Contents Introduction
With the advancement of power electronics and the emergence of energy efficiency programs in the world, there has been an increase in the use of electric and electronic equipment, introducing into the electrical system a large number of electronic loads with the purpose of making electrical efficiency, that is, replacement of conventional equipment by switched devices with the purpose of reducing the consumption of electric energy, therefore, studies to be carried out in an electrical system can no longer disregard the influence of such loads and their consequences. For example, when carrying out reactive power studies in a system, linear elements, capacitor banks and inductors have always been taken into consideration for reactive compensation and voltage level control. However, analyzing the behavior of certain electronic loads, using electric energy meters and electricity analyzers, a reactive flow can be observed that can be an injection or absorption of reactive power in the bus. [1]
This behavior of the switched electronic load presents a different characteristic of conventional data in which the concept of reactive power was born in direct connection with inductive and capacitive elements, moreover, the name “reactive” is related to the term “reactance”. For the loads of the time when this theory was developed, the idea that reactive power was related to the energy stored in the reactive elements was perfectly correct [2]. However, taking as an example the non-linear electronic load present in this work, it consists of a series electronic switch with a resistance, it is possible to obtain both types of inductive or capacitive reactive power flows, despite the physical absence of capacitors or inductors [3].
The reactive compensation through the electronic loads present in the system, impact on its performance and also on the billing of surplus reactive electric energy [4]. In Brazil, large consumers are penalized by surplus reactive energy charging, for example, in industries that operate with a significant number of induction electric motors with a power factor lower than the norm.
In this way, this paper has as objective to analyze the reactive power flux of electronic loads, isolated and in parallel, aiming the attenuation of the harmonics acting as filters by means of the self-compensation of the reactive flow, seen by the meters.
Considerations Regarding Reactive Energy Surplus in the Electrical System
Normative Resolution N°. 414 [5] of September 9, 2010, which establishes the general conditions for the supply of Electric Energy in an updated and consolidated manner, says that for consumers in group A, the reference power factor “fr”, inductive or capacitive, has the minimum allowed limit, the value of 0.92.
“To the amounts of electric energy and reactive power demand that exceed the allowed limit, the charges established in arts. 96 and 97, to be added to the regular billing of consumer units of group A, including those that opt for billing with application of the group B tariff pursuant to art. 100. ”(Normative Resolution No. 414, 2010, page 84).
Therefore, even though the amount to be paid refers to the Surplus Reactive, it is proportional to the amount of active electric energy measured in the time interval referring to the measurement and the relation of the reference power factor to the power factor of the consumer unit. [6]
For the calculation of these quantities, there is also a period of 6 (six) consecutive hours (at the power utility compa-ny's discretion) between 23:30 and 05:30 p.m., in which only the power factors below 0.92 capacitive are considered, checked at each interval of one hour. At the complementary daily period, only the power factors below 0.92 inductive, verified at the same time, are considered.
Since the nonlinear electronic load studied in this work is seen by the meter as an injector or consumer of reactive energy in the network, there is then a direct influence on the measurement of surplus reactive values described in the normative resolution, by absorbing or injecting reactive energy on the bus from the point of view of the meter, contributing to the change in the value of the power factor recorded every hour. This situation raises new discussions in the academic world about how relevant this influence will be, due to the considerable increase of electronic charges in residential, commercial and industrial buses with the purpose of achieving energy efficiency.
Considerations Regarding Active Filters
With the growth of the electronic loads inserted in the context of the electrical efficiency, the harmonics became a present part in the electrical system. This is due, in particular, to the switching of the electric current waveform in order to reduce its area and, consequently, to reduce the electrical energy consumption of the equipment. Nowadays, these low-energy devices are common in both industrial and small consumers, therefore, the presence of harmonics in the network has become inevitable.
