Problems and Solutions Concerning the Distance Protection of Transmission Lines Connected to Inverter-Based Resources

Abstract

This article presents a review of the problems and solutions concerning the distance protection of transmission lines connected to inverter-based resources (IBRs). After a brief description of IBRs and distance protection, the reported problems are classified based on their causes and effects. The causes are related to IBR behavior, and the effects are related to distance protection. The effects are classified as overall effects (observable wrong trips or an observable lack of activation of distance functions) and specific effects (related to the particular internal relay elements that failed, causing the observable overall effects). Furthermore, special attention is paid to clearly describe the research literature from relay manufacturers, since it should be closer to the current trends related to real-life problems and solutions. The causes and specific effects particularly mentioned in the reviewed literature are summarized in corresponding tables, including information about those papers where such causes and effects cannot be clearly identified.

1. Introduction

Photovoltaic and full-converter wind generation are connected to the synchronous grid through inverters, and they are the main examples of inverter-based resources (IBRs). Fault contributions from IBRs are substantially different from those from conventional synchronous generators. This fact has created reasonable concerns about the performance of transmission system protections [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101]. The distance protection of transmission lines connecting large IBRs to the grid is probably the protection that is most affected by the presence of IBRs, and this is the particular topic herein covered.

Some literature reviews concerning the impact of IBRs on transmission system protections have been published [1,2,3], and only their parts related to distance protection are herein considered due to the purpose of this article. A total of 40 papers were chosen by [1] for the literature analysis, identifying the following: (a) three categories of protection problems (accuracy, sensitivity and directionality), and four categories of IBR-related issues (controller influence, decreased inertia, need of standardization and modeling); (b) impacts on distance protection, memory polarization, negative-sequence directionality and overcurrent thresholds; (c) some solutions, mainly based on improvements related to the increase in sensitivity of overcurrent elements, the use of proper sequence quantities and proper polarizing quantities, or the use of undervoltage detectors; (d) only 9 out of 35 references verified their simulations with some tests (and only 1 of those references utilized data from real faults). In [2], two relatively large introductory sections were included before the beginning of the state-of-the-art description, and this approach seemed to be necessary because (a) IBR behavior during short-circuits in the grid is still not a well-known topic; (b) professional details of transmission line protection are typically only known by very specialized engineers in power system protections. In the case of distance protections, an interesting summary of causes and effects is shown in [2], identifying the following: (a) particular influenced functions (impedance zones, overcurrent starting, directional elements and fault-type identification elements); (b) particular causes, regarding IBR behavior during grid faults; (c) particular effects on distance protections, regarding measured impedances, overcurrent sensitivity, polarization, directionality, behavior of negative-sequence currents and faulted phase selection elements. Although the analyzed solutions in [2] are mainly related to “advanced research” proposals (distance adaptive protection schemes, use of transient signals and use of artificial intelligence techniques), the modification of the traditional protection is also included as an option. In [3], the reported problems of distance protections due to the presence of IBRs are as follows: (a) low IBR contribution of current to faults in the grid; (b) unusual negative-sequence currents; (c) atypical behavior of angle relationships in each sequence; (d) existence of different possible IBR contributions to faults in the grid due to the different possible IBR control strategies; (e) possible errors in apparent impedances due to frequency excursions in IBR currents during faults in the grid; (f) high dependence on communication channels if the distance protection is not applied without communication-assisted trip schemes; (g) related to memory polarization. Furthermore, the reported solutions are classified in 11 groups [3] (10 related to distance protection). In [3], the maturity of reported solutions is analyzed, considering only the solutions related to distance protection as follows: (a) one is a preliminary idea, far from being recommendable in practice; (b) four are considered mature topics; (c) five are considered maturing topics, and are mainly related to distance protection. In summary, the previous literature reviews show that, regarding IBR-influenced problems of distance protection, very few papers have been based on data from real faults [1], the modification of traditional protections can be still considered as a possible solution [2] and the solutions for distance protections in the presence of IBRs cannot be considered a mature topic [3].

This article particularly shows a review of problems and solutions about the distance protection of transmission lines connected to IBRs. An introduction to IBRs and distance protections is included because it is considered necessary for the proper understanding of this article. Although an exhaustive literature review is herein included for the sake of rigorousness, an important part of this article is focused on the literature from commercial relay manufacturers, which usually offer a more practical perspective. The main contributions of this article are as follows: (a) a complete literature review is presented; (b) useful descriptions of IBRs and distance protections are included to facilitate the understanding of this document; (c) the reported causes of problems due to IBR behavior and the reported specific effects on distance protections are identified and summarized in tables; (d) an understandable description of the literature from commercial relay manufacturers, which usually offer a practical perspective of the current state-of-the-art and real-life trends, is shown. The first three contributions are improvements regarding previous literature reviews [1,2,3], and the fourth one is a novel approach.

