Research on Arc Discharge Characteristics of 10 kV Distribution Line Tree Line

Abstract

Many studies have investigated tree-contact arcing ground faults. However, the effects of branch moisture content and wind speed are still not fully understood. Therefore, this paper addresses the wildfire risk caused by tree-contact arc grounding faults in distribution networks. A 10 kV distribution-line tree-contact arcing fault test platform is built. A two-dimensional multi-physics plasma model is also developed based on magnetohydrodynamics. Experiments and simulations are combined. The effects of wind speed, branch moisture content, and conductor type on arc evolution and fault characteristics are systematically studied. The results show that higher wind speed causes stronger arc-column deformation. The fault current contains more high-frequency components and sharp spikes. At 9 m/s and 16 m/s, the fault current shows strong disturbances and much lower stability. Higher moisture content increases the branch conductivity indirectly. It strengthens the carbonized conductive path and helps sustain stable arcing. For the high-moisture sample (64%), the current waveform is smooth, and its amplitude increases monotonically with fault development. For the low-moisture sample (30%), the current amplitude decreases, and spikes become more frequent. The arc tends to extinguish and reignite repeatedly, which indicates an unstable discharge process. The simulations further reveal the coupling between the arc-root temperature field and the airflow field under different wind speeds and conductivities. They also show clear differences in temperature evolution between bare conductors and insulated conductors. These findings provide experimental evidence and simulation support for identifying wildfires initiated by tree-contact arcing faults.

1. Introduction

In recent years, with the expansion of the distribution network and the implementation of the ecological strategy of ‘carbon neutrality and carbon peak’, the forest coverage rate has increased, resulting in an increase in the probability of trees touching wires. The current of the tree branch near or in contact with the wire arc discharge fault of the tree branch is very small, and the characteristics are not obvious, which makes it difficult for the protection device to respond in time and accurately. At the same time, it is also one of the most difficult faults to be detected by the patrol personnel [1,2]. China’s distribution lines are generally set up at a height of 6 to 10 m. If the surrounding trees grow rapidly, they are easy to approach or exceed the height of the wire. The branches may contact or overlap the wire under the action of wind, which may cause a transient ground fault [3]. A distribution network fault igniting a mountain fire has a high contingency. Its concealment and suddenness, and its single fire area are larger than those of other fire sources [4]. In the actual fault detection, the tree line fault is often misjudged as a disturbance and ignored because of its unobvious characteristics, resulting in frequent faults. If there are flammable materials such as trees around, it may lead to carbonization and combustion of branches, which can easily cause fire accidents and greatly threaten the safety of production and life [5,6].

For the study of arc faults in distribution networks, many researchers at home and abroad have carried out magnetohydrodynamic (MHD) arc simulations using finite-element methods. Helence Rachard, for example, developed a two-dimensional arc model to analyze the influence of an external magnetic field on arc shape and motion [7], although their model was fully enclosed. Li Xingwen et al. established a two-dimensional MHD mathematical model of air-circuit-breaker arc plasma [8], and used simulations to investigate the relationships among arc radius, electric field strength, and current, as well as the effect of an external magnetic field on arc motion. By means of MHD finite-element simulation models, the dynamic evolution of arcs under different influencing factors can be investigated and analyzed more intuitively. Meanwhile, extensive experimental studies have also been conducted worldwide on the phenomenon of transmission and distribution lines contacting trees and triggering wildfires. The research shows that with the continuous discharge of the tree line, there is a risk that the vegetation will be ignited under the action of arc and leakage current. Especially when the fault occurs in remote mountainous areas, and the protection system fails to respond effectively, it will easily lead to wildfires [9,10]. Yin et al. studied the influence of water content on the leakage current of touching trees, revealed the peak variation law, and established a prediction model [11]. Liang et al. established a time-varying model of single-phase contact tree fault resistance for wire contact tree faults [12]. Jiang et al. analyzed the arc characteristics and combustion behavior of tree-wire discharge faults, with a primary focus on exploring the stage-wise variation in leakage current and harmonic signatures [3]. Although some studies have addressed tree–line contact and flame bridging, the understanding is still incomplete. The discharge morphology, fault characteristics, and ignition mechanism under different wind speeds, branch moisture contents, and conductor types are not fully clarified. In addition, the periodic physical processes of the fault are rarely studied. This limits effective monitoring and early warning of wildfire-related faults.

