Mathematical Problems in Engineering
Volume 2022 (2022), Article ID 6389132, 43 pages
https://doi.org/10.1155/2022/6389132
Review Article

A Review of Control Techniques and Energy Storage for Inverter-Based Dynamic Voltage Restorer in Grid-Integrated Renewable Sources

Devalraju Prasad and C. Dhanamjayulu

School of Electrical Engineering, Vellore Institute of Technology, Vellore, Tamil Nadu, India

Correspondence should be addressed to C. Dhanamjayulu

Received 19 August 2022; Accepted 12 September 2022; Published 29 September 2022

Academic Editor: Albert Alexander Stonier

Copyright © 2022 Devalraju Prasad and C. Dhanamjayulu. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Power quality (PQ) is a key issue, particularly for technologically advanced process equipment, whose performance mainly depends on the quality of supply. Problems with PQ such as voltage swells/sags, interruptions, and harmonics are defined by any voltage, current, or frequency abnormalities causing damage or failure of the end-user equipment. Outages and interruptions lead to malfunctioning of end-user equipment or sensitive loads, such as diagnostic equipment in healthcare facilities, clinics, educational institutions, and detention centers, while further increasing significant economic losses. Custom power devices (CPDs) are recommended for enhancing power quality, and the best and most economical solution is considered to be the dynamic voltage restorer (DVR). Several methods are suggested to improve the PQ by using the dynamic voltage restorer; among them, most encouraging ways are to use a multilevel inverter (MLI) in the dynamic voltage restorer. This article combines the latest work of the literature, as well as a detailed discussion on PQ issues of the grid-integrated renewable energy sources (RESs), DVR principle with its operating procedures, system components, energy storage-based DVR topologies, DVR control unit, and DVR power converter-based topologies. In addition, synthesis of energy storage, control strategies, and multilevel inverters for DVR. This review benefits those interested in investigating DVR as a relevant and comprehensive reference.

1. Introduction

In the smart era, microprocessor-controlled devices or digital, electronic, and nonlinear devices are extensively used in all sectors of the industry. Nearly all these devices are sensitive, have electrical supply disruptions at any minute, and cannot be operated properly. In addition, several supplies have also been increased, which degrades power quality (PQ). Problems that happen because of inadequate power quality are data errors, automatic resets, memory losses, UPS alarms, equipment failures, software corruptions, circuit board failures, power supply problems, and overheating of electrical distribution systems. Considering these realities, PQ has become progressively critical. Not only PQ issues but also the issues related to voltage are also most important from sensitive nonlinear loads and end-users [1, 2].

The use of sensitive loads such as diagnostic equipment in health centers, educational institutions, and detention centers over several years has been fourfold, which leads to a concern with the quality of power of sensitive loads [3]. If power quality is insufficient, serious economic losses, losses in manufacturing, outage of sensitive and critical loads, and lack of information could have serious consequences [4]. Consequently, high power quality is essential for utilities, customers, and producers of electrical appliances too. The essential power quality issues include voltage swells, sags, harmonics, transients, flickers, fluctuations, and interruptions [5]. These are discussed in the next section. The sensitive and critical loads must prevent these issues in terms of power quality and voltage disturbances. In this regard, a wide range of solutions has been introduced including the best and most efficient solution for the compensation and mitigation of voltage disturbance known as custom power devices (CPDs) [6]. They act as compensating devices, each with its own control and application. CPDs such as a parallel-connected distribution static synchronous compensator (DSTATCOM), are used for correcting the power factor; for voltage compensation, the dynamic voltage restorer (DVR) is used and is connected in series; a parallel-series connected unified power quality conditioner (UPQC) can simultaneously inject voltage in series and current in parallel; however, UPQC and DSTATCOM are larger and more expensive, rather than DVR [7]. In modern power systems, the most serious and usual power quality issues are voltage sags, and DVR is used as the least expensive voltage sag solution [8].

When a voltage disturbance occurs on the supply side, the DVR supplies the required voltage to the load side. The DVR also protects from supply-side disturbances to sensitive and critical loads [9, 10]. This means that the DVR is important to compensate for voltage sags and to protect the sensitive load. The DVR is the best CPD since it has low costs, has small sizes, and can respond quickly to voltage disturbances. As an example, the DVR installation cost for the 2–10 MVA power supply is USD 300/kVA, while uninterruptible power supplies (UPSs) installation costs are USD 500/kVA. The servicing and operating costs of DVR are approximately 5% of its capital investment; however, it is much higher (about 15%) [1]. UPQC is a DSTATCOM-DVR combination with two power converters; hence, the structure of the DVR is, therefore, less than UPQC. DVR and DSTATCOM are closely related; however, DVR is used to protect the sensitive loads from supply interruptions, whereas DSTATCOM is used to protect critical loads from load-side disturbances. Furthermore, the DVR quickly (less than 1/4 cycle) responds to voltage disturbances, unlike other CPDs, such as the static VAR compensator (SVC) (2-3 cycle) [11].

Many topologies of DVR from different points of view of energy storages, power converters, and control systems have been investigated to improve power quality, cut costs, and improve the performance of DVR [12]. Furthermore, it has become widely attractive to modify DVR topology and integrate renewable energy resources with the DVR. Some general reviews on DVRs were carried out that a detailed study is lacking on modified DVR configurations and integration with renewable energy [13, 14]. Significant research is being conducted on DVR innovation and is now advanced but not many of the survey papers in the published literature are accessible. Remya et al. reviewed the DVR and reported on the challenges of the DVR systems [15]. Farhadi-Kangarlu et al. reported the combined overall DVR topologies, compensation techniques, and control strategies [16]. Significant research development in DVR technology for fifteen years after the first installation of DVR was published by El-Gammal et al. [17]. This paper intends to give a comprehensive evaluation of different components of DVR structure, as well as the integration of distributed generations into multi-inverter-based DVR. This article provides a significant contribution in the following ways:(i)Discussion of the power quality difficulties associated with RES integration into the grid(ii)To understand the working principles of DVR, the basic components, alternative DVR topologies from an energy storage approach, DVR control units, and DVR compensation techniques are provided(iii)Discussion and extensive study of different grid-connected multilevel inverters, as well as multilevel inverter-based DVR integration with distributed generation, for optimizing voltage profile

Engineers and researchers working on the issues of power quality and the mitigation of voltage disturbances will be able to use them on an extensive basis.

