© X. Li et al., Published by EDP Sciences, 2026

1 Introduction

Traditional 10 kV distribution lines primarily utilize single-source radial networks. To improve the reliability of the power supply system, ungrounded or arc suppression coil grounding methods are widely adopted for neutral points [1,2]. With the modernization of distribution networks and the development of new power systems in China, many urban distribution networks are transitioning from overhead lines to cable lines. As a result, traditional ungrounded or arc suppression coil grounding methods are gradually being replaced by low resistance grounding techniques [3–5]. In certain regions, only dual-stage zero-sequence overcurrent protection is used for single-phase ground faults in low resistance grounding distribution networks, primarily for ease of setting and maintenance. However, this approach leads to issues such as incomplete protection configuration, poor selectivity, enlarged outage areas, and reduced power supply reliability [6].

The grounding fault protection technology of low resistance grounding systems in China has evolved from traditional time-delay protection to intelligent protection. Currently, the mainstream protection scheme still adopts time-delay zero-sequence overcurrent protection (Zero-sequence Stage III Protection) as the basic configuration [7]. To ensure selectivity, the settings need to bypass the maximum capacitance current of the line, typically set to 40 A or 60 A. However, this traditional protection scheme has limited tolerance for transient resistance, with existing protection devices only able to withstand up to 140 Ω of transient resistance, making it difficult to cope with increasingly complex high-resistance ground fault scenarios. To overcome this technical bottleneck, several innovative research and practical efforts have been initiated in the industry. Guangdong Power Grid Company pioneered the application of transient zero-sequence current grouping comparison methods for detecting high-resistance ground faults in low resistance grounding systems. This method accurately identifies faulted lines by comparing and analyzing the fault zero-sequence current characteristics of all outgoing lines, reducing the zero-sequence current working threshold to 2A and increasing the tolerance for transient resistance to 2 kΩ, significantly enhancing the detection sensitivity for high-resistance ground faults [8]. However, this method has some limitations: it requires the collection of current signals from all outgoing lines at the bus, placing high demands on the data collection system and economically challenging; additionally, the action time generally exceeds 10 s, which does not meet the requirements for rapid fault clearance. To further optimize protection performance, Guangdong Power Grid Company has piloted a voltage ratio-biased zero-sequence overcurrent protection scheme. This scheme introduces zero-sequence voltage as a blocking measure, allowing for dynamic adjustment of protection settings. In the case of low-resistance ground faults, a higher zero-sequence voltage corresponds to larger current settings, effectively preventing misoperation of non-fault lines and downstream protections; conversely, during high-resistance ground faults, the zero-sequence voltage of the faulted line approaches zero, causing the zero-sequence current setting to automatically drop to a lower threshold, enhancing the protection's tolerance for transient resistance to 1 kΩ [9]. Although this scheme requires additional zero-sequence voltage measurement devices, it significantly improves the reliability of the protection.

In addition, in the literature analysis and research of scholars at home and abroad, literature [10,11] put forward the distribution network line outlet protection and branch line protection of the rectification with the program, the selectivity of the protection action has been improved, but does not involve the problem of high-resistance grounding. Literature [8] uses inverse time zero sequence current protection to improve the sensitivity of grounding protection, and gives the principle of three-level grounding protection adjustment, but the adjustment calculation is complicated and less flexible. Literature [12–14] use bus zero sequence voltage, zero sequence current and neutral current and other frequency information for grounding protection, mainly solves the problem of high-resistance fault routing, no further fault section positioning, protection selectivity is poor. Literature [15–18] use the nonlinear characteristics of the fault for high-resistance grounding fault detection, which may fail in the case of relatively large noise or nonlinear characteristics are not obvious, and can only select the line, without forming a selective protection scheme.