In this paper, the arc development process of tree-contact grounding faults is investigated by establishing a two-dimensional multi-physics model of the tree-contact arc based on magnetohydrodynamics and constructing a 10 kV distribution-line tree-contact arcing fault test platform. The characteristic behavior of tree-contact arc faults is analyzed, and the extracted fault features provide support for the identification of wildfires initiated by tree-line discharges.

2. Construction of the Tree-Line Arc Simulation Model

In this work, the tree–line discharge is described using a two-dimensional multiphysics model, as shown in Figure 1. To improve numerical stability and reduce computational cost, the geometry is simplified to a cross-section. In this section, the branch and the conductor are orthogonal. A plasma arc column is assumed to be already formed at the start of the simulation. The bare conductor is modeled as a stranded aluminum wire with a cross-sectional area of 95 mm², which is common in 10 kV distribution lines. The covered conductor is modeled as an insulated wire with a cross-sectional area of 120 mm². The branch diameter is set to 15 mm, consistent with the experimental samples. The branch and the conductor serve as the two electrodes of the arc.

A large surrounding air region is included. Open boundary conditions are applied at the outer boundaries. This reduces boundary effects on the flow around the branch, electrodes, and plasma column. It also improves convergence and better reflects real conditions. The domain is meshed with a free triangular grid. The minimum and maximum element sizes are 0.0022 mm and 15.4 mm, respectively.

In the model, the tree branch and the conductor are the two electrodes of the arc, and the materials that are mainly considered in the simulation include the density, electrical conductivity, specific heat capacity, and thermal conductivity. The specific parameter setting is listed in

Table 1. Simulation parameters of tree branch and conductor.

Parameter	Tree Branch	Conductor
electrical conductivity (S/m)	0.08~0.18	3.78 × 107
density (kg/m3)	500	2700
specific heat (J/(kg·K))	1725	908
thermal conductivity (W/(m·K))	0.22	237

.

The dynamic change in the arc plasma is the result of the joint effect of several physical fields, such as the electric field, magnetic field, flow field, and temperature field, and thus it shows very complicated physical behavior. In order to simply the model reasonably, the following basic assumptions were made [3]:

(1)

The very early igniting stage of the arc is left out. Simulation begins with an existing plasma column between the branch and conductor and considers it to be in a state of local thermodynamic equilibrium.

(2)

In addition to these species, additional species such as metal vapor, carbonaceous vapor, and similar species are not included, and they are not considered in the influence of transport and radiation properties.

(3)

The motion of the arc plasma is assumed to be laminar and not turbulent.

(4)

The arc is regarded as optically thin, meaning that self-absorption of radiation is small compared with the total radiative loss.

(5)

The electromagnetic force and magnetic pressure generated by the self-induced magnetic field are much smaller than the gas pressure and buoyancy, so their influence on the arc shape and flow field is negligible. Therefore, the self-induced magnetic field of the arc is ignored.

Under these assumptions, and within the framework of magnetohydrodynamics, the present work solves a set of coupled governing equations consisting of the electromagnetic field equations together with the conservation equations for mass, momentum, and energy.

(1) Electromagnetic field equation

$$\nabla \cdot \left(-\sigma \nabla \phi \right)=0$$

(1)

$$\mathit{E}=-\nabla \phi$$

(2)

$$J=\sigma \mathit{E}$$

(3)

$$\nabla \times (\nabla \times \mathit{A})=-{\mu }_0\mathit{J}$$

(4)

$$\mathit{B}=\nabla \times \mathit{A}$$

(5)

For the equation presented, σ denotes the electrical conductivity, φ represents the electric potential, A stands for the magnetic vector potential, E signifies the electric field strength, μ₀ represents the magnetic permeability of vacuum, and J refers to the current density.

