INTRODUCTION
Power Quality (PQ) problems have become an increasing concern with an increased
usage of critical and sensitive loads in industrial processes. Disturbances
such as voltage sags and swells, short duration interruptions, harmonics and
transients may disrupt the processes and lead to considerable economic loss.
Power Quality (PQ) problems are classified as per the international standards
and methods of classifying primary and secondary distribution problems by duration,
type and severity (Ghosh and Ledwich, 2002). Among the
disturbances, voltage sags are considered to be the most significant and critical.
Voltage sag is a momentary decrease of the voltage RMS value with the duration
of half a cycle up to many cycles. Voltage sag can cause serious problem to
sensitive loads that use voltage sensitive components such as adjustable speed
drives, process control equipment and computers and voltage sags last until
network faults are cleared. In order to increase the reliability of a power
distribution system, many methods of solving power quality problems have been
suggested (Arnold, 2001). Various voltage sag mitigation
schemes are based on inverter systems consisting of energy storage and switches.
The DSTATCOM has emerged as a promising device to provide not only for voltage
sag mitigation but a host of other power quality solutions such as voltage stabilization,
flicker suppression, power factor correction and harmonic control (Hingorani
and Gyugyi, 2000). The DSTATCOM has additional capability that provides
voltage support and regulation of VAr flow. Because the device generates a synchronous
wave form, it is capable of generating variable reactive or capacitive shunt
compensation at a level up to the maximum MVA rating of the DSTATCOM inverter.
A dynamic voltage restorer is one device having capability of protecting sensitive
loads from all supply side disturbances. The Dynamic Voltage Restorer (DVR)
employs series voltage boost technology using solid state switches to correct
the load voltage amplitude as needed (Cao et al.,
2001).
The basic concept is that during sag period, the DVR operates in boost mode and injects voltage of sufficient magnitude to maintain constant voltage throughout the sag period. This developed controllable voltage is added to the supply voltage through the use of a series transformer to get the required load voltage.
A new mitigation device for voltage sag is proposed by Lee
et al. (2007) using PWMswitched autotransformer. The performance
of the compensator for various sag conditions is presented. This study presents
mitigating device for voltage sag disturbances using PWMswitched autotransformer.
Here, the control circuit based on RMS voltage is used to identify the sag and
swell disturbances. This compensator has less switching devices and hence reduced
gate drive circuit size but has the capability to supply the required undistorted
load voltage and currents. It has only one switching device per phase and no
energy storage device. Simulation of the compensator is performed using MATLAB/SIMULINK
and performance results are presented.
MATERIALS AND METHODS
Proposed system configuration: The proposed device for mitigating voltage
sag and swell in the system consists of a PWM switched power electronic device
connected to an autotransformer in series with the load. Figure
1 shows the single phase circuit configuration of the mitigating device
and the control circuit logic used in the system. It consists of a single PWM
Insulated Gate Bipolar Transistor (IGBT) switch in a bridge configuration, a
thyristor bypass switch, an autotransformer and voltage controller.
Principle of operation: To maintain the load voltage constant an IGBT is used as power electronic device to inject the error voltage into the line. Four power diodes (D1D4) connected to IGBT Switch (SW) controls the direction of power flow and connected in ac voltage controller configuration with a suitable control circuit maintains constant RMS load voltage. In this scheme, sinusoidal PWM pulse technique is used. RMS value of the load voltage V_{L} is calculated and compared with the reference RMS voltage V_{ref}.
During normal condition the power flow is through the anti parallel thyristors. Output filters containing a main capacitor filter and a notch filter are used at the output side to filter out the switching noise and reduce harmonics. During this normal condition, V_{L} = V_{ref } and the error voltage V_{err} is zero. The gate pulses are blocked to IGBT.
Due to sudden increase or decrease in the load or due to faults voltage sag
or swell occurs. The supply voltage V_{S} and hence V_{L} decreases.
When the sensing circuit detects an error voltage V_{err} >10% of
the normal voltage the voltage controller acts immediately to switch off the
thyristors. Voltage V_{err} applied to the pi controller gives the phase
angle δ. The control voltage given in Eq. 1 is constructed
at power frequency f = 50 Hz:
Where m_{a} is the modulation index. The phase angle a is dependent
on the percentage of disturbance and hence controls the magnitude of V_{control}.

Fig. 1: 
Voltage sag/swell mitigating device 


Fig. 2: 
Voltage and current relations in an autotransformer 

This control voltage is then compared with the triangular voltage V_{tri} to generate the PWM pulses VG which are applied to the IGBT to regulate the output voltage. Hence, the IGBT switch operates only during voltage sag or swells condition and regulates the output voltage according to the PWM dutycycle. To suppress the over voltage when the switches are turned off, RC snubber circuits are connected across the IGBT and thyristor.
Voltage sag compensation: The ac converter topology is employed for
realizing the voltage sag compensator. This study considers the voltage mitigation
scheme that use only one shunt type PWM switch (Fitzer et
al., 2004) for output voltage control as shown in Fig.
2. The autotransformer shown in Fig. 1 is used in the
proposed system to boost the input voltage instead of a two winding transformer.
Switch IGBT is on the primary side of the autotransformer. The voltage and current
distribution in the autotransformer is shown in Fig. 2. It
does not provide electrical isolation between primary side and secondary side
but has advantages of high efficiency with small volume. The compensator considered
is a shunt type as the control voltage developed is injected in shunt. The relationships
of the autotransformer voltage and current are expressed in Eq.
2:
Where:
a 
= 
The turns ratio 
V_{p} 
= 
Primary voltage 
V_{L} 
= 
Secondary voltage = Load voltage 
I_{1}, I_{2} 
= 
Primary and secondary currents, respectively 
I_{S} 
= 
Source current 
I_{L} 
= 
Load current 

