Experimental Study of the Behaviour of Ring Main Unit-Type Panels in the Event of Internal Arcing in Different Compartments

Abstract

If an R.M.U. (Ring Main Unit) cubicle is installed, operated and maintained in accordance with current standards and the manufacturer’s instructions, the likelihood of internal arcing is reduced, but should not be completely ignored. This article outlines the experimental situations in which arcing can occur and the catastrophic effects seen if R.M.U. is not properly designed for this type of damage. This type of failure can be caused by a fault, abnormal operating conditions or a malfunction and represents a danger during its installation and to persons present. The situations in which this type of failure can occur are presented and experimentally analysed: bi-phase internal arcing in the connections compartment, three-phase internal arcing in the switching compartment, internal arcing in the fuse compartment and single-phase arcing between a phase and the switchgear enclosure.

1. Introduction

As energy demand has increased in recent years, smart grid networks have grown significantly, in large part due to their predominantly renewable energy supply. These smart grids require an operational exchange of information between the energy supplier and the consumer through advanced metering and bi-directional technology. In recent years, studies have been carried out to improve and develop new, more reliable metering systems based on automated systems for the management of consumer loads by energy producers. An R.M.U.-type panel is defined as a compact, fully enclosed, gas-insulated switchgear unit. Switching devices can be circuit breakers, disconnectors or a combination of disconnectors and fuses. If the switching device is a circuit breaker, it is equipped with either a self-powered primary relay or a more advanced relay with communication capabilities.

R.M.U.-type panels have the characteristics of a compact structure, flexible operation, reliable interlocking, etc., and can provide satisfactory technical solutions for various applications, satisfying the most complex user requirements by using sensors and relays together with advanced technical performance and intuitive and flexible mounting solutions. R.M.U.-type panels are suitable for distribution stations, renewable energy stations, residential areas, airports, etc.

R.M.U.-type panels operate in both medium- and high-voltage networks, controlling and measuring various variables of the electricity supply and at various points of consumption. The role of R.M.U.-type panels is to detect and isolate various faults in the electrical circuit. There are different versions of R.M.U.-type panels on the market that can connect several power lines (different power supplies), depending on the number of devices and how they are connected, according to the consumer’s needs.

Due to their reliability and durability, R.M.U.-type panels are crucial in electrical engineering and play an important role in stabilising the system by controlling and protecting the power supply [1,2,3]. However, fluctuations in power, temperature or humidity can have minor impacts on maintenance and operational safety, making equipment reliability critical. The occurrence of an internal short circuit can have a major impact on the operation of the system, which is analysed in this article, both in terms of the behaviour of the R.M.U. and in terms of operational safety.

For example, if the surface temperature of the inner wall of metal equipment falls below the dew point temperature, water droplets will condense on the surface, leading to the corrosion of metal components and reduced insulation performance. This increases the risk of arcing occurrences or equipment damage, which can lead to accidents [3]. Preventing this condensation phenomenon is therefore crucial to the safety of energy systems.

According to [4], the occurrence of condensation depends mainly on the relative humidity, the temperature and the dew point temperature in an R.M.U.-type panel. In recent years, research has focused on the prevention of condensation and the danger of high condensation pressure in the switchgear and proposed some technical measures to prevent condensation [5,6]. One of the agreed solutions is to add ventilation holes at the base of R.M.U.-type panels to solve this problem by reducing the humidity in the air [7]. However, simply adding ventilation holes without adding dehumidifiers will result in limited dehumidification efficiency.

The first versions of R.M.U.-type panels contained circuit breakers or high-voltage fuse/disconnector combinations which switched the power supply on or off in the event of a fault. Today, R.M.U.-type panels are equipped with a switch disconnector which can be operated remotely; can record and transmit voltage, frequency and current values remotely; and is configurable [8]. A special configuration that remotely sends a low battery warning message is shown in [9,10]. This unit can also measure overcurrent and voltage drop and relay system status. The same R.M.U.-type panel measures the phase and neutral intensity; voltage; active, reactive and apparent power; power factor; current; and harmonic distortion (THD) of the line and records faults that lead to the line being disconnected.