2. Some Initial Details About the Literature Review

In this literature review, 101 cited documents concern the protection problems of transmission lines connecting IBRs to the grid [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101] ((a) 3 literature reviews [1,2,3]; (b) 25 documents from relay manufacturers [4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28]; (c) 73 papers that are not from relay manufacturers [29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101]). Herein, the literature review is complemented by the following cited documents: (a) books about grid-connected inverters [102,103,104] and distance protection [105,106]; (b) grid codes that impose constraints on IBR control designs [107,108,109]; (c) publications with different details about traditional distance protections of transmission lines [110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130,131,132] and about other protections [133,134,135,136]. Although the literature about the protection problems of transmission lines connecting IBRs to the grid began in 2007, Figure 1 shows that the number of papers dealing with this topic before 2018 is negligible.

The simultaneous combination of terms about IBRs and distance protection (in the same search) was useful to find publications about this topic. In the case of IBRs, some alternative terms included inverter-based resources, inverter-interfaced generation, inverter-interfaced generators, inverter-interfaced power sources, inverter-interfaced renewable sources and inverter-interfaced energy sources. In the case of distance protection, the number of alternative terms was lower (e.g., distance relays, distance relaying), but the search was complemented by terms of related functions, such as negative-sequence currents, fault-type identification (or selection) and directional function. Articles about microgrids were not considered at all; furthermore, articles about medium voltage collector systems in wind farms or PV plants were just briefly described but not analyzed in detail. The scope of this literature review is related to the connection of large IBRs to main national grids through power transmission systems at high voltage.

3. Fundamentals About IBRs and Distance Protection

Some details about IBRs and distance protection are herein described because they are considered useful for the proper understanding of this article by more readers. These basic descriptions can be complemented by more detailed descriptions of IBRs (e.g., [102,103,104]) and distance protection (e.g., [105,106]).

3.1. Fundamentals About the Contribution of IBRs to Faults in the Grid

IBRs can be grid-following, needing a phase-locked loop (PLL) to track the grid frequency and voltage angle, or grid-forming, whose frequency is controlled to help in the power system control. There are very many options for designing IBR control, and the IBR behavior for faults in the power system is strongly dependent on the particular control designs. IBR control rapidly limits the currents during short circuits in the grid to avoid damages in the power electronics (IBR currents are often limited to around 1.2 times the rated current, unlike contributions from synchronous generators, which are in the order of 5 to 10 times the rated current), and the IBR control also drives the angles of those currents. Some IBRs always inject only positive-sequence currents, even during asymmetrical faults in the grid. Nowadays, grid codes [107,108,109] tend to impose some constraints on IBR control designs. For example, the injection of reactive currents and negative-sequence currents, proportional to the change of positive- and negative-sequence voltages, respectively, has been imposed in some grid codes; such IBR contributions to fault currents should be useful to improve the behavior of distance protection, but there are still many installed IBRs without these features. In summary, the IBR behavior for faults in the grid cannot be generalized because it depends on the IBR control. On the other hand, IBRs by themselves do not inject zero-sequence currents into the grid, but the generation plants must have grounding paths (typically through transformers with delta/wye connections), offering a constant zero-sequence impedance in their equivalent network. That is, the zero-sequence contribution from IBRs to faults in the grid is predictable, unlike the negative-sequence contribution, which is typically unpredictable.

An illustrative example of a grid-following inverter is shown in Figure 2. The PWM voltage waveshape is not sinusoidal, but the output filter and the step-up transformers provide enough impedance to connect these sources to the strong sinusoidal source voltage of the grid. The fundamental component (or first harmonic) of the PWM voltage waveshape can be seen as an equivalent voltage source (E_PWM-1h). This approach is useful to obtain similarities with the traditional phasor analysis of electric power systems. The IBR control can perform very fast changes of E_PWM-1h (in module and/or angle), which are not attainable in a synchronous generation (where internal electromotive forces change slowly). Due to this, the fact of obtaining very fast changes of E_PWM-1h has being related to the term “low inertia”, as a simplified way to describe this IBR behavior.

In the case of grid-following inverters, the PLL can fail in determining the grid frequency during faults and, consequently, the voltage waveshapes can be synthetized at a frequency different from the grid frequency. Thus, the waveshapes of IBR currents during faults in the grid can have erratic behavior regarding their modules, angles and fundamental frequencies.

3.2. Fundamentals About Distance Protection

Distance protections are based on apparent impedances; their characteristics are usually shown in the R-X plane and can have diverse shapes (mho, quadrilateral, polygonal, etc.). These protective functions use measurements of voltages and currents at relay locations. Almost all the distance protections currently in service are in microprocessor-based relays; therefore, the voltages and currents are sampled several times for each cycle, and their phasor magnitudes and angles are computed using algorithms that integrate the sampled values in a sampling window (e.g., using the last cycle of the sample). Consequently, each transition from a given condition to another one (e.g., from a steady state pre-fault condition to a fault condition) implies a transient behavior of apparent impedances seen by distance protection.

