CHAPTER 4 PV-UPQC BASED HARMONICS REDUCTION IN POWER DISTRIBUTION SYSTEMS

66 CHAPTER 4 PV-UPQC BASED HARMONICS REDUCTION IN POWER DISTRIBUTION SYSTEMS INTRODUCTION The use of electronic controllers in the electric power supply system has become very common. These electronic controllers behave as nonlinear load and cause serious distortion in the distribution system and introduce unwanted harmonics in the system, leading to decreased efficiency of the power system network and equipment connected in the network. To meet the requirements of harmonic reduction, passive and active power filters are being used in combination with the conventional converters. Presently, active power filters are becoming more affordable due to cost reductions in power semiconductor devices, their auxiliary parts and integrated digital control circuits. Resent research focuses on use of the UPQC to compensate for power-quality problems. The performance of UPQC mainly depends upon how accurately and quickly reference signals are derived. After efficient extraction of the distorted signal, a suitable dc-link current regulator is used to derive the actual reference signals. Various control approaches, such as the PI, sliding-mode, predictive, unified constant frequency (UCF) controllers are in use. Modern control theory-based controllers are state feedback controllers, self-tuning controllers and model reference adaptive controllers. These controllers also need mathematical models and are therefore sensitive to parameter variations. In recent years, a major effort has been underway to develop new and unconventional control techniques that can often augment or replace conventional control techniques. A number of unconventional control techniques have evolved, offering

67 solutions to many difficult control problems in industry and manufacturing sectors. However, at the same time, these advanced power electronics systems also help in reducing the harmonics flowing in the power systems. The THD value is the effective value of all the harmonics current added together compared with the value of the fundamental current, as discussed by John et al (2001). When excessive harmonic voltage and current are generated, filters are usually installed to reduce the harmonic distortion Fang et al (1990) and Helga et al (2004). Other methods of harmonic reduction is considered such as current injected by active power filter, discussed by Sangsun et al (2001) and Ambra et al (2003). Ideally, voltage and current waveforms are pure sinusoids. However, because of the increased usage of nonlinear loads, these waveforms become distorted. This deviation from a pure sine wave can be represented by harmonic components having a frequency that is an integral multiple of the fundamental frequency. Thus, a pure voltage or current sine wave has no distortion and no harmonics. In order to quantify the distortion, the term of total harmonics distortion is used. The inverter switches are connected and disconnected at discrete time instant to generate desired output voltage of specific magnitude and frequency. Such an output voltage in addition to its fundamental component contains a lot of other undesired harmonic components as well. Hamadi et al (2004) has discussed the minimization of the harmonic injection from an inverter into the power system. But, his work did not meet the IEEE Standard 519-1992 since the THD value high. IEEE recommended practices and requirements for harmonic control in electrical power systems address harmonics limits at the consumer and service provider interface. This standard provides procedure for controlling harmonics on the power system along with recommended limits for customer harmonics

68 injection and overall power system harmonics levels to ensure overall system voltage integrity. The utilities are responsible for maintaining the quality of voltage on the overall system and for the voltage distortion at the PCC. Excessive harmonic levels in voltage or current in the utility system can result in increased equipment heating, equipment malfunction and premature equipment failure, communication interference, fuse blowing in capacitor banks and customer equipment and process problems. Rizy et al (2003), has discussed about implementation of ANN control strategy. But unfortunately, the methodology adopted by them has some serious drawbacks. ANNs may not be suitable for a large distribution system since many smaller subsystems are required and the training time becomes excessive. However, once networks are trained, iterative calculations are no longer required and a fast solution for a given set of inputs can be provided. Suresh Mikkili et al (2011) has discussed about implementation of fuzzy control optimization approach. It is very simple and naturally fast as compared to other optimization methods. However, for some problems the procedure might get trapped in a local optimal point and fail to converge to the global (or near global) optimal solution. In order to sustain the constant frequency in the utility, utilized the Fuzzy Logic Controller based constant frequency UPFC. A Constant Frequency (CF) UPQC is composed of a UPQC and a matrix converter based frequency changer. UPQC is a combination of series active and shunt active filter. The series active filter and shunt active filter have been employed to compensate the voltage, current imbalance and harmonics. The Frequency Converter (matrix converter) has been used to control the supply frequency when it exceeds the power quality limit. Therefore, in order to overcome the flaws in the above mentioned control, hysteresis control strategies are used for better performance. When