The main scope of this paper is to provide real-life information about the behaviour of R.M.U. panels during an internal arc fault, which is the most catastrophic event that can occur. These devices provide the highest level of safety for their staff and the general public and are designed and manufactured to prevent the occurrence of internal faults, but this hazard cannot be fully eliminated. The information provided is valuable for both for designers and for final users.

2. Electrodynamic Forces

An internal arc fault is the most disastrous situation that can appear during the exploitation of electrical devices. In the event of an internal arc fault, a thermal energy up to five times the surface temperature of the sun (20,000 °C) is discharged.

An internal arc fault, which constitutes a hazard if operators are in the area, though extremely rare, might occur due to reasons such as a failure of insulation, contacts due to ageing, overvoltages in the system because of switching or lightning surges, pollution due to environmental conditions, mal-operation or insufficient maintenance.

Designs which have been successfully tested are known as IAC-classified. This classification is planned to offer a tested level of protection to operators in proximity of the equipment during normal functioning conditions and with the switchgear and control gear in their normal service position, on the occasion of an internal arc.

R.M.U.-type panel manufacturers have developed very compact designs using air insulation or gases with electro-insulating properties (SF₆, nitrogen, CO₂), which greatly reduce the distances between current paths, as well as complex R.M.U. designs. It is known that electrodynamic forces are inversely proportional to the distance between current paths and directly proportional to the square of the current value.

In Metal-Clad or switchgear-type panels, the current paths are generally parallel over long distances and the distribution of electromagnetic forces is known, whereas, in R.M.U.-type panels, the current paths are shorter and have a complicated geometry, making it difficult to anticipate the effect of their electrodynamic forces, especially in the case of an internal arc. In the case of the R.M.U.-type panel, the internal arc that is generated moves chaotically and not linearly, as it does in the case of parallel current paths (Metal-Clad), making it difficult to predict where the “foot” of the arc is located in the metal enclosure of the panel, where measures could be taken to mechanically reinforce the panels to prevent the arc from exiting with destructive effects.

The electrodynamic structure of R.M.U.-type panels is time-varying due to the fact that the current strengths vary over time. The highest values of their electrodynamic forces are due to short-circuit currents. The values of their electrodynamic forces are considerably lower in nominal operation than in short-circuit operation.

This article presents both bipolar and tripolar experiments. Electrodynamic forces occur in both two-wire and three-wire short circuits.

To calculate the electrodynamic forces in the first bi-phase short-circuit experiment, we assume that the currents i₁ and i₂ flow through the power cables (item 12 in Figure 1), which are placed at a distance from each other. We also assume that these cables are straight. We consider two parallel conductors with a distance “a” between them, fixed on one side to insulated supports at the base of the R.M.U. panel and, on the other side, to a plugin.

Figure 1 shows a schematic diagram of an MV distribution switchgear to illustrate a possible distribution of the equipment inside: 1–6 single-phase current measuring transformers; 7—support insulators with voltage dividers; 8—single-phase voltage-measuring transformers; 9—high-voltage fuses; 10—general busbars; 11—earthing disconnectors; 12—power busbars (cables); 13—circuit breakers.

The electrodynamic forces between the conductors are described in the literature [11,12,13,14,15,16,17] as

$$F_a=\frac{{\mu }_0}{2\pi }{i}_1{i}_2\frac{l}{a}{\phi }_{cD}$$

(1)

where the correction function ${\phi }_{cD}$ is expressed by Dwight curves. Equation (1) is given for rectangular conductors of infinite length. In the case of the experiments in this paper, these conductors are one-dimensional cables of finite length, so the correction factor $\phi \left(\frac{a}{l}\right)$ is added and Equation (1) becomes

$$F_a=\frac{{\mu }_0}{2\pi }{i}_1{i}_2\frac{l}{a}{\phi }_{cD}\phi \left(\frac{a}{l}\right)$$

(2)

For the purpose of writing a simplified equation, we refer to Equation (2) which, in the case of the bipolar short circuit simulated in the first experiments, is written as follows:

$$F=C{i}^2$$

(3)

with the notation $C=\frac{{\mu }_0}{2\pi }{i}_1{i}_2\frac{l}{a}{\phi }_{cD}\phi \left(\frac{a}{l}\right)$ and the observation that, in experiments, i₁ = i₂ = i.

