A Case Study of Sharing the Harmonic Voltage Distortion Responsibility between the Utility and the Consumer

Published by F. H. Costa, I. N. Santos, S. F. P. Silva and J. C. de Oliveira, Group of Power Quality Faculty of Electrical Engineering Federal University of Uberlandia Campus Santa Monica – Av. João Naves de Ávila, 2100 Uberlandia (Brazil). Phone/Fax number:+55 (34) 3239-4733, e-mail: fernandahein@hotmail.com, ivan.ufu@gmail.com, sergio@qes.com.br, jcoliveira@ufu.br

Abstract. The aim of this paper is to apply a methodology towards the sharing of responsibility between the utility and the consumer with reference to the occurrence of harmonic voltage distortions at the point of common coupling (PCC). The approach is based on the measured values of harmonic voltage and current as well as the supply and load harmonic impedance information. In addition to the general method principles, the paper focuses a specific application involving a real industrial installation, fed by 230 kV and having a large amount of rectifiers. The results are then used to verify the proposal consistency regarding the sharing of the responsibilities between the utility and the industry as far as the harmonic voltage distortion is concerned. The proposed process finds sustenance during the implementation of mitigation procedures with sights to the attendance of the standards of quality established by the regulating agencies.

Key words: Harmonic distortion, power quality, sharing harmonic responsibility, load modeling.

1. Introduction

Due to the harmful character of the harmonic distortions, standards and recommendations establish guidelines for the definition of limits for these distortions and other power quality indexes. The IEEE Standard 519 [1], EN 50160 [2] and IEC 61000-3-6 [3] are examples of documents covering such matter.

If the harmonic voltage distortion exceeds the allowed limits, mitigation procedures must be considered. The application of these procedures may cause great conflicts between the utility and the consumer due to the fact that high investments and costs are often involved. These difficulties are due, mainly, to the knowledge absence of the individual source and load contribution for the voltage distortions. In such a way, the search of technical and scientific methods to reach the trustworthy to quantify the parcels of responsibility between the parts involved would be most relevant. At the moment a few references may be found tackling this matter. Some of them are base on:

• Principles involving load modeling under distorted conditions [4]-[7];
• Harmonic active power flow [8];
• Conforming and non-conforming current components [9] and;
• Superposition principles [10

In general, such works attempt to find the main source of the distortions without worrying about the establishment of procedures toward the identification of the individual parcels of responsibility. Recognizing this limitation, this paper attempts an approach, based on site measurements and system information, that gives, at the end, the individual contribution of the utility and the consumer responsibility upon a given harmonic voltage distortion.

The general idea is based on the classic concepts of electric circuits and superposition principles. In addition to the methodology itself, a case study, using a real electric system supplying industrial installation containing a large number of rectifier units is considered. The results are given to highlight the approach utilization and method physical consistence.

2. Theoretical Fundaments

Using frequency domain techniques, it is possible to represent the utility and the consumer connected to the PCC by an equivalent Norton circuit, as given in Fig. 1.

Each harmonic order is represented by “h”. Both the representative utility and the combined linear and nonlinear loads of the consumer are shown. A more detailed arrangement is also shown in Fig. 1. The individual source and load contributions, as well as the measured variables are highlighted.

Fig.1. Frequency domain Norton equivalent circuit for a generic harmonic order “h”

In the figure:

Żhc – Consumer equivalent impedance at order h;
Żhu – Utility equivalent impedance at order h;
l̇hc – Harmonic current produced by the consumer;
l̇hu – Harmonic current injected by the utility;
V̇hm – Harmonic voltage measured at the PCC;
l̇hm – Harmonic current measured at the PCC.

Equation (1), derived from the equivalent circuit and superposition principles, gives the harmonic current generated by the nonlinear load in terms of the harmonic voltage and current measured at the PCC and the equivalent load impedance. The individual values for the harmonic voltage and current, extracted from site measurements are obtained in a similar way as given in [5], [6] and [10].

.

The harmonic current component at the PCC produced solely by the consumer nonlinear load (l̇hpcc-c) and injected in the mains is given by:

.

In these terms, the harmonic voltage, associated exclusively with the nonlinear effect of the consumer, can be determined by:

.

In a similar way to the calculations of consumer contribution, harmonic current and voltage parcels associates with the utility only are given by (4) and (5), respectively.

.

The previous equations show that, the resultant harmonic voltage originating from the consumer nonlinear load depends on the values of the impedances of the consumer and the utility. Consequently, to determine the contribution of each part, it is essential the knowledge of these harmonic impedances. To fulfill this requirement, the utility provides information related to its own harmonic impedances, thus, for may application this can be taken as a known parameter. On the other hand, it is important to observe that little or almost nothing is known about the load equivalent harmonic impedance. This guides for the necessity of the development of a strategy to the obtainment of such information. This is further discussed in the following section.

3. Consumer Load Modeling

The representation of the equivalent harmonic load impedance has motivated researchers to investigations attained to the representation of the equivalent consumer harmonic impedances. This is the case of [5] – [7]. Such references propose a parallel association of the basic elements: resistance, capacitor and inductor. Using the principles, reference [11] estimates these components based on site measurements and time domain computational techniques. This approach has been used in this paper for the necessary consumer load modeling.

Following this strategy, Fig. 2 illustrates the harmonic distorted voltage source; the load injected harmonic current and the load equivalent parameters (R, L and C). These later variables are to be calculated bellow.

Fig.2. Load equivalent circuit

The previous figure can be reorganized, generating the Fig.3.

Fig.3. Reorganized equivalent circuit.

The equivalent resistance is calculated through the total active power associated with the consumer operation and measured at the PCC. Thus, it must be pointed out that such power represents the sum of all harmonic active powers, considering only the positive ones, therefore, the ones driven by the load. In these terms:

.

Where:

P – Total active power at the PCC;
Ph – Harmonic active power at order h;
Φh -Phase angle between the harmonic voltage and current at h order.

Using the above equation and the measured rms voltage at the PCC, the load equivalent resistance can be calculated by:

.

The value of R is considered constant for the entire harmonic spectrum this hypothesis neglects the well known skin effect. Once the resistance has been found, it becomes possible to determine the individual harmonic currents flowing through the resistance branch. This current, for each harmonic order under analysis, is given by (8). It is important to remind that, this procedure must be repeated for all individual frequencies involved in the process.

By subtracting, for each frequency, the above current from the corresponding measured value, the result can be readily attributed to the combination of the three remaining components of current: the inductor (L), the capacitor (C) and the nonlinear load harmonic generation. Equation (9) expresses this relationship.

.

Where:

l̇hLCK– Harmonic current attributed to the combination of the inductor, the capacitor and the nonlinear load harmonic generation;

With this new current, the capacitance (C) can be calculated by (10). It must be detached that only the reactive powers with negative signals are considered, since the target is the capacitive element. The fundamental frequency is represented by f.

.

Once the equivalent capacitance is known, the harmonic current ( l̇hCap) can easily be determined for the distinct frequencies. Again, by subtracting this current from l̇hLCK, the result ( l̇hLK ) consists of the current associated with the inductive equivalent added with the current injected by the nonlinear load.

One again, in accordance with the previously used principles, the equivalent inductance is calculated by (11), that evidences the exclusive use of the positive values for the reactive harmonic powers.

.

By knowing the values of each harmonic current at the resistance, the capacitor and the inductor, the residual harmonic current will be that associated to the nonlinear load.

By knowing the values of each harmonic current at the resistance, the capacitor and the inductor, the residual harmonic current will be that associated to the nonlinear load.

Therefore, at this stage, in addition to the harmonic current sources, the values of linear components R, L and C are also known for all the harmonic orders under analysis. This allows the calculation of the consumer equivalent impedance for each individual frequency and, consequently, its use in the expressions (2) to (5).

4. Experimental Results

With the intention of investigating the performance of the methodology, the approach was applied to a practical situation involving an industrial installation with a large amount of rectifier load. This arrangement, in its simplified forma is shown in Fig.4 and the focused busbar corresponds to the 230 kV one. A PQ instrument was then installed at this PCC to obtain the required information as defined by the methodology. Due to the strong load behavior related to the industrial process, the equipment was configured to measure voltage and current harmonic distortions during long periods. The measurement device is a commercial product named RMS – MARH 21, capable of reading three-phase voltages and currents and calculating harmonics up to the 40th order.

Fig.4. Single line diagram of the industrial system

Although a longer period of time has been utilized, a sample of the phase to neutral voltage THD profile, over an interval of 5 minutes, is illustrated in Fig. 5. The result is related to the phase A to neutral and the other phases have shown a similar performance.

Fig.5. Phase to neutral THD voltage – measurement

Table I summarizes the above results and makes clear the information about the most relevant individual harmonic components. In addition to the minimum, maximum and average values the given summary also provides the so called P95, i.e. the level of harmonic that is associated to the probability of occurrence of 95% over the total period of measurement.

TABLE I – Phase to neutral voltage – harmonic distortion summary

.

The total harmonic distortion (THD) associated with P95 evidences that this parameter is in accordance with the European standards and the IEEE 519 limits. The same affirmation can be equally applied to the individual harmonic components. Despite the standards agreement, these values will be still used to elucidate the methodology of sharing the harmonic voltage distortion between the utility and the industry.

Using the same previous equipment in a simultaneous way as the voltage measurement, Fig.6 shows the THD current performance for line A current. This is the same phase used for the voltage result and corresponds to one of the three line currents.

Fig.6. Line THD current – measurement

Table II gives a summary of the line A current results in the same way as explained for phase A to neutral voltage.

TABLE II – Current results

.
5. Utility and Consumer Harmonic Impedances

The utility impedance was computationally obtained by supplying the required data to the HARMZS software. This is a commercial program developed and supplied by CEPEL (electrical research center – Brazil). The impedance module and angle for each frequency are given in Fig.7 and Fig.8, respectively.

The load impedance is then found in accordance with the described methodology. It must emphasized that the calculation is performed at each instant of voltage and current measurement. Therefore, the Fig.9 and Fig.10 show the time domain behavior of the calculated equivalent load impedance during the focused time interval of measurement.

Fig.7. Utility impedance module versus frequency

Fig.8. Supply impedance angle versus frequency.

Fig.9. Equivalent resistance load

Fig.10. Equivalent load capacitance and inductance over the measured time interval.

6. Results Associated to the THD Sharing

Once the necessary information is available to the use of the proposed methodology for sharing of responsibility upon the harmonic voltage distortion between the utility and the consumer, the method was applied and the final results are given in Table III. As shown, the values are related to the mentioned time interval of 5 minutes, due to this the minimum, maximum, average and P95 values are given.

TABLE III – Summary of the final sharing of responsibility at the PCC.

.

The results indicate that there are no significant problems regarding harmonic distortions. Besides, the consumer and the utility contributions to total voltage distortion are almost the same.

Fig.11 illustrates, over the 5 minutes of measurement, the instantaneous contribution of both the utility and the industry. The results are in agreement with the previous statement.

Fig.11. Utility and consumer contributions to voltage THD at the PCC over the measured period.

Focusing the individual 5th harmonic order, by applying the procedure for the measured time interval, Fig. 12 and Table IV show the contribution from the supply and the load. It can be noted that the major individual distortion is attributed to the local power authority. As the industry rectifier is composed by a 36 pulse arrangement, this is a physical expected result.

Fig.12. Utility and consumer contributions to 5th harmonic voltage distortion at the PCC over the measured period.

TABLE IV – Results to 5ª harmonic sharing in PCC.

.

If the 7th individual harmonic order is now considered, the final results are illustrated by Fig. 13 and Table V. The sharing of responsibility upon this specific frequency points out to the industry as the major generator of such component. As a matter of fact, at a first glance, this looks inconsistent. However, the existence of an industry power factor capacitor bank has been recognized as the reason for this current amplification. Thus, the final results appear physically in accordance with expected performance due to the combination of the 88 kV busbar capacitances and supply impedance.

Fig.13. Utility and consumer contributions to 7th harmonic voltage distortion at the PCC over the measured period.

TABLE V – Results to 7ª harmonic sharing in PCC.

.
7. Conclusion

This paper presented a case study related to the sharing of harmonic responsibility between the utility and the consumer. By applying the methodology here discussed throughout a real case it was highlighted the steps and the final results about the distribution of harmonic distortion between the supplier and the consumer. The results have shown that, for the present situation, both the utility and the industrial consumer have almost the same responsibility upon the final THD. As far as the process validation is concerned, due to the natural difficulties associates to the use of a real installation, no switching maneuvers were allowed. Thus the analysis was limited to physical expected performances. Using such principles it has been shown that the final indications about responsibility upon THD were found to be physically consistent. However, the authors recognize that this subject is controversial and the approach validation requires further investigation.

8. References
[1] IEEE Recommended Practice and Requirements for Harmonic Control in Electric Power Systems, IEEE Std. 519-1992.
[2] Voltage characteristics of electricity supplied by public distribution systems, European Std. EN 50160:1999.
[3] Electromagnetic compatibility (EMC) – Part 3: Limits – Section 6: Assessment of emission limits for distorting loads in MV and HV power systems – Basic EMC publication, IEC 61000-3-6, (1996).
[4] A. A. Moustafa, A. M. Moussa and M. A. El-Gammal, Separation of customer and supply harmonics in electrical power distribution systems, in: Proceedings of Ninth International Conference on Harmonics and Quality of Power, 2000, pp. 1035-1040.
[5] R E. B. Makram and S. Varadan, “Generalized load modeling in presence of harmonics and distortion,” in: Proceedings of Twenty Fifth Southeastern Symposium on System Theory, pp. 124-128, Mar. 1993.
[6] M. M. M. El Arini, “A time domain load modeling technique and harmonics analysis,” in: Proceedings of Eighth International Conference on Harmonics and Quality of Power, pp. 930-938, Oct. 1998
[7] S. A. Soliman and M. Al-Kandari, “A simple and noval technique for linear and nonlinear load modeling in the time domain”, in: Proceedings of Eighth Mediterranean Electrotechnical Conference, 1996, pp. 1616-1619.
[8] T. Tanaka and H. Akagi, “A new method of harmonic power detection based on the instantaneous active power in three-phase circuits”, in IEEE Trans. Power Del, Vol.10, pp 1737-1742, April. 1995.
[9] K. Srinivasan and R. Jutras, “Conforming and nonconforming current for attributing steady state power quality problems,” IEEE Trans. Power Del. Vol. 13, pp 212-217, Jan. 1998.
[10] Wilsun Xu and Yilu Liu, “A method for determining customer and utility harmonic contributions at the point of common coupling,” in: IEEE Trans. Power Del., pp. 804-811, Feb. 2000.
[11] S. F. P. Silva and J. C. de Oliveira, ” The Sharing of Responsibility between the Supplier and the Consumer for Harmonic Voltage Distortion: A Case Study,” in: Electric Power Systems Research, Vol. 78, pp. 1959-1968, Nov. 2008.

Source & Publisher Item Identifier: International Conference on Renewable Energies and Power Quality (ICREPQ’09) Valencia (Spain), 15th to 17th April, 2009. https://doi.org/10.24084/repqj07.327

Semi-Analytic Calculations of Overvoltages caused by Direct Lightning Strike in Buried Coaxial Cable

Published by Karol ANISEROWICZ, Renata MARKOWSKA, Bialystok University of Technology

Abstract. Results of calculations of overvoltages caused by a direct lightning strike to an underground coaxial cable are presented. Analytic formulas are used in the frequency domain. The time-domain waveforms are computed using the Inverse Discrete Fourier Transform (IDFT).

Streszczenie. Przedstawiono wyniki obliczeń przepięć spowodowanych przez bezpośrednie uderzenie pioruna w podziemny kabel koncentryczny. Wykorzystano wzory analityczne sformułowane w dziedzinie częstotliwości. Przebiegi w dziedzinie czasu obliczono z zastosowaniem Odwrotnej Dyskretnej Transformacji Fouriera (IDFT). (Pół-analityczne obliczenia przepięć spowodowanych przez bezpośrednie uderzenie pioruna w podziemnym kablu współosiowym).

Keywords: lightning; overvoltages; underground cable; analytic formulation.
Słowa kluczowe: piorun; przepięcia; kabel podziemny; sformułowanie analityczne

Introduction

Lightning discharges cause substantial threat for outdoor electronic circuits and systems. This hazard concerns both overhead and underground installations, and it was analyzed in many publications [1]-[8]. In particular, electronic systems connected to long cables spread over large areas are exposed to the lightning electromagnetic pulse (LEMP).

Long underground coaxial cables are within the scope of this paper. Buried cables are commonly used, and underground sensor cables of intrusion detection systems are among them. The coaxial cable sensors together with the co-operating equipment and devices are used in monitoring systems for protection of people and property. Low energy is necessary for proper action of such system, and relatively small amount of electromagnetic energy is enough to affect the system. Sensor cables are typically buried in soil at approximately 25-40 cm below the surface and are several hundred meters long. Cable systems may be realized as standalone or networked for much longer perimeters. Their equivalent lightning discharge collection area can be of the order of square kilometers.

Problems concerning estimation of the threat related to lightning effects are usually solved numerically [5]-[7]. Analytical or semi-analytical solutions are relatively rare. The closed-form formulations are of special value because they provide examples that may be used for testing the numerical algorithms. The aim of the present paper is to calculate overvoltages that can occur in a buried sensor cable during a typical lightning strike, and to estimate the required insulation immunity to electrical breakdown. Analytic formulas are written in the frequency domain basing on [8], and the time-domain waveforms are calculated using the Inverse Discrete Fourier Transform (IDFT). A similar problem is within the scope of paper [9], where simplified calculations of surge currents and voltages in more complex buried cable systems are described. The study introduced here was used for validation of some results presented in [9].

Analytical model of buried cable

Consider a lightning strike to ground very close to one end of an underground cable (Fig. 1). A part of the lightning current invades the cable through a metal enclosure of the cable input device.

Assume that the insulation of the system withstands the threat, so the surge current flows along the cable outer conductor to the enclosure of the device on the other end of the cable. The contribution of the cable inner conductor is neglected (Fig. 1b) [8]. Dimensions a and b of the cable cross-section are the inner and outer radius of the cable insulation, respectively. The burying depth d is not used in formulas presented further.

The transmission-line model is used here [8]. The model and its equivalent circuits are presented in Fig. 2.

Fig.1. Buried cable under study (a) and its cross-section (b)

Fig.2. Transmission-line model (a) and equivalent circuit of short segment of the line (b)

Current I(z) flows in the cable outer conductor, and U(z) is the voltage occurring between the cable outer conductor and the reference ground, in the insulating jacket. The soil propagation coefficient is equal to:

.

where µ0 and ε0 stand for the permeability and permittivity of vacuum, respectively, σg – soil conductivity, εrg – soil relative permittivity. For calculation of voltage U(z) and current I(z) (Fig. 1) it is necessary to determine characteristic impedance Z0 and propagation coefficient γ of the equivalent transmission line:

.

where Z and Y are the impedance and admittance per unit length, respectively.

Impedance Z is composed of the internal impedance of the soil (ground) Zg, the internal impedance of the cable outer conductor Zc, and the inductive impedance of the insulating jacket jωLi [8]:

.

These impedances may be calculated as follows:

.

where: δg = 1/αg – the skin depth in the soil, γ0 = 1.78107… – the Euler constant, T – the thickness of the cable outer conductor, σc – the metal (copper) conductivity,

.

The admittance per unit length Y is composed of the capacitive admittance jωCi of the insulation in series with the unit admittance of the soil Yg [8]:

.

These admittances may be approximated as follows:

.
.

where εri is the relative permittivity of the insulating layer. The grounding resistance of the equipment connected at the cable output equals Rg2 (Fig. 1). The input impedance of the equivalent transmission line is given by:

.

where l is the cable length (Fig. 2a).

Current I1 being the part of the lightning current IL invades the cable outer conductor (Fig. 1). The rest of current IL is dissipated into the ground, which is modeled by current Id flowing through the grounding resistance Rg1. The following equations are valid at the cable input:

.

The spectra of voltage U(z, jω) and current I(z, jω) at any distance z from the cable input can be calculated using the commonly known transmission-line equations:

.
Calculations of overvoltages for different waveforms of lightning current

The analyzed example concerns the underground system, so one may model the lightning impact as the surge current injection. The following grounding conditions are considered: σg = 0.01 S/m, εrg = 10, Rg1 = Rg2 = 5 Ω. Assume the following parameters of the cable: 2a = 12.73 mm, 2b = 15.5 mm, l = 200 m, T = 0.33 mm, εri = 2.3, σc = 58.6×106 S/m. These are typical for the intrusion detection sensors [10].

We apply the double-exponential approximation of the lightning current waveform:

.

Different lightning return current waveforms are used, according to [11]:

• 10/350 μs – model of the first positive stroke;
• 1/200 μs – model of the first negative stroke;
• 0.25/100 μs – model of the subsequent negative stroke.

The maximum value of the current is assumed to Im = 20 kA, which is close to typical lightning surges [12]. All the results can be easily re-calculated assuming other maximum values since the analyzed system is linear. The lightning current spectrum has the closed form:

.

The values of the coefficients are presented in Table 1 [5]. The right column contains also coefficients for waveform of 2/50 μs, which will be used in the next section. The time-domain waveforms are calculated numerically, using the IDFT algorithm.

Table 1. Coefficients for formulas (14)-(15) [5]

.

Currents IL(t), Id(t), I1(t), and I2(t) (Fig. 1a) calculated for three different surge waveforms are presented in Figs. 3-5. The associated voltages U1(t) and U2(t) are presented in Figs. 6-8, respectively.

Reflections from the cable ends are visible as smooth steps at the current and voltage waveforms.

Fig.3. Currents calculated for lightning surge of 20 kA, 10/350 μs

Fig.4. Currents calculated for lightning surge of 20 kA, 1/200 μs

Fig.5. Currents calculated for lightning surge of 20 kA, 0.25/100 μs

The calculated maximum values of currents in the cable outer conductor I1(t) and I2(t) are about 7.5-8.5 kA, which is about 37 % to 43 % of the maximum value of the lightning current IL(t). The highest value was obtained for the 10/350 μs waveform, and the lowest – for the 0.25/100 μs. The remaining current flows into ground.

The maximum voltages between the cable terminators and the reference ground are about 160-170 kV at the energized input and 70-90 kV at the output of the cable. Note that these voltages do not arise between the cable inner and outer conductors. They may be considered as the estimation of voltages in the cable insulation jacket, i.e. between the cable outer conductor and the ground.

Fig.6. Voltages calculated for lightning surge of 20 kA, 10/350 μs

Fig.7. Voltages calculated for lightning surge of 20 kA, 1/200 μs

Fig.8. Voltages calculated for lightning surge of 20 kA, 0.25/100 μs
Calculations of overvoltages for average lightning waveform and different grounding conditions

This section contains results that may be considered as estimation of typical lightning threat to the analyzed cable. Assume the surge current to be of 20 kA, 2/50 μs (right column of Table 1). These parameters are close to those of average lightning current [12].

Consider two different grounding conditions:

(a) σg = 0.01 S/m, εrg = 10, Rg1 = Rg2 = 5 Ω;
(b) σg = 0.001 S/m, εrg = 10, Rg1 = Rg2 = 10 Ω.

Fig.9. Currents in analyzed system for grounding conditions (a): σg = 0.01 S/m, εrg = 10, Rg1 = Rg2 = 5 Ω
Fig.10. Currents in analyzed system for grounding conditions (b): σg = 0.001 S/m, εrg = 10, Rg1 = Rg2 = 10 Ω
Fig.11. Voltages in analyzed system for grounding conditions (a): σg = 0.01 S/m, εrg = 10, Rg1 = Rg2 = 5 Ω
Fig.12. Voltages in analyzed system for grounding conditions (b): σg = 0.001 S/m, εrg = 10, Rg1 = Rg2 = 10 Ω

The calculated waveforms of currents and voltages are presented in Figs. 9-10 and 11-12, respectively. The results of simplified calculations for these conditions are presented in [13].

Figs. 9 and 10 show that currents depend on the grounding conditions, which is obvious, however, the current variations due to the substantial changes of the ground conductivity are not large.

In turn, Figs. 11 and 12 demonstrate that voltages for grounding conditions (b) are approximately doubled in comparison to those for case (a). The grounding resistances in case (b) are two times larger than those in case (a).

Input impedance

Frequency domain plots of input impedance Zin provide additional information for the calculation results presented in the previous section, i.e. with surge current of 20 kA, 2/50 μs and two grounding conditions

(a) σg = 0.01 S/m, εrg = 10, Rg1 = Rg2 = 5 Ω;
(b) σg = 0.001 S/m, εrg = 10, Rg1 = Rg2 = 10 Ω.

The plots of input impedances are presented in Figs. 13 and 14. At frequencies exceeding approximately 300 kHz, the modulus of the cable input impedance is approximately 1.5-2 times larger in case (b) than that in case (a). It means that the higher the ground resistivity the larger part of the high frequency components of the lightning current is dissipated by the grounding system close by the point of strike.

This observation does not concern the lowest frequency band, where the major part of the lightning energy is located. At low frequencies, the average of the modulus of the cable input impedance seems to be close to 50 Ω for both analyzed grounding conditions. This means that the low frequency components of the lightning current are distributed in the system similarly, almost irrespective to the ground resistivity.

The economically reasonable value for the grounding resistance of a buried cable sensor of the intrusion detection system is of order of 5-10 Ω in typical soil. Getting smaller values for reduction of arising potentials is usually too expensive. This means that the expected maximum voltages can be of order of tens to hundreds of kilovolts, as in Figs. 6-8 and 11-12.

Insulation coats of many cables probably cannot withstand such a stress. Additionally, the current flow of order of kiloamperes over a time exceeding 100 μs leads to significant increase in the cable temperature causing its damage. Hence, additional surge protective devices (SPDs) are necessary for protection against lightning [11].

Fig.13. Input impedance for grounding conditions (a): σg = 0.01 S/m, εrg = 10, Rg2 = 5 Ω
Fig.14. Input impedance for grounding conditions (b): σg = 0.001 S/m, εrg = 10, Rg2 = 10 Ω
Conclusion

Analytic formulation presented here may be useful for testing new numerical procedures.