69 the distortion levels on the utility system cause a problem, the proposed mitigation measures need to be implemented. In order to reduce the harmonics injected into power system by PV-UPQC, harmonics reduction technique is developed. 4.1 PROCEDURES FOR HARMONICS REDUCTION USING UPQC WITH DIFFERENT CONTROLLERS The shunt inverter filter using ANN current controller as discussed by Jeno Paul et al (2011), is essentially a cluster of suitably interconnected nonlinear elements of very simple form that possess the ability of learning and adaptation. These networks are characterised by their topology, the way in which they communicate with their environment, the manner in which they are trained and their ability to process information. Their ease of use, inherent reliability and fault tolerance has made ANNs a viable medium for control. An alternative to fuzzy controllers in many cases, neural controllers share the need to replace hard controllers with intelligent controllers in order to increase control quality. A feed forward neural network works as compensation signal generator. This network is designed with three layers. The input layer with seven neurons, the hidden layer with 21 neurons and the output layer with 3 neurons. Activation functions chosen are tan sigmoidal and pure linear in the hidden and output layers respectively. The training algorithm used is Levenberg Marquardt Back Propagation (LMBP). The compensator output depends on input and its evolution. The chosen configuration has seven inputs, three each for reference load voltage and source current respectively and one for output of error PI controller. The neural network trained for giving fundamental reference currents output. The signals thus obtained are compared in a

70 current controller to give switching signals. The block diagram of ANN compensator is shown in Figure 4.1. Source currents Figure 4.1 Block diagram of ANN-based compensator 4.1.1 Fuzzy Controller Rama Rao et al (2010), has discussed the implementation of fuzzy control optimization approach. Figure 4.2 shows the block diagram of the fuzzy logic control scheme. The control scheme consists of fuzzy controller, limiter and three phase sine wave generator for reference current generation and generation of switching signals. The peak value of reference currents is estimated by regulating the DC link voltage. In order to implement the control algorithm of a shunt active power filter in a closed loop, the dc capacitor voltage V DC is sensed and then compared with the desired reference value V DC-ref. The error signal V DC-error = V DC-ref V DC is passed through Butterworth design based LPF with a cut off frequency of 50 Hz, that pass only the fundamental component. The error signal e(n) and integration of error signal Ie(n) are used as inputs for fuzzy processing. The output of the fuzzy logic controller limits the magnitude of peak reference current I max. This current takes care of the active power demand of the non-linear load and losses in the distribution system. The switching signals for the PWM inverter are generated by comparing the actual source

71 currents (i sa,i sb,i sc ) with the reference current (i sa *,i sb *,i sc *) The output of the fuzzy controller is estimating the magnitude of peak reference current I max. This current I max comprises of active power demand of the non-linear load and losses in the distribution system. The peak reference current is multiplied with PLL output for determining the desired reference current. (a) (b) Figure 4.2 (a) and (b) Control block diagram of fuzzy controller 4.1.2 Proposed PI with Hysteresis Controller The proposed control strategy, described in section 3.3.1 and shown in figure 3.3, ensures fast elimination of higher order current harmonics of the load. Hysteresis controller is designed for controlling the switching of

72 the shunt inverter. Based on the active power demand of the load, a suitable sinusoidal reference is selected for the incoming utility current and in addition, appropriate hysteresis band is selected. Narrower hysteresis band ensures higher THD elimination, at the cost of higher switching frequency of the inverter. Suitable trade off in design is required to optimize all criteria. The constant bandwidth hysteresis current control technique is widely used in voltage-source grid connected inverters. The hysteresis current control block diagram is shown in figure 4.3 The measured phase current (i A ) is subtracted from reference current (i Aref ), and current error (i error ) is obtained. This error is compared with hysteresis band and switching pulses are generated. The bandwidth value (h) is constant and hysteresis control restricts the current in the band. (a) (b) Figure 4.3 (a) and (b) Hysteresis current control and waveforms