In Brazil, the National Agency of Electric Energy-ANEEL, through the module 8 of the Procedures of Distribution of Electric Energy in the National Electrical System-PRODIST [7], defines limits for the harmonic voltage levels in the busbars of the national electric system. The harmonic currents are generated by the non-linear loads connected to the network, when these currents pass through network impedances, they generate harmonic voltages, that is, they generate a deformation of the supply voltage of the bus. To re-duce this deformation, it is proposed to use filters to reduce the distortion of the bus voltage and prevent a certain frequency from propagating in the network.
There are several options to connect a filter to the network, among the most common are active filters and passive filters. Passive filters are circuits that have the intention of filtering a certain frequency of the system through the association of resistors, capacitors and inductors, that is, through passive components. An active filter is a type of analog electronic filter, distinguished from the others by the use of one or more active components, which may provide some form of power amplification. Typically this component may be a transistor or an operational amplifier. Since these filters act directly in order to reduce harmonic distortion and improve the Electrical Power Quality of the system, it is possible to conclude that the use of both in a nonlinear system with high harmonic distortion level, is very important to guarantee sinusoidal voltages, or more close to it as possible.
Considerations Regarding Harmonic Dis-Tortions in Electrical Systems
For the basic energy network, since 2002, the National System Operator (ONS) has established quality parameters for the voltage supplied. But from the point of view of the consumer, the restrictions to be considered are (mostly) those of the distribution system.
ANEEL, through PRODIST Module 8 [7], proposes reference values for the harmonic distortion of the voltage in the distribution system, as shown in Table I.
The Dimmer Flex uses laboratory experiments to provide non-linear characteristics, which results in waveforms of currents with significant distortions. However, for the loads supply bus voltages, no significant harmonic distortions were observed, remaining within the limits recommended by the normative resolution established by ANEEL [8].
Laboratory Measurements
The equipment used to record the electrical power in the laboratory corresponds to the MARH-21 and Fluke 434 energy analyzers, both real-time quantity meters for single phase, two phase and three phase in low, medium and high voltage electrical systems.
For the study, two electronic devices were used in the laboratory switching incandescent bulbs, called here by Dimmer Flex, one of them is single phase and the other three phase, constituting a non-linear load. Figure 1 shows the pictures of Dimmers control circuits, without the presence of incandescent bulbs.
The experiment was divided into three cases. In the first one, using the single-phase electronic device, the process starts applying voltage without any impediment to the establishment of the current and, after approximately half of the positive half-cycle (90 degrees), the current is blocked and the same occurs for the negative cycle (270 degrees). In the second case, using the three-phase electronic device switching resistors, voltage was applied to the load and the start of the current was delayed around a half of the positive half-cycle of the voltage (90 degrees), the same occurred for the negative half-cycle (in 270 degrees). Finally, in the third case, the two electronic devices, with the same switching configuration dictated previously, were placed in parallel. Subsequently, measurements were taken to evaluate the behavior of the reactive power flux of the nonlinear electronic loads in the electric system bus, as well as the harmonic distortion present in each case. Studies were carried out on both the reactive compensation and the harmonic attenuation, which refers to the quality of the energy in the bus. Figure 2 shows a line diagram of the circuit assembled in the laboratory, in which switches Ch1 and Ch2 were responsible for connecting the circuits in parallel
A. Case 1-Application of Voltage in the Load with Current Locked at 90 and 270 Degrees
The waveforms of the voltage and current in the laboratory bus, obtained with the first non-linear load switching are shown in Figure 3, from the screen prints of the Fluke 434 electric energy analyzer, where it can be seen that the voltage waveform is practically sinusoidal.
Non linear load voltage and current waveforms measured in laboratory with fluke 434.
The harmonic distortion of the voltage waveform is within the levels acceptable by the norm, as can be seen in Figure 4.
Harmonic spectrum of voltage waveform measured in tlaboratory, obtained by the data of the analyser marh 21.
In the waveform of the current, there is the presence of negative and zero sequence current harmonics, as shown by the harmonic spectrum of the current shown in Figure 5.
Harmonic spectrum of current waveform measured in tlaboratory, obtained by the data of the analyser marh 21.