The calculated voltages and currents are related to approximately average lightning current of 20 kA. In the IEC standard [11] the maximum current value is said to be of 200 kA. This means that the values displayed here can be of order larger. Cable conductors and insulation coats of many cables cannot withstand such a stress without additional protection measures.

It follows from the calculations that striving for the possibly lowest grounding resistance is of essential importance for reduction of the lightning hazard in buried cables. Note that the soil conductivity is not a critical parameter, although technical means of achieving required grounding resistance depend on the soil conductivity.

Grounding is not a sufficient measure of protection against lightning damages in buried cables. Additional surge protective devices should be installed at both cable ends.

Acknowledgment: The research was conducted within the project S/WE/1/2015, financially supported by Polish Ministry of Science and Higher Education.

REFERENCES

[1] Tominaga T., Kuwabara N., Kato J . , Ramli A., Hal im A. , Ahmad H. , Characteristics of Lightning Surges Induced in Telecommunication Center in Tropical Area, IEEE Transactions on Electromagnetic Compatibility, vol. 45, no. 1, 2003, p. 82–91
[2] Bejleri M., Rakov V. A., Uman M. A., Rambo K. J . , Mata C. T., Fernandez M. I., Triggered lightning testing of an airport runway lighting system, IEEE Transactions on Electromagnetic Compatibility, vol. 46, no. 1, 2004, p. 96–101
[3] Barbosa C. F., Zeddam A., Day P., Bourgeois Y. , Effect of guard wire in protection a telecommunication buried cables struck by rocket-triggered lightning, Proc. of 29th International Conference on Lightning Protection, 23-26 June 2008, Uppsala, Sweden, p. 6b-1-1–6b-1-6
[4] Kuramoto S., Chikai S. , Suzuki T., Tada Y. , Evaluation of lightning surge current characteristics induced on the aerial subscriber’s cable at telecommunication center and in NTT, Proc. of 28th International Conference on Lightning Protection, 2006, Kanazawa, Japan, p. 529–532
[5] Aniserowic z K., Analysis of electromagnetic compatibility problems in extensive objects under lightning threat monograph, in Polish, Bialystok 2005, pdf available at http://pbc.biaman.pl/dlibra
[6] Markowska R., Sowa A., W., Ochrona odgromowa obiektów radiokomunikacyjnych, Oficyna Wydawnicza Politechniki Białostockiej, Białystok 2013
[7] Masłows ki G., Analiza i modelowanie wyładowań atmosferycznych na potrzeby ochrony przed przepięciami, Wydawnictwo AGH, Kraków 2010
[8] Vance E. F., Coupling to shielded cables, Wiley – Interscience, 1978
[9] Mar kowska R. , Ani serowi c z K. , Exposure of underground cable intrusion detection system to overvoltages caused by lightning strike, Proc. of 24th International Conference on Electromagnetic Disturbances EMD’2017, 20-22 September 2017, Bialystok, Poland, 73-76
[10] Technical data sheet – Radiating cables, Kabelwerk, EUPEN AG, Rev.: 08/2010-10-07
[11] IEC 62305, Protection against lightning, series of standards, 2010
[12] Uman M. A., Natural lightning, IEEE Transactions on Industry Applications, 30 (1994), No. 3, 785-790
[13] Aniserowic z K., Markows ka R. , Semi-analytic calculations of overvoltages caused by direct lightning strike in buried coaxial cable, Proc. of 24th International Conference on Electromagnetic Disturbances EMD’2017, 20-22 September 2017, Bialystok, Poland, 9-12.

Authors: dr hab. inż. Karol Aniserowicz, prof. nzw. w PB, Politechnika Białostocka, Wydział Elektryczny, ul. Wiejska 45d, 15- 351 Białystok, e-mail: k.aniserowicz@pb.edu.pl; dr hab. inż. Renata Markowska, Politechnika Białostocka, Wydział Elektryczny, ul. Wiejska 45d, 15-351 Białystok, e-mail: r.markowska@pb.edu.pl.

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 93 NR 12/2017. doi:10.15199/48.2017.12.01

Three-phase Four-wire Circuits Interpretation by Means of Different Power Theories

Published by Fernando P. MARAFÃO1, Eduardo V. LIBERADO1, Helmo K. M. PAREDES2, Luiz C. P. da SILVA2, Unesp – Univ Estadual Paulista (1), University of Campinas (2)

Abstract. In order to contribute to the discussion of defining a generalized power theory, valid for unbalanced and non linear circuits, this paper discusses the relationship and discrepancies among four modern power theories. Three-phase four-wire circuits have been analyzed, since the most conflicting and intriguing interpretations take place in case of return conductor occurrence. Simulation results of different load, power supply and line conditions will be discussed in order to elucidate the author’s conclusions and to provoke the readers for additional discussions.

Streszczenie. Przedmiotem artykułu są wzajemne powiązanie i rozbieżności pomiędzy czterema aktualnymi teoriami mocy obwodów nieliniowych i niezrównoważonych. W artykule analizowane są obwody trójfazowe, czteroprzewodowe, gdyż najwięcej różnic interpretacyjnych pojawia się w związku z obecnością przewodu zerowego. Aby pobudzić dyskusję, w artykule przedstawiono wyniki modelowania różnych obciążeń i źródeł zasilania. (Interpretacja obwodów trójfazowych, czteroprzewodowych za pomocą różnych teorii mocy).

Keywords: Non Sinusoidal Systems; Power Factor; Power Theories; Unbalanced Circuits.
Słowa kluczowe: niesinusoidalne, współczynnik mocy, teorie mocy, obwody niezrównoważone.

Introduction

The search for a general applicable power theory, suitable for analysis, revenue metering or power conditioning applications has been an intriguing subject during, at least, the last 100 years. This pursuit has been motivated in the last decades by the increasing use of non linear and unbalanced load, and more recently, it has been boosted up based on the novel configurations of modern power grids, especially those with relatively low short circuit levels (such as those related to micro and smart grids).

Nevertheless, even considering the great number of important contributions [1-9], there is not a final agreement on the voltage and current decompositions and the related power definitions, which should be adopted, especially in case of multiphase circuits with return conductor [11-17].

As discussed in [13-15,17], in case of four wire circuits, some of the confusion can be explained in terms of the choice of the voltage referential and also in terms of the return conductor impedance. Moreover, most of the misunderstanding is probably based on the fact that several authors had addressed their contributions for a specific application (power conditioning, revenue metering, etc.), instead of discussing a general applicable power theory.

However, an important query remains: what is expected from a general applicable power theory? Such issue has been investigated during the last several decades and the answer is still under construction [4]. Thus, considering the need of defining a generalized power theory, the authors of this paper call the attention to a number of relevant questions, which they believe are connected to the answer of the previous query.

• Q1 Is there a more efficient domain for the analysis of power circuits, time or frequency? Which and why?

• Q2 Is it necessary to split the voltages and currents into their fundamental and harmonic components? How?

• Q3 Is it necessary and possible to relate the new concepts to the traditional and well accepted ones? How?

• Q4 Is it possible to define and associate current and power components with specific physical phenomena?

• Q5 Should different disturbing components, from different physical phenomena, be summed up into a non active current or power component? Why and how?

• Q6 How to use the active and non active current and/or power components for revenue metering, power conditioning and responsibilities assessment?

• Q7 What can be done in order to maintain the efficiency of the power system as good as possible in terms of ideal energy generation, transmission and consumption?

• Q8 How to deal with multiphase circuits with return conductors? Is it a special case in multiphase systems?

• Q9 Is it possible to use the same methodology to analyze load phenomena, as well as the entire power system?

• Q10 Is it possible or is it time to employ the novel definitions in fundamental electrical engineering courses, such as Electrical Circuits?

The only certainty is that the conventional theory is no longer able to stand for the modern non linear and/or unbalanced multiphase power circuits. Thus, assuming that answering all the previous questions has been an extremely complex work and it will certainly take some time (years or decades), but trying to contribute to the discussion of defining a generalized power theory, this paper discusses and compares the results of four modern power theories, under different conditions. The investigated proposals were the so-called STD-1459, the FBD Theory (Fryze-BuchholzDepenbrock), the p-q Theory and the Conservative Power Theory (CPT). It is important to state that the choice of these four proposals is based on a sequential work which has been realized by the authors. Other relevant proposals, especially the CPC (Currents’ Physical Components) [4], are going to be included in future analyses.

Differently from [17], in this paper the authors are mostly interested in the interpretation of the power components, instead of current parcels. Further, the STD 1459 has been added to the comparative analysis. Next section shows the most relevant power components and their respective nomenclature, for each method.

Essential definitions of the investigated power theories

In order to avoid being repetitive in terms of overview, the authors refer to the original papers mentioned afterward, for additional details. It is important to notice that all the proposals are intrinsically based on multiphase conceptions.

The IEEE Standard 1459-2010 [1,15,18]

Accordingly to the STD-1459, the interpretation of any three-phase power circuit can be done by means of the following power components:

PSTD active power;
P1 fundamental active power;
P1+ fundamental positive-sequence active power;
PH harmonic active power;
Q1+ fundamental positive-sequence reactive power;
DeI current distortion power;
DeV voltage distortion power;
S1 effective fundamental apparent power;
S1+ positive sequence apparent power;
SU1 fundamental unbalanced apparent power;
SeH effective harmonic apparent power;
SeN non fundamental apparent power;
Se total effective apparent power;
DeH non active harmonic apparent power;
NSTD non active power;
PFe effective power factor (PSTD / Se);
PF1 fundamental power factor (P1 / S1);
PF1+ fundamental positive-sequence power factor (P1+/S1+).

The FBD theory [2,3]

From the FBD current and power decompositions, the following components can be considered for the interpretation of multiphase circuits:

PΣa collective rms active power;
PΣz collective rms zero power component;
PΣv collective rms variation power component;
PΣn collective rms non active power component;
SΣ collective (total) apparent power;
PFΣ collective power factor (PΣa / SΣ).

The p-q theory [5-7]

Accordingly to the p-q Theory, the following power components could be calculated in case of three-phase four wire circuits:

pαβ instantaneous real power;
p0 instantaneous zero-sequence power;
qαβ instantaneous imaginary power;
Pαβ average value of (pαβ);
P0 average value of (p0);
Qαβ average value of (qαβ);
pαβ~ oscillating part of the real power;
p0~ oscillating part of the zero-sequence power;
qαβ~ oscillating part of the imaginary power;
Pαβ~ rms value of (pαβ~);
P0~ rms value of (p0~);
Qαβ~ rms value of (qαβ~).

For the purpose of comparisons, the above power terms are not sufficient and it has been necessary to define in this paper, the apparent power and power factor, as following:

.

The CPT theory [8-10]

Considering the CPT proposal, the following power components could be used for the interpretation of three-phase four wire circuits:

PCPT active power;
QCPT reactive power;
SCPT modulus of the complex power;
PFCPT conventional power factor (PCPT /SCPT);
V void power;
Na unbalanced active power;
Nr unbalanced reactive power;
NCPT unbalance power;
A apparent power;
λ total power factor (PCPT /A).

Considerations regarding to the voltage referential

Above and beyond numerous particular details of the previous power components calculation, it is important to point out that the voltage referential was chosen according to each author’s suggestion, i.e., for the FBD method is a virtual star point and for the other three methods, it is the return conductor (load side), as shown in Fig. 1.

Comparative analysis of STD, FBD, PQ and CPT proposals – simulation results

Fig. 1 and Table 1 illustrate the simulated power circuit, on which eleven different loads, supply and line impedances have been imposed. In the sequel, the simulated cases have been discussed in order to evaluate and compare the considered power theories. Tables 2 and 3 show the results of different power components, from each power theory, accordingly to the defined cases.

Fig.1. Simulated power circuit

Table 1. Load and line impedance’s conditions

.

Sinusoidal and balanced (127Vrms, 60Hz) voltages with neglected line impedance (strong grid)

Three-phase unbalanced resistive load – (A.1)

This simple case points for interesting discussions. The STD represents the load unbalance by means of SU1 and N components, indicating that part of the effective apparent power (Se) is not related to active power. So, in this case, the effective power factor (PFe or PF1) is smaller than one, what in some way contradicts the conventional sense of unitary PF for single-phase resistive loads. However, if just positive sequence were considered, the PF1+ would be unitary, indicating the absence of energy storage element. In this case, the non fundamental and distortion power should be zero. The nonzero values are related to computation errors (smaller than 1%).

Applying the FBD, it is possible to notice that the collective active and apparent power exactly match the active and effective apparent power from the STD. The same happens with the PFΣ, what means that the load unbalance has been considered as a non active power. Indeed, the load unbalance affects PΣz, PΣv and PΣn.

Now assuming the p-q Theory, it is possible to observe that Pαβ+P0 is equal to the active power from STD and FBD, while the imaginary average power is practically zero. In this case, the load unbalance could be observed by means of the instantaneous oscillatory behavior of real and imaginary power, as well as it could be estimated in terms of their RMS values (Pαβ~, P0~, Qαβ~).

Considering the CPT, it is possible to observe that the active power matches PSTD, PΣa and (Pαβ+ P0), from the other methods. The conventional power factor (PFCPT) is unitary. Given that there are no distorted voltages and currents, the distortion power components are zero, however, the resistive load unbalance reflects in the unbalanced active power (Na). Note that the unbalanced reactive power is zero, since there are no energy storage elements in the circuit. Such unbalance power (NCPT) influences the apparent power (A), as well as the total power factor (λ), which is smaller than one, representing the part of the currents that circulates in the resistive circuit, but does not contribute to average active power (PCPT).

Finally, other interesting comparisons could be pointed out, e.g.: the sum of the RMS real and imaginary oscillating components (from p-q Theory) results equal to the unbalanced active power from the CPT (Pαβ~ + P0~ + Qαβ~ = Na). Considering the apparent power results, it is possible to observe that the CPT value does not match the effective or collective values from FBD or STD (Se = SΣ ≠ A), as well as the total power factor. It happens particularly in case of four-wire circuits, since the FBD and STD apparent power definitions, as discussed in [1-3,13,14], intrinsically considers the power phenomena from the power system’s point of view, while the CPT addresses to the load point of view [16]. In case of four wire circuits, it means that any kind of homopolar power component, associated to the return conductor, is considered by STD and FBD approaches, while the CPT considers just the load aspects.

Single phase to phase capacitive load – (A.2)

The idea of this case has been extracted from [7], on which the authors discuss the potential of the p-q Theory in order to interpret the related physical phenomena by means of the instantaneous real and imaginary powers. Such authors call the attention to the fact that the instantaneous real power is different from zero, representing the oscillating energy flow during the charges and discharges of the capacitor, according to their terminal voltages. Moreover, the average power components Pαβ and P0 results zero (Table 2), while the Qαβ results equal (in modulus) to the reactive power from STD and CPT proposals.

If the goal is to analyze the physical phenomena by means of the power components, it is interesting to observe that Pαβ~ matches Na from the CPT and PΣv from FBD (all related to active instantaneous currents), while Qαβ~ matches Nr from the CPT (both related to reactive instantaneous currents). It makes sense if one considers the unbalanced behavior of this phase-phase capacitive load, in terms of the three-phase circuit.

The understanding of this case is based on the observation of each phase, as well as, the three-phase instantaneous power components from the p-q and CPT methods. Even if the capacitive load does not draw any three-phase active power (P), the instantaneous power, per phase, could be decomposed into active and reactive (or imaginary) components, as well as it leads to oscillatory behavior over the three-phase instantaneous components. In terms of the power phenomena interpretation, such oscillatory behavior is considered by means of the rms oscillatory components in the p-q method, as well as the N components in the CPT, the non active power in the FBD and the SU1 in the STD. The apparent power matches for STD, FBD and CPT proposals and all power factors indicate zero value.

It is also interesting to observe the negative signal of the reactive power components, for the CPT and p-q Theory, indicating capacitive behavior, such as in the conventional conception. The reactive and zero power components from STD and FBD will always result positive, in consequence of their formulations.

Single phase to neutral capacitive load – (A.3)

In this case, the CPT and STD reactive power components are equivalent and match the imaginary power from p-q Theory. Considering the CPT, it is worth to notice that the unbalance reactive power (Nr) is even greater than the balanced reactive power (Q). Besides, in the same way of case A.1, the apparent power from the CPT does not match the FBD and STD values. Observe that in this case, the influence of the homopolar components is quite severe, since it deals with a single phase load in the three-phase four-wire circuit.

Three-phase (Y connection to neutral) unbalanced capacitive load – (A.4)

Again, the CPT and STD reactive power components are equivalent and match the imaginary power from p-q Theory. Besides, (Pαβ~ + Qαβ~ = Nr). Similar to A.1 and A.3, due to the homopolar behavior of the four-wire unbalance circuit, the apparent power from the CPT does not match the FBD and STD values.

Considering the FBD, in addition to the apparent power that matches the STD definition, the variation power (PΣv) is equal to the oscillating power components from the p-q Theory. In this case, the whole FBD power is interpreted like non active power (PΣn), in terms of the zero power (PΣz) and variation power (PΣv) components.

Three-phase (Y connection to neutral) balanced capacitive load – (A.5)

In this case, all the reactive (CPT and STD), imaginary (p-q), zero and non active (FBD) power components are equivalent. The same happens with the apparent power components (Se = SΣ = A), given that the load is balanced. Besides, both the unbalance (NCPT), variation (PΣv) and rms values of the oscillating powers (p-q) results null. Three-phase balanced non linear load (three phase to neutral diode rectifiers with RC load) – (A.6) Considering such balanced non linear load, observe that the active power components are equal for all methods. The same happens to the values of QCPT, Qαβ and Q1+. The modulus of the complex power (SCPT) from the CPT also matches the p-q apparent power (Sαβ), as well as the (S1+) from STD. The rms value of the oscillating real power (Pαβ~) is equal to the variation power (PΣv) from FBD. The non active components from STD and FBD are also equivalent (NSTD = PΣn). In addition, notice that the unbalance power components from CPT (NCPT) and STD (SU1) are about zero, since the load is balanced. In this case, the STD and CPT represent the load nonlinearities by means of the distortion power (DeI) and void power (V) components, respectively. Based on the observation that this non linear load, even if balanced, leads to neutral (homopolar) currents, over again, the CPT apparent power does not match the effective or collective values from FBD or STD (Se = SΣ ≠ A). However, it can be noticed that regardless of of the neutral current, P0 and P0~ results zero, given that the voltages are sinusoidal and balanced (strong grid).

Three-phase unbalanced non linear load (three phase to neutral diode rectifiers with RC load) – (A.7)

One can observe again that the active power components from the four methods are equivalent, as well as the reactive and imaginary components from STD, CPT and p-q Theory. Moreover, the modulus of the complex power (S) from the CPT matches again the p-q apparent power (Sαβ) and the (S1+ ) from STD.

Considering the CPT, notice that the unbalanced behavior of the load appears at the active and reactive unbalanced power components (Na and Nr). On the other hand, the load nonlinear behavior results in the occurrence of void power (V). Note that since the line impedance is neglected, the voltages are not affected by the load and the voltage distortion power (DeV), from the STD is zero. As in the previous case, it is possible to observe the difference between the conventional (PF) and the total power factor (λ). Additionally, the rms value of the oscillating real power (Pαβ~) is equal to the variation power (PΣv). The non active components from STD and FBD are also equivalent (NSTD = PΣn).

Nonsinusoidal and balanced (127Vrms, 60Hz; 12.7Vrms, 180Hz; 6.35Vrms, 300Hz; 6.35Vrms, 420Hz) voltages with neglected line impedance (strong grid)

Three-phase balanced resistive load (B.1)

This other very simple case also points for interesting discussion. Note that the p-q and CPT are practically equivalents. Although a slightly oscillatory behavior of the instantaneous power (caused by the instantaneous product of distorted voltages and currents), which can be observed by means of Pαβ~ and P0~, both methods indicates an equivalent average active power (Pαβ + P0 = PCPT), which also corresponds to the active power from STD and FBD. In case of the CPT, the decompositions indicate that there is no unbalance component (NCPT), what makes sense in case of balanced load. Moreover, there is no void power (V), what means that the total apparent power is conveyed into active power on the balanced resistive load.

On the other hand, the STD proposal suggests the power decomposition into several fundamental and harmonic power components, as e.g., the harmonic active power (PH).

Regarding to STD and FBD, it is interesting to observe that the total apparent powers (Se, SΣ) practically match and they are greater than the CPT apparent power. This occurs since the imposed supply voltages have homopolar components, resulting in return conductor current circulation, even if the load is balanced.

It is valid to mention that the interpretation of non active and non fundamental components from STD and FBD and their relation with physical phenomena is not so intuitive in this case.

Three-phase balanced RL load (B.2)

In this case, all the active power components results equal (PCPT, PSTD, Pαβ + P0, PΣa). The same happens with the reactive power from CPT, STD and p-q Theory, which have practically the same values.

Besides, the inductive (RL) behavior of the load leads to a non linear condition, with different distorted phase voltages and currents. In this case, the CPT suggests the existence of void power (V), which is mainly related to the non linearity of phase voltages and currents [9,10]. There is no unbalance power (NCPT).

It is also possible to observe that the modulus of the complex power from CPT is approximately equal to the values of (Sαβ) and (S1+), from p-q Theory and STD. Besides, the non active components from FBD (PΣn) and STD (NSTD) practically match.

Table 2. Power components for cases A.1 to A.7

.

Sinusoidal and balanced (127Vrms, 60Hz) voltages with high line impedance (weak grid)

Three-phase unbalanced RL load – (C.1)

Despite of the active power components, which result in equivalent values (PCPT, PSTD, Pαβ + P0, PΣa), each method represents the power phenomena in a different way. But it is important to observe that in the case of high line impedance, the load voltages suffer the influence of the load current and consequently, it will influence various power components.

The CPT indicates active and reactive load behavior, as well as, it points to non zero unbalance power components (because of the line impedance influence on the load voltages). The p-q Theory represents the unbalanced load by means of the rms values of the oscillatory power components. Considering the FBD, it points to the zero, variation and non active power to represent the load unbalances and reactive behavior. Finally, the STD represents the load performance by means of its several power components, indicating no distortion power, but reactive and unbalance behavior.

Table 3. Power components for cases B.1, B.2, C.1 and C.2

.

Three-phase balanced non linear load (three phase to neutral diode rectifiers with RC load) – (C.2)

In this case it is interesting to observe that the resulting unbalance power components from the CPT and STD are equal to zero. On the other hand, the load current distortion results in voltage distortion by means of the line impedance, producing voltage distortion power (STD), as well as to the void power (CPT).

Except for the active power definitions, which still match, one may notice that in this case, with high line impedance (weak grid), the comparisons of all other power components does not points to any numerical equivalence. It is also very interesting to point out the differences among the power factor definitions in cases C.1 and C.2. The choice of one or other could result in very different efficiency interpretation.

Conclusions

The previous discussions indicate that the understanding of physical phenomena can be very characteristic, depending on the adopted methodology. Nevertheless, it was also possible to identify a number of similarities among the investigated power theories, as well as discussed in [19]. In some cases, the definitions are equivalent or they could be complementary.

From the point of view of quantifying the influence of specific disturbing components, the STD and CPT seem to be more suitable. However, it is important to point out that the FBD and STD formulations seem to be mainly related to the characterization of the network utilization (including sources, loads and transmission lines), while the CPT and p-q seem to be essentially related to the load phenomena.

Such difference can be mostly relevant in case of four-wire circuits, because of the homopolar current circulation and it may stimulate several discussions on which should be the most relevant approach. However, the authors believe that it is just a question of what are we interested in? Are we interested in analyzing, penalizing or compensating the load or the overall network? One or other approach should be used or modified in order to satisfy the application. For example, the CPT apparent power could be easily modified to match the STD and FBD proposals, by means of changing the voltage referential and some equations to consider the voltage and the current of return conductor [10].

Moreover, for the purpose of contributing to answer the questions of Section I and considering the forthcoming definitions of a general applicable power theory, these authors would state:

• Q1 – No. It is possible to represent the power phenomena in both, time and frequency domains, since using the correct mathematical tools and theoretical theorems. However, some physical phenomena could be easily considered by means of one or other domain, such as in case of scattered current and power definitions, using the frequency domain [4];

• Q2 – Basically, it depends on the application. It could be necessary, e.g., for compensation or revenue metering. In this case, some kind of signal processing technique should be applied. Possible formulations splitting fundamental and harmonic components can be found in [1,10];

• Q3 – Yes, in some cases. From the analyses of the CPT and STD power components, one can observe several similarities with traditional or cognitive conceptions and it certainly helps during the argumentation process of defining new quantities. The simplest idea is to make use of the novel definitions to explain very basic conditions;

• Q4 – Yes. It can be observed, e.g., in case of the distortion, void and unbalanced components from the CPT and STD, which are strictly related to physical phenomena. In case of identifying specific current and power terms associated to particular phenomena, it is important to point out that the resulting components should be preferentially orthogonal among themselves;

• Q5 – It depends on the application. If one is interested in compensating the non active components (non selectively), e.g., by means of electronic apparatus, this could be necessary. However, if one is interested in the interpretation of different physical phenomena or interested in penalizing different non ideal effects, it should not be done. In this sense, the CPT and STD would be preferable;

• Q6 – In case of active power components, it has been demonstrated that all the formulations results identical and they could be applied to energy calculation as in the common sense. Considering the non active components, a number of new indexes should be defined in order to evaluate the load performance or the network utilization. For example, it could be defined some void or unbalance factors directly from the power components (e.g., VCPT/ACPT or NCPT/ACPT from the CPT or similar relations from STD), which could be limited by specific standards, as in case of the power factor. The same components or their respective voltage and current signals could be used for power conditioning. The issue of responsibilities assessment could be explored in terms of the different power components, as well as the values of the line impedance, which is responsible for the interaction among loads and the network. In this case, the STD and CPT seem to point in a more suitable point of view. At this point, it is essential to call attention to the importance of using orthogonal current and power terms, avoiding e.g., penalization duplicity;

• Q7 – The total power factor, as defined in the FBD and STD should be assured unitary. It would represent a highly efficient load, in terms of power conversion, and minimal losses in the power system;

• Q8 – Yes. The authors believe that it is a special case and the return conductor should not be treated as an ordinary phase conductor. The condition of four-wire circuits leads to an important difference in terms of the analysis of the load or network point of view. Depending on the application, the suitable methodology should be adopted;

• Q9 – No. As discussed above, although the various similarities among the addressed power theories, it is possible to observe that the FBD and STD seem to be more convenient for the analysis of the power system, while the CPT and PQ regards to the load phenomena interpretation;

• Q10 – Yes, it is possible and it should be time of incorporating novel definitions on the fundamental courses, avoiding the situation on teaching traditional concepts based on sinusoidal and balanced conditions, which in practice, did not exist anymore in real applications.