In the experiments, the case that generates the highest forces was taken into account, i.e., the asymmetrical short circuit, using the following current expression for the circuit calculations:

$$i=Î\left(e^{-\lambda t}-cos\omega t\right)$$

(4)

where $\lambda =\frac{1}{T}=\frac{R}{L}$.

In a transient regime, the electrodynamic force is as follows:

$$F=CÎ{\left(e^{-\lambda t}-cos\omega t\right)}^2$$

(5)

The maximum value is reached when the current reaches its peak value:

$$F_{max}=C{\chi }^2{I}^2$$

(6)

For the circuit constant (the inductive character of the circuit in experiments) λ = 22.32⁻¹ and χ = 1.8. The peak factor χ depends on the X/R ratio; in practical cases there is no zero damping, therefore the highest value that can be accepted for χ is 1.8, resulting in a peak current which can be written as covered by the following value:

$$i_{peak}=\sqrt{2}I\times 1.8$$

(7)

The maximum force is obtained from Equation (6):

$$F_{max}=3.24 CÎ=3.24 F_0$$

(8)

In Equation (8), F₀ = Cβ is the maximum instantaneous value of the steady-state force.

The experiments in the next chapter continue in a three-phase mode, with the short circuit created between all three phases of the R.M.U.-type panel. The unit will be powered in the right panel, similar to the previous conditions, but this time the short is three-phase. The force applied to the centre power cable is

$$F_m=C{i}_1\left(i_2-i_3\right)$$

(9)

and the force applied to a side conductor is

$$F_l=C{i}_1\left(i_2+\frac{i_3}{2}\right)$$

(10)

with the expressions of the currents in the three cables as follows:

$$i_1=Î\left[e^{-\lambda t}sin\alpha +sin\left(\omega t-\alpha \right)\right]$$

(11)

$$i_2=Î\left[e^{-\lambda t}sin\left(\alpha +\frac{2\pi }{3}\right)+sin\left(\omega t-\alpha -\frac{2\pi }{3}\right)\right]$$

(12)

$$i_1=Î\left[e^{-\lambda t}sin\alpha +sin\left(\omega t-\alpha \right)\right]$$

(13)

The force applied to the centre cable is obtained by inserting the current expressions from (11)–(13) into Equation (9), to give

$$F_m=\sqrt{3}C{Î}^2\left[\left(e^{-\lambda t}sin\alpha +sin\left(\omega t-\alpha \right)\right)\right]\left[\left(e^{-\lambda t}cos\alpha -cos\left(\omega t-\alpha \right)\right)\right]$$

(14)

If, in Equation (14), the maximum value that can be found in the networks for λ is inserted, i.e., λ = 22.311 s⁻¹, the maximisation of the force F_m occurs for the value of the closing angle $\alpha =\pi /4$, and (14) becomes

$$F_m=\frac{\sqrt{3}}{2}C{Î}^2\left(e^{-\lambda t}+sin\omega t-cos\omega t\right)\left(e^{-\lambda t}-cos\omega t-sin\omega t\right)$$

(15)

Inserting ωt = π into Equation (15) gives the maximum instantaneous value of the force applied to the centre cable:

$$F_m=\frac{\sqrt{3}}{2}C{Î}^2{\left(e^{\frac{-\lambda \pi }{\omega }}+1\right)}^2=\frac{\sqrt{3}}{2}C{Î}^2{\chi }^2$$

(16)

if λ = 22.311 s⁻¹, then χ = 1.8, which gives

$$F_m=2.805C{Î}^2=2.805F_0$$

(17)

