowing from area #2 to area #1. Variable frequency transformer is modelled as a tie-line power flow controller which introduces the phase angle shift between interconnected areas in accordance with the power demand on account of load and wind speed variations in wind-embedded area. The purpose of VFT is to minimise the frequency deviations in area #1. Therefore, the torque applied to VFT is made proportional to tie power and rotor speed deviations in area #1 with gains t1 for rotor speed deviations and t2 for tie power deviations.
Tt ∝ (t1 ΔWm1 + t2ΔPtie) (Bhatt et al., 2010; Ngamroo et al., 1999).
Fuzzy based FESS
As already discussed in section ‘Flywheel energy storage system’, fuzzy controller is used to generate power command for FESS. To ensure continuous control, fuzzy controller is appointed with two tasks, one is to minimise the frequency deviations during disturbances, and other is to restore the nominal capacity of FESS on incessant basis, after every disturbance is tackled (Manzoor and Mufti, 2010). In the simulation model, fuzzy controller is realised by a matlab function (made up of membership functions and rule base) which takes area #2 rotor speed deviation (ΔWm2) and FESS speed deviation (ΔWfess) as inputs, and outputs the power command (Pfess) as shown in Figure 7(b).
Other auxiliaries
Tie line is realised by three-phase distributed transmission line of 350 km as shown in Figure 7(a). The steady state consumer load in each area is 50 MW which is realised by three-phase resistive load. Load variation is realised by three phase breaker and an additional resistive load. Closing and opening times of breakers are used for load increment and decrement respectively. Frequency deviation in each area is measured by phase locked loop (PLL) block.
Simulation results
Case 1: Wind speed variation in area #1
In this case study, wind speed is varied while area #1, area #2 load is kept constant. Starting with a steady state load of 50 MW in both the areas and initial wind speed of 6 m/s, area #1 is subjected to a wind speed increment of 6 m/s at t = 40 s as shown in Figure 8(a). Corresponding frequency deviation in area #1 for system ‘without VFT and fuzzy based FESS’ and ‘with coordinated control of VFT and fuzzy based FESS’ is shown in Figure 8(c). It is clearly depicted from Figure 8(c) that coordinated control of VFT and FC FESS has been effective in minimising the area #1 frequency deviation to a considerable extent, though there is no fast acting energy storage available in area #1. This is facilitated by regulated torque applied to VFT which deviates at t = 40 s as shown in Figure 8(d). Since VFT itself is not an energy storage device, the power is actually absorbed by FESS available in area #2 and VFT (available in area #1) facilitates the control of FESS locally in area #1. This is depicted by increased speed of FESS from its nominal value (i.e. Wo = 1 p.u) at t = 40 s as shown in Figure 8(e). Corresponding FESS power is shown in Figure 8(f). Convention for power flowing into FESS is positive and negative for power flowing out of FESS. After the wind disturbance is tackled and system acquires new equilibrium position, VFT torque (Tt) regains its nominal value and fuzzy controller makes sure that power flows back from FESS to the system to regain its nominal capacity. FESS regains its nominal speed (capacity) at t = 60 s as shown in Figure 8(e) thus remains prepared to handle a new disturbance afterwards.

Power system variable pertaining to wind speed variation only: (a) wind speed variation in area #1, (b) wind power, (c) frequency deviation in area #1, (d) VFT torque, (e) FESS speed deviation, and (f) FESS power.
At t = 80 s, area #1 is now subjected to a wind speed decrement of 6 m/s. Corresponding frequency deviation, VFT torque, FESS Speed and FESS power are shown in Figures 8(c) to (f) respectively. This time power flows from FESS to the system, which is depicted by its decreased speed at t = 80 s as shown in Figure 8(e) and is facilitated locally by torque (Tt) applied to VFT (deviates at t = 80 s) as shown in Figure 8(d). After the wind disturbance is tackled and system acquires new equilibrium position, VFT torque (Tt) regains its nominal value and fuzzy controller restores the nominal capacity of FESS by taking power back from system to the FESS as shown in Figure 8(f). Figure 8(c) clearly depicts that coordinated control of VFT and FC FESS has been effective in minimising the frequency deviations in area #1 during wind speed decrements as well.
Case 2: Simultaneous wind speed and load variation in area #1
In this case study, area #1 is subjected to simultaneous wind speed and load variation in area #1 while area #2 load is kept constant. At t = 30 s, wind speed is incremented by 6 m/s, at t = 60 s, area #1 load is incremented by 10% and finally at t = 100 s, both wind speed and area #1 load are simultaneously decremented by 6 m/s and 10% respectively as shown in Figure 9(a) and (b). Corresponding frequency deviation in area #1 for system ‘without VFT and fuzzy based FESS’ and ‘with coordinated control of VFT and fuzzy based FESS’ is shown in Figure 9(c). It is clearly depicted from Figure 9(c) that coordinated control of VFT and fuzzy based FESS has been effective in minimising the frequency deviations in area #1 during simultaneous wind speed and area #1 load variations. Corresponding VFT torque (Tt), FESS speed and FESS power are shown in Figure 9(d) to (f) respectively.