The remainder of the article is organized accordingly. Section 2 describes the power quality problems of RES connected to the grid, grid-integrated RES requirements, PQ standards, classifications, causes, and effects. The most important custom power device and applications have also been clarified. In Section 3, the principle of DVR and its various operating modes, DVR circuit components, and DVR topologies are presented from the point of view of energy storage. Following this, Section 4 reviews DVR control units and compensating techniques. The analysis of various grid-tied multilevel inverters with their advantages and disadvantages along with their performance indices and, finally, an elaborated review on multilevel inverter-based DVRs are provided in Section 5; the conclusions and scope of future work are given in Section 6.

2. Power Quality Issues in Grid-Connected Renewable Energy Sources

There is considerable global attention to the utilization of renewable energy sources (RESs) for electricity generation. This is because of the negative environmental effects of fossil fuels being burned to convert energy, which emits an enormous amount of CO2 and other greenhouse gases into the air. Figure 1 depicts a few of the renewable energy sources.

Figure 1: Major renewable energy sources.

Renewable energy use rose by 3 percent as the demand for all other fuels declined, according to the IEA, Paris World Energy Review 2021 [18]. The main driver was a 7 percent increase in renewable energy generation. Renewable energy accounts for a 29% share of global electricity production in 2020, up from 27% in 2019. In 2021, the production of renewable energy will be more than 8% expanding to 8,300 TWh, the fastest growth year-on-year since the 1970s. Two-thirds of the renewable energy growth will be supported by solar PV and wind. The growth of renewable energy alone in China in 2021 was almost half expected, followed by the USA, the EU, and India shown in Figure 2. The wind is set to grow by 275 TWh (almost 17 percent), the greatest increase in renewable energy production, which is significantly higher than in 2020. China will remain the biggest market for PV, and there is expansion in the United States with ongoing federal and state policy support. The new solar PV capacity additions will recover rapidly from COVID-19-related delays in India in 2021, with strong policy support in Brazil and Vietnam driven by large political support for distributed solar PV applications. The new solar PV capacities increased in Brazil and Vietnam. The total electricity generated by solar photovoltaic systems in 2021 is projected to grow to 145 TWh or almost 18 percent. Renewable sources in the electricity generation mix are expected to increase their proportion by 30 percent in 2021, all the time from all renewable sources. In combination with nuclear, carbon-free sources, the world coal plants’ output in 2021 is much higher than that.

Figure 2: Worldwide increased technology and renewable power generation in 2020-2021.

According to the IEA’s 2021 Renewable Energy Market Update [19], renewable energy is the only power source for which consumption has risen despite the pandemic by 2020, while other fuel consumption has decreased. In light of current commercial and political changes, the renewable energy market investigates new global renewable power add-ons for 2021 and 2022. It also establishes a modern biofuel production prediction for these years, as the industry experienced significant losses due to a fall in transportation demand during the pandemic. It is expected to maintain the exceptional level of renewable energy add-ons and that 270 GW will become operational in 2021 and 280 GW in 2022. This expansion is more than 50% higher than the annual capacity rise record set in 2017–2019, implying that renewables will account for 90% of global capacity growth in both 2021 and 2022, as indicated in Figure 3.

Figure 3: Net renewable capacity additions, by renewable energy market update 2021, IEA.
2.1. Requirements for Grid-Integrated RES

Although some renewable energy sources are linked to the transmission system, most of them are linked to the distribution system. The operators of both distribution (DSOs) and transmission systems (TSOs) have been facing significantly higher levels of penetration of renewable sources, especially PV and wind systems, and many other strategies to replace conventional power plants with RES have been launched [20]. These changes have forced electrical system operators to take into account the impact of that penetration on grid stability. TSOs and DSOs have introduced new regulations at PCC, a common point of connection between the power grid and RES [21, 22]. RESs have the potential to handle various disturbances, increase the stability of voltage and frequencies and reliability, and improve power quality and security of power grids; they are required to act as traditional power plants [23]. The new requirements include standards of PCC power quality [24], voltage regulation [25], frequency regulation [26, 27], voltage stability support through reactive power injection [28], frequency stability support through active power control [26, 27], and voltage ride through (VRT) [29–34]. VRT is further subcategorized into (i) LVRT (low-voltage ride through) [30], (ii) ZVRT (zero-voltage ride through) [33], and (iii) HVRT (high-voltage ride through) [34]. In [30], the energy storage system was controlled by LVRT, and reactive power supports the grid; the limitations are overshooting, high fluctuations, superimposition, additional investment costs, periodic inspection, and maintenance, as specified in the grid code, whereas in [32], STATCOM and SVC control the LVRT; it injects reactive currents and improves the VRT capacity; however, there are no data about the change of DC-link voltage during grid faults; also, the complexity and costs increase. Voltage ride-through standards in different countries are tabulated in Table 1.

Table 1: Voltage ride-through standards in different countries.
2.2. Power Quality

Power quality for electricity suppliers and their customers has become an important concern. From the point of view of customers, the consequences of disturbances in terms of financial loss can be between hundreds to millions of dollars. Power quality issues lead to losses of consumer satisfaction and also losses of load or income from the point of view of utilities. Quick incidents such as voltage transients, swells/sags, voltage impulses, high-frequency noise, and faults are power quality issues; generally, these are identified as any deviation from the standard voltage source. Hence, issues of power quality affect electrical equipment directly [35–37]. The percentage sharing of major power quality issues is shown in Figure 4.

Figure 4: Percentage quantity of major PQ problems.