Finally, assuming the increasing importance of micro grid applications for modern smart grid, on which the line impedance is expected to be higher than in case of traditional (strong) power systems, the definitions of novel power components will be a very important issue. As indicated in the previous analysis, the case of high line impedance and non linear load can be very difficult to be understood, even if considering the novel power theories formulations evaluated in this paper.

REFERENCES

[1] IEEE Trial-Use Standard Definitions for the Measurement of Electric Power Quantities under Sinusoidal, Non-sinusoidal, Balanced or Unbalanced Conditions, IEEE Standard 1459, (2010)
[2] M. Depenbrock, “Quantities of a multiterminal circuit determined on the basis of kirchhoff’s laws,” ETEP European Trans. on Electrical Power Engineering, 8 (1998), No. 4, 249-257
[3] V. Staudt, “Fryze – Buchholz – Depenbrock: A Time-Domain Power Theory,” Proc. of the International School on Nonsinusoidal Currents and Compensation, (2008)
[4] L.S. Czarnecki, “Currents’ physical components (CPC) concept: a fundamental of power theory,” Proc. of the International School on Nonsinusoidal Currents and Compensation, (2008)
[5] H. Akagi, A. Nabae, “The p-q theory in three-phase systems under non-sinusoidal conditions”, ETEP European Trans. on Electrical Power Engineering, 3 (1993), No. 1, 27-31
[6] E. H. Watanabe, R. M. Stephan, and M. Aredes, “New concept of instantaneous active and reactive powers in electrical systems with generic loads,” IEEE Trans. Power Delivery, 8 (1993), 697–703
[7] E. H. Watanabe, H. Akagi, and M. Aredes, “Instantaneous p-q power theory for compensating nonsinusoidal systems,” Proc. of the International School on Nonsinusoidal Currents and Compensation, (2008)
[8] P. Tenti and P. Mattavelli, “A time-domain approach to power term definitions under non-sinusoidal conditions,” Proc. of the Sixth International Workshop on Power Definitions and Measurements under Non-Sinusoidal Conditions, (2003)
[9] E. Tedeschi, P. Tenti, P. Mattavelli and D, Trombetti, “Cooperative control of electronic power processors in microgrids,” Proc. of the Brazilian Power Electronics Conference, (2009)
[10] P. Tenti, P. Mattavelli and H. K. M. Paredes, “Conservative Power Theory, Sequence Components and Accountability in Smart Grids”, Przeglad Elektrotechniczny, (2010), nr. 6, 30-37
[11] M. Malengret and C. T. Gaunt, “Decomposition of currents in three- and four-wire systems,” IEEE Trans. on Instrumentation and Measurement, 57 (2008), No. 5, 963-972
[12] M. Depenbrock, V. Staudt and H. Wrede, “Theoretical investigation of original and modified instantaneous power theory applied to four-wire systems,” IEEE Trans. on Industrial Application, 39 (2003), No. 4, 1160-1168
[13] A. E. Emanuel, “Reflections on the effective voltage concept,” Proc. of International Workshop on Power Definitions and Measurements under Non-sinusoidal Conditions, (2003)
[14] J. L. Willems and J. A. Ghijselen, “The choice of the voltage reference and the generalization of the apparent power,” Proc. of International Workshop on Power Definitions and Measurements under Non-sinusoidal Conditions, (2003)
[15] J. L. Willems, J. A. Ghijselen and A. E. Emanuel, “The apparent power concept and the IEEE standard 1459-2000,” IEEE Trans. Power Delivery, 20 (2005), No. 2, 876-884
[16] W. G. Morsi, and M. E. El-Hawary, “Defining power components in nonsinusoidal unbalanced polyphase systems: the issues,” IEEE Trans. Power Delivery, 22 (2007), No. 4, 2428-2438
[17] F. P. Marafão, H. K. M. Paredes, L. C. P. da Silva, “Critical evaluation of FBD, PQ and CPT current decompositions for four-wire circuits”, Proc. of the Brazilian Power Electronics Conference, (2009)
[18] F. P. Marafão, A. C. Moreira, S. M. Deckmann, “Virtual instrumentation applied to implement IEEE STD 1459-2000 power definitions,” Proc. of the IEEE Power Electronics Specialists Conference, (2005), 1712-1718
[19] J. L. Willems, “Reflections on Power Theories for Poly-Phase Nonsinusoidal Voltages and Currents”, Przeglad Elektrotechniczny, (2010), nr. 6, 11-21

Authors: Prof. Dr. Fernando Pinhabel Marafão, Unesp – Univ Estadual Paulista, Campus of Sorocaba, Av. Três de Março, 511, 18085-180, Sorocaba, SP, Brazil, fmarafao@sorocaba.unesp.br; Msc. Eduardo Verri Liberado, Unesp – Univ Estadual Paulista, Campus of Sorocaba, Av. Três de Março, 511, 18085-180, Sorocaba, SP, Brazil, eduardomeca3@gmail.com; Dr. Helmo K. Morales Paredes, School of Electrical and Computer Engineering, University of Campinas, Av. Albert Einstein, 400, 13083-970 Campinas, SP, Brazil, hmorales@dsee.fee.unicamp.br; Prof. Dr. Luiz Carlos Pereira da Silva, Av. Albert Einstein, 400, 13083-970 Campinas, SP, Brazil, lui@dsee.fee.unicamp.br.

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY (Electrical Review), ISSN 0033-2097, R. 87 NR 1/2011

Power System Harmonics

Published by Olutayo Ogunyemi, School of Science and Engineering, Atlantic International University, Honolulu, HI.

Abstract – Good power quality is essential in electrical power network. Power quality is not limited to availability of supply but also include steady frequency value and voltage magnitude and smooth waveform characteristics of the supply (Ogheneovo Johnson, 2016).The present situation of poor power quality in industrial and commercials sites, especially harmonics had created some form of attention to one of the power problems in the world; harmonics. This situation is as a result of the growth in the use of non-linear loads. The increase in the use of non-linear load had brought about the tight recommendations of IEEE standard 519 especially in the industrial and commercial sector (Robert G.Ellis, 2011).In practical terms, power quality can be defined as the rate at which the essential power parameters magnitude which are voltage, frequency and current deviates from the nominal values and can cause damage to the power infrastructure or equipment in use. In reality, there are about nine prominent power problems facing power system. These power issues are under voltage, over voltage, power sag, power surge, frequency distortion, line noise, power outage, switching transient and harmonic distortion. Power quality is said to be poor when any of the power problems is manifested within the power system. Mostly, harmonics occur as a result of alteration of the normal waveform which is generally transmitted by non-linear loads. This paper will dwell more on harmonics and its consequences in power system. Non-linear loads that cause situation of harmonics will also be reviewed in the course of this assignment.

Keywords: Total Harmonic Distortion (THD), Multipliers, Non-linear loads, Total Demand Distortion (TDD).

1. Introduction

Harmonics in three phase system power system was studied by Steinmetz in 1960 where concerns were raised on the behavior of third harmonic current which were produced from the effect of saturated iron within transformers core as well as electric machines. He was able to resolve the effect of the third level harmonics by proposing a delta connection in three phase power system which was able to block the effects of the third harmonics current (Abdelaziz, Mekhamer and Ismael, 2012).

Thereafter, rural electrification and telephone service came into existence and both the power infrastructure and the telephone circuits were mostly installed alongside with one another. Consequently, magnetizing current from transformers generated harmonic current and this created an inductive interference with the telephone circuits. Research was thereafter carried out with the purpose of eliminating the problem caused by this innovation. The interference produced is so enormous that the essence of voice communication seems defeated. Upon the completion of the research and study of the problem, a resolution emerged and the problem was addressed by filtering and placing design limits on transformer magnetizing currents (Abdelaziz, Mekhamer and Ismael, 2012)

Now that twisted pair, buried cables and fibre optics had replaced the open-wire telephone circuits and more or less, the installation of rural electrification and telephone systems are no longer placed on the same right-ofway, it is logically expected that situation of harmonics is not experienced but they still persist but not on regular basis as previously experienced (Grady, 2012).

In recent times, the usual sources of harmonics are loads that emanate from power electronics. Examples of such loads are switching power supplies, adjustable speed drives (ASD), uninterruptible power supply (UPS) and static VAR compensators. These load types have active components like the power transistors, silicon controlled rectifiers (SCR), diodes and other semiconductor electronic switches that interferes with the sinusoidal waveforms to control the power or possibly transforms the AC power network to DC power thereby creating non-sinusoidal currents from the main (Abdelaziz, Mekhamer and Ismael, 2012)

The submissions of Fourier had been of great value and had contributed to the analysis of the non-sinusoidal waveform by allowing any periodic function to be used or described in a series of sinusoidal and co-sinusoidal functions(Grady, 2012). For instance, if we are applying a fundamental frequency of 50Hz, then the second harmonics will be 100Hz, and the third, 150Hz, and so on. The respective harmonics can sum up to reproduce the original waveform and the highest harmonics of interest in power system is usually the twenty-fifth and this is in the low audible range (Soni and Soni, 2014)

Over twenty years now, there had been an appreciable and noticeable effort geared towards the analysis of power system harmonics. In this analysis, procedures for simulation methods and component models had been put in place such that the study of harmonics is becoming an important component of power system analysis and design (Durdhavale, 2016). Based on this and the knowledge of digital computers, computer simulation is now the preferred method to conduct harmonic analysis. Computer modeling for power systems for harmonic analysis and computer simulation of harmonic propagation in power systems are the two main aspects of harmonic analysis [7].

In theoretical terms, an ideal power system produces a perfect sinusoidal voltage signal at the load side but such theory is difficult to achieve practically. A distortion in power system is said to have occurred when there is a deviation from the perfect sinusoidal waveform. When this is experienced, then harmonic distortion has occurred. When electrical equipment is working in good and normal condition or when it is not loaded, only odd harmonics are produced but when transient conditions or possible equipment mal-function exist, then the system will generate even harmonics.

The present situation of poor power quality in industrial and commercials sites, especially harmonics had created some form of attention to one of the power problems in the world; harmonics. This situation is as a result of the growth in the use of non-linear loads. The increase in the use of non-linear load had brought about the tight recommendations of IEEE standard 519 especially in the industrial and commercial sector (Robert G.Ellis, 2011). Harmonics in power system should not be treated with levity as they are of great concern in the power sector. Harmonics usually affects power infrastructure and the equipment therein and most times causes a down time in operations as people may think that the problem at that point is load related. Harmonics in power system lead to situation of over-current and over-heating leaving an impression of a possible situation of overload in the power system.

2. Discussion

2.1 Fundamentals of Harmonics

Harmonics can also be known in power system as distorted waveforms and the classification of these waveforms fall in two categories; the voltage and current harmonics. There two concept also used to describe harmonics; they are the orders of harmonics and symmetrical components. Generally, harmonic component is illustrated in the equation below:

.

Where
fn = current amplitude of nth order harmonic
f1= fundamental current amplitude.

Figure 1: Decomposition example of a complex distorted signal, as addition of 50Hz

Fundamental and 3rd , 5th and 7th harmonics (150Hz, 250Hz, 350Hz respectively) (Pinyol, 2015) There is harmonics in a power network if the component of a waveform occurs at an integer multiple of the fundamental frequency. In the description of harmonic orders, the odd order harmonics and the even order harmonics are the common nomenclatures known for classification. However, there is a third classification, the triplet harmonic which is not much known. The table 1 below shows the classification of harmonics in terms of the “order”. Presently, the characteristics of the harmonic component in the power system are the odd harmonics and the odd harmonics are also represented by the waveform that symmetrical to the time axis. Even harmonics are only produced from waveforms that are not symmetrical to the time axis (Durdhavale, 2016).

Table 1: Harmonics Order (Durdhavale, 2016)

.

There are three broad categories in which harmonics can be placed and this is expressed in terms of sequence. The positive sequence harmonics, negative sequence harmonics and the zero sequence harmonics. The positive sequence harmonics are made up of the 4th, 7th, 10th, 13th and 19th , ….3k+1 order harmonics and the negative sequence harmonics are consist of the 2nd , 5th, 8th, 11th, 14th, and 17th …2k+1 order harmonics while the 3rd, 6th 9th , 12th and 15th, 3k+3 order harmonics are attributed to the zero sequence harmonics. Where k ranges from 0, 1,2,3, etc. See Table 2 below for illustration (Kamenka, 2014). In a system where three phase four wire is the wiring arrangement, the zero sequence harmonics system the zero sequence harmonics drifts through the neutral connection and causes an overheating of the conductor (Dash et al., 2014).

Table 2: Symmetrical Component of Harmonics (Kamenka, 2014)

.
Figure 2: Harmonic Distortion for a 50HZ Power System [11]

Fourier postulated a theory and in his theory, we were able to depict that any The Fourier theory tells us that any repetitive waveform is defined as the sum of the sinusoidal waveforms which are integer factor of the fundamental frequency. With the consideration of a steady state waveform with characteristics of equal positive and negative half cycles, the Fourier series can be expressed in equation 2 as shown below (Dash et al., 2014)

.

Where
f(t) = the time domain function.
n = the harmonic number (putting in consideration odd values only).
An = the amplitude of the nth harmonic component.
T = the period or length of one cycle in seconds.

It is worthy of note that harmonics being a steady state phenomenon will repeat every 50Hz or 60HZ cycle, depending on the power system in place. Spikes, dips and other forms of transient do not imply a situation of harmonics and should not be confused with harmonic condition (Robert G.Ellis, 2011).

2.2 Power Quality Indices under Harmonics

i. 2.2.1 Total Harmonic Distortion

The factor that is mostly used to determine the deviation of distorted waveforms from a sine wave is called Total Harmonics Distortion (THD) and this is described in relation with the degree of distortions in both the waveforms of the current and voltage in the power network. The calculation of distortion in the voltage and current is given in equation 3 below. As it may be seen in other publications or books, another term that can be used to describe the Total Harmonic Distortion is known as the Distortion Factor (DF) (Robert G.Ellis, 2011).

.

Where

IDn represent the magnitude of the nth harmonic as a percentage of the fundamental (individual distortion). Grady, in 2012 [4] defined THD as the ratio of the root mean square value (rms) of the harmonic above fundamental to the rms value of the fundamental. Marcuello, Arcega, Plaza, & Ibáñez, in 2011[12] termed THD as the rms value as the proportion of of all harmonic components together to the rms amplitude of the fundamental harmonics. In a white paper by Eaton in 2017 (Eaton, 2017), THD was described as the fraction of the total power of all harmonic components to the power of the fundamental frequency. Similar definition was also given by Durdhavale, in 2016.

Total Harmonic Current (THC)

The summations of current orders 2 to 40 are the cause of distorted waveforms leading to Total Harmonic Current. The value of the Total Current Harmonics (THC) is basis for the installation of active filters. Mathematically, THC can be illustrated as in equation 4 (Durdhavale, 2016).

.

Where Ih is the harmonic current of the nth order.

Total Harmonic Distortion Current (THDi)

The Total Harmonic Distortion Current (THDi) describes the magnitude of distortion present in a waveform It is derived from the fraction of the Total Harmonic Current (THC) and the fundamental current. It is expressed mathematically as shown in equation 5 below (Durdhavale, 2016)

.

Where I1 is the fundamental current

Total Harmonic Distortion of Voltage (THDv)

This is a representation of the total value of the voltage distortion in a given waveform. It can be derived from the ratio of the harmonic voltage to the non-harmonic or fundamental voltage (Pinyol, 2015). THDv can be expressed as:

.

Where Vn = voltage amplitude of the nth order harmonic, V1 = fundamental or non-harmonic voltage amplitude.

Total Demand Distortion (TDD)

Total Demand Distortion is commonly used to describe harmonics in the North America region. It is however defined as the division of the harmonic current by the full load fundamental current. The full load current is also the total non-harmonic current demanded by all load at peak time. Mathematically, it can be written as shown in equation 7 below.

.

Where In is termed as the current amplitude of the nth order harmonic, IL is the total load current demand by the system.

The definitions of THD and TDD are alike only that THD compared the harmonic content with measured fundamental current while TDD evaluates its distortion with the maximum demand current. Considering the definitions of the fundamental current (I1) and the maximum demand current (IL), the value of IL is greater than I1 for harmonic measurement purpose, hence the value of TDD and the percent of IL will tend to be lower than THD and its corresponding percent of I1 measurement. The determination of the values of THD and TDD is relevant as it assist in determining the accurate value of harmonic generated by a facility’s power network during the time where light loads are being used. (IEEE Recommended Practices and Requirements for Harmonic Control in Electric Power Systems, 1992)

Partial Weighted Harmonic Distortion (PWHD)

PWHD can be expressed as the ratio of the voltage or current within a selected group of harmonics higher order (say 14th order to 40th order) to the fundamental values of the current or voltage, as the case may be. PWHD equations for current and voltage can be given as shown in equation 8 and 9 below (Durdhavale, 2016)

.

Where I1 can be expressed as the fundamental current amplitude, and V1, the fundamental voltage amplitude.

2.3 Sources and Causes of Harmonics Distortion

Harmonics come into play as a result of non-linear loads in the power network. In recent times, enhanced power semiconductor technology coupled with power electronics devices are now used in various applications in the field of electrical and electronics engineering. These applications are in the classification on non-linear loads and thus the devices demand current with harmonic content and reactive power from the AC component of the power network (Panda et al., 2013).

For better understanding of the sources and caused of harmonic distortion, it is important to explain briefly the term “non-linear load”. We can say that a load is non-linear when the current demand from such load does is not even with the connected sinusoidal voltage. This implies that the Ohm’s law is not applicable in describing the Voltage-Current relationship as the resistance is not a constant value any longer and there will be a change in current value with each produced sinewave of the applied voltage waveform. This situation causes several positive and negative pulses. The varying current values are termed to be non-linear and they produce frequency components that are manifolds of the frequency of the power system. Besides the fact that the non-linear current produces multiples of the frequency of the power system, they also form a network with the impedance of the electrical power supply to form voltage distortion that can affect the power network and the load on it (Kamenka, 2014).

The simplest network that can be used as an illustration for a non-linear load is a diode-rectifier in its various applications such as the half-wave diode rectification, full wave diode rectification in both single phase and three phase network (Pinyol, 2015). Table 3 below shows the various classification of non-linear load responsible for the generation of harmonic distortion in a power network.

Table 3: Non-Linear Loads (Kamenka, 2014)

.

Prior to now, equipment with magnetic iron cores like electric motors, transformers and generators were known to be predominant in the cause of harmonic distortion in power network. Likewise, the arc furnaces and the arc welder’s equipment also cause harmonic distortion in power network. In recent times, where energy efficiency is crucial, there had been some introduction of power electronic equipment to enhance energy efficiency utilization and these equipment had contributed to the most serious source of harmonics within a power network in industrial and commercial facilities (Kamenka, 2014).

ii. 2.3.1 Transformers The magnetization curve explains the correlation between the input or primary voltage and current of a transformer. A transformer is working in normal condition does not generate any form of harmonic distortion unless the transformer is in a core saturation condition. At this condition, the harmonic distortion increases appreciably with the odd orders of harmonics, the third order being dominant. This kind of condition is show cased when the transformer is operating in an overload condition, especially during peak periods or when the transformer is subjected to an overvoltage condition, especially at situation of low power demand or possible switching of large reactive power load. This harmonic content is manifested in the magnetization curve of the transformer as in figure 3 below. When the transformer is working in normal condition, a little increase in the voltage usually result in a little rise in the magnetization current. Similarly, when an overvoltage condition exist, that is when the voltage is above the nominal voltage, then a small increase in voltage will result in having a large increase in the magnetization current (Kamenka, 2014).

Figure 3: Transformer Magnetization Curve (Kamenka, 2014)

iii. 2.3.2 Rotating Machines

Generators and Motors are examples of rotating machines and they also produce magnetizing field like transformer, hence they are capable of producing harmonics in power network. Though the harmonic content of produced as a result of the magnetization curve of motors is much more linear when compared to that produce by a transformer, thus, the harmonic content is not disturbing, however, the higher capacity motors have capacity of generating high harmonic content. On the other hand, generators produce observable voltage harmonics following the unpractical nature of the spatial distribution of the stator winding. The voltage harmonics produced from a generator are the 3rd order harmonics which in turn causes creates a 3rd order current harmonic in the power network (Durdhavale, 2016).

iv. 2.3.3 Arc Furnaces and Arc Welders

Arc furnaces and arc welders are high power consuming equipment that also produce corresponding high level of harmonic distortion in a power network. Arc furnaces are applicable in air refining, refining and melting applications and these phases produces different levels and gradients of harmonics. The random variation of the arc produces a combination of ignition delays and voltage changes and this situation creates some sort of harmonic spectrum with odd and even multiples of fundamental frequency. These frequencies changes intermittently with various swift levels of rise and falls (Suresh and Babu, 2015)

Figure 4: Harmonic Current Spectrum in Arc Furnace (Kamenka, 2014).

v. 2.3.4 Switched Mode Power Supplies (SMPS)

Most of the electronic devices of today are embedded with Switched Mode Power Supplies and the name is on the basis of the switching and conversion of unregulated DC input voltage to a regulated DC output voltage. Based on the SMPS principle of operation where rectifying and filtering at various stages is involved, this result in the demand of pulses of current instead of continuous current. This pulse is made of high content of harmonics of the third order and even higher orders. A typical waveform and the harmonic spectrum is shown in figure 5 below (Durdhavale, 2016)

Figure 5: Waveform and Harmonic Current Spectrum of SMPS (Kamenka, 2014)

vi. 2.3.5 Variable Frequency Drives

Variable frequency drives are applicable to equipment that uses the technology of static converters in a three-phase bridge system. We can term the bridge as six-pulse bridge or B6 bridge. This same technology is also applied in UPS application as well as inverter applications, basically AC to DC converters. The term B6 as earlier mentioned came to being from the trend of pulse generated, six voltage pulses per cycle which corresponds to the production of one pulse in a half cycle per phase. The harmonic spectrum in this case is attributed to the magnitude of pulses of the variable frequency drives. The current harmonics that emanates from a B6-bridge is in the 6n ± 1 orders which implies that we can either have orders in the form of 5th and 7th, 11th and 13th, 17th and 19th, just to mention a few. As the harmonic spectrum is dependent on the number of pulse, so the harmonic spectrum will be different if 12 or 18 pulse converter is applied for the application. See table for more information. Figures 6 illustrate how the waveform looks like; its resulting harmonic spectrum is also included.

Table 4: Pulse and Harmonic Spectra in a Variable Speed Drive (Kamenka, 2014)

.
Figure 6: Waveform and Harmonic Spectrum of a B-6 Variable Speed Drive (Kamenka, 2014)

2.4 Effects of Harmonics Distortion

Implications of harmonics within the power network can be categorized in terms of their duration, for instance, it could either be short term or long term. The failures or malfunctions of equipment or devices subjected to harmonic distortion can be categorized as short term effects while long terms effects can be linked with the thermal behavior of the devices or equipment. Harmonic situation lead to scenarios of thermal build-up or temperature rise in equipment and electrical network. When there are situations of increased temperature of electric or electronic devices, cables, motors arc, then besides the effect of higher losses, the equipment or system life becomes reduced (Suresh and Babu, 2015).

vii. 2.4.1 Power Factor

When there are situations of harmonic distortion in a power network, the power factor of the network gets affected following a higher demand in the level of current. The power factor becomes exceedingly low if there is an increase in the phase shift between the voltage and current, certainly with a situation of current harmonics within the system. Power factor can simply be explained in terms of the relationship between the active power and the apparent power. It can be defined as the ratio of the true power (in watts) to the apparent power (in VA). It determines or measures the efficiency of the energy used by a load in a power network. Power system with a high power factor demands less current when compared to that with a low power factor under the same power conditions. This implies that systems with higher power factor are more efficient than those of lower power factor. The effects of harmonic distortion which is predominant in non-linear load tend to cause a situation of bad power factor, thus lowering the efficiency of the system (Kamenka, 2014).

viii. 2.4.2 Phase and Neutral Conductors

Practically, three phase system are such that they possess phase angles and in an ideal system, the phase voltages are displaced by 1200 from one another. If the system is subjected to load and the individual phases are equally loaded, the resultant neutral current will be zero but if there is presence of current harmonic in the system, then the triplen harmonic will sum up in the neutral link so much that the total current in the neutral surpass the individual current in each of the phases, to a factor of three. This scenario may cause a situation of overload on the individual current and the neutral current. This can result in the overheating of the cables and the conductors may eventually get burnt (Kamenka, 2014).

ix. 2.4.3 Transformer

Transformers are devices that are widely known for the supply of electric power to facility loads which includes both linear and non-linear loads. Transformers are affected by harmonics in two specific ways; there are situations of additional losses within the transformer core and production of triplen harmonic current. The losses generated are a result of eddy current, resistive and magnetic losses in the core. These additional losses produce extra heat within the system which overtime reduces the operating life of the transformer insulation. For the purpose, especially in industrial applications with load that are non-linear, transformers cannot be put into use at full load because of the high level of harmonic distortion (Soni and Soni, 2014).

x. 2.4.4 Rotating Machine

Similar to the effect of harmonics on transformer, harmonics also result in additional power loss in motors and generators. The effect of the losses is such that there is high temperature build up within the devices as a result of the increase in the resistance of the system which is directly proportional to the rate of frequency increase. Hence, the harmonic current will lead to increased losses in the windings of the rotating machines. Another effect of harmonics on the rotating machine is the generation of higher vibrations inside the bearings and this can cause wear and tear within the system with an eventual earlier equipment weakness (Pinyol, 2015).

xi. 2.4.5 Circuit Breakers

Harmonic tend to cause an increase in current in power network and tend to create situations of incessant tripping in the system thereby disrupting operations. A residual current circuit breaker (RCCB) functions to sum up the current in the phase and neutral conductors and disconnects the power from load peradventure the summation does not fall within the rated limit. Harmonic situation disrupt this operation as the RCCB may not be able to add the high frequency component correctly leading to nuisance tripping that can result in shutdown in production process, loss of time and money (Ritesh Dash, Kunjan K. Mohapatra, Patrik R. Behera, 2014)

xii. 2.4.6 Power Factor Correction Capacitors (PFC)

The factors that lead to the dielectric breakdown of capacitors are temperature, voltage, current and overload in power. Situation of harmonics seriously have adverse effect on PFC capacitors such that an increase in the maximum value of voltage based on high harmonics creates additional dielectric stress, thereby causing partial discharge in the insulation and permanently damaging the capacitors. In most cases, issues that relate with capacitor performance or behavior are related to current. Also, the impedance with respect to the voltage harmonics decreases as the harmonic order increases because the relationship of the capacitive reactance to the frequency is not linear. Therefore, capacitors with voltage harmonics absorb higher current when compared with capacitors in a system without voltage harmonics. In essence, this implies that voltage harmonics in a system give rise to a situation of high current draw in capacitors and causes additional losses, quick aging of the insulation and the eventual damage of the PFC. This effects are aggravated if the harmonics are multiples of parallel or series resonance (Ciurro, 2009)

xiii. 2.4.7 Lights

Incandescent lamps are generally non-linear loads and the useful life or age decreases with the level of harmonics present. Lamps with features comprising of ballast inductor or capacitor do present a resonance problem which causes harmonic distortion; thus if the lamp is operated at about 105% of its rated voltage, then its useful life falls by an approximate value of 47% (Suresh and Babu, 2015).