73 Hysteresis current controller compares the current error with lower and upper hysteresis band. For phase- A, if i Aerror > +h, upper switch is ON, and i A increases. If i Aerror < -h, upper switch is OFF. As discussed earlier, the dc link voltage ideally should not decay, unless some active power loss occurs in the PV-UPQC. Therefore, the deviation of the DC link voltage acts as a measure of active power requirement from utility supply. The error is processed through a PI controller and a suitable sinusoidal reference signal in phase with the supply voltage is multiplied with the output of the PI controller, to generate the reference current for the supply. Hysteresis band is imposed on top and bottom of this reference current. The width of the hysteresis band is adjusted such that the supply current THD remains within that specified by the standards. As the supply current hits the upper or lower band, appropriate switching of the shunt inverter takes place so as to compel the supply current to remain within the band, by either aiding its dc link voltage to utility supply. 4.1.3 Simulation Parameter The three-phase PV- UPQC system simulation model is described in section 3.8 and shown in figure 3.6. In the simulation model, the following data parameters are used for system simulation. The source voltage is 230V in rms (325 peak voltage). A linear load of 6 KW and 4.5 KVAR is connected to the system, in addition to a nonlinear load of 5 KW and 3.75 KVAR. The DC link capacitor of 2000 F is connected between two inverters. 1.245 mh is the value of both shunt and series interface inductors. The filter capacitors of the series and shunt branches are 140 F and 20 F, respectively. A 4 damping resistor is connected in series with the shunt filter capacitor. Coupling transformers are used to connect the series and shunt active filters to the PCC.

74 4.2 RESULTS AND DISCUSSION The three-phase PV- UPQC system simulation model is described in section 3.8 and is shown in figure 3.6. The results obtained using the simulation model is presented in this section. The PV-UPQC model is simulated with following case studied: Unbalanced supply voltage and nonlinear loads with sag Unbalanced supply voltage and liner and nonlinear loads without sag Balanced supply voltage and liner and nonlinear loads with sag 4.2.1 Unbalanced Supply Voltage and Nonlinear Loads with Sag The supply voltages are unbalanced, distorted and undergoing 30% sag as shown in figure 4.4(a). The nonlinear load is connected in this case to study the PV-UPQC steady-state performance. It is investigated and the simulation results are illustrated in figures 4.4(a) to 4.4(f) and figures 4.5(a) to 4.5(d). Case Study 1: Current 400 300 No Sag Sag No Sag 200 100 0-100 -200-300 -400 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Time(s) Figure 4.4(a) Supply voltage with sag

75 Figure 4.4(b) Load current Figure 4.4(c) Load current (THD=14.3%) Figure 4.4 (d) Current injected by shunt active filter

76 Figure 4.4 (e) Supply current (THD=1.15%) Figure 4.4(f) Phase A Supply, Load and Injected currents Figure 4.4 (c) shows the load currents, which are highly distorted. The THD is 14.3%, whereas according to IEEE Standard 519-1992 they should not exceed 5% in our case. The effects of the injection of compensating currents are shown in figure 4.4 (d). Figure 4.4(e) shows the harmonics minimization sinusoidal wave (THD=1.15%), which fulfills the 5% stipulated by the IEEE standard. From figure 4.5 (e), the supply current

77 is in-phase with the load voltage, which is due to the in-phase series voltage injection. The injection is in-phase with the fundamental of the supply voltage. Thus, no reactive power is drawn from the supply. In figure 4.4 (f), the supply current, load current and injected current of are plotted together. If the injected current wave is subtracted from the load current wave, the result will be the supply current wave. Case Study 1: voltage Figure 4.5 (a) Supply voltages (THD=7.7%) Figure 4.5 (b) Voltages injected by series active filter

78 Figure 4.5 (c) Load voltage (THD=0.7%) Figure 4.5 (d) DC link voltages Figure 4.5 (e) Phase A load voltage and supply current

79 Figure 4.5(f) Phase A supply voltages, injected and load voltages From figure 4.5 (a), the supply voltages are unbalanced, distorted (THD=7.7%) and undergoing a sag. The series compensator injects voltages as shown in figure 4.5 (b). Due to this series compensation, the voltages on the load side are balanced sinusoidal. Figure 4.5(c) shows THD of 0.7%, which is much below the 5% limit stipulated by the IEEE Standard. It can be seen from figure 4.5 (c) that the load voltages are kept at the nominal level of 325 volts. The load and injected voltages for phase A are plotted together. The sum of the injected voltage and the supply voltage is the load voltage. In figure 4.5 (d), the concept of dc link voltage has been illustrated. The shunt compensator is connected at 0.02s which causes a drop in dc link voltage. After the sag from 0.02 s to 0.5s, the dc link voltage is restored. From figure 4.5 (d), it is observed that the average dc link voltage is kept constant at 400 V.