Figure 6 shows the phasor diagram of the fundamental voltage and current that feed the non-linear load obtained through the Fluke 434 power analyzer. It is noted that the current phasor of the voltage phasor is advanced.
Phasor diagram of voltage and current of the non-linear load measured in the laboratory.
Figure 7 shows that the flow of active and reactive powers seen by the meter on the non-linear load-fed bus features 130W of active power absorption and a 130VAr power injection of reactive power. Figure 7 also shows the difference between the shift factor, which represents the cosine of the voltage and current angle at 60 Hz (DPF), whose value is 0.86, and the power factor (PF) that leads in counts all frequencies present in the waveform of the current, with a value equal to 0, 72.
Characteristic of reactive power injected into the system by nonlinear load, measured in laboratory.
B. Case II-Application of the Voltage in the Load with Conduction of Current from 90 and 270 Degrees
Continuing the analysis of the energy quality behavior and the reactive flow of the switched electronic loads connected to a bus, the second case was analyzed, that is, the beginning of the establishment of the current was delayed in relation to the voltage in 90° of a Three-phase electronic device switching resistive loads.
Figure 8, 9 and 10 illustrate the voltage and current waveforms requested by the nonlinear electronic load through phases A, B and C, respectively.
Non-linear load voltage and current waveforms measured in the laboratory with the analyser fluke 434-phase A.
Non-linear load voltage and current waveforms measured in the laboratory with the analyser fluke 434-phase B.
Non-linear load voltage and current waveforms measured in the laboratory with the analyser fluke 434-phase C.
The voltage waveform did not present significant harmonic distortion, as can be seen in the harmonic spectrum of the voltage in phase A of Figure 11. Phases B and C presented the same behavior.
Harmonic spectrum of the voltage in phase A of the nonlinear load measured in the laboratory obtained by the marh 21 analyzer.
The current waveform presented a strong presence of harmonics of positive, negative and zero sequence, as illustrated in the harmonic spectrum of current of Figure 12, 13 and 14 of the phases A, B, and C, respectively.
Harmonic spectrum of the current in phase A of the nonlinear load measured in the laboratory obtained by the marh 21 analyzer.
Harmonic spectrum of the current in phase B of the nonlinear load measured in the laboratory obtained by the marh 21 analyzer.
Harmonic spectrum of the current in phase C of the nonlinear load measured in the laboratory obtained by the marh 21 analyzer.
In the phasor diagram of the fundamental voltage and current, shown in Figure 15, it is observed that the current phasor is delayed in relation to the voltage at all phases.
Phasor diagram of the voltage and current of the nonlinear load measured in the laboratory obtained from the fluke 434 analyzer.
Figure 16 shows that the flow of the active and reactive powers, seen by the bus measure that feeds the non-linear load, characterizes an absorption of 150W of active power and a reactive power absorption of 130VAr per phase. Figure 16 also shows the difference between the displacement factor, which represents the cosine of the voltage and current angle at 60 Hz (DPF), whose value is 0.87, and the power factor (PF) that leads taking into account all the frequencies present in the waveform of the current, this value being equal to 0.75.
Characteristic of reactive power absorbed from the system by non-linear load, measured in laboratory.
C. Case III-Electronic Devices in Parallel
Finally, the device of case A was placed in parallel as device of case B through phase A of the network. Figure 17 shows the waveform resulting from the current and voltage of phase A.
Voltage and current waveforms of the non-linear loads in parallel, measured in the laboratory with the fluke 434-phase a.
A small deformity is noted at the peak of the current waveform at the end of the first quarter of period and at the end of the third quarter of the period. This is due to the junction of the switches of the two electronic devices. Phases B and C presented the same behavior as in item B, as seen in figure 9 and 10, since they still have the same switching configuration as case A and B.
The waveform of the current in phase A, after the parallel connection of the electronic devices, showed a significant change in relation to the previous cases, as can be observed in figure 18.