2.5 Harmonics Mitigation Techniques

In line with the sources and causes of harmonic distortion in power system, various elimination or mitigation techniques to trap or limit the occurrence of harmonics with different levels of effectiveness and efficiency are considered below.

1. Introduction of Active Harmonic Filter The use of active harmonic filters in a power system introduces current component that can negate the effect of the harmonic contents in a non-linear load. Active harmonic filters come in different forms such as the series filter, shunt filter and hybrid filter but the most recent technology of harmonic filter available is the active filter. Active filtering technique is applicable in standalone applications or by installing the design in the input stage of a drive, UPS system or other power electronic devices. Considering all technologies in UPS application except from the use of Insulated Gate Bipolar Transistor (IGBT), the harmonics generated in these system are greater than the expected for most electrical system; hence the need for the input filter to reduce the harmonic content to a value which is less than 10 percent of total harmonic distortion of the input current. The addition of a transformer in the system will produce more inductance to the line which will additionally lower the harmonic distortion in the system (Steele, 2015).

Ideally, the active filter operates in such a manner that it observes the load current and eliminates the fundamental frequency current after which it will investigate remaining frequency content and the respective magnitude. Based on this analysis, the active filter will be able to pass in the required equal and opposite current to eliminate the different harmonics. The active filter has the capacity to cancel harmonics to the 50th order and achieve a low level of current harmonic distortion to a value as low as 5%. In order to use active harmonic filter in the power network, there is the need to take a measurement of the harmonics to be cancelled from the system after which an active filter that possesses the magnitude of harmonic current required to cancel the measured harmonic can be selected (Soni and Soni, 2014)

2. Use of K-Rated Transformers In Power Distribution Components

Generally, transformers are not manufactured and designed to operate under high harmonic content produced from the presence of non-linear load in the power network. The effect of this harmonics is such that it raises the magnetization core of the device leading to overheating and eventual sudden failure when in use. The introduction of harmonics into power system brought about the development of the K-rated transformers. K-rated transformers are not design to eliminate harmonic, rather they are designed to handle the generated heat from the harmonic current. K- Factor values are within the range of 1 to 50 whereas a standard transformer that supplies linear load has a K-factor of 1. A high value of transformer’s K-factor implies that the transformer can handle higher the degree of thermal energy generated from the harmonic current. The choice of transformer in terms of the K-factor is important in installation design and this is also dependent on the magnitude of harmonic content in the power network. Table 5 below shows the K-factor ratings that are applicable for different percentage of harmonics (Eaton, 2017).

Table 5: K Factors Rating for Different Non-linear Load in Electrical System (Eaton, 2017)

.

3. Introduction of Line Reactors

In order to control the harmonic distortion produced in a Variable Speed Drive (VFD), there is the need to connect a series reactor with non-linear load at the input line of the drive (Ritesh Dash, Kunjan K. Mohapatra, Patrik R. Behera, 2014)

A line reactor, also called input AC reactors can simply be explained as an inductor that is connected in series between a load and the source. . It functions to limit the current the current harmonic characteristic thus reducing harmonic in the system (Rockwell and Wisconsin, 2016).

A line reactor also helps to mitigate the harmonics which the VFD creates back into the line. The rating of line reactor is in horse-power (hp) and the voltage rating of the drive is applicable for use. The figure below is an illustration of a VFD circuit of motors showing the AC and DC reactors. (Lenz, 2008)

Figure 7: Circuit of a VFD of Motors Indicating the AC and DC Line reactors (Pinyol, 2015)

4. Over-sizing or Derating of the Installation

In reducing the effect of harmonics in a power system, the solution that is mostly implemented by technical personnel is to dimension for an oversized neutral conductor. However, where there is an existing installation, the solution is to reduce the capacity of the electrical distribution equipment experiencing the harmonic distortion. In modern times, the dimension of the neutral wire is always rated in the same size as the individual phase wires and may even be dimensioned (Soni and Soni, 2014).

2.6 Recommended Harmonic Limit – IEEE Std. 519™-2014

Both the end users and power system operators should be responsible for the management of harmonics in power network. Harmonic limits recommendation span through the voltage and current parameters in a power network and the recommended values are on the premises that some level of voltage distortion are generally acceptable and it is the responsibilities of all parties involved to ensure that the actual voltage distortion is kept below the generally accepted limits. The basis of this recommended limit is that in the act of limiting the harmonic current in a system, and then voltage distortion can be within the recommended level (Committee, Power and Society, 2014).

The application of the recommended limit is applicable only at the point of common coupling and cannot be used in the applications of either individual equipment or locations in the industry or facilities. The various recommended limits for voltage and current harmonics are as shown in the following tables below.

Table 6: Recommended Voltage Distortion Limit (Committee, Power and Society, 2014)

.

Table 7: Recommended Current Distortion Limits for Systems rated 120V through 69kV (Committee, Power and Society, 2014)

.

Table 8: Recommended Current Distortion Limits for Systems rated 69kV through 161kV (Committee, Power and Society, 2014)

.

Table 9: Recommended Current Distortion Limits for Systems rated >161kV(Committee, Power and Society, 2014)

.

Based on the indices on Table 6 to 9, the following are defined as:
a= Even harmonics are limited to 25% of the odd harmonic limits above.
b= Current distortions that result in a dc offset, e.g., half-wave converters, are not allowed.
c= All power generation equipment is limited to these values of current distortion, regardless of actual Isc/IL.

Where Isc relates to the peak short-circuit current at PCC, IL is defined as the peak demand load current (fundamental frequency component) at the PCC under normal load operating conditions.

2.6.1 Recommendations for Increasing Harmonic Current Limits

Following recommendations, the values as seen in Table 7, Table 8 and Table 9 can be increased by a multiplying factor when the user decides to reduce the magnitude of lower-order harmonics. The multipliers are given in Table 10 below and are put into use when steps are taken by the user to reduce the harmonic order shown in the first column of the table.

Table 10: Recommended Multipliers for Increasing the Harmonic Current Limit (Committee, Power and Society, 2014).

.
3. Conclusion

This paper presented research on harmonic distortion in the voltage and current waveform with respect to the existing level of harmonic distortion in the power system at the moment and the possible look at how harmonic will be portrayed in future.

Harmonics in power system is caused by the presence of non-linear load within the network. The different categories of equipment that give rise to harmonic distortion have numerous harmonic spectra and the peculiar harmonic spectra related to particular type of load can only be determined by having the requisite knowledge and experience in harmonics. Most of the harmonics generated from non-linear load are predominant in the electronics components which form the basis of modern technology. The demand for these types of equipment may result in serious problem in the future and the harmonic generated from these systems will significantly affect the power quality.

Harmonics in power system should not be treated with levity as they are of great concern in the power sector. Harmonics usually affects power infrastructure and the equipment therein and most times causes a down time in operations as people may think that the problem at that point is load related. Harmonics in power system lead to situation of over-current and over-heating leaving an impression of a possible situation of overload in the power system.

The problem of poor power quality relating to harmonics is not often noted by most practicing electrical engineers and consulting engineering firms and such problems if not critically analyzed would have impeded operations and establishment would have incurred a high cost in finding a solution; with an eventual damage to equipment. Good understanding of the causes, potential effects and means of mitigation can assist in the reduction of harmonic in the power system especially in the design stage of power infrastructure and the probabilities of undesired effect occurring can be reduced.

Acknowledgement – I am grateful to the school of Science and Engineering at Atlantic International University, Honolulu for giving me the required platform to be able to complete the research and analysis of this paper. I also acknowledge my team members at Powerex Limited for their understanding and cooperation while carrying out this research paper.

Profound gratitude goes to my wife and children; Ogunyemi Moyofoluwa, Oluwatofunmi and Oluwatoni for their support and understanding towards theh completion of this paper. You are greatly acknowledged.

References

Abdelaziz, A. Y., Mekhamer, S. F. and Ismael, S. M. (2012) ‘Sources and Mitigation of Harmonics in Industrial Electrical Power Systems : State of the Art’, (October).
Ciurro, N. (2009) ‘Harmonics in Industrial Power Systems’, (June).
Committee, D., Power, I. and Society, E. (2014) ‘IEEE Recommended Practice and Requirements for Harmonic Control in Electric Power Systems IEEE Power and Energy Society’, 2014.
Dash, R. et al. (2014) ‘FUNDAMENTALS OF HARMONICS AND ITS EFFECT ON’, 3(4), pp. 790–794.
Durdhavale, S. R. (2016) ‘A Review of Harmonics Detection and Measurement in Power System’, 143(10), pp. 8–11. Eaton (2017) ‘Harmonics In Your Electrical System’, What they are, how they can be harmful, and what to do about them, pp. 1–7. doi: 10.1164/ajrccm-conference.2017.A35.
Grady, P. M. (2012) ‘Understanding Power System Harmonics’, Chapter 1, (April).
IEEE Recommended Practices and Requirements for Harmonic Control in Electric Power Systems (1992) ‘IEEE std 519-1992’, Ieee, pp. 1–9. doi: 10.1109/PAPCON.2006.1673767.
Kamenka, A. (2014) ‘Six tough topics about harmonic distortion and Power Quality indices in electric power systems A white paper of the Schaffner Group Written by Alexander Kamenka’. Lenz (2008) ‘Application Note- When to Use a Line or Load reactor’, (July), pp. 1–3.
Marcuello, J. J. et al. (2011) ‘A Review of Teaching Power System Harmonics’, 1(2), pp. 1–5.
Ogheneovo Johnson, D. (2016) ‘Issues of Power Quality in Electrical Systems’, International Journal of Energy and Power Engineering, 5(4), p. 148. doi: 10.11648/j.ijepe.20160504.12.
Panda, G. et al. (2013) ‘Novel schemes used for estimation of power system harmonics and their elimination in a three-phase distribution system’, International Journal of Electrical Power and Energy Systems. Elsevier Ltd, 53, pp. 842–856. doi: 10.1016/j.ijepes.2013.05.037.
Pinyol, B. R. (2015) ‘HARMONICS : CAUSES , EFFECTS AND MINIMIZATION’, (August), pp. 1–32.
Ritesh Dash, Kunjan K. Mohapatra, Patrik R. Behera, M. R. S. (2014) ‘A REVIEW ON POWER SYSTEM
HARMONICS School of Electrical Engineering , Kiit University’, Indian Journal of Research, pp. 1991–
1992.
Robert G.Ellis, P. E. (2011) ‘POWER SYSTEM HARMONICS A Reference Guide to Causes ’, pp. 1–3.
Rockwell, A. J. T. S. and Wisconsin, M. (2016) ‘Line Reactors and AC Drives .’
Soni, M. K. and Soni, N. (2014) ‘Review of Causes and Effect of Harmonics on Power System’, 3(2), pp. 214– 220.
Steele, B. J. (2015) ‘Mitigate damaging harmonics’, (August), pp. 39–41.
Suresh, N. and Babu, R. S. R. (2015) ‘Review on Harmonics and its Eliminating Strategies in Power System’, 8(July). doi: 10.17485/ijst/2015/v8i13/56641.

Source & Publisher Item Identifier: Journal of Energy Technologies and Policy http://www.iiste.org ISSN 2224-3232 (Paper) ISSN 2225-0573 (Online) Vol.9, No.3, 2019. Publication date: March 31st 2019. DOI: 10.7176/JETP/9-3-02, https://core.ac.uk/download/pdf/234668516.pdf.

A Review of Harmonics Detection and Measurement in Power System

Published by 1. Dnyaneshwar D. Ahire, Professor, Matoshri College of Engineering and Research Centre, Nashik Savitribai Phule Pune University, Nashik, Maharashtra, India.
2. Snehal R. Durdhavale Student, Matoshri College of Engineering and Research Centre, Nashik Savitribai Phule Pune University, Nashik, Maharashtra, India.

ABSTRACT – A load is said to be “linear” when it draws a current from the supply which is proportional to the applied voltage (linear). And in the case of “non-linear” load, impedance changes with applied voltage. The current drawn from such non-linear load is also non-linear i.e. non-sinusoidal even when it is connected to a sinusoidal voltage source. Harmonic currents contents which are present in non-sinusoidal currents intermingle with the impedance of the power distribution system to create voltage distortion which affects the distribution system and the loads connected to it. The serious power-line pollution is a result of increasing use of power electronic systems and time-variant nonlinear loads in industry. Hence, power supply quality is degraded. It results in the reduction of system efficiency, apparatus overheating, and increase power. As the utilization of the number of harmonics-producing loads has increased over the years, it has become highly mandatory their influence and analysis when making any additions or changes to an installation. In this paper various harmonics detection and measurement techniques have been outlined.

Keywords: Non-linear loads, harmonic currents, power distribution system, voltage distortion, power signal quality, harmonic distortion.

1. INTRODUCTION

Power quality can be defined as a set of electrical boundaries allowing an equipment to function in its intended manner with no significant loss of performance or life expectancy. Various methodologies and techniques were proposed to improve the power quality. A power system ideal when is define when a perfect sinusoidal voltage signal is seen at load-ends. Practically, such idealism is really hard to maintain. Any deviation from the perfect sinusoidal waveform is nothing but distortion and hence harmonic distortion. It has been said about harmonics is they are voltages or currents with frequencies which are integer multiples of the fundamental power frequency. Only odd harmonics will be produced by electrical equipment’s, when working in normal or no load condition. When transient conditions or conditions of mal-function or single-phase rectification appear, even harmonics may occur. One of the parameters which affect the quality of power is harmonics current are supplied by the non-linear equipment, which disrupts the desired linear system. These distorted current pulses, due to Ohm’s law, will also instigate to distort the voltage waveforms, where these distortions would be carried back to the distribution network. Common risks of harmonics include potential fire hazard, excessive heat, false tripping of branch circuit breakers and consequently increases maintenance cost [2], [3].

1.1 Basics

In any power system, it is highly impossible to accomplish a perfect or pure sinusoidal waveform at every point of a network. The voltage and current waveform deviate massively from a sinusoidal waveform. These waveform deviations are usually called harmonic distortion.

“Harmonics” is the term which means waves having frequencies of integer multiples of one another. The harmonic component in an AC power system is nothing but the sinusoidal component of a periodic waveform that has a frequency of an integer multiple of the fundamental frequency of the system. It can be given as:

fh= n*fundamental frequency

Where, fh= harmonic order, n= integer, and the fundamental frequency is either 50Hz or 60Hz.

For example, if a system has the fundamental frequency as 60Hz then its 2nd and 3rd harmonic would have frequencies of 120Hz and 180Hz respectively[1],[[2].

Figure 1: Harmonic distortion of the electrical current waveform [2]

1.2 Classification of Harmonics

Harmonics which are nothing but distorted waveforms have two types namely voltage and current harmonics. The orders of harmonics and symmetrical components these are two concepts which are used commonly to describe harmonics. Regarding the harmonics, words odd and even harmonics are used usually but the term triplen harmonics is not much known. Table shows harmonic orders:

Table 1. Harmonic Orders

.

In the present scenario, odd harmonics are the characteristics harmonic components in the power network. Waveforms that are symmetrical to the time axis are represented by odd harmonics. In the case of even harmonics, they can only arise from waveforms that are not symmetrical to time axis [16].

1.3 Power Quality Indices under Harmonic Distortion

[16] Generally, representation of harmonic components is given with equation:

.

Where 𝑓𝑛= current amplitude of nth order harmonic, 𝑓1=fundamental current amplitude.

1.3.1 Total Harmonic Distortion (THD)

This notification is used widely in defining the harmonic content level. It is given as the ratio of the power of all harmonic components to the power of fundamental frequency.

1.3.2 Total Harmonic Current (THC)

Usually, distorted current waveform is caused by the contribution of current orders 2 to 40. THC value is used for installation of active filters. It can be written as:

.

1.3.3 Total Harmonic Distortion Current (THDi)

This value gives the total harmonic distortion of the waveform. This value can be calculated by taking the ratio of THC to the Fundamental current. It can be given as:

.

Where 𝐼1= fundamental current

Where 𝑉𝑛= voltage amplitude of nth order harmonic, 𝑣1=fundamental voltage amplitude. For the sake of good voltage quality, its value should be low.

1.3.4 Total Harmonic Distortion of Voltage (THDv)

It shows the total magnitude of the distortion in voltage. It can be calculated by calculating ratio of distorted or harmonic voltage to the non-harmonic or fundamental voltage. It can be written as:

.

1.3.5 Total Demand Distortion (TDD)

This concept is used widely used in North America regarding harmonics. It is the ratio of harmonic current to the full load fundamental current. The full load current is nothing but the total non-harmonic current consumed by all loads by the system when the system is on its peak demand.

.

Where 𝐼𝑛= current amplitude of nth order harmonic, Il= total load current consumed by system

1.3.6 Partial Weighted Harmonic Distortion (PWHD)

PWHD is the ratio of current or voltage with selected group of higher order harmonics from 14 to 40 to the fundamental value of voltage or current. PWHD for current and voltage can be given as:

.

Where, 𝐼1= fundamental current amplitude, V1= fundamental voltage amplitude.

1.4 Sources of harmonic Distortion

There are many harmonics sources are present but out of them few are listed here which play a role as the major sources of harmonics [14],[16].

1.4.1 Static Compensators

If the power source is fluctuating, static compensators are used at the ends of transmission lines or near sources of fluctuating power, static compensators manage the voltage. Reactors which are controlled by Thyristor will produce near about 1% of the 11th harmonic current.

1.4.2 Power Converters

Rectifiers give higher inductance on the dc side compared to the ac side. Hence the dc current is almost constant and then converter starts acting as a harmonic voltage source on the dc side where as the harmonics current source on the ac side.

1.4.3 Transformer

Because of saturation and hysteresis characteristics, a small level of harmonic current will get produce by transformers when they are in steady state. Initially high level of harmonics will be produced, which is 60% of the rated transformer current.

1.4.4 Rotating Machines

In the rotating machines, harmonic currents can be produced due to asymmetries in the winding pattern. Harmonics grow because of the resultant magneto motive force in the machine. Due to magnetic core saturation harmonic currents are generated.

1.4.5 Electric Arc Furnace

As the arc feed material varies, the harmonics rise up and their value cannot be predicted certainly. The electric arc furnace load gives most awful distortions a result of melting with the moving electrode and molten material.

1.4.6 Switched Mode Power Supplies (SMPS)

Latest electronic devices contain switched mode power supplies. SMPS regulates AC or DC input voltage. SMPS unit draws current pulses contain large amount of harmonics of third and above higher order harmonics.

1.5 Effects of Harmonics

Harmonics affect the power equipment’s and components. Following effects are shown on different components [14], [16]:

1.5.1 Power Factor

Harmonic distortion affects the power factor. Power factor gets worse with increasing amount of harmonic distortion. Generally, non-linear loads result in poor power factor.

1.5.2 Electric and Electronic Equipments

Basically, these equipments are considered as a source of harmonics. These devices are sensitive to harmonic distortion. They show effects as increase in supply voltage, zero crossing noise, malfunction of protective devices etc.

1.5.3 Conductors

On a regular basis, heat will be generated in the current carrying conductors due to I2R losses. As the harmonic orders increase, skin effect is produced. As the skin effect increases more I2R losses cause over heating of the conductors. Heating of conductors may also occur because of the magnetic field of harmonic currents in the neighboring conductors.

1.5.4 Transformers

Frequency causes Eddy current losses. Hence, as the harmonic order increases eddy current losses for transformers also increase. In addition to the skin effect, eddy current losses in transformer fallout in overheating and the life of the transformer would be reduced.

1.5.5 Capacitance

The Capacitors improve the power factor. They have a significant influence on harmonic levels. As the frequency of harmonics increases, the capacitive reactance decreases. As increased flow of current increase, the capacitor may get congested and impose higher dielectric stress.

1.5.6 Circuit Breakers and Fuses

Low level faults in circuit breakers caused because of the high degree of harmonic load current. High 𝒅𝒊/𝒅𝒕 ratings at zero crossings for sinusoidal waveform make the disruption complex, for load distortion. Hence, harmonic load currents results in circuit failures.

1.5.7 Lights

The distorted power supply decreases the life of the lamp gets decreased with. Harmonic currents give problems to audible noise in the case of discharging of lamps. In equipped with capacitors, together with the ballast inductor and the lamp may form a resonance problem.

1.5.8 Rotating Machines

Operating frequency plays a vital role in losses produced in the electric machines. Core and stray losses become significant for induction motor for an inverter producing high harmonic frequencies. The increase in temperature in the windings causes the lessening of life of the rotating machines. Communication between the air gap flux density and the fluxes generated by the harmonic currents in the rotor, pulsating torques are produced. By reason of the difference between time harmonic frequencies audible noises are formed. Additional problems in rotating machines caused by harmonics are equipment failure, bearing wear out, etc.

1.5.9 Telephone Interferences

Fundamental frequency doesn’t cause any serious problems but power system harmonics can cause huge problems because human audible sensitivity and telephone response peak have near 1 KHz. Inductive, capacitive and conductive interferences can be occurred between telephone line and a power line.

1.6 Why Harmonics should get detected?

Basically, harmonics are difficult to reduce. But the power quality gets reduced because of harmonics. They show economic impacts such as earlier failure of equipments, losses in distribution systems. So, they should be detected at early stage.

2. LITERATURE REVIEW

Tremendous work has been done for harmonics, their analysis and various mitigation techniques of harmonics. A brief review on this: For the reliable and efficient operation of any system a properly designed electrical system is necessary. And the system should be harmonic free.

For this purpose, capacitors in harmonic environment are applied. They are beneficial because they result in minimized THD, improved power factor and elimination of power factor penalties [3]. Lucian Asiminoae, Sergej Kalaschnikow and Steffan Hansen have discussed two harmonic detection methods. The methods are selective harmonic compensation and overall harmonic compensation [4].

An innovative method is presented for measurement of individual harmonics of a time-varying frequency. This proposed method is based on a nonlinear, adaptive mechanism. This technique offers the higher degree of accuracy, frequency-adaptivity [5]. David M. McNamara, Alireza K. Ziarani presented a new method of measurement of harmonics of time-varying frequency. This proposed method is based on the adaptive evaluation of the fundamental frequency and its harmonic components of the power signal [6].

A system made from a combination of the ARM9 chip and virtual instrument technology is designed for a real-time harmonic measurement. This system is presented in the paper [7].Frequency is a significant factor for harmonics measurement. The paper contains a review of several commonly used methods for power system harmonics measurement. And those methods are compared according to the aspect of frequency identification [8]. This paper gives a new idea for harmonic detection adopting the algorithm with combination of FFT with and wavelet transform. This instrument can obtain parameters of harmonic [9]. Hsiung Cheng Lin developed a strategy of recursive group-harmonic power minimizing for system harmonic and interharmonic evaluation in power systems. The proposed algorithm can measure integer harmonic and the interharmonics also identified accurately [10].

Harmonic components and harmonic distortion can be calculated using distortion meter. This paper presents the harmonic distortion meter based on microcontroller and its software part carries out calculations using DFT. DFT is used to find amplitude in order to measure THD in power system [11]. In this review paper an author has discussed abundant for selective harmonic detection methods in frequency domain as well as in time domain like DFT, FFT, SOGI technique and CDSP-PLL systems [12]. To estimate the fundamental frequency and to measure both harmonics and inter harmonics of any unknown frequency is not an easy task. But using the adaptive notch filter this can be done. This methodology measures fundamental frequency and harmonic and inter harmonic components fast [13].

Usage of non-linear loads in power system results in poor power quality. These loads are leading to harmonic sources; and this has become much serious problem. One of the widely used algorithms for harmonic analysis is Fast Fourier Transform (FFT). In this project, a harmonic analyzer is implemented using FFT on ARM7 core processor (LPC2138). For matching power rating the supply voltage is divided to 6V using the voltage divider. This harmonic analyzer can analyze harmonics in single phase supply and gives frequency spectrum of harmonics. This system has the advantage of being available in at low cost [15].