To calculate the force applied to the side cable, consider Equation (10), into which the expressions for the currents in Equations (11)–(13) are inserted to give the following equation:

$$F_l=-\frac{\sqrt{3}}{4}C{Î}^2\left[e^{-\lambda t}sin\alpha +sin\left(\omega t-\alpha \right)\right]\times \left[e^{-\lambda t}\left(\sqrt{3}sin\alpha -cos\alpha \right)+\sqrt{3}sin\left(\omega t-\alpha \right)+cos\left(\omega t-\alpha \right)\right]$$

(18)

The maximum of Function (18) is obtained for $\alpha =\frac{7\pi }{12}$; with λ = 22.311 s⁻¹ we obtain the maximum value exerted on the side cable:

$$F_l=-2.621C{Î}^2$$

(19)

The equations expressed above are valid in the transient regime in which the experiments in the following chapter are carried out to simulate the most severe conditions encountered during R.M.U. panel operation.

According to Equations (1)–(19), currents to the order of tens of kiloamperes and distances between relatively small current paths (100 mm) result in forces to the order of hundreds of Newtons, which deform electrical circuits. These deformations cause the supporting insulators to break and internal arcing to occur. The electrodynamic forces that occur displace the arc that has formed and, interacting with ferromagnetic walls, relocate it to the furthest point of the circuit, producing thermal and mechanical effects that are destructive and dangerous to operators.

3. Materials and Methods for the Internal Arc Experiments

Metal-enclosed high-voltage switchgear and control gear for AC and rated voltages above 1 kV, including R.M.U.-type panels, are widely used in distribution systems. When arcing occurs in metal-enclosed high-voltage switchgear and control gear due to a short-circuit fault, a large amount of energy is released inside the panel, causing rapidly propagating heating. Arcing puts service personnel at risk but can also have a serious impact on the entire electrical system.

The internal arc tests presented below are carried out in accordance with the relevant international standards [18] and evaluate the behaviour of R.M.U.-type panels, with respect to internal arcs, under the most severe operating conditions and the consequences of this phenomenon. The main contribution of this paper is the creation of bi-phase and single-phase arcs inside an R.M.U. in its connections compartment, with the aim of developing efficient evaluation methods to investigate the effects of internal arcing, which will also provide a significant contribution to the early planning and design phases of R.M.U.s.

The R.M.U.-type panel used in the experiments connects or disconnects the circuit by means of a switch disconnector for the first three experiments and by means of a circuit breaker for the last two experiments and is thus designed to feed a transformer down-stream and provide short-circuit or overload protection. An R.M.U.-type panel with a switch disconnector and high-voltage fuses is shown in Figure 2a–c and its wiring diagram in Figure 2d.

Figure 2 shows the components of the R.M.U.-type panel: 1—three-position switch disconnector; 2—high-voltage fuse compartment; 3—arrester with linear variable resistance; 4—pluggable connectors in the cable compartment; 5—earthing switch; 6—general busbars.

If an R.M.U.-type panel is installed, operated and maintained in accordance with the manufacturer’s instructions, the occurrence of internal arcing will be reduced, but cannot be ignored. The occurrence of damage inside an R.M.U.-type panel, either due to a fault or an abnormal operating situation, may result in arcing and pose a hazard to equipment and personnel. When selecting an R.M.U.-type panel, the possibility of arcing is always taken into account to ensure an acceptable level of protection for the operator. Conventionally, the selection of appropriate equipment in relation to internal arcs is governed by the procedure [19], which is based on the assumption that the user plays an important role in risk reduction.

The purpose of these experiments is to deliberately create fault situations in the compartments of R.M.U.-type panels that, based on statistics and dielectric requirements, are most vulnerable. In addition, the international standards for quality certification require tests to verify arc resistance in all compartments: the connections compartment, switching compartment and fuse compartment.

Recently, abnormal situations have been observed in R.M.U.-type panels used equipped with a combination of swich disconnectors and fuse-links, where the fuse used for short-circuit-protection malfunctions and, when it breaks, creates an electric arc which, by ejecting the porcelain body, becomes extremely dangerous to the outside. In general, the connection compartment of an R.M.U.-type panel has plug-in-type connections which provide reliable insulation for the terminals, but single-phase or bi-phase arcing faults occur due to insulation damage. In the switching compartment, internal arcing is caused by an incorrect switching of the switch disconnector and is always three-phase.

