Research on New Characteristics of Power Quality in Distribution Network

Published by He Li, Chao Lv, Yinfan Zhang, Inner Mongolia Electric Power Research Institute (China)

Abstract - In recent years, power systems have gradually become electronically powered. The quality of power in the distribution network presents new features. Although power electronics technology brings a qualitative leap to renewable energy, microgrid technology, distributed power supplies, electric vehicles, etc., it also affects the power quality of the distribution network. This paper discusses the new characteristics of power quality in the distribution network, such as rapid voltage fluctuation, frequent three-phase unbalance, subsynchronous harmonics caused by disturbance frequency variation, and ultra harmonics. At the same time, governance recommendations are elaborated. Combined with existing problems, we will look forward to the development direction of power quality in distribution networks in the future.

Keywords - Distribution network, Power electronics, Powerquality, Ultra harmonics, MapReduce

Introduction

Power quality is one of the latest branches in power system research. It refers to maintaining the amplitude and frequency of the distribution bus voltage and current at the rated value, and the waveform is similar to the quality of the sine wave. Power quality problems occur in actual load equipment and components or transmission and distribution subsystems, which are not conducive to the normal operation of various equipment and power systems, affecting the stability, continuity, and reliability of the power system. In recent years, the increasing popularity of clean energy technologies and the development of distributed power sources, microgrid technologies, and electric vehicles are inseparable from power electronics technology. By using power electronic devices, flexible transformation and control of electric energy can be realized, bringing life to people conveniently. The distribution network gradually moves toward power electronics, but it also makes the power quality of the distribution network show new characteristics[1]. This paper will discuss the new characteristics of power quality around the power grid of the distribution network.

NEW CHARACTERISTICS OF POWER OUALITY IN DISTRIBUTION NETWORK

A. Fast Voltage Fluctuation

High-power, large-capacity, and impact loads in power systems are the main causes of rapid voltage fluctuations, such as electrified railways, large-scale rolling mills, variable frequency speed control devices, and large-scale steelmaking electric arc furnaces. Rapid voltage fluctuations can cause problems such as power equipment not working properly or causing unqualified production products. For the voltage fluctuation IEC 61000-4-30, three parameters are described: maximum voltage fluctuation, steady-state voltage fluctuation, dynamic voltage fluctuation, wherein the maximum voltage fluctuation is defined as the maximum and minimum voltages appearing during voltage fluctuations. Poor steady-state voltage variation is defined as the difference between two adjacent steady-state voltage rms values containing at least one voltage variation characteristic[5]. Voltage fluctuation is a series of voltage fluctuations. The determination of a voltage fluctuation needs to identify two periods of at least one stable holding of 1 second. The magnitude of one voltage variation is the difference between two extreme values. The extreme value must be every half of the fundamental voltage period. The root means square value. Rapid Voltage Change (RVC) refers to two physical quantities of maximum voltage fluctuation and steady-state voltage variation in the above voltage fluctuations, which are defined as the same as the voltage fluctuation characteristics. In order to ensure the stable operation of the power system and the good power quality of the user side, it is necessary to define the limits of the rapid voltage fluctuations and establish an effective algorithm for accurate detection and analysis.

B. Frequent three-phase imbalance

The three-phase voltage imbalance of the traditional three-phase three-wire distribution network is mainly caused by the asymmetry of the three-phase load. The three-phase voltage imbalance of the three-phase four-wire distribution network is mainly caused by the imbalance of a large number of single-phase load arrangements. The negative sequence and zero sequence current are caused by the unbalanced voltage drop across the line impedance [6]. Three-phase voltage imbalance in the distribution network. In addition to these reasons, the interaction between the smart home system and the photovoltaic system makes the three-phase imbalance problem of the low-voltage power distribution system more and more frequent. At the same time, the microgrid contains a large number of nonlinear and unbalanced loads. The non-linear load will bring positive, negative, and zero-order harmonic interference to the converter. Unbalanced load will bring negative and zero-order harmonics to the converter. Interference. Additionally, in AC/DC hybrid transmission and distribution, the negative sequence current component will be generated when the grid fails, and the three-phase current imbalance will occur in the distribution system under the action of the negative sequence component [7].

C. Disturbance frequency change

In the context of power electronics, with the use of frequency conversion adjustment devices, thyristor rectification power supply devices, synchronous cascade speed devices, cyclo-converters, and other renewable energy sources such as wind power and photovoltaics, the frequency of the power frequency disturbance shifts, causing the current amplitude, phase, and waveform to change. When the integer harmonics are generated, the interharmonics with both continuous and discrete components are generated, subsynchronous harmonics appear. Question [4].