Power system variable pertaining to simultaneous wind speed and load variation in area #1: (a) wind speed variation in area #1, (b) load variation in area #1, (c) frequency deviation in area #1, (d) VFT torque, (e) FESS speed deviation, and (f) FESS power.
Case 3: Simultaneous wind speed and load variation in area #1, and load variation in area #2
In this case study, both area #1 and area #2 are simultaneously subjected to a variety of wind speed (area #1) and load (area #1 and area #2) variations as given in Figure 10(a) and (b) respectively. Corresponding frequency deviations in area #1 and area #2 for system ‘without VFT and fuzzy based FESS’ and ‘with coordinated control of VFT and fuzzy based FESS’ are presented in Figure 10(c) and (d) respectively. VFT torque (Tt) and FESS speed are given in Figure 10(e) and (f) respectively. Frequency deviation in area #2 is tackled locally by FESS only as is depicted by its deviated speed shown in Figure 10(f) while VFT has no role in minimising the frequency deviations of area #2 as is depicted from Figure 10(e) in which VFT torque (Tt) shows no deviation for load variations of area #2. It is clear from Figure 10(e) and (f) that VFT torque (Tt) deviates for disturbances (load and wind) of area #1 only while FESS speed deviates for disturbances of both area #1 and area #2. Thus, the purpose of coordinated control is effectively achieved as power to minimise the frequency deviations in area #1 and area #2 is absorbed/supplied from FESS while VFT only assists the control of FESS in area #1.

Power system variables pertaining to simultaneous wind speed and load variations in area #1 and area #2: (a) wind speed variation in area #1, (b) simultaneous load variation in area #1 and area #2, (c) frequency deviation in area #1, (d) frequency deviation in area #2, (e) VFT torque, and (f) FESS speed deviation.
Observation
It is obvious from above figures that system performance has considerably improved with the proposed scheme. Tables 2 and 3 summarise the performance in two areas.
Frequency deviation comparison in area #1.
Frequency deviation comparison in area #2.
Conclusion
In this paper, coordinated control of VFT and fuzzy based FESS for frequency stabilisation of wind penetrated two-area power system has been investigated. Wind turbine generator is installed in area #1 and fuzzy controlled FESS is installed in area #2. VFT is modelled as a tie line power flow controller between area #1 and area #2 whose task is to assist the control of FESS locally in area #1. To validate the proposed scheme, modelling of VFT, FESS, wind turbine generator and two areas has been done in matlab. VFT is realised by wound rotor induction machine mechanically coupled to dc drive motor system. Two areas are realised by synchronous generator coupled to diesel governor, and wind turbine generator is realised by squirrel cage induction generator coupled to wind turbine. Fuzzy controller is used to generate the power command for FESS. To ensure the continuous control, fuzzy controller is appointed with two tasks, one is to minimise the frequency deviations during disturbances, and other is to restore the nominal capacity of FESS on incessant basis, after every disturbance is tackled. The developed model is investigated for different area load and wind speed variations. Frequency deviations in area #1 on account of wind speed and load variations are minimised by coordinated control of VFT and FC FESS, in which power is supplied from FESS and VFT only assists the control of FESS in area #1. While frequency deviations in area #2 are minimised locally by FESS without the assistance of VFT. Simulation results reveal that frequency deviations in wind-embedded two area power system are effectively minimised (around 70%) by coordinated control of VFT and FC FESS.
Footnotes
Appendix A
Appendix B
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.