Apparent impedances should be internally computed by distance protections according to the phase(s) involved in the fault; otherwise, the apparent impedance of a non- faulted loop could cause the wrong operation of distance functions. Due to this fact, a faulted-phase detection element (or algorithm) is usually necessary for distance protection. Unfortunately, information concerning this element is not always clearly given by relay manufacturers. Nowadays, faulted-phase detection elements based on magnitudes of currents do not seem to be in use. Two types of conditions for the faulted-phase detection algorithms, reported by relay manufacturers, are shown in Figure 3 (Figure 3a, based on angles between negative- and positive-sequence currents [110]; Figure 3b, based on angles between negative- and zero-sequence currents [111]). Notice that both algorithms need the negative-sequence currents.

Negative-sequence quantities have also been very useful for determining the fault directionality of unbalanced faults. Figure 4 helpfully explains this point, showing only the negative-sequence networks for the sake of simplicity. Z_2M and Z_2N are negative-sequence source impedances at both line ends; Z_1L is the positive-sequence impedance for the total line length, Z_1X is the positive-sequence impedance between the relay and the fault point and Z_1Y is the positive-sequence impedance between the fault point and the remote line end (i.e., Z_1L = Z_1X + Z_1Y); Z_2I is the negative-sequence equivalent impedance that represents the additional interconnections between buses M and N (i.e., in meshed networks, substations at both line ends are also usually interconnected by other paths, at the same voltage level and/or at a different voltage level, and the net effect can be summarized by equivalent impedances); V_2R and I_2R are the negative-sequence voltage and current at the location of the analyzed relay. Figure 4a shows the case of forward faults (in the protected line, for the sake of simplicity). Considering that the phases of I_2R and I_2M are similar to each other (as tends to happen for systems with only synchronous generation), the quotient V_2R/I_2R is an inductive impedance with a negative sign (if Z_2I tends to infinite, I_2R is equal to I_2M and V_2R/I_2R is −Z_2M). Figure 4b shows the case of reverse faults (in the busbar of relay location, for the sake of simplicity). The quotient V_2R/I_2R is an inductive impedance (if Z_2I tends to infinite, then I_2R is equal to I_2N and V_2R/I_2R is Z_1L + Z_2N). Having different signs in the result of V_2R/I_2R is key for this directional detection method. The assumption of Z_2I tending to infinite is not the general case, but it is very useful to show the tendency in a much simpler way (Z_2I is included in this explanation for the sake of rigorousness, but it is often omitted in the literature regarding distance protection in order to show simpler deductions).

The effect of fault resistances on the apparent impedance seen by distance protection is influenced by the current from the remote line end. Figure 5 shows the case of three-phase faults, for the sake of simplicity. Z_1M and Z_1N are positive-sequence source impedances at both line ends; Z_1I is the positive-sequence impedance that represents the additional interconnections between buses M and N (i.e., similar to Z_2I, previously described for the case of the negative-sequence network), V_1R and I_1R are the positive-sequence voltage and current at the location of the analyzed relay; E_M and E_N are the source voltages behind the impedances for buses M and N, respectively. Considering that the apparent impedance (Z_app) is simply V_1R/I_1R, Equation (1) can be easily obtained by the application of Kirchhoff’s law. This equation indicates that the effect of fault resistance is amplified and phase-shifted due to the influence of I_1Y/I_1R. The mathematical analysis of asymmetrical faults is conceptually similar but slightly more complex [112]. However, there are diverse polarization methods for ground faults [113] from different manufacturers of real-life distance protections, and this fact implies that the apparent impedances seen by distance protections are dependent on the relay model. In general, the effect of fault resistance on the apparent impedance seen by the distance protection depends on the pre-fault load flow [112,113] (which determines the values of E_M and E_N, and the circuit solution under fault conditions depends on E_M and E_N).

$${\mathrm{Z}}_{\mathrm{app}}=\left({\mathrm{Z}}_{1\mathrm{X}}\right)+{\mathrm{R}}_{\mathrm{F}}\left(1+{\displaystyle \frac{{\mathrm{I}}_{1\mathrm{Y}}}{{\mathrm{I}}_{1\mathrm{R}}}}\right)$$

(1)

Memory- and cross-polarized distance functions are usually applied to avoid wrong directional detection for faults very near to the relay location. These polarization methods are often applied with offset mho characteristics; in these cases, the offset impedance (Z_OF) is not a relay setting and it can be seen as a quantity dependent on the fault direction. Figure 6a shows that Z_OF is in the third quadrant for forward faults, whereas Figure 6b shows that Z_OF is in the first quadrant for reverse faults [114,115,116,117] (Z_R is the reach setting of the distance protection). The value of Z_OF for forward faults is dependent on different variables [115]. For the sake of simplicity, Z_OF for forward faults is often roughly approximated to −Z_M (Z_M is the source impedance at the relay location); that is, the offset mho characteristic expansion can be very large if Z_M corresponds to a weak infeed.

In general, there are diverse polarization methods for distance functions. For the purpose of this article, it is worth mentioning that negative- and zero-sequence currents have, for a long time, been utilized for the reactance reach line of quadrilateral characteristics [126]. Nowadays, this type of method has been utilized by different manufacturers (e.g., [110,127,128]).