(2) Mass conservation equation

$${\displaystyle \frac{\partial \rho }{\partial t}}+\nabla ·\left(\rho v\right)=0$$

(6)

For the equation presented, ρ denotes the density, v represents the flow velocity, and t stands for the time.

(3) Momentum conservation equation

$$\mathit{F}=\mathit{J}\times \mathit{B}$$

(7)

$$\rho {\displaystyle \frac{\partial v}{\partial t}}+\rho \left(v·\nabla \right)v=\nabla ·\left[-\rho \mathit{I}+\mu \left((\nabla v+{\left(\nabla v\right)}^T)\right)\right]+F$$

(8)

For the equation presented, I denotes the unit tensor, μ represents the dynamic viscosity of the plasma, p stands for the pressure, J refers to the current density, and B signifies the magnetic flux density.

(4) Energy conservation equation

$$\rho c_p{\displaystyle \frac{\partial T}{\partial t}}+\rho c_{p}v·\nabla T=\nabla \left(k\nabla T\right)+Q$$

(9)

$$Q={\displaystyle \frac{\partial }{\partial t}}\left({\displaystyle \frac{5k_{B}T}{2q}}\right)\left(\nabla T·J\right)+E·J+Q_{rad}$$

(10)

$$Q_{rad}=4\pi {\epsilon }_n$$

(11)

It has c_p for specific heat at constant pressure, T for temperature, k for thermal conductivity, k_B for the Boltzmann constant, Q for electric charge, and ε_n for net volumetric emission.

In the established arc plasma model, the conductor is specified as the anode, and the bottom boundary is grounded with the electric potential set to 0. The tree branch is defined as the cathode. The boundaries of the surrounding air domain are treated as open boundaries, where the left side is specified as the inlet for natural wind and the right side as the outlet. The gas pressure is set to one standard atmosphere, and the temperature to 293.15 K.

In the simulation, both the bare conductor and the insulated conductor are selected as commonly used types for 10 kV distribution lines. The simulation parameter comparison is shown in

Table 2. Comparison of simulation parameters between the bare conductor and the insulated conductor.

	Bare Conductor	Insulated Conductor
radius	5.88 mm	5.8 mm
insulation thickness	0 mm	1 mm
cross-sectional area	95 mm2	120 mm2

below.

3. Simulation Results and Analysis of the Tree-Line Arc Discharge

3.1. Variation in Arc Morphology Under Different Wind Speeds

To elucidate the evolution of arc morphology, the flow-field distributions under different wind speeds are analyzed, as shown in Figure 2.

From Figure 2, it can be seen that the maximum temperature at the arc root is mainly located in regions with relatively low flow velocity. Due to the flow characteristics of a cylinder in cross-flow, a distinct low-velocity zone appears on the leeward side of the conductor, where the arc-column temperature is higher than that in the high-velocity region. This indicates that the temperature distribution in the arc-column region is closely related to the local flow-velocity distribution in the vicinity of the tree-contact arc discharge.

Because the conductor has a circular cross-section, it strongly affects the airflow in the gap between the conductor and the branch. When wind passes the curved surface, the local velocity increases due to aerodynamic effects. The curvature can also cause flow separation. This creates vortices and a low-pressure region on the leeward side. At the same time, an accelerated flow forms on the windward side. As a result, the flow speed in the arc region changes. This may influence the arc shape and its stability. Therefore, we compare the external wind speed with the local flow velocity in the arc region. The results are summarized in Table 2.

As shown in

Table 3. Relationship between wind speed and flow velocity in the arc region.