80 Case Study 2 4.2.2 Linear and nonlinear loads without sag The supply voltages, as shown in figure 4.6(a), are unbalanced and distorted (the 5th harmonic is present) but there is no sag this time. The nonlinear load is permanently connected and the linear load is connected at 0.2s and disconnected at 0.5s. In this case study, the PV-UPQC dynamic performance at load change is investigated and the simulation results are illustrated through the figures 4.6 to 4.10. Case Study 2: load currents Figure 4.6 (a) Load current dynamics

81 Figure 4.6 (b) Load current with nonlinear load (THD =14.4%) Figure 4.6 (c) Load current with Nonlinear and linear loads (THD =7.9%)

82 Figure 4.6 (d) Load current with Nonlinear load (THD = 14.4%) The simulated duration is composed of three sub-intervals: 0 s to 0.2 s, only with the nonlinear load connected to the system 0.2 s to 0.5 s, with both the linear and nonlinear loads connected to the system 0.5 s to 0.8 s, only with the nonlinear load is connected to the system For clarity, the load current waveforms, shown in figure 4.6(a), for each of these three subintervals is zoomed in figures 4.6 (b) to 4.6 (d). The injected and supply currents are shown in figures 4.7 and 4.8. The load current changes are reflected in both injected and supply currents. A change in the load causes the system to go through a transient period after which the supply currents become balanced sinusoids, as shown in figures 4.8(b) to 4.8(d), in phase with the load voltage as shown in figure 4.10. Thus, the PV-UPQC adjusts the injected current according to the load condition, ensuring that, in steady-state, the supply currents are always balanced sinusoids and no reactive power is drawn from the supply.

83 25 Linear and Nonlinear loads 20 15 10 5 0-5 -10-15 -20 Nonlinear load Nonlinear load -25 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Time(s) Figure 4.7 (a) Injected current dynamics Figure 4.7 (b) Injected current with nonlinear load

84 Figure 4.7(c) Injected current with linear and nonlinear loads Figure4.7 (d) Injected current for nonlinear load Figure 4.8(a) Supply current with dynamics

85 Figure 4.8 (b) Source current with nonlinear load (THD =2.4%) Figure 4.8 (c) Source current with nonlinear and linear loads (THD =1.3%) Figure 4.8(d) Source current with nonlinear load (THD=2.4%)

86 During the entire simulation period (from 0s to 0.8 s), the supply voltages are unbalanced and distorted (THD=5.05%), as shown in figure 4.9 (a). The injected voltage by the series compensator is shown in figure 4.9 (b). Due to this series compensation, the voltages on the load side are balanced sinusoidal waveforms, as shown in figure 4.9(c), with THD 1%, which is much lower than the 5% limit recommended by IEEE Standard 519-1992. From figure 4.9(c), it can be seen that the load voltages are kept at the nominal level (325 V). Figure 4.9 (a) Supply voltage (THD = 5.05%) Figure 4.9 (b) Inverter injected voltage

87 Figure 4.9 (c) Load voltage (THD=1%) Figure 4.9 (d) DC link Voltage The DC link voltage dynamics is shown in figure 4.6(d).The load current increases at 0.2s and decreases at 0.5s. This is due to the connection or disconnection of the linear load. Since it takes a finite time interval to calculate the new reference current, the shunt compensator cannot immediately respond to the load change. Some settling time is required to stabilize the controlled parameter around its reference. During transient periods, the DC link capacitor is supplying active power in order to ensure the power balance.