Faults very near to one line end tend to be in zone 2 of the distance protection of the other line end. To avoid the zone 2 delay, the distance protection is usually complemented by communication-assisted trip logics, for instance, permissive underreach transfer trip (PUTT), permissive overreach transfer trip (POTT) and directional comparison unblocking (DCUB). Weak infeed logic, in conjunction with echo logic, is available to avoid the non-operation of the communication-assisted trip logic if one line end has a weak source. The communication-assisted trip logic coexists with the noncommunication-based trips of distance protection zones; thus, a loss of communication does not imply the total loss of distance protection.

The analysis of distance protection has many details. For the purpose of this article, two additional details must be mentioned as follows:

(a)

The source-to-line impedance ratio (SIR) is a simplified parameter that has often been utilized to describe some features of distance relays. Although there are different possible ways to describe SIR, if Z_1I is neglected (as shown in Figure 5), the SIR for three-phase bolted faults at the remote line end is simply Z_1M/Z_1L. For instance, a high SIR can imply that the voltage at the relay location is very low for faults at the remote line end and, consequently, substantial errors in the voltage measurement can be expected.

(b)

Transmission voltages are often measured with the help of capacitive voltage transformers (CVTs), which consist of a capacitive voltage divider with an inductance at the low-voltage side and ferro-resonance suppression circuits. CVT transient behavior should be considered for the proper setting of distance protections [132].

4. Reported Problems of Distance Protection of Transmission Lines Connected to IBRs

The approach regarding causes and effects related to these problems, originally shown in [2], is herein modified in some extent to present the points in a more organized understandable and useful way.

4.1. Reported Causes of the Problems Analyzed

The overall cause of the problems analyzed is the behavior of the short-circuit current of the IBR (SCC-IBR) for faults in the grid. The reported specific causes of the problems analyzed are as follows:

(a)

Low magnitude of SCC-IBR.

(b)

Control of angles of SCC-IBR.

(c)

Absence or erratic behavior of negative-sequence in SCC-IBR.

(d)

Frequency of SCC-IBR is different from grid frequency.

(e)

“Low inertia”.

(f)

Variation in pre-fault load flow of the IBR.

Note 1: as explained in Section 3.1, the term “low inertia” means, in this context, that the controlled SCC-IBR could be seen as the result of a fast change in the internal voltage of an equivalent source (and, by analogy, that fast change could be seen as the result of having “low inertia”).

Note 2: the documents that analyze the variation in the pre-fault load flow of the IBR deal with the effect of fault resistances on the apparent impedance seen by the distance protections, but their analysis is based on assuming that the IBR has the same behavior during short circuits as the traditional synchronous sources (and, nowadays, this premise cannot be kept).

4.2. Reported Troublesome Effects of SCC-IBR on Distance Protection

The possible undesired effects of SCC-IBR on distance protection are herein divided into overall effects and specific effects. There are basically three possible overall effects as follows: (a) wrong operation of zone 1 for faults outside the protected zone; (b) loss of detection in zone 1 for faults that should be detected in zone 1; (c) loss of detection in zone 2 for faults that should be detected in zone 2. The possible specific effects are diverse and are listed in the next paragraph. Overall effects can essentially be caused by the same specific effects. The approach regarding effects shown in [2] is only about the (herein called) specific effects. A detailed analysis of these problems and solutions should be performed from the perspective of these specific effects, but the overall effects are useful to offer a simpler and non-detailed overview regarding this topic.

The reported troublesome specific effects of IBRs on distance protection are as follows:

(a)

Apparent impedance (Z_app) is far away from protection zones.

(b)

The dynamic path of Z_app enters zone 1 for external faults.

(c)

SCC-IBR is below relay thresholds.

(d)

Delay in the operation of distance protection.

(e)

Wrong operation of memory- or cross-polarized distance protection.

(f)

Wrong operation of reactance reach line, polarized with zero- or negative-sequence currents.

(g)

Wrong phase directional detection.

(h)

Wrong negative-sequence directional detection.

(i)

Wrong fault-type detection.

(j)

Failure in the communication-assisted trip.

(k)

Need to update the way to compute SIR.

(l)

Wrong trips due to transient CVT behavior.

5. Solutions Proposed in the Literature

5.1. Avoiding the Use of Distance Protection

Nowadays, the main practical solution to the problems of distance protection of transmission lines connected to IBRs is to avoid the use of distance protection, and the main available alternative is the line differential protection (87 L). The 87 L drawback of depending on communication channels is often diminished by increasing the investments in more reliable communication links. Distance protection with communication-assisted trip logic (e.g., POTT with weak infeed and echo logic) is another practical solution to complement the 87 L, and it has the same drawback related to the dependence on communication channels. These practical solutions have also been reported in the literature [15,23,61,98].

The use of traveling wave-based protections is theoretically a possible alternative for lines connecting IBRs to the grid [69,87,99,100]. These protections are also dependent on communication channels (in the commercial relays that have these protections), and there is no reported application of this solution in real-life cases of lines connecting IBRs to the grid.

5.2. Improving the Settings Related to Distance Protection

Some proposed improvements for settings related to distance protection are as follows:

(a)

Inclusion of time delay for zone 1 [36]. This improvement can be useful to (a.1) avoid wrong trips due to the transient path of Z_app; (a.2) wait for the trip of the distance protection on the strong source side (grid side), avoiding the influence of the remote-side current on the distance protections on the weak source side (IBR side).