According to the Fourier series decomposition theory, periodic non-sinusoidal quantities can only decompose into integer harmonics. In fact, many non-linear loads are fluctuating. Electrified railways and industrial electric arc furnaces, such as thyristor rectified power supplies, are rapidly changing shock loads, and their electrical quantities vary in milliseconds or microseconds. In these few milliseconds or a few microseconds, the premise of "periodicity" does not exist for the power frequency. In this case, interharmonics will occur, and the results analyzed by Fourier techniques are not consistent or exactly in line with reality. Interharmonics refer to harmonic components whose frequency is an integer multiple of the non-fundamental frequency, also called non-harmonic, and the simple harmonic whose frequency is lower than the fundamental frequency becomes the subsynchronous harmonic.

Due to the problems of power quality and stability of power distribution systems caused by sub-synchronous harmonics, it has caused widespread concern in the power industry. The problem of subsynchronous oscillation caused by harmonics between subsynchronization is particularly prominent. A sub-synchronous harmonic of a large-scale wind farm in a certain area of Xinjiang, China, caused frequent fluctuations in the voltage of the ground. On July 1, 2015, the sub-synchronous harmonic component crossed 5 voltage levels to cause proximity. The unit's relay protection action eventually led to the unit's cutting machine accident. Subsynchronous oscillations in the power grid are mainly characterized by large amplitude continuous growth (divergence) or constant amplitude subsynchronous or ultrasynchronous current, voltage, and power harmonics randomly. The oscillation frequency will change affecting the safety of the power grid and equipment operation, even causing serious stability accidents or equipment damage. Because the subsynchronous harmonics have frequency time-varying characteristics, the traditional suppression method for a single fixed frequency is no longer applicable. It is necessary to study its time-varying and randomness as well as the propagation path and causes, and then develop control measures [8].

D. Ultra high harmonic problem

Since the third-generation semiconductors SiC and GaN have been widely used in recent years, the frequency of power electronic switching devices has evolved from several kilohertz to several tens of kilohertz or even hundreds of kilohertz. With the intelligent development of power systems and the application rate of power electronic devices, the ultra-high harmonics of 2 to 150 kHz have become a new problem in power quality of distribution networks[9]. In particular, the incorporation of renewable energy and the large-scale use of switching power supplies inject large amounts of harmonics into the distribution network, triggering new features of power quality in distribution networks.

Studies have shown that the transmission dispersion of ultra-high harmonics is different from ordinary harmonic emission, which is a new phenomenon of power quality. The main feature is the ability to have both primary and secondary emissions. The definition of primary emission refers to the emission of the detected electrical equipment in its own operation, which is related to its own topology structure, connection impedance, and other factors. The main feature is that the amplitude is very small [10]. The definition of secondary emission refers to the emission generated by the detected equipment after receiving the nearby equipment, which is related to the grid impedance and equipment impedance of the nearby excitation equipment, and its amplitude is several times that of the original transmission [11]. When the primary emission and the secondary emission are mutually osmotic, the emission source will be difficult to discriminate. At present, this research still can not simulate the actual emission at the laboratory level, which brings great research and standard development for the ultra-high harmonics. Large complexity [12].

Ultra high harmonics will bring many adverse effects to electrical equipment and communication in distribution networks [12], which are mainly divided into the following four categories:

• Interference with the normal use of electrical equipment in the distribution network, such as causing the electric vehicle charging station to not work properly, interfering with contact lighting.

• Causes equipment failure or damage, affecting the service life of electrical equipment.

• The noise at the equipment or installation point increases. It is shown by experiments that the noise decibel increases with the increase of the frequency and amplitude of the Ultra-high harmonics, which affects people's lives.

• Interference communication, due to the resonance effect, the end-user equipment forms a low impedance path, resulting in power line communication failures, such as data information transmission errors, causing measurement errors.

With the development of power grid electronic distribution, the hazards and impacts of ultra-high harmonics will become more and more serious. The relevant power departments need to conduct research and testing, and collect ultra-high harmonic data of power grids and equipment, and rationally formulate relevant standards, normative indicators, limits, measurements, and calculation methods.

ANALYSIS METHOD OF NEW CHARACTERISTICS OF POWERQUALITY IN DISTRIBUTION NETWORK

Although the power quality of the distribution network has presented many new features in recent years due to power electronics, in the context of power big data, data throughout the entire distribution network can be used for power quality analysis. The combination of power distribution quality and big data technology provides a reliable technical basis for the efficient operation of the power market and plays a certain role in supporting the safe and economic operation monitoring of the power grid. It will greatly promote the development of artificial intelligence technology in the development of the distribution network. Power quality monitoring data is a typical multi-dimensional massive dataset. Massive, multi-source, heterogeneous, low-value density power quality data makes traditional power quality analysis methods incapable, and the MapReduce parallel processing framework makes this problem feasible.