Given the extremely violent nature of R.M.U.-type panels during an arc, these experiments aim to reproduce the most severe conditions for the development of mechanical and thermal energy inside them. The usual electrical parameters of an R.M.U. are 24 kV or 36 kV, and its fault currents are 16 kA_RMS and 20 kA_RMS, such that short-circuit powers of 600 MVA and 1200 MVA develop high mechanical and thermal energies inside them for one second. The causes of arcing—insulation damage, the incorrect operation of a circuit element, switching surges, a short circuit downstream of the R.M.U. or the incorrect operation of a switching device—cannot be completely eliminated. Precautions must therefore be taken to ensure that, in the event of an arc, its external effects do not injure the operator or unauthorised persons in the vicinity. Individual R.M.U.-type panels integrated into substations or transformer stations must have a minimum risk of accident, and even if they are equipped with intelligent protection devices that quickly disconnect the power supply from the fault, the metal enclosure must withstand the effects of the arc for one second. In the rush to achieve small dimensions, manufacturers have created small compartment volumes, where the same energy generates high mechanical pressures. Although measures have been taken to limit faults and their effects (fault detectors, fast current-limiting devices, overpressure valves), the robustness of their metal enclosures must be verified by experiments under extreme conditions. In terms of personnel access, R.M.U.-type panels are classified as Type A—restricted to authorised personnel—and Type B—unrestricted, accessible to the general public.

With regard to the arc-protected parts of an R.M.U.-type panel, they can be protected at the front (F), side (L) and rear (R). For example, the IAC-AFLR is a restricted-access panel that has arc protection at the front, lateral and rear side.

The switchgear-forming part of the main circuit and the earthing switches of the switchgear and control gear in R.M.U.-type panels have been tested to verify their rated making and breaking capacities, in accordance with the relevant standards and under appropriate conditions for their installation and use. This means that they have been tested as they are normally installed in switchgear and control gear, with all associated components whose arrangement may affect their performance, such as connections, brackets, etc. When determining which components are likely to affect the performance of an R.M.U.-type panel, particular attention should be paid to the mechanical forces caused by the short circuit, the evacuation of pressure, the possibility of disruptive discharge, etc. Admittedly, in some cases, this influence can be quite negligible.

The equipment under test was arranged to achieve the most demanding conditions in terms of maximum unsupported busbar lengths, conductor configuration and connections within the equipment. Terminal connections have been arranged to avoid unrealistic stresses or unrealistic terminal support. Interlocks are provided in all positions to prevent the operation of switching devices, access to control interfaces and the insertion or removal of removable parts.

The tests carried out include the case of an arc fault inside the metal enclosure or inside the components forming part of the R.M.U.-type panel, performed under normal operating conditions. The tests carried out take into account the effects on all parts of the enclosure, such as internal overvoltage, the thermal effects of the arc, the effects of hot gases, and glowing particles. These tests do not focus on and do not reveal the influence of internal arcing between compartments, damage to internal partitions, external connections or the presence of gases with potentially toxic properties or the risk of fire propagation to materials or equipment located in the vicinity of the control and switchgear.

The tests were carried out with an R.M.U.-type panel installed under normal operating conditions, i.e., with the position of the switchgear and removable components set to create a power circuit. Other equipment, such as measuring and monitoring equipment, was also in place for normal operating conditions. The equipment used in the experiments did not require any panels or doors to be removed or opened to perform switching operations; if this were necessary, the experiments would be performed with them open. In particular, the removal or replacement of high-voltage fuses is not considered normal operation, even for maintenance purposes [20]. The combination of fuse-links and switch disconnectors can limit the short-circuit current and minimise the duration of the fault, and it is well known that the arc energy transferred cannot be predicted by calculation based on I²t. In this case, the maximum arc energy can occur at current levels below the maximum rated breaking capacity [21,22,23,24,25].