This research proposes different protection schemes for various levels of transient resistance ground faults based on the complex characteristics of grounding faults in low resistance grounding systems. Conventional zero-sequence overcurrent protection methods can rapidly isolate larger fault currents. As the transient resistance increases and the fault current diminishes, a high-sensitivity grounding protection scheme is employed for outlet lines, branches, and boundary protections, which includes reduced setting values and extended operating times to achieve accurate fault isolation. For high-resistance ground faults exceeding 1500 Ω, the scheme synchronously analyzes the magnitude ratios of zero-sequence currents at the exits of each line and the neutral line to accurately identify and reliably isolate the faulted line, thereby improving fault location accuracy. This scheme provides a practical and effective multi-level protection solution for addressing grounding protection issues in low resistance grounding distribution networks.

2 Characterization of single-phase grounding faults in small resistance grounding systems

In power systems, the zero-sequence current of low-resistance grounded systems exhibits significant characteristic differences under different operating conditions. During normal operation, an unbalanced zero-sequence current of small magnitude is generated due to asymmetry in parameters such as line-to-ground capacitance. However, when a single-phase-to-ground fault occurs, the zero-sequence current in the line increases significantly. The research scenario of this paper is the typical topology of short-distance and small-capacity medium-low voltage distribution networks. In this scenario, the line impedance is small, and most transformers are of low capacity grade. The influence of their leakage reactance on fault current is within the engineering allowable error range and will not significantly change the core coordination law of multilevel protection. Therefore, the assumption of ignoring transformer and line impedances is adopted in the modeling. This phenomenon is studied in detail using a typical 10 kV low-resistance grounded system, as illustrated in Figure 1, with a focus on analyzing the relationship between the fault point's zero-sequence current and the transition resistance. The specific parameters of the system are set as follows: the neutral grounding resistor is 10 Ω, the line's per-unit positive-sequence impedance is 0.27 + j0.08 Ω/km, and the per-unit zero-sequence impedance is 2.7 + j0.348 Ω/km. To simplify the analysis, the impact of the grounding transformer reactance is neglected.

When analyzing the zero-sequence current at a single-phase ground fault point, the symmetrical component method can be employed to transform the complex asymmetric fault into a superposition of symmetrical components. To simplify calculations, the influence of line-to-ground capacitance can be neglected in the analysis since its capacitive reactance is significantly larger than the line impedance and neutral grounding resistance. In the fault analysis modeling, complete positive-, negative-, and zero-sequence networks must first be established, as shown in Figure 2, where the fault point k₁ is marked in Figure 1. The composite sequence network includes the following parameters: the equivalent impedances on the power supply side are denoted as R₁ (positive sequence), R₂ (negative sequence), and R₀ (zero sequence), while the equivalent impedances from the fault point k₁ to the busbar side are represented by X₁ (positive sequence), X₂ (negative sequence, equal to X₁), and X₀ (zero sequence). The system's rated voltage is Ė_s, and the ground fault transition resistance is R_g.

From Figure 2, the effective value of the zero sequence current 3I₀ is:

$$3I_0=\frac{3E_{\text{S}}}{\left|2Z_{\text{S}1}+2Z_{\text{L}1}+Z_{\text{S}0}+Z_{\text{L}0}+3R_{\text{g}}+3R_{\text{n}}\right|}.$$(1)

From equation (1), the relationship between the zero sequence current 3I₀ and the magnitude of the transition resistance R_g can be plotted when a ground fault occurs at the first end of the line via the transition resistance, as shown in Figure 3.

The calculation results in Figure 3 demonstrate that when a non-transition-resistance grounding fault occurs at the line head, the zero-sequence current at the fault point reaches 600 A, indicating a high fault current under metallic grounding conditions. After the fault occurs, the zero-sequence current exhibits a rapid decreasing trend as the transition resistance increases. This is because the increased transition resistance alters the distribution of fault current, imposing greater limitations on the current path through the fault point, thereby gradually attenuating the zero-sequence current.

According to NB/T 10650-2021 ⟨⟨Technical Guide for Single-Phase Grounding Fault Detection in Distribution Networks》, a high-resistance grounding fault in medium-voltage distribution networks (10 kV–35 kV) is defined as having a transition resistance exceeding 1000 Ω. As shown in Figure 3, when the transition resistance surpasses this critical value, the zero-sequence current stabilizes and no longer responds significantly to variations in line impedance [19]. This phenomenon suggests that under high transition resistance conditions, the current propagation characteristics undergo a notable change, causing the zero-sequence current to remain nearly constant and unaffected by line impedance or fault distance. Theoretical calculations using equation (1) confirm that under such conditions, the zero-sequence current drops to an extremely low level of less than 5 A.