88 Figure 4.10 (a) Load voltage and load current with nonlinear load Figure 4.10(b) Load voltage and load current with nonlinear and linear loads Due to this transient supply of active power, the dc link voltage decreases during the sag at the time interval of 0.2s 0.4s and the swell in the voltage occurs during the interval 0.5s 0.7s. After clearing the

89 interruptions, as illustrated through the figure 4.6 (d), the dc link voltage is restored back to its reference value (400 V). The shunt compensator is controlled in such a way that the supply current is in-phase with the fundamental of the supply voltage, which means that no reactive power is drawn from the supply. The fundamental of the series injected voltage is in-phase with the supply current, as shown in figure 4.10. Case study 3 4. 2.3 Balanced linear and nonlinear loads with sag During the time interval from 0s to 0.8 s, both the linear and nonlinear loads are connected with 30% of supply voltage sag, which is generated at 0.2 s and cleared at 0.5 s. During the same time, the supply voltage becomes balanced and distorted (the 5th harmonic is present). The supply voltages are as shown in figure 4.11(a) and figure 4.11 (e) Supply current during the sag with fuzzy logic control. In this case study, the dynamic performance of the PV-UPQC system, at the occurrence or clearance of the supply voltage sag, is investigated and the simulation results are illustrated through figures 4.11 to 4.15. The supply current is increased considerably during the sag, as shown in figure 4.12. During the sag, the load should draw the same amount of power from the supply as it does in the normal condition (nominal power). Since the supply voltage is decreased during the sag, the supply current should be accordingly increased in order to provide the same nominal power to the load.

90 Figure 4.11 (a) Supply current with dynamics Figure 4.11 (b) Supply current before sag (THD=3.8%) Figure 4.11(c) Supply current during the sag (THD=1.1%)

91 Figure 4.11 (d) Supply current after the clearance sag (THD =3.8%) Figure 4.11 (e) Supply current during the sag with fuzzy logic control (THD=1.2%) The supply voltage dynamics are shown in figure 4.12(a). During the sag condition from 0.2 s to 0.5s, the supply voltages undergo 60% sag. They are unbalanced and distorted with THD of 7.7%, as shown in figure 4.12 (c). During the other two sub-intervals from 0s to 0.2s and from 0.5s to 0.8s, the supply voltages are balanced sinusoidal waveforms at nominal level (325V), as shown in the figures 4.12 and 4.13.

92 400 300 No Sag Sag No Sag 200 100 0-100 -200-300 -400 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Time(s) Figure 4.12 (a) Supply voltage with dynamics Figure 4.12 (b) Supply voltage before sag (THD=0.55%)

93 Figure 4.12 (c) Supply voltage during sag (THD=7.7%) Figure 4.12 (d) Load voltage after the clearance of sag (THD=0.55%) Figure 4.13 (a) Voltages injected during sag

94 Figure 4.13(b) DC link voltage during sag 400 300 No sag Sag No sag 200 100 0-100 -200-300 -400 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Time(s) Figure 4.14 (a) Load voltage with dynamics Figure 4.14 (b) Load voltage before sag (THD=1.4%)

95 Figure 4.14 (c) Load voltage during sag (THD=0.75%) Figure 4.14 (d) Load voltage sag with ANN controller (THD=1.2%) During the supply voltage sag, the series compensator injects voltages, as shown in figure 4.14(a). Due to this series compensation, the voltages on the load side are balanced sinusoidal waveforms with THD of 0.75%, which is much lower than the 5% limit recommended by IEEE Standard 519-1992, as shown in figure 4.14(c) and figure 4.14 (d) shows the load voltage sag with ANN controller. Figure 4.15 shows that the load voltages are kept at the nominal voltage throughout the simulation time interval.

96 The DC link voltage dynamics are shown in figure 4.13 (b). Due to the occurrence of the supply voltage sag at 0.2 s, the supply current is increased. With the clearance of sag at 0.5s, the supply current decreased. Since it takes a finite time interval to calculate the new reference current, the shunt compensator cannot immediately respond to this supply current change. Some settling time is required to stabilize the controlled parameter around its reference. Consequently, after the occurrence or clearance of the sag instants (0.2s and 0.5s), there exist some transient periods during which the DC link capacitor is supplying active power in order to ensure the balanced power. Due to this transient supply of active power, the DC link voltage is undergoing sag during the time interval 0.2 s 0.4 s and a swell during the interval 0.5 s 0.7 s. From figure 4.14 (d), it can be observed that after clearing the transients, the DC link voltage is restored back to its reference value (400 V). Figure 4.15 (a) without sag the voltage and current

97 Figure 4.15 (b) during sag the voltage and current Figure 4.15(c) after clearance of the sag voltage and current Due to unity power factor compensation throughout the entire simulated period, the supply current is in phase with the fundamental of the supply voltage, as shown in figure 4.15, which implies that no reactive power is drawn from the supply. A complete summary of the results with and without the proposed system is given in table 4.1 and in figures 4.16 to 4.21.