A constant wave Terahertz spectrometer is integrated with 1X2 LiNbO3 which is fiber coupled and customized optical phase modulator which allows direct modulation of Terahertz (THz) beam and measurement of the 1st and 2nd harmonics of modulation. Thus, using optical phase modulation rather than bias modulation harmonics measurement is carried out [17].

We know that harmonics is a very basic property of power quality. So it has become necessary a thing to measure these harmonics. Instead of using traditional measurement device a new method to detect and measure harmonics is presented. This device consists of the analog to digital converter, FFT unit, LCD display unit, and network communication unit. This methodology adopts FPGA and DSP processor. Experimental results show that using presented device more accuracy is obtained and harmonic power flow is also analyzed [18].

3. PROPOSED WORK

The various harmonic mitigation strategies adopted in the last three decades have been reviewed. Based on this survey a new methodology to control harmonic distortion in power system is introduced. In the proposed method harmonics get detected using ARM7 core processor (LPC2478). The software side performs FFT calculations for getting the amplitude of the fundamental frequency and the nth order harmonic. The distortion is calculated using the ratio of the amplitude of measured harmonic to the fundamental frequency.

The benefits of the proposed optimization method are:

1. Detection of harmonics in easier way
2.Correct measurement of harmonics and THD

4. CONCLUSIONS

Non-linear loads result in harmonic distortions in the power system and the associated problems were discussed briefly. Comprehensive review related to various methodologies to detect and measure harmonics in power system that was mentioned in the literature is done. Based on this review a new hybrid optimistic method to detect and measure harmonics is introduced in this paper. The system is designed for detection and measurement of harmonics on ARM 7 platform. Proposed system uses FFT algorithm for measuring total harmonic distortion.

5. REFERENCES

[1] Harmonics Detection and Filtering, Low Voltage Expert Guides by Schneider Electric.
[2] Harmonics in Your Electric System, A White Paper of Eaton Corporation.
[3] Douglas Andrews, Martin T. Bishop, John F. Witte,(May-June 1996), Harmonic Measurements, Analysis, and Power Factor Correction in a Modern Steel Manufacturing Facility, IEEE Transactions on Industry Applications, Vol. 32, No. 3.
[4] Lucian As iminoae,, Sergej Kalaschnikow and Steffan Hansen, Two harmonic detection methods used in industrial shunt active filters
[5] Masoud Karimi-Ghartemani, and M. Reza Iravani,( January 2005), Measurement of Harmonics/Inter-harmonics of Time-Varying Frequencies, IEEE Transactions On Power Delivery, Vol. 20, No.3.
[6] David M. McNamara, Alireza K. Ziarani, Thomas H. Ortmeyer(January 2007), A New Technique of Measurement of Nonstationary Harmonics, IEEE Transactions On Power Delivery, Vol. 22, No.1
[7] Weicheng XIE, Xia YANG, (2010), A Power Harmonic Measurement System Based on Wavelet Packet Transform and ARM9, IEEE.
[8] Gary W. Chang, Senior Member, IEEE and Cheng-I Chen, (2010), Measurement Techniques for Stationary and Time-Varying Harmonics, IEEE
[9] Shouxi Zhu, Wenlai Ma, (2012), Design of Power Harmonic Detection Instrument Based on DSP and ARM, CISME Vol.2 No.2
[10] Hsiung Cheng Lin, (February 2012),Power Harmonics and Interharmonics Measurement Using Recursive Group-Harmonic Power Minimizing Algorithm, IEEE Transactions On Industrial Electronics, Vol. 59, No.2
[11] Jaipreet Kaur Bhatti, Deepak Asati, (June 2012), Harmonic Detection using Microcontroller, International Journal of Computer Technology and Electronics Engineering (IJCTEE) Volume 2, Issue 3.
[12] Yi Fei Wang,Yun Wei Li,(2013), An Overview of Grid Fundamental and Harmonic Components Detection Techniques, IEEE
[13] Zhaobi CHU, Ming DING, Shaowu DU, Xueping DONG, (2013), Normalized estimation of fundamental frequency And measurement of harmonics/interharmonics.
[14] Alexander Kamenka, (2014) Six Tough Topics about Harmonics Distortion and Power Quality Indices in Electric Power System, A White Paper of Schaffner Group.
[15] Jeena Joy, Amalraj P.M., Aswin Raghunath, Nidheesh M.N Vinu Joseph,(August 2014), Harmonic Analysis of 230 V AC Power Supply Using LPC2138 Microcontroller, Transactions on Engineering and Sciences ,Vol.2, Issue 8.
[16] Manish Kumar Soni, Nisheet Soni, (February 2014), Review of Causes and Effect of Harmonics on Power System , International Journal of Science, Engineering and Technology Research (IJSETR), Volume 3, Issue 2.
[17] J. R. Demers, B. Kasper1, D.R. Daughton, (2015), Simultaneous measurement of the 1st and 2nd harmonics of a phase modulated coherent frequency-domain THz spectrometer
[18] Feng Guihong, Zhang Jing, Zhao Yisong , Ying Yong, Zhang Bingyil, Harmonic Power Detection and Measurement Device Based on Harmonic Power Flow Analysis.

Source & Publisher Item Identifier: https://ijcaonline.org/archives/volume143/number10/durdhavale-2016-ijca-910394.pdf , International Journal of Computer Applications (0975 – 8887) Volume 143 – No.10, June 2016.

Assessment of Flange Diffuser Structures to Improve the Power Generation of a Diffuser Augmented Wind Turbine

Published by 1. Yiyin KLISTAFANI, 2. A. M. Shiddiq YUNUS, 3. Muhammad ANSHAR, 4. Sri SUWASTI, State Polytechnic of Ujung Pandang, Indonesia (1) ORCID: 1. 0000-0003-3064-9612; 2. 0000-0002-9599-6941; 3. 0000-0001-5687-5023; 4. 0000-0002-7070-9651

Abstract. Wind energy has become the most popular renewable based power plant for the last decades due to its environment benighted and large natural availability. Although modern wind turbine successfully installed worldwide, some areas with low speed wind characteristic might require a special innovation to increase the amount of conversion of extracted wind energy into electric power. One of among popular techniques for the low speed wind turbine is Diffuser Augmented Wind Turbine (DAWT) which are continued to develop from time to time for example by using numerical simulation as an early stages before manufacturing. In this paper a numerical simulations are performed to investigate the effect of attached flange on wind velocity characteristics. Numerical simulations were carried out for the flow field around various flange diffuser type structures to improve the performance of a DAWT. The present studies specifically investigate the effect of attached flange to outlet diffuser with various flange’s angle (0°, 10°, 20°, 30°) on the wind velocity characteristics. Numerical studies were conducted using the Computational Fluid Dynamics (CFD) method. The studies demonstrate that the curved diffuser with flange 10° generates the strongest increment of the wind velocity compared to the other configurations. The maximum velocity inside the diffuser increases up to 115.14%. It is found that the wind velocity at the diffuser centreline is not capable to represent the overall velocity of each section. The curved diffuser with flange 10° shows the highest increment of the average wind velocity along diffuser with the greatest increment of 102.4 % at x/L = 0.36, and the highest increment wind velocity at the diffuser centreline section at x/L = 0.18 is 115.14%.

Streszczenie. Energia wiatrowa stała się najpopularniejszą elektrownią wykorzystującą odnawialne źródła energii w ciągu ostatnich dziesięcioleci ze względu na zaciemnione środowisko i dużą naturalną dostępność. Chociaż nowoczesne turbiny wiatrowe są z powodzeniem instalowane na całym świecie, niektóre obszary o niskiej prędkości wiatru mogą wymagać specjalnej innowacji w celu zwiększenia ilości konwersji wydobytej energii wiatru na energię elektryczną. Jedną z popularnych technik dla turbin wiatrowych o niskiej prędkości jest turbina wiatrowa z dyfuzorem (DAWT), która jest od czasu do czasu rozwijana, na przykład przy użyciu symulacji numerycznej jako wczesnych etapów przed produkcją. W artykule przeprowadzono symulacje numeryczne w celu zbadania wpływu przymocowanego kołnierza na charakterystykę prędkości wiatru. Przeprowadzono symulacje numeryczne pola przepływu wokół różnych konstrukcji typu kołnierzowego dyfuzora, aby poprawić wydajność DAWT. Obecne badania w szczególności badają wpływ zamocowania kołnierza do dyfuzora wylotowego o różnym kącie kołnierza (0°, 10°, 20°, 30°) na charakterystykę prędkości wiatru. Badania numeryczne przeprowadzono metodą obliczeniowej dynamiki płynów (CFD). Z przeprowadzonych badań wynika, że zakrzywiony dyfuzor z kołnierzem 10° generuje najsilniejszy przyrost prędkości wiatru w porównaniu z innymi konfiguracjami. Maksymalna prędkość wewnątrz dyfuzora wzrasta do 115,14%. Stwierdzono, że prędkość wiatru w osi dyfuzora nie jest w stanie przedstawić całkowitej prędkości każdej sekcji. Zakrzywiony dyfuzor z kołnierzem 10° wykazuje największy przyrost średniej prędkości wiatru wzdłuż dyfuzora z największym przyrostem 102,4% przy x/L = 0,36, a największy przyrost prędkości wiatru w środkowej części nawiewnika przy x/L = 0,18 to 115,14%. (Ocena konstrukcji dyfuzorów kołnierzowych w celu poprawy wytwarzania energii w turbinie wiatrowej ze wspomaganiem dyfuzorem)

Keywords: CFD, DAWT, Diffuser, Flange, Wind energy, Wind turbine
Słowa kluczowe: turbina wiatrowa, dyfuzor, .

Introduction

The potential for renewable energy in the world is quite large and has the potential to be developed. One of the potential renewable energy that can contribute significantly to energy needs is wind energy. Wind energy is one of the very clean and sustainable energy sources that abundantly available naturally. Currently, wind energy covers about 6% of the global electricity demand (https://wwindea.org/worldwind-capacity-at-650-gw/). The potential of wind energy is huge and study shows if 20% of the possible wind resources are able to be utilized [1]. One of the problems in the utilization of wind energy conversion technology is that the wind speed is too low for the application. It is well known that wind turbines usually operate for the rated wind speed of around 8-11 m/s [2], [3]. The power of the wind is proportional to the cubic power of the wind velocity approaching a wind turbine. This means that even a small amount of its acceleration gives large increase on the energy generation [4]. Therefore, wind turbine innovation is very important to optimizing the utilization of wind energy, especially in areas with low wind speed characteristic. One of the developments in wind turbine innovation is the DAWT (Diffuser Augmented Wind Turbine) concept which is equipped with a diffuser sheath on the rotor. The use of diffuser is intended to increase the effective wind speed, therefore, the power produced by wind turbines increases.

There are many studies that focus on wind turbine innovations in increasing wind speed, for example the studies from Refs [5]-[8]. Studies that focusing on finding ways to increase the wind speed are introduced by Kannan et al [5], Lipian et al [6], [7], and Khamlaj and Rumpfkeil 8. Yadav and Kumar [9] have also reviewed related shrouded wind turbines with low wind speeds. In the previous study, Ohya et al. [10] developed the diffuser structures by attached flange at the exit periphery to the diffuser body. It was confirmed in the study that the diffuser structure with flange was effective for collecting and accelerating the wind than diffuser without flange. In addition, the power output coefficient increase five times greater than conventional wind turbines. The development study related of DAWT has also been carried out in the previous study by Yiyin et al [11] where in the study the diffuser type structures are modified into four types, namely flat diffuser, curved diffuser, flat diffuser with inlet shroud, and curved diffuser with inlet shroud. The results obtained from the study that the curved diffuser showed the highest improvement of the centreline and average wind velocities along diffuser. The greatest observed increment was 76.99 % at with the maximum average wind velocity of 8.85 m/s.

With the development of computer technology and engineering software evolution, it is possible to model engineering problems using Computational Fluid Dynamics (CFD) approaches for example the CFD simulation for Darrieus Type Wind Turbine for performance investigation [12], [13]. The simulations range from the simplest 2D Reynolds-Averaged Navier-Stokes (RANS) approach to the most complex Direct Numerical Simulation (DNS) approach. A good agreement of the CFD computations using the SST turbulence model was obtained in several computations, for example Pape and Lecanu [14], Sørensen et al [15], Bangga et al [16] and [17], Weihing et al, [18], and Jost et al [19]. These encourage the use of CFD for predicting the fluids engineering problems especially with the help of the Menter SST k-ω model. Having considered the above background, the development of a wind power system with high output aims at determining how to collect wind velocity efficiently and what kind of diffuser design can generate energy effectively from the wind speed. In the present studies, several numerical investigations will be carried out for the flow field around diffuser structures aiming to identify the optimized configuration.

Numerical Methods

The CFD studies mainly concern about the flow development around four types of diffuser with attached flange at the exit diffuser.

A steady two-dimensional approach was employed for the present studies. It will be shown that this is sufficient for predicting the main flow features, but not the wake behavior of the flow. However, the latter is not of interest as the focus of the present studies is only for estimating the flow acceleration inside the diffuser. The geometry was created using the Ansys workbench 2019 R1. The curved diffuser was generated according to the geometry specified in the numerical studies carried out by Klistafani et al [11], where the curved diffuser is a geometry that can provide the best performance improvement for DAWT compared to a flat diffuser. The diffuser has a thickness of 1.25 cm. This was designed based on the recent studies by Hu and Wang (2015) who employed ten layers of plate with each has a thickness of 1.25 mm. The flange length (h) used is 0.2 m referring to previous studies [5], [8]. Four different types of the diffuser were introduced, namely curved diffuser with flange 0°, 10°, 20°, and 30°. These structures are illustrated in Figure 1. Detailed information about their dimension is given in Table 1 and Figure 2.

Fig 1. Curved diffuser with flange: (a) 0°, (b) 10°, (c) 20°, and (d) 30°.

Table 1. Diffuser type structure (2D) dimensions

.

The domain of the simulation is illustrated in Figure 3. The inlet of the flow is located at 5 times the inlet diameter of the diffuser (D). The velocity inlet boundary condition was applied at this location. The flow leaves the computational domain at 8.5D distance from the outlet plane of the diffuser with the outflow boundary condition. The side walls were set as a non-slip wall that are sufficiently far away from the area of interest to ensure the minimal effect on the flow characteristics near the diffuser. The computations were carried out using the commercial software Ansys Fluent 2019 R1. The flow was assumed to be steady and the incompressibility effect was neglected. This is reasonable because wind turbines usually operate at a much smaller velocity than the speed of sound. An initial undisturbed wind velocity of 5 m/s was prescribed at the velocity inlet plane. The same velocity was employed by Ohya et al [10] in their experiment. The turbulence closure was modelled using the two-equation SST k-ω model according to Menter [20]. This model combines the the standard k-ε model [21] in the freestream and the Wilcox k-ω model [22] for the wall bounded flow. The model is good for predicting flows with a strong adverse pressure gradient as demonstrated already in [11], [15]-[19], [23]-[25]. The pressure velocity coupling uses the SIMPLE method. All the variables were solved using the second order discretization. The computations were carried out for 10,000 iterations, otherwise convergence was achieved if the residual of the momentum reaches 1e-6.

Fig 2. Detailed dimensions of the curved diffuser with flange 30°.

Fig 3. Computational domain and its associated boundary conditions of the curved diffuser with flange 0°.

The mesh was generated using ANSYS Workbench 2019 R1 software. Mesh parameters and controls are shown in Table 2. An enlarged view of the mesh near the curved diffuser wall is shown in figure 4. Grid independence studies were carried out in advance to ensure that the results are independent of the mesh resolution. The results are shown in table 3 where the streamwise velocity ratios (U/U∞) of the five meshes are compared. It can be seen that Grid 3 has an optimal grid size with a number of cells is 50,310. Adding the number of cells as in Grid 4, it doesn’t give too much computational results, with a prediction difference value of 3.7%.

Fig 4. Zoom of the mesh near the curved diffuser with flange 0° velocity for curved diffuser with flange 0°.

Table 2. Mesh parameters and controls

.

Table 3. Grid Independence – Difference value of streamwise flow

.

Table 4. Wind velocity at midline for all curved diffuser type structures compare with curved diffuser without flange [11]

.
Results and Discussion

The dimensionless streamwise velocity U/U∞ at midline diffuser plots for five diffuser type structures are presented in figure 5. In case of the diffuser type structures, the distribution of the axial velocity reveals that the maximum velocity occurs for curved diffuser with flange 10°. All of the diffuser structures by attached flange at the exit periphery to the diffuser body give a positive impact on increasing wind speed. The difference in increased velocity generated by the curved diffuser flange 10° compared to diffuser without flange [11] is 30.96%. The curved diffuser with flange 10° shows a better performance, although curved diffuser with flange 20° also give great increment of wind speed. The difference of its increment is very small (0.85%). Maximum wind speed of curved diffuser with flange 10° is not occurs at entrance position, but at x/L = 0.18. Detailed information regarding the comparison of the wind velocity at midline of all diffuser type structures is shown in table 4.

Further comparison of the dimensionless streamwise velocity U/U∞ for four diffuser type structures and curve diffuser without flange [11] are presented in figure 6, in which the average velocity data are taken at each section of diffuser. As shown in figure 6, diffusers equipped with the flange have the bigger average wind velocity through inside diffuser than curved diffuser without flange. At the inlet diffuser section (x/L = 0), the highest value of the averaged wind speed occur in curved diffuser with flange 10° and 20°. However the curved diffuser with flange 10° has the highest maximum average wind speed than others in the inside diffuser (x/L = 0.36). The difference increment value of maximum averaged wind speed generated by curved diffuser 10° compare with curved diffuser without flange is 25.47%. the highest maximum average wind speed is 10.12 m/s with the increment value is 102.45%.

Comparison of velocity contour on curved diffuser without flanges and with flanges 10° can be seen in Figure 7. Vortices flow at downstream diffuser with flanges 10° larger (indicated by blue area) than flow through curved diffuser without flange. The large vortexes in the downstream area have suction effect in the upstream areas; as a result the wind that crosses the upstream diffuser increases the wind velocity (indicated by orange contour). The velocity contour strengthen the previous discussion (Figures 5 and 6), namely curved diffuser with a flange 10° giving better performance than curved diffuser without flange. As informed in table 4, the difference wind increment of both geometries is 30.96%.

In line with the discussion result of the velocity contour, the pressure contour also shows that the curved diffuser with flange10° provides the best performance. This is indicated by the high pressure in the downstream region and at area around the diffuser wall, thereby strengthening the evidence that the suction effect caused by the curved diffuser with the flange 10° is very strong compared to the diffuser without the flange. Pressure contour regarding the comparison of curved diffuser without and with flange 10° is shown in figure 8.

In line with the discussion result of the velocity contour, the pressure contour also shows that the curved diffuser with flange 10° provides the best performance. This is indicated by the high pressure in the downstream region and at area around the diffuser wall, thereby strengthening the evidence that the suction effect caused by the curved diffuser with the flange 10° is very strong compared to the diffuser without the flange. Pressure contour regarding the comparison of curved diffuser without and with flange 10° is shown in figure 8.

Fig 5. Wind velocity distributions at the midline axis along the axial positions compare with diffusser without flange (Klistafani et al, 2018)

Fig 6. Average wind velocity distributions along the axial positions compare with diffuser without flange [11].

Fig 7. Velocity contour of curved diffuser (a) without flange and (b) with flange 10°

Fig 8. Pressure contour of curved diffuser (a) without flange and (b) with flange 10°

Figure 9 presents the velocity profiles for different flange angle of diffuser compared with curved diffuser without flange and without diffuser at all. It’s to clarify the rate of wind flow within various configuration of diffuser. It can be seen that the wind velocity at the upstream zone is same for all the configurations of curved diffuser. The wind velocity at section x/L = -0.54 (far away from inlet diffuser) not influenced by the presence of the curved diffuser. It becomes evident that the wind velocity slightly increases at the near inlet diffuser (x/L = -0.18), although the difference of wind velocity increases for all the curved diffuser within and without diffuser is small. The increase in wind velocity is clearly visible when entering the diffuser (x/L = 0), where the curved diffuser equipped with the flange 10° and 20° have the good performance than the others. However, the greatest increase in wind velocity along the diffuser (x/L = 0.36 until x/L = 1) is actually generated by the curved diffuser with flange 10°. Overall, it can be clearly seen that Curved diffuser equipped with flange have wind velocity increment bigger than curved diffuser without flange (x/L = – 0.18 until x/L = 1).

The research vertical axis wind turbine (VAWT) investigated by Saedi et al [26] is considered to estimate the generated power production of the turbine equipped with the curved diffuser with flange. The turbine has a radius of 2 m and a height of 1.38 m. The estimated power curves of the turbine for various wind speeds and curved diffusers with flange can be seen on figure 10. In these plots the turbine is assumed to be located at x/L = 0.36 where the maximum average wind speed takes place. Flange that equipped at outlet diffuser can improve generated power of turbine significantly than diffuser without flange. Curved diffuser with flange 10° has the greatest estimated generated power production than others.

A diffuser with a flange angle 0° has 12% lower power (2.9369 kW) than the power generated by a flange diffuser 10° (3.3395 kW). However, the addition of the flange angle is not in line with the increase in power generated, when the flange angle is enlarged to 20°, the resulting power decreases by 2.6% to 3.2519 kW. A diffuser with a flange angle 30° produces less power than a diffuser with another flange, which is 2.6421 kW. The research vertical axis wind turbine (VAWT) investigated by Saedi et al [26] is considered to estimate the generated power production of the turbine equipped with the curved diffuser with flange. The turbine has a radius of 2 m and a height of 1.38 m. The estimated power curves of the turbine for various wind speeds and curved diffusers with flange can be seen on figure 10. In these plots the turbine is assumed to be located at x/L = 0.36 where the maximum average wind speed takes place. Flange that equipped at outlet diffuser can improve generated power of turbine significantly than diffuser without flange. Curved diffuser with flange 10° has the greatest estimated generated power production than others. A diffuser with a flange angle 0° has 12% lower power (2.9369 kW) than the power generated by a flange diffuser 10° (3.3395 kW). However, the addition of the flange angle is not in line with the increase in power generated, when the flange angle is enlarged to 20°, the resulting power decreases by 2.6% to 3.2519 kW. A diffuser with a flange angle 30° produces less power than a diffuser with another flange, which is 2.6421 kW.

Fig 9. Velocity profiles for different flange angle of diffuser along y-coordinate at several axial positions

Fig. 10. Power curve of the considered wind turbine for various angle flange at x/L = 0.36.

Conclusion

Numerical simulations have been carried out for flow fields around curved diffuser with various angle of flange. The main conclusions derived from the study are as follows:

1. All of the diffuser structures by attached flange at the exit periphery to the curved diffuser body give a positive impact on increasing wind velocity.

2. Curved diffuser flange 10° shows the highest improvement of wind velocities, not only the centreline wind velocity but also the average wind velocity. The highest increment of the wind velocity at the diffuser centerline section is 115.14% with the maximal velocity is 10.76 m/s.

3. The curved diffuser with flange 10° has the greatest maximum average wind speed than others in the inside diffuser (x/L = 0.36). the highest maximum average wind speed is 10.12 m/s with the increment value is 102.45%.

4. Curved diffuser with flange 10° has the greatest estimated generated power production around 3.3395 kW.

5. The curved diffuser with flange 10° very suitable to be used as a wind turbine shroud to improved wind turbine performance.

Acknowledgements – The authors would like to acknowledge the funding support of directorate of research and community service, directorate general of strengthening research and development, ministry of research, technology and higher education of Indonesia.