Based on the zero-sequence network of single-phase grounding faults shown in Figure 4, this study analyzes the magnitude relationships of zero-sequence currents in various outgoing lines and the neutral line when grounding faults occur at different locations in a low-resistance grounding system. The diagram includes the following parameters: bus zero-sequence voltage (U̇₀), the negative value of pre-fault phase voltage (−U̇_f), line-to-ground capacitance of each outgoing line (C_0i, i ranging from 2 to n), line-to-ground capacitance of branch lines in Line I (C_01,j, j ranging from 2 to m), zero-sequence current at each line outlet (I_0i, i from 2 to n), zero-sequence current at branch line outlets in Line I (I_01,j, j from 2 to n), zero-sequence current at the fault point (I_0Rg), and zero-sequence current in the neutral line (I0_TN). To simplify analysis, the model neglects the influence of line impedance and distribution transformer impedance. The analysis considers two fault scenarios: when switch S is at position 1, it represents a single-phase grounding fault at point k₁ on main Line I; when switch S is at position 2, it indicates a single-phase grounding fault at point k₂ on branch line m of Line I.

When a grounding fault occurs at any location on Line I (either on the main line or branch lines), the healthy lines unaffected by the fault will generate capacitive zero-sequence currents at their terminals, with magnitudes equal to their respective line-to-ground capacitive currents:

$$3I_{0i}\left(i=2,3,\cdots ,n\right)=3\omega C_{0i}{\dot{U}}_0.$$(2)

Based on the characteristic parameter analysis of a typical 10 kV low-resistance grounding system, the line-to-ground capacitance per unit length is approximately 0.28 μF/km. Considering the structural characteristics of distribution networks with their compact radial power supply range, the line length is typically limited to within 15 km. Theoretical calculations using equation (2) show that the maximum possible line-to-ground capacitive current of the system is about 22.8A. Although this value is significantly smaller than the zero-sequence current during metallic grounding faults (typically up to 600A), being only 3.8% of it, it is notably larger compared to the zero-sequence current under high-resistance grounding conditions (typically less than 5A).

As can be seen from Figure 4, when the distribution transformer impedance is neglected, the zero-sequence current in the neutral line can be expressed in terms of the bus zero-sequence voltage and the neutral grounding resistance:

$$3{\dot{I}}_{R_{\text{n}}}=\frac{{\dot{U}}_0}{R_{\text{n}}}$$(3)

where Rn represents the neutral grounding resistance.

By comparing the magnitude of the calculated zero-sequence current in the neutral line with that in any healthy line i (i≠ 1), we obtain:

$$\left|\frac{3{\dot{I}}_{R_n}}{3{\dot{I}}_{0i}}\right|=\frac{1}{3R_{\text{n}}\omega C_{0i}}\ge 25.4$$(4)

Through phasor analysis of zero-sequence currents at different locations following ground faults in low-resistance grounding systems, several critical electrical characteristics can be identified. From a vector relationship perspective, the zero-sequence current at the faulted line outlet constitutes the inverse vector sum of two components: the zero-sequence currents from healthy lines and the neutral line. Theoretical analysis reveals a 90° phase difference between healthy line zero-sequence currents and neutral line zero-sequence currents, ensuring that the resultant vector magnitude always exceeds its individual components. Consequently, the zero-sequence current amplitude at the faulted line outlet significantly surpasses that in the neutral line. Precise calculations demonstrate that the faulted line's zero-sequence current magnitude exceeds healthy line currents by more than 20 times - a remarkably stable ratio that remains essentially unaffected by variations in transition resistance or system configuration parameters.