Disturbances that could lead to problems with power quality may be an operation of unbalanced and nonlinear loads, start or switch off huge loads such as motors, energization of transformers and capacitor banks, or failure of devices such as transformers and wires, lightning, and natural events. The two main standards for power quality issues are the IEC and the IEEE. Table 2 contains the latest revisions to these standards [38–48], and the time required for RES to clear the abnormalities when exposed to abnormal voltage and frequency [49] is listed in Table 3.

Table 2: Power quality standard and their recommendations.
Table 3: Response to abnormal grid voltage and frequency.

Table 4 lists the most important issues of power quality and their definitions, and Table 5 provides reasons and effects of problems of power quality together with their duration and magnitude. IEC 61000-3-2 (1995-03), IEEE-519, and IEC/TS 61000-3-4 (1998-10) establish guidelines for limiting harmonic problems. IEEE P1564 and P1547a address the voltage sag issues. The first deals with the impacts of voltage sag, while the second deals with ways to stabilize a system through voltage sag mitigation. The flicker problem characterizes IEC 61000-4-15. IEEE 1159–1995 characterizes general problems with power quality. IEEE 1250–1995 and IEEE P1409 discuss the impacts and corresponding solutions on power quality issues. IEEE Standard P1547 discusses microgrid characteristics and their interconnections with the power system.

Table 4: Description of power quality issues.
Table 5: Characteristics, causes, and effects of power quality issues.
2.2.1. Power Quality Improvement (PQI) Techniques

Various PQI research aimed at minimizing problems with PQs are reported, and the ideal grid integration of renewable energy sources is promoted; however, every mitigation technique creates certain difficulties; hence, it will continue to be an active research sector in the future too. Current quality improvement (CQI) and voltage quality improvement (VQI) strategies are two types of PQI techniques that emphasize renewable energy integration, as shown in Figure 5. The VQI technology addresses voltage variations and frequency mitigation in DGs. Further subclassifications of VQI techniques may include energy storage (ES) [50], custom power devices (CPDs) [51], optimization of energy conversion [52, 53], spinning reserve [54], and a few additional unique technologies based on the variable frequency transformer (VFT) [55] and the virtual sync machine [56]. Further CQI technologies can be divided into passive filters (PFs) [57], active power filters (shunt and series) [58], smart impedance [59], hybrid filters [60], and multifunctional DGs [61].

Figure 5: Power quality improving techniques for integration of renewable energy.
2.2.2. Custom Power Devices (CPDs)

Critical equipment protection against voltage sags and interruption is supplied with storage units. Examples of storage systems include flywheel energy storage system (FESS), superconducting magnetic energy storage (SMES), uninterruptible power supply (UPS), ultracapacitors (UCAPs), and batteries. They are used to mitigate the required energy due to faults and voltage drops. The most efficient method of mitigating voltage slags is by using custom power devices (CPDs); they ensure that the quality and reliability of supply are guaranteed to customers [62, 63]. Table 6 contains the most important CPDs to mitigate the issues related to power quality.

Table 6: Applications of the CPDs.

3. Dynamic Voltage Restorer

DVR is a compensating device having a series impedance that is serially connected between the PCC (point of common coupling) and the load as shown in Figure 6. It supplies the required voltage during sag to synchronize the load voltage and permits the switching of real and reactive power between DVR and distribution systems [64]. Westinghouse manufacturing company (part of Siemens) introduced the first DVR in August 1996, with the support of the Electric Power Research Institute (EPRI), with an installed capacity of 2 MVA at Duke Power Company located in North Carolina (USA) on 12.47 kV distribution systems having three voltage source inverters and three injection transformers with the in-phase compensation technique [10, 15].

Figure 6: Basic structure of the dynamic voltage restorer.

The equivalent circuit of DVR is obtained by connecting a voltage source VComp in between the source (VS) and load (VL) with their respective impedances ZS and ZL as shown in Figure 7. At PCC source, the current IS is divided into IL and IoT., where IL is the sensitive load current and IOT is another load current. The voltage at PCC is represented by VG, and the voltage compensated by DVR is VDVR. Resistance R and inductance L are obtained from the impedance Z of the filter and injection transformer, and the values of RDVR and XDVR are related to VDVR. The impedance of the source, load, and DVR are ZS, ZL, and ZDVR, respectively.

Figure 7: Equivalent model of the dynamic voltage restorer.

P S is the real power, and QS is the reactive power of supply.

P L is the real power, and QL is the reactive power of the load.

P DVR is the real power, and QDVR is the reactive power supplied by the DVR.

The voltage across sensitive load VL is given by

The percentage of voltage handling capacity and current handling capacity iDVR of DVR is given by [65]

I DVR is the current rating of DVR, VS, rated is the rated supply voltage, and IL,rated is the rated load current of DVR [66].

DVR mainly consists of a bypass switch, injection transformer, filter, inverter, and DC-link capacitor or energy storage as shown in Figure 8. DVR can be categorized into 1 − ϕ DVR shown in Figure 8 and 3 − ϕ DVR shown in Figure 9. A bypass switch is used to connect the DVR to the line during the injection mode and disconnect the DVR under standby operation. The injection transformer will adjust the DVR output voltage and isolate DC to AC systems. High-frequency harmonics present in DVR output voltage to attenuate this LC filter are used. The voltage is maintained as constant during transient by using a DC-link capacitor; hence, a steady-state operating range of DVR will improve. For deep voltage sags, the external energy storage supplies the desired real power to the load. The most essential part of DVR is the inverter; it produces the required controllable voltage for compensation.

Figure 8: Schematic representation of 1 − ϕ DVR.
Figure 9: Schematic representation of 3 − ϕ DVR.