(b)

Increasing the pickup of the negative-sequence directional function to give preference to the use of the zero-current directional function [4,5].

(c)

Modification of settings related to fault-type identification logic to give preference to the use of undervoltage-based logic [4,5,7,8].

(d)

Activating a “best choice” algorithm to automatically change the way of polarizing the quadrilateral characteristic (or to substitute the quadrilateral characteristic with a mho characteristic) [10,11,12,13].

5.3. Development of New Protection Algorithms or Functions

Some proposals for new protection algorithms are as follows:

(a)

A novel distance function, controlled by weak infeed logic [7,8]. Details of this option are described in Section 6.

(b)

New fault-type identification algorithms, for instance, based on (b.1) phase-shift between zero- and negative-sequence voltages and/or phase-shift between positive- and negative-sequence voltages [77]; (b.2) superimposed positive- and/or negative-sequence voltages [17]; (b.3) calculating the angle difference between the sequence voltages [83]; (b.4) phase-shift between voltages in the αβ plane [79]; (b.5) angles of pure fault currents [28]. Fault-type identification algorithms for the weak infeed side of transmission lines, based on the angles of sequence voltages, are already implemented in commercial relays [110,130]; they have shown good performance in systems with IBRs.

(c)

New directional detection algorithms, for instance, based on (c.1) phase currents and voltages, as the directional protection in electromechanical relays [71]; (c.2) the equivalent admittance measured by the relay [70].

(d)

New ways to compute the apparent impedance, for instance, based on (d.1) estimation of the Thevenin equivalent circuit for the synchronous grid [32]; (d.2) estimation of equivalent impedances for both line ends, using only local data [15,30,51,88,90] or using data from both line ends [27]; (d.3) estimation of IBR pure fault impedances at each instant [35]; (d.4) the use of phase-to-ground loop also for double line to ground faults [29]; (d.5) analysis of homogeneity in negative- and zero-sequence networks [33]; (d.6) the open-circuit property of the negative-sequence network on the side of some inverters in order to analytically determine the error in the apparent impedance calculation [26].

(e)

New adaptive characteristics in the R-X plane, for instance, based on pre-fault IBR conditions [38,39,40,41,43].

A sound comparison between new proposed protection algorithms and commercially available protective functions should be performed, considering many details (e.g., accuracy, speed and dependability) in a very wide spectrum of possible cases. This type of comparison is usually only available for relay manufacturers that have worldwide information about the behavior of power system protections; probably, due to this fact, many research ideas (relatively naive) in this field are never installed in real-life protective relays. For instance, although adaptive distance protection has been proposed for more than 30 years, for traditional power systems to consider the changes of Z_app loci [131,132], and for protective relays to have different available setting groups to perform these adaptive changes, no report on real-life applications of this technique have been found.

5.4. Development of New IBR Control Algorithms

New IBR control algorithms have been specifically proposed to improve the behavior of transmission system protections, and a review of them is shown in [101]. The main objectives of these new IBR control algorithms are as follows:

(a)

Inclusion of grid code requirements about reactive current injection and/or negative-sequence current injection [28,45,47,48,49,50,64,72,84].

(b)

Avoiding phase shift between fault currents at both line ends [45,47,48,49,72,96].

(c)

Improving the fault-type identification [28,64,84,96].

(d)

Improving directional detection [72].

(e)

Improving the behavior of the PLL to avoid errors in the internal reference of frequency for the inverters [65].

Some proposals for changes in IBR control algorithms are dependent on the inclusion of changes in the design of the distance protection [28,45,46,47,48,49,50,64,65,72,84,96]. On the other hand, other proposals for changes in IBR control algorithms are not dependent on changes in the design of distance protection [46,49,72].

The development of new IBR control algorithms to fulfill the new grid code requirements regarding reactive current injection and/or negative-sequence current injection could be an important step in the mitigation of problems caused by IBR contribution to distance protections. Thus, the analysis of this topic should surely be updated in the near future due to the inclusion of new findings related to the adaptation of IBR control algorithms to the new (more restrictive) grid codes.

5.5. Development of Protections Based on Artificial Intelligence Algorithms

The research on power system protection based on artificial intelligence algorithms has been reported for more than 30 years (e.g., [133,134]). This type of solution has also been proposed to solve problems related to distance protections in IBRs, for instance, (a) the use of artificial neural networks [42] or support vector machines [78] for distance protection; (b) the use of a mixture of machine learning methods for fault-type identification [79]; (c) the use of decision trees and support vector machines for directional detection [85].