A. Introduction to MapReduce

With the rapid development of science and technology, multi-modal data collection methods have caused explosive growth in the volume of data in various industries. Traditional parallel computing based on HPC clusters has not been able to meet the actual needs of parallel computing analysis processing in a big data environment. In order to solve this problem, MapReduce came into being. MapReduce is a model and method proposed by Google for large-scale data parallel processing. It can perform tasks in parallel on Hadoop clusters composed of many common and inexpensive PCs [13]. "Split and combination" is the core idea of MapReduce. By decomposing the operations on large-scale data sets to different PC nodes, MapReduce performs the corresponding calculation and processing analysis tasks by each PC node, and finally obtains the final result by merging the intermediate results of each node.

The MapReduce framework is a parallel computing model, as shown in Fig. 1. MapReduce realizes large-scale data mining by utilizing the Hadoop platform and achieves parallel distributed computing by automatically segmenting the input massive data. In the MapReduce model, two abstract programming interface calls, Map and Reduce, are implemented by user programming. Each function uses a key-value pair (Key/Value) as input and output to realize massive data parallel computing tasks.

B. MapReduce-based distribution network power quality analysis method

Power quality analysis involves a large number of monitoring indicators and the past algorithms do not meet accuracy requirements. With the explosive growth of distribution network power quality monitoring data, powerquality analysis will be constrained by performance bottlenecks. Therefore, big data technology is urgently needed to solve the problems caused by huge power quality monitoring data. Faced with this problem, the HadoopMapReduce architecture performs relatively well in big dataanalytics, providing a convenient condition for big dataprocessing. Through MapReduce's core idea of "divide andconquer", the computing task is divided into multiple nodes for parallel computing, which speeds up processing efficiency. Therefore, the power quality monitoring data can be divided into multiple nodes for parallel statisticalanalysis.

According to the characteristics of power quality monitoring data of the distribution network, the power quality data of the distribution network, including voltage deviation, voltage fluctuation, three-phase frequency deviation, unbalance, harmonic content, etc., are constructed. According to MapReduce's idea of "divide and conquer," the power quality monitoring data of the distribution network is analyzed and processed in parallel. The basic idea is shown in Fig. 2.

Define monitoring data  sequence  sets Ti={Pij },Where Ti represents the ith monitoring point among, Pij indicates the jth monitoring data of the ith monitoring point, i is incremented in the order of monitoring points, i =1,2,3, ...n,n represents the number of monitoring points, j=1,2,3,...m,m represents the total number of monitoring points at each monitoring point,incremented by monitoring time. Pij = (Vij1,Vij1,...,Vij1),L indicates the number of monitoring indicators. For the new characteristics of power quality of distribution networks, it generally includes four indicators: rapid voltage fluctuation, disturbance frequency variation ultra-high frequency three phase unbalance and harmonic, that is L=4[14]. Decompose the monitoring data sequence Tj, all triplet monitoring point data fragmentation records are represented in triples <Nid,k,Tik>, where Nid is the monitoring point number, k is the serial number of the fragment, and Tki is the monitoring data sequence k of the monitoring point Tki= {Pij l j[bk, ek]}, where bk, and ek are calculated by the following formulas:

<Nid , k> is used to mark each piece of power quality data after fragmentation. Nid represents the power quality monitoring point number, and k represents each serial number. Map will use all power quality monitoring point data fragments as the input of the Map function. Power quality data is distributed to different nodes for data processing. The corresponding form of each Map function output <k,v> is

<(Nid,k,c),(a1,a2,a3,a4,a5)> ,where c ∈{1,2,3,4} represents four indicators such as voltage fluctuation deviation, frequency disturbance variation, three-phase unbalance, harmonic content, a1 is the number of unqualified, a2 is the average value, a3 is the number of overruns, a4 is the maximum value, and a5 is the minimum value [15]. After the calculation is completed, each parameter is compared with the national standards of power quality. According to the analysis and calculation, various indicators are obtained. If frequent three-phase imbalance occurs, the a3 indicator shows abnormality and finally outputs the result.