The formation mechanism and characteristics of zero-sequence currents in low-resistance grounding systems are fundamentally governed by system impedance properties. The decisive factor lies in the system's inherent characteristic where line-to-ground capacitive reactance substantially exceeds neutral grounding resistance. During single-phase ground faults, zero-sequence currents establish two distinct circulation paths: the primary path flows from the fault point through substation buses to the neutral grounding resistor before returning to earth; the parallel path forms through line-to-ground capacitances of healthy lines or branch circuits. This current distribution mechanism explains why the primary loop current magnitude exceeds healthy line/branch outlet zero-sequence currents by over 20 times. In contrast, during normal system operation - even under three-phase load imbalance conditions - no such dramatic amplitude disparities (exceeding 20:1 ratios) are observed among measured zero-sequence currents in either lines or the neutral conductor.

Fig. 1 10 kV low resistance grounding system diagram.

Fig. 2 Single-phase grounding fault composite sequence network diagram.

Fig. 3 The relationship between zero sequence current and transition resistance.

Fig. 4 The zero-sequence network diagram of single-phase-to-ground fault in low-Resistance grounded system.

3 Multi-level protection scheme for ground faults

In low-resistance grounding systems where ground fault currents are typically limited to relatively low magnitudes, conventional multi-level protection configurations predominantly utilize highly sensitive zero-sequence current stage III protection [20]. By implementing properly coordinated time-delay settings, selective tripping of protective devices can be achieved at different levels including outlet lines, branches, and boundaries. The relatively low fault current amplitudes make the protection scheme less demanding in terms of operating speed requirements, which provides technical feasibility for establishing a time-graded multi-level protection system. This section conducts an in-depth investigation into optimized protection configuration strategies for ground faults, with a specific focus on the distribution line employing typical low-resistance grounding as illustrated in Figure 5.

In the proposed system topology, circuit breakers QF1, QF2, and QF3 are installed at each feeder outlet of the bus, while branch lines B1 and B2 are controlled by circuit breakers QF11 and QF12 respectively, with QF13 serving as the boundary circuit breaker. To achieve precise fault zone isolation, this study develops a multi-level protection scheme featuring a three-tier configuration of feeder outlet-branch- boundary protection.

Fig. 5 Multi-level protection scheme for ground faults in low resistance grounding systems.

3.1 Outlet protection

According to the principles of GB/T 14285-2023 «Technical Regulations for Relay Protection and Security Automatic Equipment》, the stage III zero-sequence current protection configured for line outlet circuit breakers must meet selectivity requirements to ensure localized and rapid fault clearance while preventing protection maloperation. The setting value should be higher than the maximum possible capacitive current of the line to avoid maloperation caused by system three-phase unbalance. Theoretical analysis results in Section 1 show that the line-to-ground capacitive current in typical 10 kV low-resistance grounding systems generally does not exceed 22.8A. In engineering practice, a reliability factor of 1.3 is adopted for setting calculations [20]. Therefore, when using zero-sequence current transformers as detection devices, the line protection setting value can be uniformly set at 30A.

In power systems, the operating time delay of line protection needs to consider multiple factors. Firstly, its operating time must coordinate with both the upstream distribution transformer protection and maintain a time difference Δt with downstream branch line protections. The time delay setting should neither be too long nor too short, typically set between 2 to 4 s to ensure thermal stability of electrical equipment [21]. When determining time coordination between protection levels, factors such as the inherent operating time of circuit breakers and arc extinction time must also be considered. Generally, the time difference between protection devices is set at 0.3 s to ensure rapid and localized fault isolation when faults occur.

3.2 Branch and boundary protection

For branch line protection settings, the maximum downstream line-to-ground capacitive current must be comprehensively considered. Based on the distribution network topology, the supply range of branch lines is typically limited to within 5 km. When zero-sequence current transformers are employed as measurement elements, theoretical calculations combined with engineering experience suggest a unified setting value of 10A [22]. This value ensures both the sensitivity requirements of protection devices and effectively prevents maloperation. Regarding time coordination, the operating time delay of branch line protection needs to coordinate with both upstream and downstream protections. Compared to the downstream boundary protection devices, a standard time difference Δt should be added to ensure protection selectivity; while compared to adjacent upstream line protections, its operating time delay should be reduced by one time difference Δt.