In general, the research on the possible application of these techniques to power system protection is important because it usually offers new perspectives regarding the nature of the problems and their solutions. However, the possible application in real-life protective relays is not at all obvious. An example of this fact is the training of artificial intelligence algorithms to discriminate between transformer magnetizing inrush and internal faults. This option was theoretically proposed many years ago [133] and the number of published papers about this topic is huge, but there is no real-life protective relay using this type of algorithm. On the other hand, an academic research on high-impedance fault detection in overhead distribution circuits was the origin of an algorithm based on expert systems for a real-life protective function many years ago [134], and some relays of that manufacturer still offer the corresponding legacy algorithms [135]; however, this case can be considered a rare exception (actually, one of the original authors described some implementation issues, many years ago [136], and this function has often been recommended for monitoring purposes instead of tripping due to the need of avoiding wrong trips).

6. Analysis of Articles from Relay Manufacturers

The following four relay manufacturers have published articles related to the distance protection of transmission lines connected to IBRs: SEL [4,5,6,7,8,9], GE [10,11,12,13], Hitachi [14,15,16,17,18] and Siemens [19]. Two of them (SEL and GE) cooperated with Sandia National Laboratories (USA) in a research [20] that consisted of using inverters’ black-box type software models (provided by four independent manufacturers of commercial inverters) to simulate an exhaustive set of relevant faults in a power system model taken as an example. The simulations were performed with electromagnetic transient software, and the waveshapes of voltages and currents corresponding to selected relay locations were obtained in order to be injected into real commercial relays from the manufacturers participating in the research. Relatively similar research was developed in Europe, as part of a larger research project (MIGRATE project [21]), and the test results on relays from four different manufacturers have been shown [22,23]; however, the manufacturers’ reports concerning the causes of the wrong behavior of distance functions in those tests have not been found during this literature review. These multi-institutional research efforts, in the USA and Europe [20,21], with the participation of electric utilities, relay manufacturers and other research institutions, are clear signs of the treatment of this issue as a recognized problem to be solved.

6.1. Articles from SEL

In [4], the aforementioned research project led by Sandia National Laboratories is briefly described as an introduction to show the found problems, which can be explained with the help of Figure 7. The faults in D and G are very near to bus 2, and they are the main fault locations of interest.

The following overall effects on distance protections are described in [4]: (a) wrong trip of R1 in zone 1 for faults in D or G; (b) wrong behavior of zone 2 of R1 for faults in D or G, because the apparent impedance is not continuously seen in zone 2; (c) wrong trip of R2 in zone 1 for faults in G.

The main reported specific effects on distance protections [4] are as follows: (a) wrong fault-type detection causing Z_app to enter zone 1 of R1 for faults out of its zone 1; (b) erratic fault-type detection causing Z_app to leave zone 2 of R1 for faults within its zone 2; (c) wrong directional detection causing the wrong trip of zone 1 of R2 for reverse faults; (d) transient behavior of capacitive voltage transformers causing Z_app to enter zone 1 of R1 for faults out of its zone 1. It is worth highlighting that these specific effects are clearly shown in [4] because the tested relays have the capability of showing the instantaneous activation/deactivation of their internal logic signals, and this feature is useful for a clear analysis of fault records.

In [4], the main causes of the reported problems are the features of SCC-IBR (low magnitude, behavior of angles, absence or erratic behavior of negative-sequence), and it is also shown that the frequency of the measured negative-sequence current can be different from the frequency of measured voltages.

The proposed initial solutions for the problems described in [4] are shown in [5]. Such solutions consist of the following: (a) raising the negative-sequence current pickup, with a consequent improvement in the fault-type identification logic (which changes to an algorithm based on undervoltage) and in the directional detection (which changes to an algorithm based on zero-sequence current); (b) improvements in reach and delay settings for zone 1 to prevent overreaching; (c) enabling the capacitive voltage transformer’s transient blocking logic, available in the relay.

A new distance protection, whose behavior can be unaffected by the troublesome features of SCC-IBR, is proposed in [7,8]. This solution is based on a weak infeed logic that is applied to the internal control of the distance function (instead of being applied to the communication-assisted trip logic, which is the traditional use of the weak infeed logic). The zone reach is provided by a nondirectional offset characteristic (i.e., covering the origin of the R-X plane), with an independent directional restraint element based on weak infeed logic (the source is strong or weak, depending on the directionality). The use of incremental quantity directional elements is also proposed as an option for the directional restraint of that nondirectional offset characteristic of the distance function [7,8].

In [6], the problem of SIR calculation for lines near to IBRs is analyzed, emphasizing that the erratic behavior of negative-sequence currents must be considered. In [9], a novel polarization method is proposed, which is based on a state machine to select the polarizing voltage among self-polarization, memory-polarization and a novel PLL-based polarization (that creates the polarizing voltage from the frequency measurement).

6.2. Articles from GE

The results from this manufacturer, related to the aforementioned research project led by Sandia National Laboratories, are shown in [10,11]. The following overall effects on distance protections are identified: (a) wrong trip of R1 in zone 1 for faults out of zone 1; (b) loss of detection in zone 1 for faults that should be detected in zone 1.

The main reported specific effects on distance protections [10,11] are as follows: (a) the dynamic path of Z_app enters zone 1 for external faults; (b) wrong operation of reactance reach line, polarized with zero- or negative-sequence currents; (c) wrong operation of memory-polarized mho characteristic; (d) wrong fault-type identification; (e) wrong negative-sequence directional detection and wrong phase directional detection; (f) the loss of detection in zone 1 is due to the off-nominal frequency of IBR currents.