The dynamic voltage restorer works in three operating modes. They are (i) voltage compensating or active or injecting mode, (ii) standby mode, and (iii) fault current limiting or protecting mode. When the voltage disturbance occurs within the operating range of the DVR. at that instant. the bypass switch is open and DVR switches to active mode and feeds the grid with the required voltage. Once the voltage is in its normal range, the bypass switch will close and end the compensation mode. It is not necessary to inject voltage into the grid under normal conditions; hence, the DVR is left out through a bypass switch to reduce the power loss in the DVR by restricting the inverse effect on line voltage; this mode is known as the standby mode. Sometimes, high current flows through DVR due to SC fault on the load side which will cause damage to the injection transformer and DVR components; hence, the identification of downstream fault is necessary to protect the DVR [67]. The protection scheme of DVR is presented in Figure 10. It provides an alternative path to the fault current through breakers, thyristors, and varistors and ensures that the current path should be present. If the current path is not present, a severe overvoltage appears at the terminals of the injection transformer [66, 68].

Figure 10: Protection scheme of the dynamic voltage restorer.

By lowering the fault current, the DVR disables both main and backup protection. As a result, the duration of the defect may be extended. As a result, using a DVR to reduce the fault current to zero, interrupt it, and send a trip signal to the upstream relay or circuit breaker is preferred (CB). The ability to inject 100% of the voltage is required for the FCI function. As a result, the series transformer and VSC power ratings would be almost three times higher than those of a standard DVR with 30%–40% voltage injection capabilities. As a result, the DVR will cost extra. The economic feasibility of such a DVR is determined by the cost of the DVR, and the importance of the sensitive load is protected by the DVR.

3.1. DVR Topologies

The DVR can be categorized into various groups, related to the configuration of energy storage, location of the filter, presence of the injection transformer, and structure of the converter. Figure 11 shows various categories of system topologies.

Figure 11: Classification of DVR system topologies.
3.1.1. Location of the Filter

Usually, the high-frequency or carrier frequency switching technique is used in DVRs. Hence, the output voltage is embodied with higher-order harmonics; thus, it is required to attenuate the harmonics, and for that, generally a low-pass LC filter is used [69–73]. As indicated in Figure 12, the filter is connected to the inverter side [72] or the supply side [73]. The converter side filter has the advantage that it does not allow the higher harmonic currents into the series transformer; hence, the rating of a transformer will decrease; also, the stresses on the transformer are reduced. However, the presence of a passive element will produce an extra drop over the transformer. A capacitor is connected to the series transformer in the case of a grid-side filter scheme. It will solve the issues related to the presence of the inductor in the inverter side filter. However, this topology has poor efficiency as compared to the previous one for harmonic reduction.

Figure 12: Different locations of the filter in DVR.
3.1.2. Presence of an Injection Transformer

In DVR, the inverter generates the required compensation voltage and is injected through the injection transformer. Based on injection transformers, there are two topologies: injection transformer-based DVR topologies and transformerless DVR topologies.

(1) Transformerless DVR Topologies. In some DVR topologies, the injection transformer is removed to overcome the issues associated with core saturation and inrush currents present in the transformer; such topologies are named transformerless topologies [74] shown in Figure 13. The benefits of transformerless DVR topologies are less weight and a significant reduction in volume and price [75, 76]. However, the converter must allow the full voltage; hence, for better efficiency of higher voltage applications, multilevel inverters are used. A transformerless DVR presented in [77] consists of a buck-boost converter with five bidirectional switches. It has the advantages of compact structure and lightweight. However, due to bidirectional switches, this converter experiences a commutation problem and high cost. To overcome the tough commutation problem, Zhou et al. [78] proposed a DVR with an indirect AC/AC converter heaving six unidirectional switches.

Figure 13: Transformerless DVR [15].

(2) Transformer-Based DVR Topologies. For isolation, an injection transformer is required; in the case of DC-link in DVR, it is activated from the supply or load side via a rectifier; however, if energy is received from battery storage in DVR, the injection transformer is not required because the battery works as a floating DC source [79]. The major issues associated with injection transformers are saturation and protection of the transformer. The solutions to the saturation problems in the injection transformer have been addressed in [80–83]. Figure 14 represents transformer-based DVR.

Figure 14: The transformer connected to DVR [15].
3.1.3. Energy Storage

In DVR, energy storage means external energy devices (not for DC-link capacitors) are used to inject real power into the grid. Depending on energy storage, there are two DVR topologies: (i) without energy storage topologies and (ii) with energy storage topologies.

(1) Without Energy Storage. By connecting a series converter, a shunt converter (mostly rectifier), or an AC-AC converter to the grid, the required compensating energy is directly received in this method. The converter is added to either the grid side or load side which are presented in Figures 15 and 16 , respectively. Preliminary research was carried out on DVR without DC-link before 1996 [84]. A few developments in this topology were reported in [85–87] after the development of bidirectional switch escalating the progress of AC-AC converter-based DVR [88]. According to the nature of voltage sags, a control scheme is proposed to compensate for the sag with phase jump in [89]. In these topologies, a considerable amount of supply voltage is present at the time of sag, and this residual voltage is utilized to boost and protest the full load at rated voltage. However, during fault conditions, it draws more current from the line; hence, upstream loads have higher voltage drops.

Figure 15: DVR connected at the grid side.
Figure 16: DVR connected at the load side.

(2) With Energy Storage. The required compensation energy is drawn from a variable DC-link (capacitor supported or self-charging) shown in Figure 17 or from a constant DC-link (external energy storage supported) shown in Figure 18. In capacitor-supported topology, the discharging energy from the DC-link capacitor is injected into the grid during the active mode. In the standby mode, the DC-link capacitor is charged and stores the required energy. It is more economical since there is no need for external storage. However, the stored energy in the DC-link capacitor is limited; hence, it is inadequate to compensate the deep sags for high pf loads in a lengthy period. Constant DC-link topology requires direct energy storage devices, such as SMES, supercapacitors, and batteries, and also an extra high-rated energy converter is connected to transfer the large stored energy to a low-rated DC-link storage to maintain a constant voltage during sag, but the size and capital cost of the DVR get increased.

Figure 17: DVR with variable DC-link.
Figure 18: DVR with constant DC-link.