3.1. Test, Layout and Assembly Conditions

The test circuit fully complies with the requirements of international standards [20] for the certification of the quality of metal-enclosed switchgear. In addition to the electrical parameters, great importance was attached to the layout and mounting of the R.M.U.-type panel for testing. It is placed in an enclosure with wall and ceiling dimensions that are well defined in [20] and is surrounded at a 300 mm distance by vertical and horizontal cotton indicators that simulate protective equipment that could be affected by hot gases or flames coming from the metal casing of the R.M.U.-type panel.

In general, all types of panels are designed to limit the pressure generated by an arc and direct it to areas that are safe for the operator and the public. Air-insulated models have an exhaust system at the top and rear and gas-insulated (SF6) models have pressure valves at the bottom. In some cases, substation panels are also fitted with upper or lower gas ducts to remove hot gases from the operating area.

In order to fully evaluate the R.M.U.-type panel, these tests take into account all compartments where arcing can occur: the switching compartment; connections compartment; and fuse compartment. A simplified single-wire diagram of the electrical circuit used in the experiments is shown in Figure 3.

Figure 3 shows the main equipment used to generate the parameters required for the experiments, parameters that can only be obtained experimentally in a high-power laboratory: 1—2500 MVA AC source; 2—Making Switch; 3—master breaker; 4—current-limiting coils; 5—step-up transformers; 6—Rogowski coils; 7—R.M.U.-type panel; 8—voltage divider; 9—data acquisition system.

According to [20], the experiments can be performed at a given voltage, current and duration to be considered valid for all lower values of current, voltage and duration. The experiments can be performed, according to the standard, at any suitable voltage up to and including the rated voltage. The experiments were performed at 6 kV, so the following conditions must be met:

-

The current value during the experiments, computed by a digital recording device, must remain constant or the test shall be extended until the integral of the a.c. component of the current (I*t) equals the specified value, with a tolerance of +10%. In this case, the current should be equal to the specified value at least during the first three half-cycles and should not be less than 50% at the end of the test. The r.m.s. values for these experiments are 16 and 20 kA.

-

The arc must not be extinguished prematurely in any of the phases in which it has been initiated. Temporary single-phase extinguishing is permitted, as long as the cumulative duration of the intervals without current does not exceed 2% of the experiment’s duration and the single events last no longer than to the next prospective current zero, provided that the integral of the a.c. component of the current equals at least the value specified in the standard for the relevant phase.

The instant of closing shall be chosen for the experiments so that the peak current will flow into one of the outer phases and a major loop will also occur in the outer phase. Because the voltage during the experiments is lower than the rated voltage, the peak value of the prospective current is irrelevant; this is the reason why oscillograms are not presented in this article. However, during the experiments, the peak value of the current must not drop below 90% of the peak value. At the bi-phase initiating of the arc experiments, the instant of closing was chosen to provide the maximum possible d.c. component. The frequency chosen for the experiments was 50 Hz, according to the rated frequency.

The experiments were carried out in an enclosure simulating a room, represented by a floor, a ceiling 2000 mm above the floor and two walls perpendicular to each other. The height of the model was determined by the highest part affecting the gas flow, including the compression release flaps, which must not touch the ceiling when opened.

In all experiments, the R.M.U.-type panels were surrounded by indicators on the front and side wall of the compartment, where the arc was initiated. They were made from 150 g/m² black cotton fabric, laid so that the cut edge did not face the panel. To prevent them from igniting each other, they were mounted on sheet steel frames with a depth of 2 × 30 mm and a size of 150 × 150 mm each. The mounting of the indicators in relation to the panel has taken into account protruding elements that are not expected to affect the hot gases, simulating, as realistically as possible, the position that a person would normally take in front of the equipment.

3.2. Experiments in the Switching Compartment

In the switching compartment, arcing is enhanced by an unsuccessful attempt to break by the switch disconnector or by damage to the arc extinguishing medium. In general, any switching of inductive or capacitive loads generates an arc between the fixed and moving contacts, which is extinguished when the current crosses zero and the rated voltage is restored under transient conditions.