REFERENCES

[1] Qasim, A.Y., Usubamatov, R., Zain, Z.M. and Quad, G.A., 2012. The parameters affect on power coefficient vertical axis wind turbine. IIUM Engineering Journal, 13(1).
[2] Bangga, G., Hutomo, G., Wiranegara, R. and Sasongko, H., 2017. Numerical study on a single bladed vertical axis wind turbine under dynamic stall. Journal of Mechanical Science and Technology, 31(1), pp.261-267.
[3] Bangga, G., Kusumadewi, T., Hutomo, G., Sabila, A., Syawitri, T., Setiadi, H., Faisal, M., Wiranegara, R., Hendranata, Y., Lastomo, D. and Putra, L., 2018, January. Improving a twoequation eddy-viscosity turbulence model to predict the aerodynamic performance of thick wind turbine airfoils. In J. Phys.: Conf. Ser (Vol. 974, No. 1, p. 12019).
[4] Abe, K.I. and Ohya, Y., 2004. An investigation of flow fields around flanged diffusers using CFD. Journal of wind engineering and industrial aerodynamics, 92(3-4), pp.315-330.
[5] Kannan, T.S., Mutasher, S.A. and Lau, Y.K., 2013. Design and flow velocity simulation of diffuser augmented wind turbine using CFD. journal of engineering science and technology, 8(4), pp.372-384.
[6] Lipian, M., Karczewski, M. and Olasek, K., 2015. Sensitivity study of diffuser angle and brim height parameters for the design of 3 kW Diffuser Augmented Wind Turbine. Open Engineering, 1(open-issue).
[7] Lipian, M., Karczewski, M., Molinski, J. and Jozwik, K., 2016. Numerical simulation methodologies for design and development of Diffuser-Augmented Wind Turbines–analysis and comparison. Open Engineering, 1(open-issue).
[8] Khamlaj, T.A. and Rumpfkeil, M.P., 2018. Analysis and optimization of ducted wind turbines. Energy, 162, pp.1234-1252.
[9] Yadav, A.K. and Kumar, D., 2017. Review of a Shrouded Wind Turbine for Low Wind Speeds. International Digital Library of Technology & Research, Vol. 1, Issue 5.
[10] Ohya, Y., Karasudani, T., Sakurai, A., Abe, K.I. and Inoue, M., 2008. Development of a shrouded wind turbine with a flanged diffuser. Journal of wind engineering and industrial aerodynamics, 96(5), pp.524-539.
[11] Klistafani, Y., Mukhsen, M.I. and Bangga, G., 2018. Assessment of various diffuser structures to improve the power production of a wind turbine rotor. Technische Mechanik. Scientific Journal for Fundamentals and Applications of Engineering Mechanics, 38(3), pp.256-266.
[12] Damian Mazur., Analiza CFD pionowej turbiny wiatrowej typu Darrieus, Journal Przeglad Elektrotecnichzny, 08/2013 Page no. 77. 2013
[13] Damian Mazur., Static analysis of the Darrieus wind turbine including loads from the numerical fluid mechanics, Journal Przeglad Elektrotecnichzny, 10/2013 Page no. 293.2013.
[14] Pape, A.L. and Lecanu, J., 2004. 3D Navier–Stokes computations of a stall-regulated wind turbine. Wind Energy: An International Journal for Progress and Applications in Wind Power Conversion Technology, 7(4), pp.309-324.
[15] Sørensen, N.N., Michelsen, J.A. and Schreck, S., 2002. Navier–Stokes predictions of the NREL phase VI rotor in the NASA Ames 80 ft× 120 ft wind tunnel. Wind Energy: An International Journal for Progress and Applications in Wind Power Conversion Technology, 5(2-3), pp.151-169.
[16] Bangga, G., Lutz, T. and Krämer, E., 2017. Root flow characteristics and 3D effects of an isolated wind turbine rotor. Journal of Mechanical Science and Technology, 31(8), pp.3839-3844.
[17] Bangga, G., Lutz, T., Jost, E. and Krämer, E., 2017. CFD studies on rotational augmentation at the inboard sections of a 10 MW wind turbine rotor. Journal of Renewable and Sustainable Energy, 9(2), p.023304.
[18] Weihing, P., Letzgus, J., Bangga, G., Lutz, T. and Krämer, E., 2016, September. Hybrid RANS/LES capabilities of the flow solver FLOWer—Application to flow around wind turbines. In Symposium on hybrid RANS-LES methods (pp. 369-380). Springer, Cham.
[19] Jost, E., Fischer, A., Bangga, G., Lutz, T. and Krämer, E., 2017. An investigation of unsteady 3-D effects on trailing edge flaps. Wind Energy Science, 2(1), p.241.
[20] Menter, F.R., 1994. Two-equation eddy-viscosity turbulence models for engineering applications. AIAA journal, 32(8), pp.1598-1605.
[21] Launder, B.E. and Spalding, D.B., 1972. Mathematical models of turbulence (No. BOOK). Academic press.
[22] Wilcox, D.C., 1993. Turbulence Modelling for CFD, DCW Industries Inc., La Canada, California. IBSN 0-9636051-0-0.
[23] Hu, J.F. and Wang, W.X., 2015. Upgrading a shrouded wind turbine with a self-adaptive flanged diffuser. Energies, 8(6), pp.5319-5337.
[24] Bangga, G. and Sasongko, H., 2017. Dynamic stall prediction of a pitching airfoil using an adjusted two-equation URANS turbulence model. Journal of Applied Fluid Mechanics, 10(1), pp.1-10.
[25] Bangga, G., Weihing, P., Lutz, T. and Krämer, E., 2017. Effect of computational grid on accurate prediction of a wind turbine rotor using delayed detached-eddy simulations. Journal of Mechanical Science and Technology, 31(5), pp.2359-2364.
[26] Saeidi, D., Sedaghat, A., Alamdari, P. and Alemrajabi, A.A., 2013. Aerodynamic design and economical evaluation of site specific small vertical axis wind turbines. Applied energy, 101, pp.765-775.

Authors: Yiyin Klistafani, ST,MT, Energy Conversion Study Program, Mechanical Engineering, State Polytechnic of Ujung Pandang, Indonesia, E-mail: yiyin_klistafani@poliupg.ac.id; A. M. Shiddiq Yunus, ST, MEngSc, PhD, Energy Conversion Study Program, Mechanical Engineering, State Polytechnic of Ujung Pandang, Indonesia, E-mail:shiddiq@poliupg.ac.id;Prof. Ir. Muhammad Anshar, M.Si, PhD, Power Engineering Study Program, Mechanical Engineering, State Polytechnic of Ujung Pandang, Indonesia, E-mail: muh_anshar@poliupg.ac.id; Sri Suwasti, ST,MT, Energy Conversion Study Program, Mechanical Engineering, State Polytechnic of Ujung Pandang, Indonesia, Email: sri_suwasti@poliupg.ac.id;

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 98 NR 4/2022. doi:10.15199/48.2022.04.05

A New Technique to Detect Harmonic Sources in Polluted Power Systems

Published by Pietro Vincenzo Barbaro, Antonio Cataliotti, Valentina Cosentino, Salvatore Nuccio, Dipartimento di Ingegneria Elettrica, Elettronica e delle Telecomunicazioni, Università di Palermo, Palermo, Italy. E-mails: barbaro@diepa.unipa.it, acataliotti@ieee.org, cosentino@diepa.unipa.it, nuccio@unipa.it

Abstract: This paper presents a comparative analysis among different nonactive power quantities proposed in literature in nonsinusoidal conditions; with respect to this, a new single-point approach is proposed, for the detection of the dominant harmonic sources in polluted power systems. It is based on the observation that in the same distorted working condition the analyzed power quantities present a different behavior. In order to verify the theoretical assumptions, some simulations tests were carried out on a standard IEEE test system, proposed as a benchmark for harmonic propagation studies. Simulation results show how the approach based on a comparison of different definitions of nonactive powers can give some useful information for the detection of dominant harmonic sources.

Keywords: harmonics, nonactive powers, harmonic sources.

1. INTRODUCTION

In the last years current and voltage distortion is ever increasing and the problem of the detection of harmonic sources has become more urgent, because of the proliferation in distribution systems of a number of loads that draw non-sinusoidal currents. In practical situations, harmonic sources can be located both upstream and downstream the metering section, so that both supply and load may be responsible for harmonic distortion. Thus, it has became a very important target to determine customers and utilities polluting contributions to the disturbances affecting the supply voltage at the metering section.

The traditional billing quantities do not allow one to achieve any useful information about the responsibility for the disturbances affecting the power system. They are related to the concepts of active, reactive and apparent powers, and power factor, that are well known in sinusoidal conditions, but that are not meaningful anymore in nonsinusoidal situations (with the exception of the active power). On the other hand, the international standards concerning the measurement on polluted power systems, refers to the measurement of the amplitudes of single harmonics and of some traditional parameters, such as the Total Harmonic Distortion Factor (THD), but do not provide any piece of information about the detection of harmonic sources.

With respect to this problem, several approaches have been proposed in literature for harmonic sources detection. They can be generally divided into multi-point and single-point methods [1-4]. The multi-point methods are based on the elaboration of more than one measurements performed in different metering sections; these methods can give a complete information about the harmonic state of the power system, but they require the implementation of a distributed and synchronous measurement system, with a complex and expensive measurement instrumentation. On the contrary, the single-point methods have many advantages, e.g. easy implementation and low cost, but in some conditions they can report imprecise information about the harmonic state of the system. Some of single-point strategies are based on the evaluation of harmonic active power flow at the metering section. However, it has been demonstrated that in some practical situations this approach cannot provide a correct information about the location of the dominant harmonic source, upstream or downstream the metering section. On the other hand, it could be interesting to study the behavior of “nonactive” components of the apparent power. The interpretation of these power terms has been widely discussed in literature [5-8]. Several “nonactive” power definitions have been formulated, starting from different approaches for the grouping of the terms of instantaneous power that do not contribute to the net transfer of energy.

In this paper the authors have investigated if a comparative evaluation of different definitions of nonactive powers, already proposed in literature, could give useful information about the non-linearity degree of the power system and the location of the dominant harmonic source. Thus, a new approach is proposed for the detection of the dominant harmonic source in power systems, that is based on the simultaneous evaluation of three nonactive power quantities at the metering section. The new approach starts from the theoretical observation that in the same distorted working condition the considered power quantities present a different behavior. This is due to the different grouping of the components of the terms of instantaneous power that do not contribute to the net transfer of energy. The proposed strategy was firstly validated on a simple test system, developed by the authors, that is able to simulate different working conditions, with both sinusoidal and distorted supply and both linear and non linear loads. Further simulations were carried out on a IEEE standard test power system [9], proposed, by other authors, as a benchmark system for the analysis of multi-point measurement techniques for harmonic pollution monitoring. Simulation results show how the approach based on a comparison of different definitions of nonactive powers could give some useful information for the detection of dominant harmonic sources.

2. THE CONCEPT OF REACTIVE POWER IN NONSINUSOIDAL CONDITIONS

It is well known that, for steady-state conditions, in a single-phase system affected by harmonics, the instantaneous voltage and current can be expressed as follows [5]:

.

where v1 and i1 are the power system frequency components of voltage and current, and the remaining terms vh and ih contain all the remaining harmonic components. V1 and I1 are the rms values of the fundamental components of voltage and current, Vh and Ih are the rms values of the h-harmonic components of voltage and current, α1 and β1 are the phase angles of the fundamental components of voltage and current, αh and βh are the phase angles of the h-harmonic components of voltage and current, ω=2πf is the angular frequency, t is the time, V0 and I0 are the direct voltage and the direct current terms, obtained for h = 0.

The instantaneous power is the product of the instantaneous voltage and current; it can be written as:

.

(The angle θh = βh −αh is the phase angle between the phasors Vh and Ih).

The first term, pa, contains all the components that have non-zero average value; the total average value is the active power, which is equal to the sum of harmonic active powers:

.

The second term, pq, contains all the components whose average value is nil, thus it does not contribute to the net transfer of energy. The first addendum of pq contains the terms related to harmonic components that are present in both voltage and current; the other addenda contains the terms related to harmonic components that are not common to both voltage and current, including the direct components.

As it is well known, in the sinusoidal case, the first addendum is a sinusoidal term, with a frequency double of the power system frequency, while the other addenda are nil; the amplitude of the double-frequency component is the reactive power.

Finally, the apparent power is given by the product of the rms values of voltage and current:

.

The interpretation of pq in nonsinusoidal case has been widely discussed in literature. Several “nonactive” power definitions have been formulated, starting from different approaches for the grouping of the components of pq. The developed power theories can be mainly classified into time-domain and frequency-domain approaches [6-8].

The time domain approach is based on the concept of splitting the load current into two or more components, that are meant to be responsible for different energy phenomena. The most general time-domain power theory is due to Fryze. Its approach is essentially based on the separation of the current i into two components; the first one, namely the “active” current ia, is in phase with the voltage and has the same waveform, the second one, namely the “nonactive” or “reactive” current, ir = i – ia is the remaining part of the current. Starting from this approach, the apparent power can be divided into active and reactive power (in analogy with the sinusoidal case). On the other hand, the first power theory in the frequency-domain was formulated by Budeanu: it was based on the assumption that in nonsinusoidal situations, a power system can be ideally decomposed into a number of elementary sinusoidal systems, each one corresponding to a singular harmonic of the spectrum of the voltage or of the current . In this sense for each elementary sinusoidal system, the traditional electrical quantities can be defined (rms values of voltage and current, active, reactive and apparent powers and power factor). Starting from the Fryze and Budeanu approaches, many other power theories have been developed and different definitions of reactive power have been formulated (Kusters-Moore, Page, Shepherd-Zakikhani, Sharon, Czarnecki, etc).

However, it can be observed that none of the proposed definitions is able to maintain all the properties of the reactive power in the sinusoidal case. Moreover, many of the above mentioned power theories were developed for the single-phase case; their extension to the three-phase system is not a trivial question, and it is strictly related to the generalization of the apparent power concept, that, in the three-phase case, is not uniquely defined. In this sense, different concepts for the apparent power were introduced and discussed in literature [10-11]. From the physical point of view, the expression (7) can be interpreted in different ways. A first concept is that the apparent power is considered as the maximum active power that can be transmitted under ideal conditions, (i.e. sinusoidal symmetric voltages and sinusoidal balanced currents) with the same voltage impact (insulation and no-load losses) and the same current impact (line losses). This is the approach of the IEEE Standard 1459-2000 [5]. A second concept is that the apparent power is considered as the maximum active power that can be transmitted for the given voltage waveform and the given current rms value of the current (line losses). This definition corresponds to the concepts developed by many authors (Buchholz, Fryze, ecc.) [10]. On the other hand, other definitions of apparent power were introduced, such as the “Arithmetic Apparent Power”, given by the sum of the phase apparent powers, or the “Vector Apparent Power” [11].

It was demonstrated that, in the sinusoidal and balanced case, all these concepts for the apparent power lead to the same results On the contrary, in the most general case of a distorted and unbalanced system, the definitions of apparent power lead to different results.

3. THE PROPOSED APPROACH

It can be observed that, in sinusoidal conditions, all the definitions of reactive power lead to the same result, that is the sinusoidal reactive power. On the contrary, in the presence of harmonic distortion they assume different values. This difference can be very significant, depending on the working conditions, i. e. depending on the amount of the harmonic distortion in both voltage and current and on the presence of common and uncommon harmonics in voltage and current. This is due to the different grouping of the components of the terms of instantaneous power that do not contribute to the net transfer of energy. In some cases only the harmonic components that are common to both voltage and current are considered. In other cases, like in the Fryze approach, both common and uncommon harmonics are considered. In other cases an intermediate situation is considered.

Starting from these considerations, the authors have investigated if a comparative evaluation of different reactive powers, already defined in literature, could give useful information about the non-linearity degree of the power system and the location of the dominant harmonic source. More in detail, the following quantities are considered:

.

(S is the apparent power, P is the active power, V is the rms value of the voltage, SC the complementary power [12] and k is the index related to the harmonic components that are common to both voltage and current).

Q1 can be considered as a minimum reference value, since it is the only nonactive power component in the sinusoidal condition; on the other hand, QF is the maximum value for the nonactive power, since it is related to all terms of pq, and it is the only nonactive component of the apparent power. It can be easily observed that the higher is the amount of distortion, the higher is the difference between Q1 and QF. On the other hand, the expressions of SQ lead to a nonactive power value that is intermediate between Q1 and QF, due to the fact that SQ is not the only nonactive component of the apparent power as shown by (10). SQ depends on both common and uncommon harmonics of the voltage and common harmonics of the current.

Therefore, in case of a non sinusoidal supply voltage and a linear load, the harmonic content of the current corresponds to the one of the voltage and its contribution, in terms of power, is small if compared with the fundamental one. Thus, SQ is closer to Q1 than to QF. On the contrary, in case of a sinusoidal supply voltage and a non linear load the amount of the distortion of current is higher than the one of the voltage. In this case, SQ is closer to QF than to Q1. in this sense, SQ can be considered as an indicator of the nonlinear behavior of the load. Finally when the supply voltage is sinusoidal and the load is linear SQ, Q1 and QF have comparable values.

Therefore, a comparison among Q1, QF and SQ, calculated in a PCC in the same working conditions, could give a piece of information on the detection of disturbing loads. In a single point strategy, the proposed approach could be combined with the one based on the sign of active power harmonic components [4], allowing one to avoid misleading results.

The proposed approach was developed for three-phase systems, considering each of the nonactive powers Q1, QF and SQ as the sum of the respective phase quantities. This was preliminarily made for the balanced case, where the different approaches for apparent power resolution lead essentially to the same results. On the other hand, the validity of the proposed approach was investigated also in the unbalanced case. In this case, the separation of the effects of the unbalance and nonlinearity is not easy to achieve, because of the overlapping of the effects due to such disturbances.

4. SIMULATIONS

4.1. Preliminary validation

In order to carry out a preliminary validation of the proposed approach, a simple three-phase balanced test system was implemented on a calculator, by means of the software package POWER SYSTEM BLOCKSET® of MATLAB®. The test system was realized with: a three phase symmetrical voltage supply, either sinusoidal or distorted (THD = 6,9%) with a known harmonic content; an equivalent network impedance; a linear and balanced load (resistive-inductive load) and a non linear and balanced load (a diode bridge rectifier feeding a dc load; it was dimensioned in order to absorb a fundamental active and reactive power equal to the one absorbed by the linear load). The implemented test system is reported in figure 1.

Fig.1: Three-phase balanced test system

Simulations were carried out for different working conditions:

1. sinusoidal supply voltage and linear load (switches 1 and 3 closed, 2 and 4 open);
2. sinusoidal supply voltage and non linear load (switches 1 and 4 closed, 2 and 3 open);
3. nonsinusoidal supply voltage and linear load (switches 2 and 3 closed, 1 and 4 open);
4. nonsinusoidal supply voltage and non linear load (switches 2 and 4 closed, 1 and 3 open).

The proposed approach, based on the comparison of the nonactive powers Q1, QF and SQ, was implemented by means of the software package SIMULINK® of MATLAB®; each phase was considered as a single phase system and the three-phase quantities were evaluated as the sum of the respective reactive powers obtained for each phase.

In figure 2 some simulation results are reported, that are referred to the above mentioned working conditions.

Fig.2: Simulation results for the test system of fig. 1 for different working conditions.

As expected, the nonactive powers Q1, SQ and QF have different values in the same working condition, with the exception of the case of sinusoidal supply and linear load. More in detail, it can be observed that Q1 and QF assume respectively the minimum and the maximum value for nonactive power. The difference between these values is more significant when the harmonic distortion is present and when the load is nonlinear; in this sense, the difference between Q1 and QF can be considered as a global indicator of the non linearity degree of the system. Further considerations can be made with respect to the value of SQ that is intermediate between Q1 and QF; its value depends on the nonlinearity degree of the load. For example, in the case of a sinusoidal supply and a non linear load SQ is close to QF, while in the case of a linear load and a nonsinusoidal supply SQ is close to Q1. Finally in the case of nonsinusoidal supply and non linear load the values of SQ is between Q1 and QF.

4.2. IEEE Test System

In order to test the proposed strategy on a more complex real size network, further computer simulations were carried out on the IEEE Test System n. 2 proposed in [9] (see figure 3). This system was already used as a benchmark for the analysis of some multi-point measurement techniques for harmonic pollution monitoring by other authors [1]. It is based on the IEEE 13 bus radial distribution test feeder; it contains voltage, regulators, three and single phase line configurations, shunt capacitors, spot and distributed loads. Phase-ground and phase-phase connected loads are included. For harmonic studies, load compositions are specified to include harmonic producing loads. Three types of loads are considered for test purposes: fluorescent light banks, adjustable speed drives, and composite residential loads. Complete data of the system are reported in [9].

Fig.3: IEEE Test System n. 2

The IEEE benchmark network was implemented by means of the PSCAD/EMTDC software. With respect to the original network configuration reported in [9], the following simplifying assumptions were made [1]: the distributed load between nodes 32 and 71 were modeled as two spot loads connected to the above mentioned nodes; all the three-phase lines were considered as transposed and they were modeled by means of balanced π branches; the loads supplied by single-phase and two-phase feeders were aggregated to the closest three-phase node, thus obtaining a complete three-phase network. With these assumptions, the network under test essentially consisted of a sinusoidal and balanced power source (at node 50), a transformer (between nodes 50 and 31) and the following five loads:

– L1 (at node 33, including the single-phase load 34);
– L2 (at node 32, consisting of the single-phase load 45, the phase-phase load 46, and half the distributed load between nodes 32 and 71);
– L3 (at node 71, consisting of half the distributed load between nodes 32 and 71, the phase-phase load 92 and the single-phase loads 52 and 911, with shunt capacitors);
– L4 (at node 71, consisting of a three-phase load);
– L5 (at node 75, consisting of a three-phase load, with shunt capacitors).

The simulations were carried out considering several different network configurations obtained by substituting some of the nonlinear and/or unbalanced loads with equivalent linear and balanced loads having the same power characteristics of the original ones. Moreover, further simulations were carried out on a modified the test system, that was obtained by substituting the nonlinear and unbalanced loads with nonlinear but balanced loads (they were obtained by reproducing the harmonic content of phase A on the other phases). This was made in order to verify the validity of the proposed strategy when the power system is contemporary affected by both harmonic distortion and unbalance.

In each test, the simulation on the PSCAD/EMTDC environment were run and the instantaneous values of voltages and currents were calculated for each considered load. The obtained data were saved in a MATLAB file and they were used as input data for the evaluation of the nonactive powers Q1, SQ and QF for each metering section.

The first series of tests were performed on modified test systems, where only one load at time was considered in its original configuration, while all other loads were substituted with their linear and balanced loads. Moreover, when the disturbing load was both nonlinear and unbalanced, the tests were repeated by substituting the original load with a nonlinear and balanced load, as described before.

The simulation results showed that the comparison of the powers in each metering section led to the correct individuation of the dominant polluting source.

For example, figure 4a shows the simulation results in the case of loads L1, L2, L3, L4 linear and load L5 (globally capacitive) non linear and unbalanced, in its original configuration reported in [9]. It can be observed that for the loads L1, L2, L3 and L4, the nonactive powers Q1, SQ and QF are very close in all cases; thus, it can be deduced that these load have a linear behavior and the harmonic distortion at the metering section is due to the supply. On the contrary, for the load L5, the difference between the considered nonactive powers are more significant and SQ is closer to QF; thus, it can be concluded that L5 is the disturbing load. The same considerations can be made with respect to the case reported in figure 4b, where the original load L5 was substituted by a nonlinear and balanced load. Also in this case, the comparative analysis of the nonactive powers Q1, SQ and QF leads to the correct individuation of the disturbing load L5.

The second series of test were performed considering more than one nonlinear loads at time. For example, figures 5a shows the obtained results in the case of loads L1, L3 and L5 linear and loads L2 and L4 non linear (original configurations). The analysis of the nonactive powers in each metering section led to the correct location of the disturbing loads. Also in this case, the tests were repeated by substituting the original nonlinear loads with the nonlinear and balanced loads; figure 5b shows the obtained results. In both cases, the analysis of the nonactive powers in each metering section led to the correct location of the disturbing loads.

Finally, also in the case of all nonlinear loads the proposed strategy based on nonactive powers led to the correct location of the disturbing loads. In Figure 6a and 6b the simulations results are reported; in detail, figure 6a is referred to the original configuration of the test system, while figure 6b is referred to the modified test system, obtained by substituting all the nonlinear and unbalanced loads with the nonlinear balanced loads.

Fig.4a: Simulation results of the proposed approach in the case of L1, L2, L3, L4 linear loads and L5 non linear load (original configuration).

Fig.4b: Simulation results in the case of L1, L2, L3, L4 linear loads and L5 non linear balanced load.

Fig.5a: Simulation results in the case of L1, L3, L5 linear loads and L2 and L4 non linear loads (original configuration).

Fig.5b: Simulation results in the case of L1, L3, L5 linear loads and L2 and L4 non linear and balanced loads.

Fig.6a: Simulation results in the case of all non linear loads (original configuration).

Fig.6b: Simulation results in the case of all non linear balanced loads.

5. CONCLUSIONS

In this paper a new single-point strategy is proposed, for the detection of the dominant harmonic source in a polluted power systems. It is based on the comparison among different reactive power quantities proposed in literature, that, in the same working conditions, assume different values at the metering section. Several simulation tests were carried out on a standard IEEE test system, proposed, by other authors, as a benchmark system for the analysis of multi-point measurement techniques for harmonic pollution monitoring. The obtained results show that the proposed approach can give useful indications for the detection of the dominant harmonic source in a metering section, in both balanced and unbalanced situations. On the other hand, the proposed approach can be used also in multi-point strategy to detect disturbing loads, performing a comparison among Q1, QF and SQ in each metering section and. In this sense, the combination of different strategies could be useful, in order to achieve a better information on the harmonic state of the system and on the location of harmonic sources.

REFERENCES

[1] C. Muscas, L. Peretto, S. Sulis, R. Tinarelli, “Implementation of multi-point measurement techniques for PQ monitoring” IEEE Instrumentation and Measurement Technology Conference, IMTC 2004, 18-20 May 2004, Como, Italy, pp. 1626-1631.
[2] A. P. J. Rens, P. H. Swart, “On Techniques for the Localization of Multiple Distortion Sources in Three-Phase Systems. Time Domain Verification” ETEP, Vol. 11, No 5, Sept.-Oct. 2001.
[3] E. J. Davis, A. E. Emmanuel, D. J. Pileggi, “Evaluation of Single-Point Measurements Method for Harmonic Pollution Cost Allocation” IEEE Trans. On Power Delivery, Vol. 15, No 1, January 2000.
[4] M. Aiello, A. Cataliotti, V. Cosentino, S. Nuccio, “A Self-Synchronizing Instrument for Harmonic Sources Detection in Power Systems”, IEEE Transactions on Instrumentation and Measurement, Vol. 54, No 1, February 2005, pp. 15-23.
[5] IEEE Std 1459-2000, “IEEE Trial-use standard definitions for the measurement of electric power quantities under sinusoidal, non sinusoidal, balanced or unbalanced conditions”, September 2002
[6] P. S. Filipski, P. W. Labaj, “Evaluation of reactive power meters in the presence of high harmonic distortion”, IEEE Trans. On Power Delivery, Vol. 7, No. 4, October 1992.
[7] A. E. Emanuel, “Powers in nonsinusoidal situation. A review of definitions and physical meaning”, IEEE Trans. On Power Delivery, Vol. 5, No. 3, July 1990
[8] L. S. Czarnecki, “Budeanu and Fryze: two frameworks for interpreting power properties of circuits with nonsinusoidal voltages and currents”, Electrical Engineering, vol. 80, n. 6, 1997, pp. 359-420.
[9] IEEE Task Force on Harmonics Modeling and Simulation, “Test Systems for Harmonic Modeling and Simulation”, IEEE Transactions on Power Delivery, vol. 14, n. 2, 1999, pp. 579-587
[10] J. L. Willems, J. A. Ghijselen, A. E. Emanuel “The apparent power concept and the IEEE Standard 1459-2000”, IEEE Transactions on Power Delivery, Vol. 20, No. 2, pp. 876-884, April 2005, pp. 876-884
[11] A. E. Emanuel, “Apparent power definitions for threephase systems”, IEEE Transactions on Power Delivery, vol. 14, n. 3, July 1999, pp. 767-772
[12] D. Sharon, “Reactive power definitions and power factor improvement in nonlinear systems”, Proc. IEE, Vol 120, n. 6, 1973, pp. 704.706

Source URL: https://citeseerx.ist.psu.edu/document?repid=rep1&type=pdf&doi=08a680e923b0baff60eaa25dc17109aee0456c4e , XVIII IMEKO WORLD CONGRESS Metrology for a Sustainable Development September, 17 – 22, 2006, Rio de Janeiro, Brazil.