The energy storage of intermittent renewable sources is an extensive area of research since energy storage is utilized in several applications in the grid, including energy shifting, electricity supply capability, supporting frequencies and voltages, and the management of electricity bills [90–92]. Energy storage technologies are pumped hydroelectric (PHS) [93], compressed air energy storage (CAES) [94], flywheel energy storage (FES) [95], battery energy storage (BES) [96], thermal storage [97], and hydrogen [98]. Table 7 provides a technical and economic summary of key energy storage technologies. According to the International Hydroelectricity Association (IHA), PHS is the most advanced and well-established storage technology in the world, with a total installed capacity of 153 GW in 2018 and is expected to increase by almost 50 percent to about 240 GW by 2030 [112].

Table 7: Technical parameters of different energy storage technologies.

A PHS system can support seasonal management, time-shifting, peak lopping, back-up, filling of the valley, and energy management. It is, however, geographically constrained, capital-intensive, and frequently delayed owing to environmental permission. Technology in CAES can provide services of flexible power quality, but it also suffers from the same problem as PHS such as geographically limited. During the industrial revolution, flywheel technology was quickly developed as the mechanism for smoothing steam engines [95]. Energy storage batteries are an electrochemical storage system that delivers quality services in power and were recently used to supply variable renewable storage systems such as solar PV and wind, in part driven by reductions in battery costs. Several technologies for batteries are covered in [113] from advanced lead acids widely used in automotive and heavy goods vehicles to the development of flow battery technologies that could be used in the power storage market. However, the most popular technology for portable and power storage systems has emerged from lithium-ion battery systems. Lithium-ion batteries have a low self-discharge rate, have a high energy density, are flexible and lightweight, and require little maintenance when compared to other battery technologies, making them the most popular. However, depending on the ambient conditions, lithium-ion systems necessitate temperature monitoring, and certain installations have cooling systems. The worldwide demand for batteries is increasing, with the use of batteries for portable and energy storage devices.

The benefits of various energy storage (ES) technologies such as energy density, cycle lifetime, and specific power can be combined with those of hybrid energy storage systems (HESS). Furthermore, combined technologies can differ in their electricity characteristics significantly; for example, cut-off voltage prevents overload. A PE interface is thus important in controlling the both terminal voltage and power flow of individual devices in the HESS between various ES devices. Several configurations have been proposed and discussed that connect each HESS to ES device to the common DC-link [114–119].

HESS integration can be implemented by utilizing cascaded modular converters and existing PE devices. For example, the typical HESS composed of batteries and supercapacitor can be integrated into the MMC by connecting the DC-link to the battery, and connecting the submodules to the supercapacitors or connecting the batteries and supercapacitors to various submodules. Advanced control algorithms can control the active power supplied by DC-links/batteries and submodules/supercapacitors in the AC power [120]. Table 8 lists major electronic power suppliers to the commissioned BESS utilities and their solutions.

Table 8: Major power electronic unit providers for commissioned utility BESS [121].

A capacitor-supported DVR was proposed by Nielsen and Blaabjerg [122]; it gives comparatively less performance for deep sags. Narrow sags, that persist not more than one minute, are compensated by using an ultra-capacitor-based DVR [123]. Wang and Venkataramanan [88] reported that flywheel is an effective technology for short-term energy storage. Kim et al. [124] proved that SMES is capable to mitigate the narrow voltage sags for MJ/750 kVA DVR. Shi et al. [125] presented that SMES is effectively mitigating the sags persistence of 100 ms. Superconducting magnetic energy storage (SMES)-based DVRs are analyzed for voltage sag compensation and technically validated in [126–128]. However, as compared to DVR supported with regular battery energy storage (BES), the SMES-based DVR systems are uneconomical because of the high capital cost of SMES coils [129]. To overcome this, hybrid energy storage (HES) was introduced by Shim et al. [130]; in that, they reported SMES/HES is suitable for getting smooth output from renewable energy sources. Gee et al. [131] presented that SMES or battery energy storage systems based on DVR are well suited for three-phase loads. MW-class SMES integrated with SFCL-based DVR for voltage sag compensation is proposed and demonstrated in [132].The potentials and constraints of various DVR topologies are listed in Table 9.

Table 9: Potentials and constraints of various DVR topologies.

4. Control Systems in DVR

The control system plays a vital role in the DVR, and it goes through several stages; these include (a) detecting voltage disturbance, (b) generating the reference voltage, and (c) controlling the converter. The systematic procedure of the control system and different types of techniques used in the detecting voltage disturbance unit, reference generation unit, voltage and current controllers, and PWM pulse generator are presented in Figure 19.

Figure 19: Different techniques at various stages in the control strategy of DVR.

The efficiency of the control algorithm is completely dependent on the accuracy and quality of the techniques used in voltage detection. Input data received by the detecting unit is the measured voltage on the supply or load or both sides. The operating mode of DVR is decided by this stage whether DVR operates as standby mode or active mode or protection mode, by using traditional or advanced detection techniques. In traditional voltage peak detection [133], RMS detection [134] are used, whereas in advanced detection, wavelet [135], Kalman filtering [136], discrete Fourier transform(DFT) [74], fast Fourier transform (FFT) [136], numerical matrix [137], missing voltage [135], rotating DQ reference frame [90], and synchronously rotating frame (SRF) [89] are used. If there is no voltage disturbance, then the bypass switch is turned ON and DVR acts as the standby mode. If any load side fault is detected, then DVR comes into protected mode. Once it detects voltage disturbance, then the DVR comes into active mode and generates the magnitude and phase angle of the reference voltage in the reference generation unit by using a suitable compensation technique (explained in a further subsection) and is injected through the converter. This information is computed by using different phase-locked loops (PLLs).

Figure 20 depicts a rotating DQ reference frame controller. The error signal and change in error signal drive the PI controller, which analyses the input and generates controller output. The PWM receives the controller output as a reference voltage. The inverter is controlled by the pulses generated by the PWM pulse generator. The magnitude and phase angle of the reference voltage are generated using a correction approach and are fed into the multilayer inverter. This data are derived using phase-locked loops (PLLs).