If this arc is not extinguished after a maximum of 15 ms, it will develop into a free, uncontrolled, generalized arc, which is a fault whose manifestation outside the panel is undesirable. For the purpose of this experiment, the arc is created by means of a fuse wire between all phases that is energized for one second at a nominal voltage with a current of 16 kA_R._M._S.

The R.M.U.-type panel in Figure 2 was used in the first experiment and consists of side panels only, without a panel containing high-voltage fuses. The special enclosure, the panel and the used indicator, shown in the photograph in Figure 4a, are subjected to an arc of 16 kA for 1 s using the circuit in Figure 3.

The oscillograms obtained by applying a current of 16 kA for one second at a voltage of 6 kV are shown in Figure 5.

Figure 5 shows the oscillograms obtained from the first experiment, in which the R.M.U.-type panel was subjected to an internal arcing test in the switching compartment, initiated with a ϕ 0.5 mm² copper wire at the point furthest from the power supply, as shown in Figure 4b. The power supply was three-phase, with all switching devices in the R.M.U. power-on position and all doors closed and secured. The R.M.U.-type panel was fully equipped with its original equipment; no experimental model was used and it was earthed through the point provided by the manufacturer. The switching compartment was filled with air at a relative pressure of 1.55 bar, for environmental reasons, as permitted by [20]. The parameters obtained are the applied voltage values between phases: 6.1/6.2/6.2 kV; peak current values: 32.7/33.4/−42.6 kA; r.m.s. values over the whole short-circuit current range: 16/16.3/16.5 kA; between-voltage drop values: 0.573/0.607/0.588 kV; and duration: 1 s. After 400 ms from the start of the arc, during which time the pressure and temperature in the compartment increased, the pressure valve activated, releasing the gas inside. This is confirmed by the increased voltage drop across the arc for 250 ms. The experiment is considered successful as the electrical parameters are within the 5% limit given by [20], there was no arc interruption during the experiment and there was no distortion of the current. In addition, the behaviour of the R.M.U.-type panel is considered adequate as the doors did not open, the pressure valve worked correctly and the gases were released to the ground, the horizontal and vertical indicators were not affected, no parts were detached from the R.M.U.-type panel and the arc did not create holes in the metal enclosure.

3.3. Experiments in the Connections Compartment

In R.M.U.-type panels, the connections compartment has solid plug-in-type insulation and the probability of a three-phase arc is extremely low because it is unlikely that the solid insulation on all three phases will be damaged in the same place at the same time. Therefore, ref. [20] imposes a bi-phase arc in this compartment and, in the case of networks with insulated neutrals, the current value of the two phases on which the arc occurs is 0.87 I_sc. Thus, in a symmetrical and balanced electrical system, the short-circuit current I_sc is expressed as a function of the phase voltage U_f and the phase short-circuit impedance Z_sc:

$$I_{sc}=\frac{U_f}{Z_{sc}}=\frac{U}{\sqrt{3}{Z}_{sc}}$$

(20)

In this bi-phase short-circuit experiment, the system becomes symmetrical but unbalanced, so (20) becomes

$$I_{sc}=\frac{\sqrt{3}U}{2\sqrt{3}{Z}_{sc}}$$

(21)

because the bi-phase short-circuit voltage is $U=\sqrt{3}{U}_f$, and the impedance becomes Z = 2Z_sc. So, in this experiment, a short circuit was created between two phases in the connections compartment, as shown in Figure 6.

As can be seen from the photograph in Figure 6, the initiation point of the arc was between the most-exposed phases, close to the outer side panel. Indicators were placed to the left and to the front of the R.M.U.-type panel, and a wall was placed 100 mm to the rear and right side of the R.M.U. In order to assess the possibility of the bi-phase arc developing into a three-phase arc on the third phase, only a single insulated terminal head was fitted, without the cable. The resulting oscillogram is shown in Figure 7.