Elspec becomes PQSynergy™ & Power Quality Blog Sponsor

We are pleased to announce that Elspec was a Sponsor for PQSynergyTM International and Conference Exhibition 2024 and Power Quality Blog. It is a privilege to be supported by Elspec.

We are grateful to the Elspec team.

.

Elspec is a leading global provider of advanced power quality analyzers, power quality solutions and power monitoring software. We focus on developing technology that shifts the power quality market to the next level. With over 35 years of experience, we develop tailor-made power quality solutions along with state of the art power quality analyzers to empower our customers to succeed their power and energy challenges.

Elspec’s advanced technology captures, monitors, and records all waveform signals continuously at up to 1024 samples per cycle. Every parameter, nothing missed.

• Gain complete visibility into your power quality for proactive decision-making.
• Ensure compliance and avoid costly penalties by exceeding IEC 61000-4-30 Class A and IEEE-519 2022 standards.
• Minimize equipment wear and optimize performance by detecting harmonics up to the 511th harmonic.
• Maximize control with insights from over 5,000 power parameters.
• Quickly identify and resolve issues with unparalleled root-cause analysis.

Learn More – Elspec’s PQ Analyzers & Recorders
.
.
.
.

For more information, you can visit elspec-ltd.com

Validation of Aperiodic and Oscillatory Stability Calculations in a Practical Power Systems

Published by Igor Razzhivin, Aleksey Suvorov, Mikhail Andreev, Aleksandr Gusev, Tomsk Polytechnic University

Abstract. The stability of electric power systems is one of its most important properties. This article discusses small-disturbance rotor angle stability: aperiodic and oscillatory. The authors consider, typically for stability analysis, the numerical integration methods by modeling in known numerous digital software simulation tools and propose a method for validating of simulation results by benchmark tool instead of field data. The feasibility of the proposed approach is clearly illustrated by the given fragments of the corresponding experimental studies.

Streszczenie. Stabilność systemów elektroenergetycznych jest jedną z jego najważniejszych właściwości. W artykule omówiono stabilność kątową wirnika o małych zakłóceniach: aperiodyczną i oscylacyjną. Autorzy rozważają, typowo dla analizy stabilności, metody integracji numerycznej poprzez modelowanie w znanych licznych narzędziach do symulacji oprogramowania cyfrowego i proponują metodę walidacji wyników symulacji za pomocą narzędzia wzorcowego zamiast danych terenowych. Wykonalność proponowanego podejścia wyraźnie ilustrują podane fragmenty odpowiednich badań eksperymentaln. (Walidacja aperiodycznych i oscylacyjnych obliczeń stabilności w praktycznych układach elektroenergetycznych)

Keyword: aperiodic and oscillatory stability, electric power system, simulation, validation.
Słowa kluczowe: stabilnośc systemu elektroenergetycznego, oscylacje, walidacja

Introduction

The stability of electric power systems (EPS) is one of its most important properties. The systematic basis for classifying power system stability was developed into appropriate categories by the CIGRE Study Committee 38 and the IEEE Power System Dynamic Performance Committee, according to which the stability is determined by the main system variable in which the instability can be observed and the size of the disturbance considered [1]. This article deals with the small-disturbance rotor angle stability that in the result can be of two forms: the increase in the rotor angle through a nonoscillatory or aperiodic mode due to the lack of synchronizing torque, or rotor oscillations of increasing amplitude due to the lack of sufficient damping torque.

In general the rotor angle stability problem involves the study of the electromechanical oscillations inherent is extremely important and relevant given the fact that modern EPS are constantly being updated: continuing growth in interconnections, the use of new technologies and controls, and the increased operation in highly stressed conditions. All of these have a significant effect on the properties of EPS. Therefore, with prospective and detailed design, the development of special automatic control devices, changes in EPS operating conditions, etc., its ability to regain a state of operating equilibrium is checked, that is the keep of stability. The aperiodic stability is associated with a change in the active power balance in EPS, the system must restore equilibrium between the electromagnetic torque and the mechanical torque of each synchronous machine in the system. Otherwise, the perturbation will increase the angle δ, as a result, the machine may fall out of synchronism.

Oscillatory stability is associated with the settings of automatic voltage regulators (AVR) of generators, since in some combinations of the circuit state condition and settings of excitation regulators, fluctuations in the control system can occur, causing increasing fluctuations in the angle δ until the machine drops out of synchronism [2, 3].

There are different approaches in estimating the static stability of EPS [4, 5], qualitative methods are widely used, for example, the use of Lyapunov functions, estimating the eigenvalues of the matrix. However, finding a suitable Lyapunov function has always been a difficult task, requiring significant mathematical calculations and transformations [6, 7]. Therefore, it is usually analyzed using numerical integration methods by modeling in numerous well-known digital software modeling (ST) tools, for example, Eurostag, PSS\E, ETAP, DIgSilent PowerFactory and ect. [8]. The authors [9] describe in detail the fundamental problems of numerical methods for solving differential equations, show that in this regard there is a problem of obtaining reliable modeling information that is inherent in all ST for calculating EPS modes, as a result of which the reliability of such calculations is often unsatisfactory.

Thus, the use of mathematical simulation results, with the unknown completeness and reliability of information about modes and processes in EPS, can lead to incorrect design and operational solutions related to the analysis of EPS dynamic stability, and the development of events and means for its preservation and improvement. This necessitates validation of such information [9]. It is obvious that in general, the validation of the calculation of modes and processes in EPS should be carried out using full-scale measurement data. However, the published results of the validation demonstrate the differences between the obtained process information in EPS and the full-scale data, which confirms the existence of the above problem of numerical modeling of large EPS. The authors in [10] described challenges related to existing validation approaches, in which the problem of mismatch of simulation results in ST is solved by adapting the calculation results to full-scale data by varying model parameters, mainly static characteristics of loads and regulators. It is important to understand that this approach allows you to adapt the model to a specific disturbance, but also does not solve the problem of comprehensive validation [11].

The article proposes an alternative approach to comprehensive validation of the calculation of aperiodic and oscillatory stability of EPS, based on the use of data modeling from benchmark tool (BT) as the source information, instead of full-scale data. An article is devoted to this issue, which is organized as follows: Methodology of comprehensive validation of ST describes the proposed approach to comprehensive validation of EPS stability calculations. Case studies presents the results of pilot studies confirming the feasibility of the proposed approach. The conclusions summarize the main findings.

Methodology of comprehensive validation of ST

To perform a comprehensive validation of ST calculation of the aperiodic and oscillatory stability of EPS as a source of the complete and reliable information, a model standard is attached – created on the basis of the Hybrid Real-Time Power System Simulator (HRTSim), which provides the non-compositional reproduction of a single continuous spectrum of normal and abnormal quasi-stable and transient processes in real time over an unlimited interval with the guaranteed acceptable accuracy in a particular equipment and simulated three-phase EPS in general [10, 12]. The validation of a BT created on the basis of HRTSim can be performed according to any state or process, for example, according to a quasi-stable one obtained using SCADA. Since the HRTSim uses the same detailed mathematical model EPS for all states and processes and uses a methodically accurate solution method. Therefore, the validation of one state can be guaranteed to extend to the entire spectrum and transients, including switching overvoltage. Thus, given the HRTSim properties, it can be used as a BT. Accordingly, a comprehensive validation of ST technique is proposed in the small-disturbance rotor angle stability, which is determined by a sequence of actions:

1. Setting and reproducing the circuit state condition and validating the reproduced data.

On the basis of the normal electrical scheme of a specific EPS and its database of equipment parameters and process automation settings in HRTSim and validating ST, within its capabilities, the initial circuit state condition of the simulated EPS is reproduced (if there are data from measuring devices of the simulated EPS, the possibility of using this information is not excluded). In the absence of some simulation data, their automatic calculation is carried out, based on the equations of current balances, active and reactive powers in adjacent nodes, taking into account power losses and voltage drops in transmission lines, transformers. Thus, the original circuit state condition of the simulated EPS is set.

The next step is the data validation, which is based on the evaluation of reliability of circuit state condition parameters of EPS model reproduction: current and power balances in nodes, state of switching equipment, the validity of active and reactive power values, currents and voltages in power lines and transformers. Based on the results of checking the validity of the PMU/SCADA data and detecting errors, they are automatically corrected.

2. Formation and implementation of validation scenarios.

To implement scenarios to evaluate the small-disturbance rotor angle stability of the simulated EPS, it is necessary and sufficient to reproduce a number of perturbations that can lead to aperiodic or oscillatory loss of stability. Due to the variety of factors, conditions and processes leading to the violation of the rotor angle stability, their complete validation is the subject of separate studies. Usually, the validation scenario for assessing the reliability of calculations of steady-state operation used to determine the aperiodic stability limit consists in load power increase representing the same increase in the generation and consumption of active power. In this case, generators mutual angles, the angles between voltage vectors at the terminals of study area are controlled.

The validation scenario for assessing the reliability of the calculations of transients used to determine oscillatory instability is focused on the most significant analysis currently associated with the work of automatic control systems: AVR with Power System Stabilizer (AVR with PSS), as well as frequency and power. This analysis can be performed either by methods of the automatic control theory or by the results of calculation of a transient process at small disturbances [4]. Due to the inapplicability of classical mathematical methods of the automatic control theory to evaluate the oscillatory instability of real EPS, this assessment is carried out based on the results of calculation on ST of the corresponding transients caused by small perturbations. For this purpose in states close to limit values of internal angles of synchronous machines taking into account the standard margin, load changes are created, leading to changes of mutual angles of generators (10-30 deg.). The evaluation of the oscillatory instability is carried out by waveforms of changes in mutual angles, the excitation voltage, the active and reactive power (RP), the frequency and voltage of the stator of synchronous machines.

Case studies

Experimental studies were carried out according to the proposed methodology. As the ST adopted common in the world practice the complex calculation state and electromechanical processes to EPS. A real three-phase normal electrical circuit of the Tomsk region was adopted as a model of the power system. The problem of obtaining the parameters of the real power system is widely known; moreover, the data in the power system is constantly changing. Therefore, it is not always successful to compare the operating mode of a real power system with a ST. In our case, the parameters of all models, all network elements of the simulated EPS, are set based from the data of dispatcher measurements, the configuration is set on the basis of the dispatch diagram. The electrical machines (EM) parameters, their excitation systems with AVR and PSS and prime movers, taking into account their control systems, as well as the characteristics of the mechanisms driven by electric motors, are set averaged according to the corresponding reference data. The transmission lines parameters mutual induction, the automatic control system laws of controlled shunt reactors and the characteristics of the magnetization of transformers (autotransformers) and EM are set similarly. The model includes 200 three-phase units, 42 electrical machines, 42 transformers, 97 transmission lines, 63 loads. At the first stage, the actual validation of the BT – HRTSim with SCADA was performed, the results showed a high level of coincidence. Also quasisteady-state compared process of short term electromagnetic transient as a result of which also get a good match [13]. At the second stage, a similar state of the all the EPS elements models and their parameters in the ST was formed as BT.

1. Validation of aperiodic instability

An intersystem transmission line of 500 kV was chosen as the investigated cross-section with a heavier state to assess the reliability of calculations of the aperiodic instability limit. Waveforms of the controlled operating variables in the critical stability state of the simulated EPS are shown in Figure 1.

According to the results presented, with the same load power increase, the processes calculated using ST and reproduced in BT (HRTSim) are fundamentally different and the violation of aperiodic instability in EPS in this experiment using ST is not detected. In fig. 1a, the waweforms of the δ angles demonstrates the rotation of all three EPS generators starting from 7 sec., then G2 and G3 are pulled into synchronism for about 8 sec., and G3 continues to asynchronous state. The HRTSim software (device) allows oscillographing the generators angles only in the range of 0 – 360 deg., therefore, discontinuities are visible on the graph. In fig. 1b, asynchronous state is not observed.

Fig.1. The waveforms of generators mutual angles and voltage, and voltage angles at 500 kV substation (a) HRTSim, (b) ST

The results of their validation are presented in the table 1.

Table 1. Controlled operating variables in the critical state

.

The greatest differences were obtained in the amplitude and phase of the voltage at the 500 kV substation and are caused by the discrepancy between the flux distribution of the PM in the network and the EM load by RP with their identical active power in the critical state and the initial steady-state (Table 2).

In particular, when the state becomes heavier, the loading of the generators according to RP occurs individually, in accordance with their sensitivity coefficients to various changes in the network, which significantly differ in static and dynamic modeling.

Table 2. Reactive powers of EM and their difference

.

2. Validation of oscillatory instability calculations

To determine the oscillatory instability, a 5% load power increase is performed on the PS. At the same time, for the monitored generator equipped with AVR with PSS, ST and BT (HRTSim) are initially set to the same average statistical AVR with PSS settings.

Figure 2 shows waveforms of the angle δ, voltage, and frequency, and excitation voltage, active and reactive power of the generator of one of the power plants.

Fig.2. The waveforms of generator processes at the first combination of AVR with PSS settings HRTSim, (b) ST

According HRTSim reproduced synchronous oscillation processes occur which are missing in similar waveforms obtained with ST for the following reasons:

1. The stator voltage of the generator on the ST waveform changes instantly with load power increase, due to simplified models of EM and network elements, therefore, the effect of AVR with PSS does not appear after the voltage drops to a level of ∆U = 0.18 kV, corresponding to a new flux distribution with a load power increase, and the corresponding the response of the main channel to the voltage deviation forms the excitation voltage of the critical value. Continuous operation of AVR with PSS in HRTSim contributes to a lower voltage drop at the generator terminals ∆U = 0.03 kV. Therefore, despite the identical mathematical models of automatic control systems in both means, in particular AVR with PSS, on the adequacy of which the results of the oscillatory instability assessment also depend, the completeness and reliability of the calculation of operating parameters, to a change in which automatic control systems respond, determine the nature of the course of transient processes, and, accordingly, the completeness and reliability of the assessment of oscillatory stability.

2. The influence of the transformer electromagnetic force (EMF) on the transient process of changing the variable equations of the stator voltage along the d and q axes is demonstrated by the waveforms presented in Figures 3.

The error in calculating the stator voltage along the d axis with the exclusion of the transformer EMF is associated with a large number of circuits along the d axis. The largest oscillation amplitude is characteristic of the transformer EMF, it mainly determines the transformation of oscillations between the rotor and stator circuits, leading to their occurrence in the stator voltage and the corresponding action of AVR with PSS, which contributes to the occurrence of synchronous oscillations of the generator. Therefore, the neglect of the transformer EMF leads to a significant distortion of transient processes reproduction at small disturbances, amplified by the incorrect functioning of automatic control systems, which does not allow, in most cases, to carry out a reliable assessment of oscillatory instability.

Fig.3. The waveforms of changes in the variables of the stator voltage equation along the d (a) and q (b) axes of the station generator

3. Significant and constant in magnitude inductive reactance of static network models elements for only one frequency value (50 Hz) distort the propagation of oscillations, especially low-frequency ones, which, together with a low probability of their occurrence in EM with their simplified modeling, significantly reduces the possibility of oscillations at small disturbances.

Conclusion

An alternative way for solving the problem of validation of calculations of small-disturbance rotor angle stability: aperiodic and oscillatory instability proposed, which consists in using the benchmark tool – HRTSim, as a source of initial data.

The results of experimental studies of the developed comprehensive validation ST tools in terms of calculating the aperiodic and oscillatory instability have confirmed the theoretically and practically grounded properties and capabilities of HRTSim, allowing for a guaranteed comprehensive validation of existing STs. The error in calculations of EPS aperiodic stability in ST is associated with inadequate EM loads on RP and its network flow distribution due to the use of static models. Validation of small-disturbance rotor angle stability calculations performed using ST revealed differences from processes reproduced using HRTSim, due to distortions of the generation and propagation of waveforms, especially low-frequency ones, associated with the simplification of EM models and the use of static models.

Acknowledgment – This work was supported by the Ministry of Science and Higher Education of the Russian Federation, project No. MK-3249.2021.4

REFERENCES

[1] Kundur P., Paserba J., Ajjarapu V., Andersson G., Bose A., Canizares C., Hatziargyriou N., Hill D., Stankovic A., Taylor C., Van Cutsen T. and Vittal V., Definition and classification of power system stability, IEEE Trans. Power Syst., 19 (2004), No. 3, 1387-1401, doi: 10.1109/TPWRS.2004.825981
[2] Komkov A. L. et al., Implementing the System Functions of the Automatic Proportional-Derivative Excitation Control of Synchronous Generators, Power Technol. Eng., 53 (2019), 356–359, doi:10.1007/s10749-019-01084-y.
[3] Hannan M. A. et al., Artificial Intelligent Based Damping Controller Optimization for the Multi-Machine Power System: A Review, IEEE Access, 6 (2018), 39574-39594, doi: 10.1109/ACCESS.2018.2855681.
[4] Kundur P., Power system stability and control. New York, NY, USA: McGraw-Hill, 1994
[5] Burghetti A. Handbook of electrical power system dynamics: Modeling, stability and control Canada: Institute of Electrical and Electronics Engineers, 2013
[6] Jastrzębski M., Kabziński J., Mosiołek P., Adaptive Motion Control with State Constraints Using Barrier Lyapunov Functions, Przegląd Elektrotechniczny, 92 (2016), No. 4, 112-119, doi:10.15199/48.2016.04.24
[7] Cifci A., Uyaroglu Y., Energy function analysis of a twomachine infinite-bus power system by Lyapunov’s second method, Przegląd Elektrotechniczny, 88 (2012), No. 2, 270-273
[8] Hatziargyriou N. et al., Definition and Classification of Power System Stability Revisited & Extended, IEEE Trans. Power Syst, doi: 10.1109/TPWRS.2020.3041774
[9] Guo H. et al., A critical review of cascading failure analysis and modeling of power system, Renewable and Sustainable Energy Reviews, 80(C) (2017), 9-22. doi:10.1016/j.rser.2017.05.206
[10] Suvorov A. et al., Comprehensive Validation of Transient Stability Calculations in Electric Power Systems and Hardware Software Tool for Its Implementation, IEEE Access, 8 (2020), 136071-136091, doi: 10.1109/ACCESS.2020.3011207
[11] Huang Z., Nguyen T., Kosterev D., Guttromson R., Model Validation of Power System Components Using Hybrid Dynamic Simulation, 2005/2006 IEEE/PES Transmission and Distribution Conference and Exhibition, Dallas, TX, USA, 2006,153-160, doi: 10.1109/TDC.2006.1668475
[12] Suvorov A. et al., Potential Application of HRTSim for Comprehensive Simulation of Large-Scale Power Systems with Distributed Generation, International Journal of Emerging Electric Power Systems, 20(5) (2019), doi: 10.1515/ijeeps2019-0075
[13] Suvorov A., Gusev A., Andreev M., Askarov A., A validation approach for short-circuit currents calculation in large-scale power systems, Int. Trans Electr Energ Syst., 30 (2020) No 4, e12276, doi: 10.1002/2050-7038.12276

Authors: PhD, senior lecturer of Department of Electric Power Systems, Igor Razzhivin, Tomsk Polytechnic University, 30, Lenin Avenue, Tomsk, Russia, E-mail: lionrash@tpu.ru; PhD, associate professor of Department of Electric Power Systems, Aleksey Suvorov, Tomsk Polytechnic University, 30, Lenin Avenue, Tomsk, Russia, E-mail: suvorovaa@tpu.ru; PhD, associate professor of Department of Electric Power Systems, Mikhail Andreev, Tomsk Polytechnic University, 30, Lenin Avenue, Tomsk, Russia, E-mail: andreevmv@tpu.ru; doctor of science, professor of Department of Electric Power Systems, Aleksander Gusev, Tomsk Polytechnic University, 30, Lenin Avenue, Tomsk, Russia, E-mail: gusev_as@tpu.ru

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 97 NR 6/2021. doi:10.15199/48.2021.06.26

Energy Management for a New Power System Configuration of Base Transceiver Station (BTS) Destined to Remote and Isolated Areas

Published by 1. Abdallah MOUGAY1, 2. Mohamed KHATIR1, 3. Mohamed FLITTI1, 4. Sid-Ahmed ZIDI1, 5. Ahmed Ganoune2, ICEPS Laboratory, Department of Electrical Engineering, Djillali Liabes University, Sidi Bel Abbes, 22000, Algeria (1) Department of Electrical Engineering, Taher Moulay University, Saida, 20000, Algeria (2) . ORCID: 1. : 0000-0001-6676-4466; 2. 0000-0002-3906-9131; 3. 0000-0002-7469-20174. : 0000-0001-7224-8812; 5. 0000-0003-2024-4275.

Abstract. This paper discusses the energy management for the new power system configuration of the telecommunications site that also provides power to electric vehicles. The modeling and control of the proposed system, composed of hybrid energy sources that are photovoltaic panels and a diesel generator with batteries, are also presented. The hybrid system will provide energy to a telecommunications site located in an isolated area. The management algorithm used in this work aims to significantly reduce the investment costs of the power system. From an environmental point of view, the hybrid system can reduce gas emissions from the diesel generator, while maintaining spaced maintenance targets compatible with the operation of the isolated site.

Streszczenie. W artykule omówiono zarządzanie energią w nowej konfiguracji systemu elektroenergetycznego obiektu telekomunikacyjnego, który zapewnia również zasilanie pojazdom elektrycznym. Przedstawiono również modelowanie i sterowanie proponowanym systemem, składającym się z hybrydowych źródeł energii, którymi są panele fotowoltaiczne oraz generator spalinowy z bateriami. System hybrydowy dostarczy energię do zakładu telekomunikacyjnego zlokalizowanego na odizolowanym obszarze. Zastosowany w pracy algorytm zarządzania ma na celu znaczne obniżenie kosztów inwestycyjnych systemu elektroenergetycznego. Z punktu widzenia ochrony środowiska. System hybrydowy może zmniejszyć emisje gazów z generatora Diesla, przy jednoczesnym zachowaniu rozmieszczonych w odstępach celów konserwacji zgodnych z działaniem odizolowanego miejsca. (Zarządzanie energią dla nowej konfiguracji systemu zasilania bazowej stacji nadawczo-odbiorczej (BTS) przeznaczonej do odległych i odizolowanych obszarów)

Keywords: Hybrid system, Base transceiver station (BTS), Photovoltaic system, Diesel generator, Electric vehicle, Batteries.
Słowa kluczowe: hybrydowy system zasilania, bazowa stacja nadawczo-odbiorcza BTS.

Introduction

Algeria is a developing country where many households are located in isolated areas or at a significant distance from the power grid. The costs of connection to the power grid are high and sometimes connection is simply not possible. This is why independent systems are interesting to meet the energy needs of the population in these areas. Generating electricity from renewable energy sources gives consumers greater assurance that their electricity is environmentally friendly. However, the random nature of these sources forces us to establish rules for the design and use of these systems to get the most out of them. On the other hand, the global expansion of cell phone base stations is increasingly taking place in areas where the power grid is often subject to relatively long outages or where access to the power grid is not available. Diesel generators are used to supply power to one or more base transceiver stations (BTS) also in these areas. These require extensive maintenance and consume relatively high levels of diesel fuel [1], [2]. Diesel generators therefore generate high operating costs and mobile network operators face the challenge of limiting the total cost of ownership. In this case, solar photovoltaic energy (PV) seems to be the most attractive solution to meet the energy needs of a case station in many parts of Algeria [3], [4]. Algeria is located between 36°42′ north latitude and 03°13′ east longitude, making it an ideal location for the use of solar energy. The daily solar radiation varies between 3.8 and 6.5 KWh/m², and it should be noted that Algeria has one of the largest solar deposits in the world. The average annual rate of sunshine exceeds 3000 hours. It is also the most important of the whole Mediterranean basin with 169440 TWh/year. The average solar energy received is 1700 KWh/m²/year in the coastal regions, 1900 KWh/m²/year in the highlands and 2650 KWh/m²/year in the Sahara. Our country can therefore cover part of its energy needs with photovoltaic systems. This work introduces a new algorithm that manages and clarifies the transit of energy according to priorities to manage our hybrid system (PV panels + diesel generator + batteries) to ensure the continuous and sustainable reliability of energy to supply the Telecom site in isolation and electric vehicles, and the algorithm determines the optimal size of the photovoltaic generator equipped with batteries and the diesel generator for the BTS [5], [6]. However, regardless of the methodology used and the accuracy with which the different elements of the PV array are taken into account, two types of estimates are still confronted. The first one requires a large climate database and uses an accurate prediction based on complex simulations. The second uses an algorithmic sizing method. The later is the most common method for telecommunication stations characterized by low power [7].

Materials and Methods Presentation of the Telecommunications site

The site is a BTS station, owned by one of the Algerian cell phone network operators, located on the side of the national road No. 6 in an isolated area of a city in southern Algeria called Bechar. The average annual solar radiation in this region is estimated at 5.52 KWh/m2/day. The total power of the instantaneous communication equipment is evaluated from the standby generator screen (power generated), throughout the day because the communication equipment operates 24 hours a day. The site has diesel generators that operate to ensure efficient use and extend the life of the equipment. To avoid the need to refuel every 15 days, a fuel tank is installed with automatic transfer of diesel fuel to the units. This study aims to add solar panels and batteries to the previous system for several reasons; firstly, the presence of year-round solar radiation on the site, secondly to save fuel consumption, thirdly to reduce gas emissions, and fourthly to power electric vehicles in the area. To this end, a hybrid system consisting of solar panels, batteries and a diesel generator was developed.