A transformer with N1:N2 = 1:1 ratio is used as an autotransformer to boost the voltage on the load side when sag is detected. With this the device can mitigate up to 50% voltage sag. As the turns ratio equals 1:2 in autotransformer mode, the magnitude of the load current I_{L} (high voltage side) is the same as that of the primary current I_{1} (low voltage side). From Eq. 2, it is clear that V_{L} = 2V_{P} and I_{S} = 2I_{L}. The switch is located in the autotransformer’s primary side and the magnitude of the switch current equals the load current.
The voltage cross the switch in the offstate is equal to the magnitude of the input voltage. When sag is detected by the voltage controller, IGBT switched ON and is regulated by the PWM pulses. The primary voltage V_{P} is such that the load voltage on the secondary of autotransformer is the desired RMS voltage.
Ripple filter design: The output voltage V_{P} given by the
IGBT is the pulse containing fundamental component of 50 Hz and harmonics at
switching frequency. Hence, there is a necessity to design a suitable ripple
filter at the output of the IGBT to obtain the load voltage THD within the limits.
A combination of notch filter to remove the harmonics and a low pass filter
for the fundamental component is used. Capacitor C_{r1} in combination
with source inductance and leakage inductance form the low pass filter.
The notch filter is designed with a center frequency of PWM switching frequency
by using a series LC filter. A resistor may be added to limit the current. The
impedance of the filter is given by Eq. 3:
Where, R, L_{r} and C_{r2 }are the notch filter resistance, inductance and capacitances, respectively. The resonant frequency of the notch filter is tuned to the PWM switching frequency. The capacitor is designed by considering its kVA to be 25% of the system kVA. Capacitor value (C_{total}) thus obtained is divided into C_{r1} and C_{r2} equally. The notch filter designed for switching frequency resonance condition is capacitive in nature for frequencies less than its resonance frequency. Hence, at fundamental frequency, it is capacitive of value C_{r2 }and is in parallel with C_{r1 }resulting to C_{total}.
RESULTS AND DISCUSSION
Simulation analysis is performed on a threephase, 115/11 kV, 100 MVA, 50 Hz
system to study the performance of the PWM switched autotransformer in mitigating
the voltage sag. The SIMULINK model of the system used for analysis is shown
in Fig. 3. An RL load is considered as a sensitive load which
is to be supplied at constant voltage.

Fig. 3: 
MATLAB/SIMULINK model of a 3phase system used for voltage
sag studies 


Fig. 4: 
Model of 3phase PWM switched autotransformer 

Table 1: 
System parameter use for simulation 


Table 1 shows the system parameter specifications used for
simulation. Under normal condition, the power flow is through the antiparallel
SCRs and the gate pulses are inhibited to IGBT. The load voltage and current
are same as supply voltage and current. When a disturbance occurs, an error
voltage which is the difference between the reference RMS voltage and the load
RMS voltage is generated.

Fig. 5: 
The simulation waveforms of the load voltage for voltage
sag of 27% at 0.1 sec 

The PI controller thus gives the angle δ. Control voltage at fundamental
frequency (50 Hz) is generated and compared with the carrier frequency triangular
wave of carrier frequency 1.5 kHz.

Fig. 6 : 
Simulation waveform for per phase error voltage 1 in RMS value 


Fig. 7: 
Simulation waveform for per phase error voltage 2 in RMS
value 


Fig. 8: 
Simulation waveform for per phase error voltage 3 in RMS
value 


Fig. 9: 
Simulation waveform for voltage sag mitigated using pwm switched
autotransformer 

The PWM pulses now drive the IGBT switch. The simulation modeling of PWM switched
autotransformer used as mitigating device along with its control circuit is
shown in Fig. 4.
The autotransformer rating in each phase is 6.35/6.35 kV (as line voltage is
11 kV) with 1:1 turns ratio. The effective voltage available at the primary
of autotransformer is such that the load voltage is maintained at desired RMS
value (6.35 kV or 1 pu). Voltage sag is created during the simulation by sudden
application of heavy load of P = 10 MW and Q = 50 Mvar for a period of 0.1 sec
(5 cycles) from t_{1 }= 0.1 sec to t_{2} = 0.2 sec, Fig.
69 shows the simulation waveforms of the load voltage
for voltage sag of 27%.
CONCLUSION
A new voltage sag compensator based on PWM switched autotransformer has been presented in this study. Control circuit based on RMS voltage reference is discussed. The proposed technique could identify the disturbance and capable of mitigating the disturbance by maintaining the load voltage at desired magnitude within limits. The proposed technique is simple and only one IGBT switch per phase is required. Hence, the system is more simple and economical compared to commonly used DVR or STATCOM. Simulation analysis is performed for 27% voltage sag for three phase system and simulation results verify that the proposed device is effective in compensating the voltage sag disturbances.