The proposed solutions [10,11] are the use of (a) a controllable polarized mho, with a restricted mho expansion; (b) voltage-based fault-type identification algorithms; (c) a directional element based on zero-sequence current instead of negative-sequence current; (d) an algorithm to determine the best polarization selection for the reactance reach line; (e) a delay in the zone decision if frequency excursions are detected. More extensive sets of tests are shown in [12,13], and the manufacturer indicates that the proposed solution has good performance in the case of ground distance elements. In the case of phase distance elements, the manufacturer indicates that an enhanced security feature has been applied to solve the problems, providing a promising performance.

6.3. Articles from Hitachi

Five articles regarding the effect of IBRs on distance protection are found from this manufacturer [14,15,16,17,18]. The IBR effect on distance protection is particularly analyzed in [14,15] (mainly overall effects), whereas different transmission system protective functions are analyzed in [16] (highlighting that the main problems are related to distance protection).

Extensive tests are performed in [14] using (a) a system with a transmission line connecting the IBR with the grid; (b) a PSCAD model of the IBR, representing a full-converter wind generation plant that follows the North American grid code. In [14], (a) the overall effects on distance protections are reported by percentages of detected dependability and security problems during the tests; (b) the reported specific effects are mainly related to the behavior of the apparent impedances and, consequently, the proposed solutions lead to improvements in the corresponding relay algorithms. On the other hand, the IBR effect on fault location algorithms is also analyzed in [14], recommending the use of fault locators based on information from both line ends.

Extensive tests are also performed in [15] using (a) the well-known IEEE 39-bus test system; (b) the inclusion of full-converter wind generation plants in two buses; c) two IBR penetration scenarios. In [15], (a) the overall effects on distance protections are reported by tables of apparent impedances and plots of apparent impedance trajectories in the R-X plane; (b) the reported specific effects are mainly related to the behavior of apparent impedances and, consequently, the proposed solutions lead to improvements in the corresponding relay algorithms. In [15], the authors indicate that their solution is generic, irrespective of the nature of the system (i.e., systems with only synchronous generation, or systems including IBRs). On the other hand, this test system is also utilized in [17] to analyze the voltage-based fault-type classification methods, and an improved algorithm is proposed to solve the problems related to the use of traditional methods.

In [18], the distance protection of wind farms’ medium voltage collection systems is analyzed. Although this particular topic certainly is related to the distance protection of IBRs connected to the grid, it is not within the scope of the present literature review because the power system topology is quite different. This literature review covers the case of high-voltage transmission lines with distance protection at both line ends; however, in the case of medium-voltage collection systems [18], the distance protection is only located at the collector substation, and the main analyzed problems are usually related to fault detection in ungrounded systems (or with high-impedance grounding).

6.4. Article from Siemens

An article from this manufacturer [19] concerns the distance protection of wind farms’ medium-voltage collection systems (as this paper describes in the last paragraph of Section 6.3). Although this particular topic is certainly related to the distance protection of IBRs connected to the grid, it is not within the scope of the present literature review because the power system topology is quite different. In the mentioned case of medium-voltage collection systems [19], the distance protection can be located at both line ends of one feeder under certain circumstances, but several wind generators are directly connected along that feeder (that is, there are intermediate infeeds of fault currents along the collector). This literature review covers the case of high-voltage transmission lines with distance protection at both line ends, without any generator connected along the line length.

6.5. Articles from Schneider Electric

Schneider Electric cooperated in the aforementioned MIGRATE project [21]. Although manufacturers’ reports regarding the behavior of distance functions during the project tests were not found, a person from Schneider Electric was the co-author of two papers derived from that project [23,25]. Furthermore, a proposed algorithm for faulted-phase selection was coded into a Schneider Electric relay [24] as part of that project.

7. Summary

The causes of problems in the distance protection of transmission lines connected to IBRs, along with the specific effects on distance protections, are shown in an organized way in Section 4. In this context, the causes are related to the particular features of short-circuit currents from IBRs and the specific effects concern distance protection.

Table 1. Causes of analyzed problems according to reviewed references.

Cause	Mentioned by
	Non-Manufacturers	Manufacturers
Low magnitude of SCC-IBR	[38,44,45,52,61,63,68,76,77,78,79,89,95]	[-]
Control of angles of SCC-IBR	[26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,44,46,47,48,49,50,53,54,55,56,57,58,62,64,68,77,78,79,80,81,82,83,84,88,89,90,95,97]	[4,5,7,8,10,11,12,13,14,15,16,17,18,23]
Erratic behavior of negative-sequence of SCC-IBR	[37,45,58,64,68,70,71,72,73,74,75,79,85,99]	[4,5,7,8,10,11,12,13,22,23,24,25]
Frequency of SCC-IBR different from grid frequency	[65,89]	[4,5,7,8,9,10,11,12,13]
“Low inertia”	[52,63,65,67]	[7,8,9]
Variation in pre-fault load flow of IBR	[33,42,43,67,94]	[-]
Not specified	[51,59,60,66,69,85,86,87,91,92,93,96]	[-]

“Not specified”: this information is not directly mentioned in the reference.

and

Table 2. Specific effects of IBRs on distance protection according to reviewed references.