The setting value for boundary line protection can reference the branch line protection configuration, i.e., adopting the same setting value of 10A. This selection is based on the following considerations: firstly, the line-to-ground capacitive current characteristics downstream of boundary switches are fundamentally consistent with branch lines; secondly, identical setting values help simplify protection configuration schemes and improve system operation and maintenance efficiency. For time delay settings, the operating time of boundary protection should be reduced by one standard time difference Δt (0.3 s) compared to branch line protection. This time coordination strategy not only ensures the selectivity requirements of protection device operation but also maximizes the reduction of fault clearance time, effectively enhancing the system's fault isolation speed.

4 High sensitivity ground fault protection

An in-depth analysis of fault characteristics in typical 10 kV low-resistance grounding systems reveals that conventional multi-level ground fault protection methods exhibit significant limitations. Specifically, when the zero-sequence current protection setting at line outlets is configured at 30A with a sensitivity factor of 2, the protection device may fail to operate for ground faults with transition resistance exceeding 86Ω, necessitating improvements to existing multi-level ground fault protection schemes.

Regarding optimization of ground protection device performance: while reducing protection settings can enhance device sensitivity to some extent, this approach simultaneously compromises the selectivity of protection at all levels. This occurs because excessively low setting values may cause protection devices to respond to faults in adjacent zones, leading to maloperation. To resolve this contradiction, the protection selectivity can be ensured by appropriately extending the protection operation time delay. This graded time-delay protection strategy ensures that the protection device closest to the fault point operates first, thereby significantly improving protection reliability.

For single-phase ground faults in 10 kV low-resistance grounding systems, if the fault current amplitude falls below conventional protection settings, the protection devices will be unable to accurately detect and operate. As demonstrated by the theoretical analysis in Section 1, under such fault conditions the maximum capacitive current in the zero-sequence main loop does not exceed 30A, while in non-fault zones (i.e., healthy lines or branches), the zero-sequence current is generally below 1.5A. To prevent maloperation of protection devices in non-fault zones, a reliability factor Krel is introduced in setting calculations, from which the lower limit of protection settings should be greater than:

$$I_{\text{set}}=K_{\text{rel}}{I}_{max}=1.3\times 1.5=1.95A.$$(5)

The protection setting value also needs to take into account the maximum unbalance current of the system under normal operating conditions and the measurement error of the protection device. The maximum unbalance current generated by the system mainly comes from the zero sequence current caused by the asymmetry of the line parameters. Measured data show that the 10 kV small resistance grounding system in the cable line and overhead line in normal operation of the maximum zero sequence current were 0.26A and 0.37A. Considering the normal operation of zero sequence current transformer has 0.5% measurement error, and 10 kV small resistance grounding system commonly used in the current transformer can measure the maximum value of 600A, that is, the allowable size of the protection error is:

$$I_{\text{set}}=600\times 0.5\%=3A.$$(6)

After comprehensively evaluating the aforementioned technical factors, the setting value for high-sensitivity ground protection is determined to be 3A. This value not only satisfies all technical requirements but also provides appropriate margin for potential system disturbances.

When a system fault triggers the high-sensitivity ground fault protection, an additional time delay Δt should be incorporated into the existing protection operating time to prevent maloperation in non-fault zones. This protection setting strategy - combining reduced setting values with extended operating time - effectively enhances protection sensitivity while simultaneously ensuring system operational stability and reliability.

5 High resistance ground fault routing

The theoretical analysis in Section 1 demonstrates that when the transition resistance of a single-phase ground fault exceeds 1000Ω, the system's zero-sequence current drops below 5A. As the transition resistance surpasses 1500Ω, the zero-sequence current further decreases to an extremely low level of less than 3A, at which point the high-sensitivity ground protection devices may fail to operate. However, even under such high-resistance fault conditions, significant characteristic differences persist in the zero-sequence current amplitudes between the faulted line, healthy lines, and the neutral conductor. Based on these distinctive characteristics, we propose a novel fault line selection scheme employing transverse comparison, with the detailed selection flowchart presented in Figure 6.