Wine Speed (m/s)	Maximum Flow Velocity in the Arc Region(m/s)	Percentage Increase
1	1.51	151%
3	4.82	161%
5	10.8	216%
10	23.4	234%

, when the incoming wind speed is 1, 3, 5, and 10 m/s, the maximum flow velocity in the arc gap reaches 1.51, 4.82, 10.8, and 23.4 m/s, corresponding to increase rates of 151%, 161%, 216%, and 234%, respectively. Overall, the local flow velocity is consistently higher than the incoming wind speed, and its amplification becomes more pronounced as the wind speed increases, indicating a significant local acceleration effect in the gap between the conductor and the tree branch.

3.2. Effect of Different Moisture Contents on Arc Temperature

Arc discharge is produced when current passes through ionized gas in air. Tree branches with higher moisture content can more readily sustain arc discharge and therefore are more likely to initiate arcing under contact fault conditions. According to the relationship between ion concentration and migration rate [], the electrical conductivity σ of vegetation can be expressed by (12):

σ = VμgF

(12)

where V is the concentration of the electrolyte or salt, μ is the ionic mobility, g is the degree of dissociation of the conductive ions, and F is the Faraday constant.

According to the above equation, a reduction in moisture content leads to decreases in μ and g, thereby lowering the electrical conductivity σ of vegetation. In the simulations, the effect of moisture content on tree-contact arcing for the same tree species is indirectly represented by varying the electrical conductivity at the branch end. Typical conductivity values of fresh branches, i.e., 10⁻³ S/m, 10⁻² S/m, and 10⁻¹ S/m, are selected to simulate different moisture-content gradients and to investigate their influence on arc temperature. The simulation results are shown in Figure 3.

As shown in Figure 3, increasing the branch moisture content indirectly raises its electrical conductivity, which makes the arc easier to sustain and allows it to expand toward both electrodes. The temperature in the arc-root region then rises rapidly, accelerating carbonization and further strengthening the conductive path. Although water vaporization is an endothermic process that can locally reduce the temperature in the short term, under sustained discharge, the enhanced conductivity and increased energy input dominate, leading to an overall increase in arc temperature.

3.3. Analysis of the Influence of Conductor Type on Arc Characteristics

In this study, numerical simulations are carried out to analyze tree-contact arc discharges involving insulated conductors, and the corresponding isothermal contour distributions of the arc are shown in Figure 4.

As shown in Figure 4, when the arc is extinguished, a discontinuity appears in the arc segment on the branch side, because the reduced temperature increases the arc resistance and thereby breaks the current path. Distinct isotherms can also be observed in the insulation layer at the end of the insulated conductor and near the branch side, indicating that these regions are influenced by the arc and that heat accumulation has left ablation traces. In the following, we compare the simulation results for tree-contact arc discharges on damaged insulated conductors and on bare conductors, and the corresponding arc-temperature profiles are presented in Figure 5.

From Figure 5, we can see that the development trend of the arc discharge between insulated conductors and branches is similar to that of bare conductors, and it is also harder for arc re-ignition to occur in the insulated conductors. When an arc fault occurs on an insulated conductor, the arc plasma area is more significantly influenced by wind speed. Due to the existence of the insulation layer, compared with the bare conductor, it is slow under no wind or low wind conditions, which makes the temperature of the arc root on the conductor side and branch side higher. However, under high wind speeds, the insulation layer extends the duration of the arc, and thus the heat accumulated in the arc plasma region increases less rapidly.

4. Tree Wire Discharge Arc Grounding Fault Test Platform

4.1. Test Platform and Equipment

In this test, a 10 kV distribution line tree discharge arc grounding fault test platform was built. The HRYDJW test transformer made by China Wuhan Huarui Yuanda Power Equipment Co., Ltd. was used during the test, and the rated capacity was 50 kVA. The oscilloscope used in the test is a SIGLENT SDS5034X oscilloscope made by China Shenzhen Dingyang Technology Co., Ltd., and the highest sampling rate is 5 GSa/s.