Figure 20: Rotating DQ reference frame controller.

Many papers [138] have recently introduced a variety of PLL schemes such as the synchronous reference frame PLL (SRF-PLL) [139, 140], the dual second-order generalized integrator PLL (DSOGI-PLL) [141], the double synchronous reference frame PLL (DSRF-PLL) [142], and the enhanced PLL (EPLL) [143]. Table 10 illustrates the comparison between the strong and weak sides of the selected PLL algorithms.

Table 10: PLL synchronization schemes and performance comparisons.

For reference generation, the coordinate transformation (Park and Clark) method [144], the symmetric component estimation method [145], and the instantaneous power theory (PQR) method [146] are used. The converter is controlled either by an open-loop (feed-forward) controller or a closed-loop (feedback) controller [147]. In open-loop control, the generated voltage reference is straightly given as a reference to the PWM. However, in the closed loop, the generated reference voltage is fed to the controllers such as passive-based controller [148], two degrees of freedom (2DOF) [149], sliding mode [150], H-infinity [151], repetitive [152], neural [153], fuzzy logic [154], multiloop [88], hysteresis voltage [155], state feedback [156], predictive [157], and proportional-resonant (PR) [133]. The controller output is given as a reference voltage to the PWM. The operation of the controller is controlled by the pulses generated by the PWM pulse generator. Sinusoidal PWM [158], space vector PWM (SVPWM) [159], multilevel fundamental frequency [160], selective harmonic elimination (SHEPWM) [161], hysteresis modulation [162], multicarrier PWM [163], phase-shifted PWM [164], level-shifted PWM [165], reduced carrier PWM [166], subharmonics PWM [167], and discontinuous multilevel space vector modulation [168] are the various modulation techniques used to generate the pulses. Table 11 represents the comparison of different linear controllers, and Table 12 represents the comparison of different control strategies used in grid-connected three-phase four-leg inverters.

Table 11: Comparison of various linear controllers.
Table 12: Comparison of different control strategies.
4.1. Compensation Techniques Used in DVR

The required DVR output voltage is achieved by using a suitable compensating technique. The selection of compensation strategy depends on the reference produced by the reference generation unit because it influences the phase-locked loop (PLL), which leads to a key task in synchronization of grid voltage. Hence, by using the proper compensation technique, the PLL output is controlled. Different compensation strategies are presented in Figure 21.

Figure 21: Classification of DVR compensation techniques.
4.1.1. Presag Compensation (PSC)

In this strategy load voltage is maintained with the presag voltage; therefore, no voltage disturbance is sensed by the load because the load voltage is having the same magnitude and phase angle [66]; hence, it is also known as the voltage quality optimized technique. The vector representation of PSC is shown in Figure 22. During sag, the DVR is controlled by adding more real power which affects the rating of direct energy storage or energy received from the grid; hence, the requirement of energy source to supply active power will increase apart from reactive power injected by the inverter. It is acceptable for both balanced and unbalanced sensitive loads heaving phase jump or not. (3) gives the magnitude of VDVR, and the phase angle of VDVR is obtained from equation (4).where VDVR is the DVR injected voltage, ɸ is the phase angle between VL and IL, is the grid voltage at sag, δ is the corresponding angle of phase jump to , and p is the corresponding phase of the supply voltage (a, b, or c).

Figure 22: Presag compensation technique.
4.1.2. In-Phase Compensation (IPC)

If any disturbance in the supply voltage concerning the magnitude, then DVR restores the same voltage in phase with the supply voltage [194]; hence, it is known as voltage amplitude optimized control. The vector representation of IPC is shown in Figure 23. This technique reduces the requirements of supplying real power but is unable to compensate for the phase angle of load voltage which may cause harm to some sensitive loads. The injected voltage VDVR is given by

Figure 23: In-phase compensation technique.
4.1.3. Minimum Energy Compensation (MEC)

DVR voltage is controlled by adding voltage at 90° to the load current [195]. The vector representation of MEC is shown in Figure 24. This technique minimizes the capacity of energy storage by drawing more real power from the grid, and this minimization is inversely proportional to the sag depth. Equations (6) and (7) represent the magnitude and phase angle of VDVR, respectively.where α is a shifted phase angle between VDVR and IL.

Figure 24: Minimum energy compensation technique.
4.1.4. Hybrid Compensation

The benefits of previous compensation techniques are mixed to develop a hybrid compensation technique. This technique avoids a large DC-link capacitor [146]. In the proposed compensation strategy in [196], first, the load voltage restores via the PSC strategy and catches a transition to the MEC strategy. Figure 25 shows the transformation between presag to in-phase compensation. In the beginning, the presag technique is applied to compensate for the voltage sag as shown in Figure 25(a). Once the DC-link voltage reaches a specific point, then it starts to change the compensation strategy, and the systematic synchronization of grid voltage was carried out by phase look loop (PLL) as shown in Figures 25(b) and 25(c). In this way, the phase angle is gradually changed, till it is in phase with the grid voltage as shown in Figure 25(d). The mapping of different compensation techniques used for various applications is shown in Figure 26, and Table 13 compares the performance of several compensating strategies.

Figure 25: Hybrid compensation technique: (a) presag compensation, (b, c) change from presag to in-phase, and (d) in-phase compensation.
Figure 26: Mapping of compensation strategies.
Table 13: Comparison of various compensation techniques.

5. Inverter Topologies

In many industrial applications, MLIs have found their extensive influence such as UPFC, drives with high power and medium voltage, DSTATCOM, electric vehicles (EV), active power filters, DVR, microgrid, grid integrated or standalone photovoltaic (PV) systems, and numerous other fields [197]. Because of the power and voltage rating limits on power-semiconductor devices, two-level voltage sources (VSIs) were mostly limited to low voltage and medium power applications. Furthermore, these pulse width (PWM) inverters have been affected by excessive switching losses caused by high-frequency switching. These constraints motivated the introduction of a multilevel inverter (MLI), which generates many voltage levels on the inverter output using a variety of voltage sources, capacitors, and power semiconductor devices [198]. To keep MLIs economical for grid-connected renewable energy applications, current advancements in MLIs have focused on reducing switch counts, gate driver circuits, and DC supply, as well as improving the quality of power and fault-tolerant capability [199].