The specific implementation steps are as follows:

• Data Acquisition and Activation Criteria:

  The detection devices collect zero-sequence current (I_0i) from each feeder and neutral line zero-sequence current (I_0Rn). The activation threshold is set at the maximum unbalanced current value (0.5A). When collected zero-sequence currents exceed this threshold, the fault line selection procedure is triggered.

• Fault Type Identification Algorithm:

  • Bus Fault Identification: Calculate the amplitude ratio between neutral line zero-sequence current and each feeder's zero-sequence current. A bus fault is confirmed when the neutral current exceeds all feeder currents by more than 20 times.

  • Line Fault Identification: After excluding bus faults, calculate amplitude ratios between feeder zero-sequence currents. A line fault is identified when one feeder's current (I_0i) exceeds other feeders' currents (I_0j) by more than 20 times.

This scheme employs a low activation threshold design strategy, ensuring high protection sensitivity. The selection accuracy remains stable even under severe three-phase load imbalance conditions. Particularly for high-resistance ground faults, the relative ratio-based criterion not only enhances fault identification sensitivity but also effectively prevents both protection failure to operate and maloperation issues.

Fig. 6 Fault line selection flow chart.

6 Discussion

6.1 Simulation parameters

The experimental simulation model of 10 kV small resistance grounding system constructed based on Figure 5 is shown in Figure 7. In order to make the simulation closer to the engineering reality, the neutral grounding resistance RN is set to 10Ω, and the model contains a bus and three outlet lines (the lengths of lines I, II, and III are 10 km, 6 km, and 8 km, respectively), as well as two branch lines (B₁ is 2.5 km and B₂ is 2 km). The line parameters are configured as follows: positive sequence resistance is 0.27 Ω/km, capacitance is 3.395 × 10⁻⁷ F/km, inductance is 0.255 × 10⁻³ H/km, and the negative sequence parameters are the same as those of the positive sequence; zero sequence resistance is 2.7 Ω/km, inductance is 1.109 × 10⁻³ H/km, capacitance is 2.8 × 10⁻⁷ F/km.

Fig. 7 Improved ground fault protection configuration diagram.

6.2 Analysis of simulation results

According to the ground fault multilevel protection scheme in this paper, conventional grounding protection (zero sequence current Ⅲ protection) and high sensitivity ground fault protection are configured at the line exit circuit breaker, branch line circuit breaker and boundary circuit breaker respectively, and the difference between the upper and lower levels is set to Δt = 0.3 s, and Table 1 shows the specific protection calibration scheme.

In the simulation, single-phase ground faults were set at three locations on Line I: main line k₁, branch line k₂, and boundary line k₃, with transition resistances of 10Ω, 100Ω, and 1500Ω respectively. Table 2 presents the detected zero-sequence currents (3I₀) at various protection installation points under different fault locations and transition resistance conditions.

The simulation data analysis reveals that for ground faults with 10Ω transition resistance occurring at different sections (k₁, k₂, k₃) of Line I, all conventional ground fault protection devices - including feeder circuit breaker QF₁, branch circuit breaker QF₁₁, and boundary circuit breaker QF₁₃ - operate reliably and achieve selective isolation of faulted sections. This observation validates the effectiveness of conventional ground fault protection in low-resistance fault scenarios. When the fault occurs at k1 with increased transition resistance (100Ω), the zero-sequence current amplitude decreases significantly to 25.6A, falling below the 30A threshold and causing conventional protection failure. In this case, the fault zero-sequence current (exceeding 3A) triggers the high-sensitivity ground protection, which achieves selective tripping through a 4.5 s time-delay strategy. For ultra-high resistance faults (1500Ω), the system zero-sequence current drops below 3A, exceeding neither the conventional protection threshold (30A) nor the high-sensitivity protection threshold (3A). To address such high-resistance faults, an intelligent line selection scheme based on transverse comparison of zero-sequence currents can be employed to isolate the faulted line.