The schematic diagram of the simulation experiment platform is shown in Figure 6 and Figure 7. The transformer is linked to the steel-cored aluminum strand, and voltage is measured with a capacitive voltage divider. The oscilloscope is collecting the fault current signal as well as the leakage current through a parallel 10 Ω non-inductive sampling resistor. The leakage current flowing through the vegetation is obtained from the voltage across the resistor. The scope to collect leakage current with a sample frequency of 12.5 kHz and a sample frequency of 125 MHz to investigate the fault waveforms.

The FLUKE infrared thermometer shown in Figure 8 is used to monitor the temperature of the end of the branch and the side of the contact wire of the branch. The temperature measurement range is −40 °C~500 °C. The wind speed is measured by the force anemometer shown in Figure 9, and the simulation test is carried out. The thermometer is compared with a calibrated reference thermometer in a constant-temperature water bath to correct its zero point and scale. The anemometer is compared with a calibrated reference anemometer in a standard air-flow field to adjust its calibration coefficient. Consistency across multiple experiments is ensured by strictly controlling the boundary conditions and operating procedures. All experiments in this study were conducted in a high-voltage laboratory, where the influences of environmental temperature, relative humidity, and background airflow on the tests were negligible.

4.2. Selection and Treatment of Test Branches

The equivalent resistance at the contact interface between branches and wires accounts for approximately 90% of the total equivalent resistance of the tree, and this can essentially reflect the discharge ignition performance of vegetation under actual fault conditions [10]. The eucalyptus that was used as the sample is a common tree in the forest area, and the collected branch specimen is about 80 cm long and 1.5 cm in diameter. This study adopts the relative moisture content to determine the moisture level of branches. Firstly, some branches of the fresh branches used in the experiment are intercepted to measure the weight, and then the test branch samples are dried in the blast drying box. Drying at 60 °C in the blast drying box until the quality of the branch samples does not change. Finally, the weight of the branches of the dried moisture is measured. The moisture content of the test branches can be measured by comparison [9], and the calculation formula is shown in Formula (13).

$$W={\displaystyle \frac{M-M_G}{M}}\times 100\%$$

(13)

In the equation, W denotes vegetation relative water content; M refers to branch fresh weight, and M_G represents branch dry weight.

5. Test Results and Analysis

5.1. Analysis of Tree Wire Arc Discharge Test Phenomenon

During the test, the increase in the sag of the simulated wire is connected horizontally and vertically to the branch sample. The relative water content of the branch sample is controlled at about 50%, the tree line spacing is set to 0.6 cm, and the ambient temperature is maintained at about 25 °C. The number of repeated experiments under the same conditions was not less than 10 times. Combining multiple sets of tests, the whole ignition process of the eucalyptus wire arc discharge test is shown in Figure 10.

According to the experimental observations, the ignition process can be roughly divided into four stages: initial arc generation, carbonization-channel development, water-vapor loss, and arc extinction. According to Figure 4, in the initial stage of the fault, a weak discharge appears between the branch and the conductor, accompanied by transient luminescence, electric sparks, and a sharp sound. As the fault develops, a stable arc channel is formed between them. The arc column gradually thickens and chars the branch, and the surface carbonizes to form a carbonized channel, which enhances conductivity; as a result, the arc brightness and discharge intensity increase, and the insulation capability of the vegetation continues to decline. With increasing discharge duration and expansion of the conducting area, a large amount of internal water vapor is lost, the branch temperature rises rapidly, and moisture evaporates along the carbonized channel, accompanied by a piercing squeal, while flame combustion intensifies. When the moisture content drops to a low level, and the carbonized channel extends to its limit, the highly unstable arc is finally extinguished when the branch is instantaneously broken down.

5.2. Tree-Line Arc Discharge Fault Tree Branch Temperature Monitoring

Due to the limited temperature measurement range of the infrared thermometer, it is impossible to accurately monitor the high-temperature arc of the tree branch. Therefore, this paper mainly monitors the temperature of the tree branch and the wire discharge poles. The temperature change during the fault development is shown in Figure 10.