Depending on supply DVR is classified into single phase and three phase. Based on the inverter, each one is again divided into different topologies, and they are shown in Figure 27. Half-bridge inverters [200] and H-bridge (full-bridge) inverters [201] are familiar inverter topologies in single-phase DVR. Besides that, numerous multilevel inverters, matrix converters, and impedance-fed inverters [202] are used for both single and three-phase DVRs. AC-AC converter-based DVRs [78, 203] are used to mitigate the PQ problems without a DC-link capacitor as shown in Figure 28. However, during voltage sag, AC-AC converters draw huge current from the grid. Thus, these are not suitable for long-duration voltage sag mitigation in weak grids. For deep voltage sag, Z-source converter-based DVR with less DC-link voltage was presented in [204], although it requires more storage elements and also suffers from shoot-through.

Figure 27: Different inverter topologies in DVR.
Figure 28: AC-AC converter-based DVR [78].

Full-bridge, four-leg, six-switch, and six-switch split capacitor topologies are the most common traditional inverter topologies used in three-phase DVR. In [205], the authors implemented a DVR by using a three-phase full-bridge PWM voltage-source inverter supplied by lead-acid batteries for real power support as shown in Figure 29; the DVR can compensate the voltage swells/sags using the voltage-space-vector PWM (VSVPWM) approach and the software phase-locked loop (SPLL), thus keeping the load voltages at 1.0 p.u. A 4-leg VSC-based DVR shown in Figure 30; the DVR provides closed-loop control for producing unbalanced 3-phase voltages with a zero-sequence component, and the DVR’s response during voltage sags is instant, with less than 8 ms delay. The extraction of sequence components accounts for the majority of the time delay [206].

Figure 29: 3 − ϕ full-bridge inverter-based DVR [205].
Figure 30: 3 − ϕ four leg inverter [206].

Six-switch two-level voltage-source inverter-based DVR with a common DC capacitor (Cdc) shown in Figure 31 was proposed in [207]. In this topology, to safeguard unbalanced and nonlinear loads, a simple generalized technique for generating instantaneous reference compensating voltages for managing self-supported DVR based on basic SRFT was devised. A fundamental positive-sequence extractor has been suggested, which extracts three fundamental positive-sequence phase voltages from just two unbalanced and/or distorted line voltages. A six-switch split capacitor shown in Figure 32 was implemented in [208]; here, a 3-D voltage space vector PWM algorithm has been implemented for the control of the three-phase four-wire inverter, and the positive, negative, and zero sequence components of the terminal voltages were controlled instantaneously. A DVR with a three-phase split capacitor inverter was studied, which can inject three distinct voltages in series with the main circuit to maintain the voltage waveform at the sensitive load. Usually, two-level voltage source inverter-based DVR [201–207] is suitable for low voltage. Merits and demerits of various inverter designs of DVR are reported in Table 14, and a comparison of various conventional two-level inverter topologies of DVR is tabulated in Table 15.

Figure 31: 3 − ϕ six-switch two-level inverter-based DVR [207].
Figure 32: 3 − ϕ six-switch split capacitor inverter-based DVR [208].
Table 14: Merits and demerits of various inverter topologies in DVR.
Table 15: Comparison of various two-level inverter topologies in DVR.

However, for higher power, two-level voltage source inverters are not suitable because the switches will block large voltage, and more creates electromagnetic interference; to get better from these issues, multilevel inverter (MLI) is the best option. The benefits of MLIs are lower output voltage step, high power quality, fewer switching losses, minimum harmonics, and better electromagnetic compatibility.

MLIs are useful for high-voltage applications due to their capacity to synthesize output-voltage waveforms with an expanded harmonic spectrum and achieve a higher voltage with a limited maximum device rating. A multilevel output can be produced by properly arranging power-switching semiconductive devices and voltage sources. MLIs are classified into two types: classical MLIs and recent hybrid MLIs.

5.1. Classical Multilevel Inverter Topologies

Conventional MLI topologies are commonly employed in a wide range of manufacturing and grid-connected renewable energy systems. Classical topologies are popular because they have a modular structure, are easy to manage, have low harmonic distortion, are fault-tolerant, and have low switching losses. The familiar classical multilevel inverter topologies are diode-clamped (or) neutral-point clamped (NPC), flying capacitor (FC), and cascaded H-bridge (CHB) inverters shown in Figure 33. Certain performance metrics and terms used often in MLIs are as follows: Nsw is the switch count, Nso is DC sources, Nlel is the number of voltage levels, Ncc is clamping capacitors count, Nd is the diode count, Ncd is the count of clamping diodes, Ndc is the number of DC-link capacitors, Vblk is the total blocking voltage, PIV is the peak inverse voltage, and TSV is the total standing voltage of MLI. Table 16 presents quantitative and qualitative comparisons between classical MLI topologies. All these topologies have their positives and difficulties. Capacitor voltage balancing is difficult when the voltage level increases in the case of diode-clamped MLI; hence, these are restricted to three levels. However, most of the industries use three-level NPC inverter. Flying capacitor MLI requires more DC capacitors for higher voltage levels. However, there is flexibility to set the switching combinations feasible for DC capacitor voltage balance [164]. Due to its modularity characteristic, CHB MLI topology becomes more reliable and popular. However, each bridge needs an isolated DC source, and for higher levels, the requirement for switches also increases [144].