DISTRIBUTION NETWORK POWER QUALITY DEVELOPMENT DIRECTION

With the future development of power grid electronic power supply and source-network-load storage technology, as well as non-electrical factors such as climate, marketization, and public awareness, the power quality problem of the distribution network is more and more affected. In the future, the power quality problem of the distribution network becomes a complex problem of time-space correlation characteristics, which needs to be solved by global optimization configuration and reliable monitoring. The problem of three-phase unbalance and equipment utilization reduction caused by single-phase load, especially after the single-phase electric vehicle charging equipment is connected to the distribution network, the problem will be further highlighted: the distributed electronic energy access and isolated network operation. The resulting short circuit capacity reduction problem will cause harmonic distortion to become more prominent, new load and component access will lead to new power quality problems, such as the sensitivity of LED lighting fixtures to flicker. Grid-side access to different filter circuits increases the capacitive characteristics of the power grid and increases the risk of grid resonance. The simplification of the power electronic circuit during the power quality analysis problem leads to some power quality problems such as the impact of LED lighting. Current problems include harmonic characteristics of new equipment and harmonic superposition problems, such as the influence of different numbers of electric vehicle charging on the harmonic characteristics of the distribution network; the accuracy of the sensor for voltage measurement, especially for the measurement of ultra-high harmonics. If the frequency response characteristics cannot meet the requirements, the measurement results are not credible. Analysis of power quality monitoring data visual display problems, such as a large database of the monitoring, and how to effectively polymerize.

CONCLUSION

The power electronicization of the distribution network is the premise of an intelligent distribution network. Based on the power electronicization of the distribution network, this paper discusses the sub-synchronous harmonics and ultra-high frequency caused by rapid voltage fluctuation, frequent three-phase unbalance, and disturbance frequency variation. New features of power quality in distribution networks, such as harmonics, are designed by MapReduce, and the parallel analysis algorithm is provided. Finally, the development direction of power quality in the distribution network is forecasted.

REFERENCES

[1] Zhao Min, Li Shunwei, Yue Yunli, etal. Performance and InfluenceAnalysis of Power Electronics in Distribution Network[1. ChinaHigh-tech, 2017(1): 149-150.

[2] Math H 」 Bollen, Sarah K Ronnberg Primary and secondarynarmonicsemiss1on:harmonicinteractionasetofdefinitions[C]//International Conference on Harmonics & Quality ofPower, Belo Horizonte, Brazil.2016:703-708.

[3] YUAN Xiaoming, CHENG Shijie, HU Jiabing. Dynamic StabilityProblem of Multi-Scale Voltage Power Angle in Power ElectronicPower System[1.Proceedings of the CSEE, 2016, 36(19): 5145-5154

[4] Dong Xiaoliang, Tian Xu, Zhang Yong, etal. Analysis of typical events and influencing factors of subsynchronous resonance in series compensated transmission system of Wuyuan Wind Farm[J]. HighVoltage Engineering 2017, 43(1): 321-328.

[5] Guo Shanghua, Huang Chun, Wang Lei, etal. Detection and ControlMethod of Voltage Fluctuation and Flicker[1]. Relay, 2004, 32(3): 45-48.

[6] Qi Hao, Wang Xiangsheng, Gao Shen, etal. Research and Application of Power Quality Integrated Management System []. ElectronicDesign Engineering, 2017, 25(5): 170-174.

[7] Qi Hao, Wang Xiangsheng, Gao Shen, etal. Research and Application of Power Quality Integrated Management System [. ElectronicDesign Engineering, 2017,25(5):170-174.

[8] Aurora Gil-de-Castro, Sarah K Rönnberg, Math HI Bollen. A study about harmonic interaction between devices[C]//2014 EEE 16thInternational Conference on Harmonics and Quality of Power(CHOP),Bucharest, Romania, 2014: 728-732

[9] Narendrak,FedirchukD,MidenceR,etal.Newmicroprocessorbasedrelay to monitor and protect powersystems against sub-harmonics[C]//2011 lEEE Electrical Power and Energy Conference(EPEC),Winnipeg, MB, Canada, 2011:438-443.

[10] Lin Haixue. New Development of Harmonic Problems in Power GridOn Superharmonics[1 .Supply and Electricity, 2016(1): 35-3 8.

[11] A Larsson. An overview of the origin and propagation ofsupraharmonics(2-150kHz)[J]. Brazilian Journal of Physics, 2014, 32(2b): 483-494.

[12] Ana-Maria Blanco, Jan Meyer, Sarah RÃnnberg, Math Bollen. Surveyof supraharmonic emission of household appliances[C]//InternationalConference on Electricity Distribution, Glasgow, Scotland, 2017:1 -5.

[13] Zhang Bin.Study on several key technologies based on MapReducebig data parallel processing[D]. Donghua University, 2017.

[14] Ana-Maria Blanco, Jan Meyer, Sarah Ronnberg, Math Bollen. Survey of supraharmonic emission of household appliances[C]//Internationa!Conference on Electricity Distribution, Glasgow, Scotland, 2017: 1-5.

[15]Song Jie,Sun Zongzhe,Mao Keming etal.Research Progress ofMapReduce Big Data Processing Platform and Algorithm[1.Journalof Software,2017.28(3):514-543.

[16] Singh U, Singh S N. Application of fractional Fourier transform forclassification of powerquality disturbances[. IET Science,Measurement&Technology, 2017,11(1 ):67-76.