Harmonic spectrum of the current in phase A of nonlinear loads in parallel, measured in laboratory by the marh 21 analyzer.
In the phasor diagram between the fundamental voltage and current shown in Figure 19, it is observed that the phasor of the current is in phase with that of the voltage in phase A, while phase B and C have the same expected characteristic.
Phasor diagram of the voltage and current of the nonlinear load measured in the laboratory obtained from the fluke 434 analyzer.
Figure 20 shows the evident difference of the active and reactive power flows seen by the meters from phase A where the devices are in parallel and in phases B and C where they are switched on in the second half of the positive half-cycle and in the second half of the negative half cycle. In the bus that feeds the non-linear load, in phase A we obtained an active absorbed power of 270 W and reactive of 30VAr. The phases B and C, presented an absorption of 140W of active power and a absorption of reactive power of 140VAr. Figure 20 also illustrates the shift factor, whose value is 1 for phase A and 0.86 for phases B and C, and the power factor (PF), which is equal to 0.99 for phase A and 0.74 and 0.75 for phases B and C.
Characteristics of active and reactive power absorbed from the system by non-linear load measured in laboratory.
D. Analysis of Results
Analyzing the results obtained in relation to the energy quality, it was observed that non-linear electronic loads in both case A and in case B, presented a total current harmonic distortion around 60%. This can lead to a significant increase in the true effective value of the current, due to the strong presence of positive, negative and zero sequence harmonics generated by electronic load. A significant increase in zero sequence harmonics can also lead to an increase in the neutral current in the bus where loads with these characteristics are connected. However, when analyzing case C, where the two electronic devices with opposing switches were placed in parallel on the bus, the total harmonic distortion of current was reduced to 10%. The harmonic attenuation of the most significant orders is observed, as seen in figure 18, verifying the actuation of the switching device as an active filter. Thus, it can be concluded that in a bus where there is a device with a 0° angle of firing and an extinction angle of 90° for the first half-cycle of the waveform in parallel with another with 90° firing angle and extinction angle of 180° in the first half cycle, the harmonics generated by one device can be attenuated by the other, as verified in the analyzers and electric energy registers used in the present work.
When the reactive power absorption and injection ratio is observed in the evaluated cases, it can be verified that the reactive power injected into the bus in case A, which generated a power factor of 0.72 capacitive and the reactive power absorbed of the bus in the case case B, whose power factor was 0.75 inductive, were compensated when they were put in parallel, generating a power factor of 0.99 inductive. Thus, from the meter's point of view, there was a self-compensating power factor of the bus.
Conclusion
In this paper, the influence of the nonlinear electronic loads with two types of switches was analyzed, both in about the energy quality, and about the compensation of the reactive flow in the bus. In the laboratory experiment two electronic devices were used, each one switching a resistor constituting a set with strong nonlinear characteristics. Through the data recorded in the energy analyzers it was possible to observe a reactive power flow of the power bus, depending on the type of switch used. The nonlinearity of the electronic load provided a high current distortion, with the presence of negative and positive sequence harmonics, which could increase the effective value of the currents in the phases of the bus and in the neutral conductor, by means of the harmonics of zero sequence. However, by placing the two devices in parallel, a self-compensation of these harmonic orders was observed with an expressive decrease of the current THD seen by the meters. Therefore, if electronic loads with current switching are placed in a bus of an electric system in relation to the half-cycles, they can provide an automatic attenuation of the harmonic distortions present in the network. The injection and the absorption of reactive power in the bus by the electronic devices studied here can also contribute to the improvement of the power factor of the bus, that is, the reactive injected by a type of switching can compensate the reactive absorbed by the device with another type of switching, according to the meters used. It is important to remember that the characteristics of non-linear electronic loads used in this study can be found in several electronic loads on the market, recommended to consumers of groups A and B, with the aim of contributing to a greater efficiency, with regard to active electric power consumption. Finally, there is a need for more in-depth studies on the impact of the injection of reactivate the electronic loads in Consumer Units of group A and B.