Figure 33: Classical multilevel inverter topologies [198]. (a) NPC. (b) FC. (c) CHB.
Table 16: Quantitative and qualitative comparison of classical MLI topologies.
5.2. Hybrid MLI Topologies

Recently, many hybrid topologies have been proposed by researchers mainly emerging from conventional topologies to attain power quality issues and high grid code standards economically [214–220]. Recent hybrid MLI topologies are mainly classified into reduced components with H-bridge topologies [221–238]; they have separate blocks for polarity and level generation and are best suited for LV applications, as shown in Figure 34(a). Reduced components without H-bridge topologies [239–247], capable of generating bipolar waveform, consisting of several unit cells connecting in serial, are mainly used in medium voltage (MV) applications, as shown in Figure 34(b). Miscellaneous hybrid topologies are primarily designed for high power and MV/HV applications [248–257], as shown in Figure 35(a), whereas, some other topologies [258] use variable turn ratio transformers to generate more voltage levels while improving the LV supply from fuel cells, and standalone PV arrays are as shown in Figure 35(b). In MV grid-connected power systems, traditional topologies are still commonly employed. However, the worsening penetration and compliance with power quality and the high grid code standards measures of renewable power systems have led scientists to invent new RC topologies for MVs and high-power applications in modern times. In these topologies, there is no separate H-bridge to create polarity; other benefits include modularity, fault tolerance, high reliability, and reduced space requirements. The comparison of various hybrid MLI topologies is presented in Table 17.

Figure 34: Reduced components topologies (a) with H-bridge [219]and (b) without H-bridge [197].
Figure 35: Miscellaneous hybrid topologies: (a) MV/HV and high-power applications [256] and (b) LV supply from standalone PV arrays [258].
Table 17: Comparison of various hybrid MLI topologies.

In [77], a new DVR topology based on a buck-boost AC/AC converter was proposed. It contains five bidirectional switches, an inductor, and a capacitor, and the topology’s most notable feature is the lack of an injection transformer, allowing for a direct connection to the grid without the need for storage elements. As a result, as compared to conventional topologies, this topology has less physical volume, mass, and cost. A DVR with a cascaded H-bridge multilevel converter was proposed in [144] that could connect directly to the MV network without the use of an injection transformer as shown in Figure 36. The voltage restoration is achieved by the capacitors as energy storage using the zero active power compensation technique.

Figure 36: DVR with a cascaded H-bridge multilevel converter [144].

DVR with five levels reduced the number of power circuit components. TCHB inverter [259] was used to mitigate the voltage sag using two voltage compensation schemes, in-phase and presag compensation. In [260], the authors proposed an S4L inverter-based DVR with a single DC power source and reduced switch count; thus, it is more cost-effective. Furthermore, it generates seven levels, which greatly aids in the reduction of the system’s harmonic problem. Interline DVR with a CHB multilevel inverter was proposed by Shahabadini et al. [261] to mitigate the voltage sag with better THD. An adjustable DC-link connected MLI-based DVR is suggested in [160] , which is suitable for compensation of both long- and short-period sag. DVR with an open-end winding transformer having reduced inverter loss and lower harmonics is proposed in [165]. A cascaded OEW transformer-based DVR is reported in [262] with better voltage levels and reduced THD even though it does not require extraclamping diodes. The number of switches required for n-level voltage in conventional MLI is Nsw = 2(n − 1). For a definite voltage level, these topologies require a high number of switches; thus, the required gate drivers, size, and cost are increased. To overcome these, a T-type MLI-based DVR is proposed for medium- and high-power applications in [263] which is shown in Figure 37, and a comparison of various multilevel inverter topologies in DVR is tabulated in Table 18. A comparison of various multilevel inverter topologies in DVR is shown in Figure 38.

Figure 37: Reduced switch count (T-type) MLI-based DVR [263].
Table 18: Comparison of various multilevel inverter topologies in DVR.
Figure 38: Comparisons of various MLI-based DVR topologies.

In addition to mitigating the sag and swelling, the DVR can execute other functions such as compensation of selective voltage harmonics [86, 270, 271], mitigation of voltage flickers [272–274], reactive power compensation [275], and frequency compensation [276]. Protection of the DVR [277] is very important during the fault at the load side of the DVR; otherwise, the same downstream fault current is present in the DVR, and it may damage the DVR. This protection is achieved by making an alternative path for the fault current through breakers, thyristors, and varistors. However, other proposals [85, 270, 278–280] also actively limit the downstream fault current. Various functions and applications of DVR are presented in Figures 39 and 40, respectively.

Figure 39: Functions of DVR.
Figure 40: Applications of DVR.

6. Conclusions and Future Scope

A systematic review of different types of DVR systems and the future scope of the relevant literature are discussed in this article. Studies reviewing the DVR include many areas, but specifically, power quality issues, energy-storage topology, absence of energy, and controlled strategies are covered in this paper. DVR configurations based on power converters and control units at different stages are described in detail based on the latest literature. In the orientation towards the integration of renewable energy sources, certain updated and upgraded DVR configurations are also presented. This review supports the selection of the best, most cost-effective, and high-performance DVR configuration based on the requirements of researchers and scientists working on this prospective research [281].

Future research could be conducted in several areas of the literature. The following are some essential considerations; however, they are not exhaustive:(i)The efficiency of the DVR circuit is limited by VSI, filter, and transformer losses. The buck nature of the VSI output voltage necessitates the use of a boost converter between the energy storage and the inverter, which adds more switches, controls, and complexity. By using a multilevel inverter in place of VSI partly or entirely, the need for filters can be eliminated, resulting in fewer switching losses. This allows us to increase DVR’s efficiency, LVRT capacity, cost performance, deep sag/swell compensation, reliability, and harmonic compensation.(ii)The depth of the sag and duration of the sag determines the storage capacity. If there is a long-term sag, the energy storage capacity declines and the DVR’s compensating property gets reduced. The strategy used during the detection stage unit may be modified, particularly when it comes to deep/long voltage sags/swells to the extent that the detection process is highly reliable.(iii)DG and DVR integration has gained popularity due to its reliable features. Future research should take into account the integration of DVR with microgrids and smart grids which will improve power quality for the end-users.

Data Availability

No data were used to support this study.

Conflicts of Interest

The authors declare that they have no conflicts of interest.