The ratio of line Ⅰ outlet zero sequence current to line Ⅱ outlet zero sequence current is:

$$K_{\text{I}/\text{II}}=\frac{2.14}{0.03}\approx 71$$(7)

The ratio of line Ⅰ outlet zero sequence current to line Ⅲ outlet zero sequence current is:

$$K_{\text{I}/\text{III}}=\frac{2.14}{0.06}\approx 36$$(8)

From equations (7) and (8), it can be seen that when a high-resistance ground fault occurs at k1 (transition resistance of 1500 Ω), it is found that the amplitude of the zero sequence current at the exit of line Ⅰ, Ⅱ and Ⅲ is significantly greater than that at the exit of line Ⅰ (more than 20 times), according to which it can be accurately determined that a ground fault has occurred in line Ⅰ, and the time-delayed protection strategy is finally realized through the Fault removal. This result fully verifies the effectiveness of the proposed fault routing method in the case of ultra-high resistance ground fault.

6.3 Comparative analysis with existing methods

According to DL/T 584-2017«3 kV ∼ 110 kV power grid relay protection device operation and adjustment regulations》 in the provisions of the small resistance grounding system grounding fault using zero sequence current protection, and its adjustment value to avoid the unbalance current and single-phase grounding fault capacitance current during normal operation of the system. However, in the high resistance grounding fault will be due to the fault current is lower than the existing protection rectification value and make it refused to act. For this reason, this paper proposes a protection strategy that reduces the protection setting value and extends the protection action time, and on the basis of retaining the original setting value, a high sensitivity ground fault protection is added to form a dual multilevel protection system. For high resistance ground faults exceeding 1500Ω, the ground fault routing strategy based on the amplitude ratio of the zero sequence currents of the outgoing and neutral lines is adopted, which effectively solves the problem of protection blindness of the existing multilevel protection scheme in various transition resistance size ground fault scenarios.

7 Conclusion

This study proposes a multi-level protection scheme for grounding faults in small resistance grounding systems. Adopting the well-configured multi-level zero sequence overcurrent protection scheme, it can quickly and selectively remove the grounding faults with large fault currents, and will not expand the scope of outage by overstep tripping; adopting the time-delayed, low fixed-value and highly sensitive grounding protection to selectively remove the grounding faults with small fault currents; and for the high-resistance grounding faults with the transition resistance as high as 1500 Ω and above, the magnitude of zero sequence currents of the line outlets and the neutral line can be compared horizontally to correctly select the faulted line by using the zero sequence current amplitude of the line outlets and neutral line. For high resistance grounding faults with transition resistance up to 1500Ω and above, the amplitude of zero sequence current of each line outlet and neutral line is utilized to correctly select the faulted line by horizontal comparison, and the fault is removed with delay. This protection scheme makes up for the problems of poor selectivity, low sensitivity and insufficient detection capability of high resistance grounding, which is of great significance for improving the reliability of power supply and safeguarding equipment and public safety. There are certain limitations in this study; in the future, we can further optimize the protection criteria under complex working conditions and integrate intelligent technologies to improve the intelligence level of the system. At the same time, the actual implementation of the scheme faces challenges such as large on-site commissioning workload, insufficient compatibility of old equipment, and communication anti-interference. In the follow-up, we will conduct targeted optimization and improvement to promote the engineering implementation of the scheme and ensure the safe and stable operation of the system.

Funding

This research is supported by: China Southern Power Grid Corporation, No. GZKJXM20232476.

Conflicts of interest

The authors declare that they have no conflicts of interest.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Author contribution statement

Xinhao Li, Xiaobing Xiao, and Xiaomeng He are responsible for designing the framework, analyzing the performance, validating the results, and writing the article. Feng Wang, Wei Du, Wei Huang, and Xinyi He are responsible for collecting the information required for the framework, providing software, critical review, and administering the process.

References

All Tables

All Figures

	Fig. 1 10 kV low resistance grounding system diagram.
In the text

	Fig. 2 Single-phase grounding fault composite sequence network diagram.
In the text

	Fig. 3 The relationship between zero sequence current and transition resistance.
In the text

	Fig. 4 The zero-sequence network diagram of single-phase-to-ground fault in low-Resistance grounded system.
In the text

	Fig. 5 Multi-level protection scheme for ground faults in low resistance grounding systems.
In the text

	Fig. 6 Fault line selection flow chart.
In the text

	Fig. 7 Improved ground fault protection configuration diagram.
In the text