It can be seen from Figure 11a that the fault point will heat up rapidly in a short time, the temperature near the fault point is high and accompanied by Mars scintillation, and the temperature at the end of the branch grounding also gradually increases, which is significantly higher than the temperature in the middle of the branch. The arc fault discharge occurred about 20 s, the overall temperature of the branch reached 100 °C, and a large amount of water vapor began to evaporate. After 30 s, with the development of the carbonization channel, the overall temperature of the branch gradually increased, and the main high-temperature area was concentrated near the carbonization channel, which could reach hundreds of degrees Celsius. As shown in Figure 11b, when the arc fault stops, the maximum temperature of the branch can reach 179.1 °C, which is closely related to the type and moisture content of the branch. As shown in Figure 11c, during the development of the carbonization channel, when the temperature of the fault point rises to about 250 °C, the branches begin to burn, and the temperature can reach 469.3 °C.

5.3. Effect of Wind Speed on Tree-Wire Arc Discharge

The wind has a considerable impact on the arc discharge of the tree line. High wind speeds will lead to branches burning and fire spreading quickly. As shown in Figure 12, the increase in wind speed will aggravate the arc deformation, such as extension, distortion, or splitting. The movement of electrons and ions in the plasma channel driven by wind may lead to the expansion and splitting of the arc channel, forming multiple interlaced channels, increasing the heat dissipation area, reducing the temperature, and affecting the arc stability.

In this paper, the wind speed is controlled by adjusting the gear of the fan and the distance between the air outlet and the fault point. The wind speed is adjusted to 1 m/s, 4 m/s, 9 m/s, and 16 m/s to simulate the strength of natural wind. The simulation test is carried out to explore the influence of wind speed on the leakage current in the process of tree-line arc discharge. The waveform characteristics of tree-line arc discharge leakage current with different wind speeds are shown in Figure 13.

Due to the wind effect, the current waveform shows spikes and deformation to different degrees. At wind speeds of 1 m/s and 4 m/s, the current amplitude is about 0.8 A. At 9 m/s, it increases to about 2 A. At 16 m/s, it reaches about 5 A. As wind speed increases, more high-frequency components appear in the waveform. This is especially clear at 9 m/s and 16 m/s. Sharp disturbance pulses and spikes become much more frequent, which indicates increased current instability. Besides the pulses and spikes, a distorted peak may appear near the “zero-crossing” under wind. This arc-related peak is associated with residual plasma between the electrodes. The arc channel contains hot and highly ionized gas. Under stronger wind, the ionized gas around the arc is more easily blown away. This makes the plasma density more non-uniform. Fluctuations in the ionized gas density can cause the arc resistance to change rapidly. As a result, the current becomes unstable, and peaks appear in the waveform. Stronger wind intensifies these processes. The peak amplitude becomes larger, and the waveform instability rises markedly.

5.4. The Influence of Water Content on Fault Current Waveform

Moisture content belongs to the category of influencing factors of tree wire arc discharge. The paper investigates the influence of moisture content on the tree wire discharge. In the process of testing the development of tree line discharge fault with water content, it is found that the development process of tree line arc discharge is obviously different when vegetation with different water content occurs. In this paper, the leakage current and fault waveform are analyzed and studied.

For the branches with 64% high water content, the specific fault leakage current trend and waveform characteristics are shown in Figure 14.