Supplying electric vehicles with electrical power in a BTS station

The role of a BTS is to convert the electrical energy of a signal into electromagnetic energy carried by an electromagnetic wave (or vice versa). To ensure their operation, GSM mobile relays need a continuous and reliable power supply. This energy often comes from an electrical distribution network with a back-up source [8], [9]. This paper addresses the possibility to power electric cars through a BTS relay. Electric vehicles save energy in a storage unit such as a battery. Electricity is used to drive the wheels of an electric vehicle by means of an electric motor. They have a specific energy storage capacity, which must be replenished by connecting them to an electric charger. Electric cars do not emit pollutants into the atmosphere when driven. Thus, no NOx, fine particles, unburned hydrocarbons or other carbon monoxide, often blamed for their health impact, are released into the environment. There are still particulate emissions from the tires and brakes of all vehicles, but the switch to electric vehicles has an immediate benefit for air quality in cities and near roads. Figure 1 represents an electric vehicle and an electric charger.

Fig.1. Electric Vehicles

The Technical and economic study of a hybrid system

Hybrid systems are technically, economically and ecologically advantageous compared to conventional diesel and renewable photovoltaic systems combined with diesel generation solutions. They provide energy for the telecommunications site with the possibility of charging some electric cars passing through the site and recharging them with electricity via the site’s electric charger. The dimensional rigidity of each electrical installation ensures a better efficiency of the last test, so we simulated all the phases of the photovoltaic panel hybrid system with batteries/Diesel to determine their reliability before installation. For this purpose, a general simulation program of the system over 24 hours was developed. The results of the simulation are represented to visualize the passage of electrical energy from the sources to the load and to the general system.

Presentation of the hybrid system

In Figure 2, the hybrid system is composed of four essential parts: a diesel generator operating as a core power generator and a photovoltaic panel field producing renewable energy, and a storage system placed next to the load and the telecommunications site. The electric vehicles represent the load and finally the charger of the electric vehicle.

Fig.2. Configuration of system telecommunications equipment and charger for electric vehicle

Case Studied

The unexpected increase in the number of subscribers and the demand for high-speed data has led to enormous growth in cell phone networks in recent years. Indeed, cell phone networks (GSM relay, radar) have evolved to meet the needs of mobile subscribers and the extension of the coverage area. The role of a GSM relay is to convert the electrical energy of a signal into electromagnetic energy carried by an electromagnetic wave (or vice versa). To ensure its operation, the GSM relay needs a stable and reliable power supply. To this end, we prepared a hybrid system consisting of solar panels, a diesel engine and batteries to power the Telecom site and charge the electric cars with an on-site electric charger [10], [11].

Peak power calculation of the photovoltaic panels array

As the irradiation varies from month to month, the peak power of the studied photovoltaic field varies during the months of the year, and the calculation of this power is given by Equation (1):

.

With: PP: Peak power of the photovoltaic field (W); EReq: Daily requirement (Wh/day); SSTC: Sunshine in STC conditions (SSTC = 1KW/m2); SMonthly: Sunshine scaled annual average (KWh/day/m2); CL: Correction factor applied to take account of the different losses (CL = 0.7).

Thus, the numerical application for this case study is as follows:

.
Choice of modules

Depending on the total power required by the loads as well as the type of our installation (not connected to the grid), we opted for photovoltaic modules with a power of 580 WP each.

Calculation of the number of photovoltaic panels

The number of photovoltaic panels modules is determined by Equation (2):

.

with: NP: Number of photovoltaic panel; PP: Total power of photovoltaic fields; PUnitary: Power of a photovoltaic module. Thus, the numerical application for this case study is as follows:

.

So, if we opt for a NP = 29 panels, the peak power of the field will be:

PP = 29 x 580 = 16820 W

Diesel generator

The proper size of the diesel generator is very important to avoid low load or energy shortage, and the power produced by the diesel generator is represented by Equation (3):

.

with: PN: Rated output power of the diesel generator [KW]; ηDg: Efficiency of the diesel generator [%]; TDg: Diesel generator running time [h].

The generator set is generally sized to cover peak consumption. In our case, the power reaches 5.8 KW. We therefore choose a diesel generator with a power of 6KW.

Sizing of the battery bank

Energy storage plays an important role in a stand-alone hybrid energy system. In most cases, batteries remain the most cost-effective technology.

Choice of voltage and calculation of capacity

We choose batteries with a voltage of 2V each. Knowing that in the case of our system it is the storage batteries which impose the voltage on the PV field. The battery with 250 Ah storage capacity. For the case of our load, we want to have autonomy of 3 days. The field capacity of standard batteries is given by the relation:

.

with: CB: Total battery capacity (Ah); RD: Daily requirement (Wh/day); DAuto: Number of days of autonomy; VB: Battery voltage (V); MDD: Maximum depth of discharge (80%); KB: Battery temperature coefficient (0.85).

Thus, the numerical application for this case study is:

.

The number of batteries is 24 in series which keeps the same capacity of CB = 5895 Ah, which ensures a voltage of V = 48V.

Use and proper functioning

The use of batteries is subject to constraints that must be respected to ensure their proper functioning and longevity. They cannot remain unused for long periods without negative consequences on their lifespan. Repeated random charge/discharge cycles must be avoided. Their state of charge must not reach extreme values to avoid premature degradation. The role of this storage system is to provide the charge for a relatively long period of time (hours or even days). In this work, we seized the different elements that make up our PV/Diesel hybrid system with storage batteries. We studied the energy conversion chain (DC/AC) and proceeded to the choice of the module to be used as well as the power of the generator needed as an emergency source. The objective is to limit the intervention of the generator in the most unfavorable months, for this we dimensioned an efficient storage system to overcome this disadvantage. The optimization of the energy produced by the panel requires the installation of an MPPT regulator, to maximize and force its operation at its maximum power.

Table 1.The hybrid system sizing

.
The hybrid system architecture

The concept of decentralized electricity has encouraged the development of means of production from renewable sources. The current trend shows that the integration of this type of resource in isolated electrical systems is done in association with the use of a conventional source, such as diesel generators. Thus, the photovoltaic generator operates either in parallel or alternately with the diesel generator. Thus, there are several configurations of PV/Diesel hybrid systems [12], [13].

Fig.3. Hybrid system topology

Figure 3 shows the solar panels will keep the system powered and the batteries charged. In the absence of optical radiation, the storage system will intervene to compensate for the lost energy, but in the absence of optical radiation and the storage system, the generator will compensate for the lost energy when the electrical circuits switch automatically (through the switch) [14].

Economic analysis

It is important to study the economic importance of the hybrid system consisting of solar panels, a diesel generator and batteries to ensure that it is more cost-effective over the life of the project and the environment in terms of reducing gas emissions. To this end, we used the HOMER program to determine all the details of the climate and the amount of energy the site needs every hour, as well as the current price of fuel and the amount of energy that can be stored to continue developing the communication site and electric vehicles with the necessary energy. The goal is to use the hybrid system consisting of a clean renewable energy source, a diesel generator and batteries, and finally we see the installation of the hybrid system. The energy management is carried out according to an algorithm that guarantees a permanent supply of energy at the lowest cost, taking into account the economic aspect.

Energy management strategy

The flowchart is a schematic representation, to show and visualize the transit of energy, in order of priority and according to its usefulness. Figure 4 presents an energy management flowchart of our overall PV/Diesel system. The schematic representation of an autonomous electrical power generation system via a flowchart allows good energy management to ensure continuous and permanent energy reliability and longevity of our PV/Diesel system [15], [16]. Algorithm of the hybrid system

A. Case n°1: PVPower > LP (load power)

If the PVPower supplied by the photovoltaic system is superior to the load power (LP) and the state of charge of the batteries is less than 80%, the surplus will be supplied to the storage batteries. If the state of charge of the batteries is at 80%, the excess produced by the photovoltaic field will be supplied to the dissipative load.

B. Case n°2: PVPower < LP (load power)

If the PVPower supplied by the photovoltaic system is less than the load power (LP) and the state of charge of the batteries is greater than 20%, the batteries provide the energy deficit. On the other hand, if the state of charge of the batteries is less than 20%, the generator will provide the load and the surplus will be used to charge the batteries.

Fig.4. Flowchart of the algorithm

Load characteristic Scenario

The hybrid system (photovoltaic fields with batteries and diesel generator) supplies the Telecom site with power for 24 hours. The presence of sufficient radiation causes the photovoltaic field to produce high power to power the load; otherwise, the storage system powers the load. In the absence of sufficient radiation and the inability of the storage system to provide energy, the diesel generator powers the load. As for the electric vehicles, they are charged at different times by an electric charger located at the telecommunications site: the first vehicle is charged at 2 am with a 3 KW charge, the second vehicle is charged at noon with a 3.5 KW charge and the third vehicle is charged at 5 pm with a 2.5 KW charge (see the table 3). There, it can be argued that our system has the potential to power the telecommunications site and electric vehicles.

Assessment of the electricity consumption of a GSM site and electric vehicles

The relay is equipped with a photovoltaic field with batteries and a diesel generator, and this voltage is converted to DC voltage to power telecommunications equipment. The consumption of GSM relays varies depending on the operating system, the air conditioning requirements of the communication equipment and the site lighting. By estimating the total consumption of a GSM site for this purpose, using measurements (metering) of the sites in service, this consumption can be as high as a few kilowatts. Thus, it will be possible to recharge electric vehicles passing through the site using an electric charger [17], [18].

Table 2. Detail of the Telecom site load during the day

.

Table 3. Detail of the electric vehicles during the day

.

Table 4. Global load of Telecom site and electric vehicles in the day

.
Results and discussion

Simulation results highlight the importance and role of the hybrid system in isolated regions as well as from an economic and environmental perspective.

Load curve for Telecom site

The load characteristic (Telecom site) represents the variation of the energy used over time. In Figure 5, the load curve graph contains three consumption peaks for our isolated site. It can be observed that the phase that takes the interval from 6 pm to 9 pm, its consumption is very high at 7 pm (estimated at 5,167 KW) [19].

Fig.5. Graph of the daily consumption of the telecom site
Load curve for Telecom site and electric vehicles
Fig.6. Graph of the daily consumption of the telecom site and electric vehicles

Figure 6 represent the total load curve graph (Telecom site + electric vehicles) shows four consumption peaks, and it can be seen that the phase that takes the time interval from 10 am to 2 pm has a very high consumption at noon (estimated at 5.8 KW).

Sunlight profile

Figure 7 shows typical sunshine values over a 24-hour period. Starting at 7 am, sunshine values begin to increase, with a peak at noon, and then decrease to zero at 6 pm.

Fig.7. Graph of the daily sunshine profile
Battery power curve
Fig.8. Graph of the power produced by the batteries

Batteries are an additional source to the photovoltaic field, at the end of the charge. Figure 8 illustrates the curve of energy supplied by the batteries. The battery charges from 8 am to 4 pm, while from 6 pm to 10 pm it provides energy to cover the energy deficit. Depending on the peak of the energy supplied by the battery, there are three important intervals:

Between midnight and 8 am, power consumption decreases and the battery begins to provide energy to the site;

Between 8 am and 4 pm, the battery is charging, so it does not provide energy;

From 4 pm to midnight, the battery provides power to the load to cover the energy deficit in the system.

Power curve (PV, LOAD, BATT)

To analyze the power supplied by either the PV field or the batteries on the one hand and the power consumed by the load on the other hand, the power, PV field, load and storage power diagrams of the system are recorded on the same graph in Figure 9.

Fig.9. Graph of the Power (PV, LOAD, BATT)

From midnight to 6 am, the batteries exclusively power the charge. Between 6 am and 7 am, the charge is provided by the photovoltaic system and the storage system. From 8 am to 4 pm, the field produces enough energy to power the load and a surplus is supplied to the storage batteries or the dissipating load. From 4 pm to 5 pm, the PV system and the storage system provide the charge. In the absence of sufficient radiation, the power produced by the field is zero and the storage system supplies energy to the load.

The role of a diesel power generator in the hybrid system

In a hybrid system, the photovoltaic field injects energy immediately into the communication site, and stores the surplus in batteries for later use. In the event that the PV system and the storage system do not provide energy, the diesel generator will provide the necessary energy to consider it as a source of energy available at all times [20].

Simulation with HOMER

Enter the necessary data in the software HOMER

• The data of equipment
• The data of the PV generator
• The data of the batteries
• The data of the converter
• Data from the diesel generator
• Data on the fuel
• Control data and system constraints
• Launch of the calculation
• Results

Assessment of the energy resource available on the site

For the data, simply enter the longitude and latitude of the desired location, and a simple click on the “Get Data from the Internet” icon gives the results. Figure 10 shows the solar radiation data for the study area. It can be seen that the radiation varies between 3.200 KWh/m2/ day for the month of December and 7.450 KWh/m2/day for the month of June with an annual average of 5.52063 KWh/m2/day. The monthly brightness index is defined as the ratio of terrestrial radiation to extraterrestrial radiation. The values of the latter vary by location and season. In Figure 10, it is observed that the monthly clarity index varies between 0.590 in December and 0.672 in April and that the annual clarity index is equal to 0.635.

Fig.10. Solar radiation data for the study area

Energy demand assessment (load profile)

In our case study, we imported a data file from the site is a BTS station, owned by one of the Algerian cell phone network operators to present the load profile, as shown in Figure 11 and 12.

Fig.11. Load profile 1
Fig.12. Load profile 2
Starting calculations

Once all data entered, we obtain the architecture of the system presented in Figure 13. Now it is enough to launch the calculation. A click on the button ‘calculate’ will display the results.

Fig.13. System architecture after entering the required data

Considering all inputs, HOMER simulates repeatedly to get suitable solution. Optimization results are displayed in terms of categorized and overall, showing most feasible architecture which satisfied all inputs and constraints that designers give. After simulating all possible configurations, we obtained the overall results shown in Figure 14. We can see the best solution by type of system.

Fig.14. Results obtained after the simulation

We simulated our system with the HOMER program and obtained the same results as in previous simulations. HOMER simulates system configurations with all combinations of components specified in the entry data. It eliminates the results of all infeasible system configurations, which do not meet the electricity demand and are not compatible with the specified resources and constraints [21]. Results are ranked from highest to lowest in terms of most cost-effective to least cost-effective based on the current net cost of the system. The hybrid system is the most cost-effective over the life of the project, with a cost of 152771 $ and a levelized energy cost of 0.357$/KWh. For a 17 KW PV array system, a 6KW diesel generator, a 24-cell storage system, a 6KW transformer, with a capital cost of 56728$. The results obtained in the simulation results window, represent the detailed technical and economic data on each system installation that HOMER simulates [22], [23].

The figure 15 shows the details of the annual electricity production and consumption for the system. The results indicate in the table 5 that the cost of electricity for a hybrid system with batteries and EV charger is 0.357 $/KWh versus 0.570 $/KWh for the hybrid system without batteries and diesel generator alone 0.640 $/KWh, the hybrid system with batteries and EV charger can be economically viable. It is noted that the hybrid solution with batteries and electric vehicle charger is more reliable and more cost-effective than the other solutions (diesel generator only and hybrid without storage). The analysis of the environmental impact of the studied system allows to determine the emission of air pollutants, we can see that carbon dioxide and nitrogen oxides and carbon monoxide and particles are the principles of combustion gases, the reduction of these emissions is an objective of this study, the hybrid system is the optimal solution because the emissions are lower than the conventional solution (generator). Thus, the hybrid system allows to reduce the use of the generator, which means less gas (CO2 and CO and NOx and PM). The dimensioning of the PV/Diesel electric hybrid energy production system is done based on the knowledge of the energy potential of the site and after evaluation of the daily needs of the isolated studied site. To model the proposed hybrid system, we chose a photovoltaic conversion chain, a model of the cells that make up the panels, to push our field to operate at its maximum power whatever the weather conditions [24], [25]. According to the simulation results, all the characteristics of the hybrid system are more advantageous than diesel alone. Fuel consumption should be increased by switching from the conventional solution (generator only). Considering the technical performance of the hybrid solution, it can be seen that the electric power production is higher in the hybrid system than in the diesel-only system. This is mainly due to the share of renewable energy used in the system, which results in a higher energy surplus than the conventional solution. Thus, the hybrid system can be said to provide a higher load than diesel alone [26], [27]. As a result, the lifetime of the generator set decreases from the hybrid to the conventional solution. Finally, since the emissions of all gases for a hybrid system are much lower than for the conventional diesel solution alone, it can be concluded that from a technical, economic and environmental point of view, the hybrid system is more cost-effective than conventional diesel alone [28], [29].

Fig.15. Statistics on the yearly electrical power production and consumption of the system

Table 5. Comparison of the results obtained

.
Conclusion

The installation of telecommunications networks requires a permanent and uninterrupted power supply and very expensive wiring. Some regions with no means of communication, low population density and difficult access hope one day to be able to communicate with the outside world. Moreover, the aim is to allow operators to cope with a wide variety of constraints during exceptional events. The supply of energy to these facilities is always a delicate issue and the choice of this energy must satisfy both economic and technical conditions. This work focused on the simulation of a photovoltaic system with batteries and a diesel generator to power a telecommunications site and electric vehicles passing through the site via an electric charger. The site is isolated from the electricity distribution network. In addition, details on simulation and dimensioning were provided. To this end, an algorithm was implemented that aims at a good and close management of energy transit to ensure a permanent supply of energy while taking into account the economic aspect of the system. This will make it more profitable over the lifetime of the project and from an environmental point of view in terms of reducing the emissions of gases that cause air pollution. To use HOMER, the user enters the inputs (information on loads, components and resources), HOMER then calculates and displays the results, and the user can examine the results in tables and graphs. HOMER is primarily an economic model and can be used to compare different combinations of component sizes and quantities, and to explore how variations in resource availability and system costs affect the cost of installing and operating different system designs. The installation of a battery photovoltaic generator and a diesel generator for the remote site allows the load to be matched to demand and the power output to be split between the battery photovoltaic generator and the diesel generator. The objective of the simulation is to study and test these performances before its installation. The simulation models, which are sufficiently accurate, are used to create scenarios of conditions closer to practical reality. In perspective, we hope that our simulation and sizing work will be complemented by validation tests in the field to find out the real performance of our hybrid system and that the modeling we carried out will be enriched.

Acknowledgments: This work was supported by Intelligent Control and Electrical Power System (ICEPS) Laboratory of Sidi-Bel- Abbes University, Algeria.

REFERENCES

[1] Bitterlin.I.F., Modelling a reliable wind/PV/storage power system for remote radio base station sites without utility power, Journal of Power Sources, 162 (2006), No.2, 906-912
[2] Afif, B., Merabet, B., Benhamou, A., Chaker, A., Standalone hybrid power system using homer software optimal case sizing of ferraguig (Northwest of Algeria), International Journal of applied Power Engineering, 8 (2019), No.3, 287-298
[3] Yahiaoui, A., Benmansour, K., Tadjine, M., A new sizing algorithm of renewable hybrid systems PV-diesel generator battery: Application to the case of Djanet city of Algeria, 8th International Conference on Modelling, Identification and Control (ICMIC), (2016)
[4] Samer, R., Contribution to the modelling of reversible electrolyser and hydrogen fuel cell systems for coupling to photovoltaic generators, PhD thesis, University of Toulouse, France, 42 (2008), No.8, 183
[5] Rezzouk, H., Mellit, A., Feasibility study and sensitivity analysis of a stand-alone photovoltaic-diesel-battery hybrid energy system in the north of Algeria, Renewable and Sustainable Energy Reviews, 43 (2015), 1134-1150
[6] Mathieu, D., Diagnosis of diesel generator sets by analysis of crankshaft rotation speed, Doctoral thesis, Jean Monnet University, France, 43 (2010), No.1, 215
[7] Issaadi, W., Khireddine, A., Issaadi, S., Management of a base station of a mobile network using a photovoltaic system, Renewable and Sustainable Energy Reviews, 59 (2016), 1570-1590
[8] Kaldellis, J.K., Optimum Hybrid Photovoltaic-Based Solution for Remote Telecommunication Stations, Renewable Energy, 35 (2010), 2307-2315
[9] Alsharif, M., Jeong, K., Hybrid Off-Grid SPV/WTG Power System for Remote Cellular Base Stations Towards Green and Sustainable Cellular Networks in South Korea, Energies, 10 (2017), No.1, 1-23
[10] Benatiallah, A., Mostefaoui, R., Boubekri, M., A comparison of Simplified and Statistical Methods of Sizing Photovoltaic Installation for a Remote Dwelling in the Saharan Region, Journal of the Human-Environmental System, 8 (2005), No.1,1-6
[11] Ike, Dike U., Adoghe, Anthony U., Ademola, A., Analysis of Telecom Base Stations Powered by Solar Energy, International journal of scientific & technology research, 3 (2014), 369-374
[12] Notton, G., Muselli, M., and Louche, A., Autonomous hybrid photovoltaic power plant using a back-up generator: a case study in a Mediterranean Island, Renewable Energy, 7 (1996), No.4, 371-391
[13] Rekioua, Djamila, Matagne, Ernest., Optimization of photovoltaic power systems: Modelization, simulation and Control, Green Energy and Technology Series, Springer, (2012)
[14] Idda, A., Slimani, M.E., Bentouba, S., Hammaoui, Y., Different PV System Configurations for Uninterruptible Power Supply (UPS): Application to GSM Relay, Journal of Advanced Research in Science and Technology, 4 (2017), 574-582
[15] El Khadimi, A., Bchir, L., Zeroual, A., Dimensioning and optimisation technico-economic to a system of energy photovoltaic hybrid wind – with the Storage System, Rev. Energ. Ren, 7 (2004), 73-83
[16] Borrowy, BS., and Salameh, Z.M., Methodology for Optimally Sizing the Combination Battery Bank and PV Array in a Wind/PV Hybrid System, IEEE Transaction on Energy Conversion, 12 (1997), No.1, 73-78
[17] Shafiqur, R., Mahbub Alam, Md., Meyer, J.P., Al-Hadhrami, Luai M., Feasibility study of a wind PV diesel hybrid power system for a village, Renewable Energy, 38 (2012), No.1, 258-268
[18] Shaahid, S.M., El-Amin, I., Rehman, S., Al-Shehri, A., Ahmad, F., Bakashwain, J., Al-Hadhrami, Luai M., Techno-Economic Potential of Retrofitting Diesel Power Systems with Hybrid Wind-Photovoltaic-Diesel Systems for Off-Grid Electrification of Remote Villages of Saudi Arabia, International Journal of Green Energy, 7 (2010), No.6, 632-646
[19] Data file from the site is a BTS station, owned by one of the Algerian cell phone network operators: http://www.algercity.com/economie/operateurs-telephoniques, (2020)
[20] Azah, M., Tamer, k., Design of Hybrid PV/Diesel Generator Systems at Minimum Cost: Case study for kuching, Malaysia, Transmission and Distribution Conference and Exposition (T&D), IEEE PES ,Orlando, FL, (2012), 1-6
[21] Gourbi, A., Bousmaha, I., Brahmi, M., Tilmatine, A., Numerical study of hybrid photovoltaic Power supply system, Journal of power Technologies, 96 (2016), No.2, 137-144
[22] Dekker, J., Nthonotho, M., Chowdhury, S., Economic analysis of PV/diesel hybrid power systems in different climatic zones of South Africa, Electrical Power and Energy Systems, 40 (2012),104-112
[23] Ying-Yi, H., Yuan-Ming, L., Yung-Ruei, C., Yih-Der, L., and Pang-Wei, L., Optimization Capacities of Distributed Generation and Energy Storage in a Small Autonomous Power System Considering Uncertainty in Renewables, Energies, 8 (2015), No.4, 2473-2492
[24] Jacobus, H., Evaluating the impact of adding energy storage on the performance of a hybrid power system, Energy Conversion and Management, 334 (2011), 928-935
[25] Phuangpornpitak, N., Kumar, S., PV hybrid systems for rural electrification in Thailand, Renewable and Sustainable Energy Reviews, 11 (2007), No.7, 1530-1543
[26] Deshmukha, MK., and Deshmukh, SS., Modeling of hybrid renewable energy systems, Renewable and Sustainable Energy Review, 12 (2008), No.1, 235-249
[27] Mondal, A.H., Manfred, D., Hybrid systems for decentralized power generation in Bangladesh, Energy for Sustainable Development, 14 (2010), 48-55
[28] Pirhaghshenasvali, M., Behzad, A., Optimal modeling and sizing of a practical hybrid wind/PV/diesel generation system, The 5th Annual International Power Electronics, Drive Systems and Technologies Conference (PEDSTC), (2014)
[29] Makhija, Satya, P., and Dubey, S.P., Techno-economic analysis of standalone hybrid energy systems to run auxiliaries of a cement plant located in Jamul, Chhattisgarh, India , Environmental Progress & Sustainable Energy, 35 (2015), 221-229

Authors: PhD. Abdallah Mougay, Department of Electrical Engineering, ICEPS Laboratory, Faculty of Electrical Engineering, University of Djillali Liabes, Sidi Bel Abbes, Algeria, E-mail: mougayabdallah.27@gmail.com; Professor. Mohamed Khatir, Department of Electrical Engineering, ICEPS Laboratory, Faculty of Electrical Engineering, University of Djillali Liabes, Sidi Bel Abbes, Algeria, E-mail: med_khatir@yahoo.fr; Dr Mohamed Flitti , Department of Electrical Engineering, ICEPS Laboratory, Faculty of Electrical Engineering, University of Ain Temouchent, Algeria, E-mail:flitti_med@yahoo.fr , Professor Sid-Ahmed Zidi, Department of Electrical Engineering, ICEPC Laboratory ,Faculty of Electrical Engineering, University of Djillali Liabes, Sidi Bel Abbes, Algeria, E-mail: sbzidi@yahoo.fr, PhD. Ahmed Ganoune, Department of Electrical Engineering, Faculty of Electrical Engineering, University of Taher Moulay, Saida, Algeria, E-mail: ahmed.ganoune@univ-saida.dz.

Source & Publisher Item Identifier: PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 97 NR 8/2021. doi:10.15199/48.2021.08.01