Specific Effects	Mentioned by
	Non-Manufacturers	Manufacturers
Apparent impedance (Zapp) is far away from protection zones	[26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62]	[4,5,7,8,10,11,12,13,15]
The dynamic path of Zapp enters zone 1 for external faults	[15,29,30,31,33,34,35,36,37,40,41,48,55,56,57,58,59,60,89]	[4,5,7,8,14,16,23,24]
SCC-IBR below relay thresholds	[44]	[4,5,7,8,10,11,12,13,15]
Delay in the operation of distance protection	[45,51,54,60,63,88]	[-]
Wrong operation of memory or cross-polarized distance protection	[52,63,64,65,66,67]	[7,8,9,10,11,12,13]
Wrong phase directional detection	[50,68,69]	[-]
Wrong negative-sequence directional detection	[50,68,69,70,71,72,73,74,75]	[4,5,7,8,10,11,12,13,15]
Wrong fault-type detection	[50,75,76,77,78,79,80,81,82,83,84,85,86,87]	[4,5,10,11,12,13,17,22,23,24,25]
Failure in the communication-assisted trip	[61,62,75,89]	[-]
Not specified	[90,91,92,93,94,95,96,97]	[-]

“Not specified”: this information is not directly mentioned in the reference.

complement that information with particular references that mention each cause and each specific effect. Furthermore, Table 1 and Table 2 also particularize whether the cited references are documents from manufacturers or not. Similarly,

Table 3. Proposed solutions according to reviewed references.

Proposed Solution	Mentioned by
	Non-Manufacturers	Manufacturers
Avoiding the use of distance protection (e.g., using line differential protection).	[15,23,61,69,87,98,99,100]	[-]
Improving distance protection settings (e.g., inclusion of time delay for zone 1, avoiding negative-sequence directional functions, preferring undervoltage-based fault-type identification).	[-]	[4,5,7,8,10,11,12,13]
New protection algorithms (e.g., distance functions, fault-type identification, directional detection, apparent impedance calculation, adaptive functions).	[15,17,26,27,28,29,30,32,33,35,51,70,71,77,79,80,88,90]	[-]
New IBR control algorithms.	[28,45,46,47,48,49,50,63,64,65,72,84,98]	[-]
Protections based on artificial intelligence algorithms.	[42,78,79,85]	[-]

A row for “not specified” is not included in this table for the sake of simplicity.

summarizes the proposed solutions according to the reviewed references. Furthermore,

Table 4. Classification of reviewed references according to the method of identifying causes and effects and testing the proposed solutions.

Method to Identify Causes
Using Relay Fault Records	Using Simulations	Neither of These Options
[7,8,9,10,11,12,13,18,19,23,92]	[4,5,6,10,11,12,13,15,16,17,20,21,22,24,25,26,27,28,30,31,32,33,34,35,36,37,38,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,93,94,95,96,97,98,99,100]	[14,29,39,101]
Method to Identify Effects
Tests on Real Relays	Tests on Simulated Relays	Neither of these Options
[4,5,6,9,10,11,12,13,18,19,20,21,22,23,99]	[15,16,17,24,25,26,27,28,30,31,32,33,34,35,36,37,38,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,98,100]	[7,8,14,29,39,90,91,92,93,94,95,96,97,101]
Method to Test the Solutions
Real Relays	Simulated Relays	Neither of these Options
[4,5,6,9,10,11,12,13,18,19,20,21,22,23,99]	[16,17,24,25,26,27,28,30,31,32,33,34,35,36,37,38,40,41,42,43,44,45,46,47,48,49,50,51,52,61,62,63,64,65,66,67,68,69,70,71,72,73,74,76,77,78,79,80,81,82,83,84,85,86,87,88,89,98,100]	[7,8,14,15,29,39,53,54,55,56,57,58,59,60,75,90,91,92,93,94,95,96,97,101]

Some references utilize different methods; for instance, in the case of causes, references [10,11,12,13] are simultaneously in the row of relay fault records and the row of simulations.

classifies the reviewed references according to the method of identifying causes (based on real fault records or on simulations) and effects (based on testing real relays or simulated relays), and the method of testing the solutions (based on real relays or simulated relays).

8. Conclusions

A review of problems and solutions regarding the distance protection of transmission lines connected to inverter-based resources (IBRs) is shown. An introductory description about IBRs and distance protection is included because both topics are relatively complex, and their knowledge is necessary for the proper understanding of this article. An exhaustive literature review is shown for the sake of thoroughness, and special attention is paid to clearly summarize the research literature from relay manufacturers. This approach should be useful for researchers and protection engineers. The description of the reported problems is split into causes (related to IBR behavior) and effects (on distance protection). Causes and effects of problems, and the proposed solutions, are carefully extracted from the reviewed literature, and summarized in the corresponding tables. Consequently, the final information is summarized in a very organized way.