From Figure 14, it can be seen that the overall waveform is a power-frequency AC signal, approximately sinusoidal, with a basically stable frequency. For the branch with 60% moisture content, the leakage current amplitude is about 1.25–2 A and remains relatively stable. In the period of 11.0–11.4 s, the current amplitude is small, and the waveform is relatively smooth and symmetrical. By 16.0–16.4 s, as the carbonization channel develops, the current amplitude increases significantly, and the waveform becomes slightly distorted. During 30.0–30.4 s, the current is nearly saturated, with the largest and stable amplitude, and small spikes appear locally, indicating a stable discharge state after the formation of the carbonization channel. In the process of arc discharge fault of the branch tree line with 64% high water content, the carbonization channel will gradually show an upward trend of current amplitude with the continuous and stable development of the fault. This is because the high moisture content of the branches in the arc discharge process causes the water gradually evaporate, forming a carbonization channel, which makes the arc more stable. In this case, the fault current waveform shows the characteristics of smooth and no burr. This is because the carbonization channel of the high moisture content branch can effectively guide the arc, so that the current waveform will not exhibit obvious distortion during the development process. In addition, the current change at ‘zero’ is not a flat straight line, but a curve similar to the current waveform development trend. This shows that in the process of fault, the change trend of current is continuous and smooth, and there is no obvious waveform distortion.

For the branches with a low moisture content of 30%, the trend and characteristic waveform of fault leakage current are shown in Figure 15.

From the figure, the overall current still appears as a power-frequency, approximately sinusoidal waveform, but its amplitude is significantly smaller than that under the high-moisture-content condition. For the branch with 30% moisture content, the leakage current amplitude is about 0.1–0.8 A. During −9.6–−9.0 s, the current amplitude gradually increases from a small value, and the waveform is relatively smooth and symmetrical, indicating the progressive establishment of the arc. In 8.0–8.4 s, the current amplitude decays rapidly and exhibits obvious intermittency and fluctuations, corresponding to an arc-flickering, easily extinguished state. During 30.0–30.4 s, the current stabilizes at about 1 A, with a regular waveform and little distortion, representing a stable leakage-current stage under low moisture content. During tree–line arc faults with a 30% low-moisture branch, the carbonized channel develops intermittently. The arc flickers continuously, and the fault current waveform shows many burrs. However, unlike the wind-affected cases, there is no obvious pulse spike after the “current zero crossing”. When the current amplitude is small, the “zero-rest time” becomes longer, which reflects a weaker arc condition. Compared with high-moisture branches, the current amplitude is much lower. The waveform is also less smooth. Overall, the tree–line arc discharge is more unstable. During fault development, the arc is more likely to extinguish and reignite repeatedly.

6. Conclusions

In this paper, a two-dimensional multi-physics model of the tree-line arc based on magnetohydrodynamics is established, and a 10 kV distribution-line tree-line arcing ground-fault test platform is constructed to investigate the characteristic behavior of tree-line arc faults. The main conclusions are as follows:

(1)

The tree-line ignition process can be divided into four stages—arc initiation, development of the carbonized channel, moisture evaporation, and arc extinction—each associated with distinct temperature distributions and current waveform characteristics, which provides a physical basis for stage-wise identification of tree-line arc faults.

(2)

Wind speed has a pronounced influence on tree-line arc discharges. As wind speed increases, arc deformation becomes more severe, the high-frequency components in the current waveform increase, and arc stability deteriorates. At wind speeds of 9 m/s and 16 m/s, the number of spikes in the current waveform rises significantly, indicating stronger instability. Under high wind speeds, the arc channel tends to extend or split, which enhances heat dissipation, reduces the arc temperature, and affects its stability and physical properties.

(3)

The moisture content of the tree branch significantly affects the fault current waveform. With the branches with 64% moisture, the fault current waveform will be a smooth wave without many spikes, the amplitude gradually increases as the fault develops, and the current varies smoothly during zero intervals. In contrast, for branches with a low moisture content of 30%, the fault current waveform exhibits pronounced spikes and reduced amplitude, and the arc is prone to repeated extinction and reignition, resulting in a highly unstable discharge process.

(4)

The conductor type alters the temperature evolution and reignition characteristics of tree-line arcs. For insulated conductors, the arc-root temperature is higher under low wind speeds, and arc reignition becomes more difficult under high wind speeds compared with bare conductors. The combined current and temperature signatures under different wind speeds, moisture contents, and conductor types can serve as a reference for online identification of high-impedance tree-line arc faults in distribution networks and for establishing wildfire early-warning criteria in forested distribution-line corridors.