98 Table 4.1 Comparative percentage THD voltage and current with and without PV-UPQC system. Case study No s Events Without PV- UPQC Current Voltage %THD %THD With PV-UPQC Current Voltage %THD %THD 1 2 3 Unbalanced supply voltage nonlinear loads with sag Unbalanced supply voltage linear and nonlinear loads without sag Balanced supply voltage linear and nonlinear loads with sag 14.3 7.7 1.15 0.7 14.4 5.05 1.3 1 14.18 7.7 1.1 0.55

99 Figure 4.16 Percentage THD with and without PV-UPQC for nonlinear loads with sag and unbalanced supply voltage Figure 4.17 Percentage THD with and without PV-UPQC for unbalanced supply voltage linear without sag and nonlinear loads

100 Figure 4.18 Percentage THD with and without PV-UPQC for balanced supply voltage with sag and linear and nonlinear loads Figure 4.19 Percentage THD with and without PV-UPQC for unbalanced supply voltage with sag and nonlinear loads

101 Figure 4.20 Percentage THD with and without PV-UPQC for unbalanced supply voltage without sag with liner and nonlinear loads Figure 4.21 Percentage THD with and without PV-UPQC for balanced supply voltage with sag and nonlinear loads

102 Table 4.2 THD values of load current with fuzzy logic control Rama Rao et al (2010) and proposed PI with hysteresis control THD load current without controller Fuzzy logic control THD load currents Proposed PI with hysteresis control THD load currents 14.3% 1.2% 1.1% Figure 4.22 Percentage THD for fuzzy logic control and proposed hysteresis control

103 Table 4.3 THD value of load voltage Artificial Neural Networks (ANN) Jeno Paul et al (2011) and proposed PI with hysteresis control THD for load voltage without controller ANN control THD load voltage PI with Hysteresis control THD load voltage 7.7% 1.2% 0.55% Figure 4.23 Percentage THD for ANN control and proposed hysteresis control The performance evaluation of PV-UPQC system associated with linear and nonlinear loads with sag using hysteresis control, in terms of the THD of the load current, is given in table 4.2. Comparing with the results of THD of 1.2% in the load current obtained using an ANN control by Jeno Paul et al (2011), the proposed hysteresis control has achieved a lower THD of 1.1%, which is shown in figure 4.22. Comparing with the results of THD of 1.2% in the load voltage obtained using a fuzzy logic control by Rama Rao et al (2010), the proposed hysteresis control has achieved a lower THD of 0.55% as shown in table 4.3 and figure 4.23. The analysis shows that the proposed

104 approach is more efficient with respect to ANN and fuzzy logic control strategies. On the basis of the research work presented in this chapter, a paper entitled Photovoltaic based improved power quality using unified power quality conditioner has been published in the International Journal of Electrical Engineering Vol.4, No.2, pp.227-242, July 2011. On the basis of the research work presented in this chapter, a paper entitled Power Quality Improvement using Photovoltaic Compensation Techniques has been published in the International Journal of Power Engineering Vol.2, No.1, pp. 33-46, June 2010. 4.3 CONCLUSION In this chapter, three significant case studies of an unbalanced system with linear and nonlinear loads are presented. The first case is supply voltage sag with nonlinear loads. The second case is linear and nonlinear loads without sag. The third case is linear and nonlinear loads with sag. The simulation results shows that the linear and nonlinear loads are connected the voltage becomes unacceptably distorted due to the switching frequencies in the supply current. For comparative analysis of the proposed PI with hysteresis control, the fuzzy logic and ANN control strategies of other researchers have been considered. Among various control techniques, the proposed hysteresis current control is the most preferable technique for shunt compensation. The hysteresis control method has simpler implementation, enhanced system stability and fast response. The proposed approach ensures that the unwanted harmonics are reduced compared to the other control strategies.