The test was carried out as described in [20] for outer-cone connections, where the third phase has a plug similar to the one used during its operation, which can be energised with a voltage at least equal to the voltage applied. The voltage applied was 6 kV_RMS, which was considered sufficient to ensure that the arc did not extinguish prematurely. Temporary single-phase arc extinguishing is permitted, provided that the cumulative duration of the no-current intervals does not exceed 2% of the duration of the test and that individual events do not last longer than until the next zero crossing, and provided that the integral of the AC component of the current is at least 87% of the value of the current, the rationale being that 87% is the value of the bi-phase current. This means that each phase of a bi-phase short circuit carries less current than a three-phase short circuit, i.e., 13% less load. Once the arc has been created, as can be seen in Figure 7, it develops into a three-phase arc. The rated peak current is I_peak = 41.3/−41.8/- kA, the r.m.s. bi-phase current is I_r._m._s._bi-phase = 18/18.1/- kA_RMS, the r.m.s. three-phase current is I_r._m._s._three-phase = 19.8/20/20.1 kA, the measured voltage between phases is U = 0.761/4.9/4.6 kV_RMS and the voltage drop is U_drop = 0.668/0.613/0.702 kV_RMS. In Figure 7, the voltage between the phases on which the arc started is shown on a different scale to make it easier to follow its progression. The arc developed into a three-phase arc after 49 ms. The effects of this progression are shown in the photograph in Figure 8.

Statistically, the insulation of the connections in the connection compartment is strong enough that the bi-phase arc that occurs will not become three-phase 95% of the time.

The tests are continued under the same conditions on an identical new panel, but this time the cable is also inserted into the terminal head, the phase insulation is increased and the current value is changed to three-phase 16 kA. The connection compartment is generally small, so the pressure developed inside can exceed 3 bar and, combined with the temperatures of over 2000 °C generated by the arc, can have a devastating effect on this compartment. Under the same conditions as in the previous experiment, the power supply is three-phase and a 0.5 mm² diameter fuse wire is placed on two phases whose insulation had been previously damaged. The three-phase supply maintained an assumed bi-phase fault current of 0.87 × 16 kA = 14 kA for one second. The oscillogram is shown in Figure 9.

The parameters obtained are applied voltage values between phases of 6.2/6.2/6.2 kV; peak current values: 41.3–41.3/- kA; r.m.s. value over the whole short-circuit current range: 17.3/17.3/- kA; between-phase voltage drop values: 0.686/4.8/4.3 kV; and duration: 1 s. The layout of the panel and the indicators after the experiment is shown in Figure 10.

3.4. Single-Phase Testing

In the case of this experiment, the situation was reproduced of, in a grounded neutral network, an arc occurring in an R.M.U.-type panel between one phase and the metallic enclosure, with the other phases on the sides being energised and insulated accordingly.

In addition to the mechanical and thermal effects of the arc, this experiment also examines whether the insulation of the other two phases from the earth is compromised, thus generalising the arc.

At the rated voltage, the arcing single-phase fault current is I = 16 kA and develops less energy for 1 s, so the effects are significantly less than for a three-phase or bi-phase fault.

The situation when the central phase is earthed to the door by means of a 0.5 mm² diameter copper wire between the central phase and a welded screw, as shown in Figure 11b, by means of an arc with a current value of 16 kA, when the other two phases are energised at a nominal voltage with their currents limited to 25 A, if any. The oscillogram is show in Figure 12.

In this experiment, we will observe the occurrence of the arc current I₂ at the time set by the automatic programmer, Figure 3, point 9, with the voltage on phases U₁ and U₂ on the sides. The values obtained in the experiment are as follows: the voltages U₁ and U₃ are 14.5/14.5 kV for 420 ms, until the current is generated, and then 2.97/2.97 kV for 862 ms; the voltage U₂ on the phase on which the arc is initiated is generated after 185 ms and has a value of 0.471 kV for 786 ms and a value of 1.08 kV for 75 ms when there is no current on this phase; the currents I₁ and I₃ appear at the time of the arc’s development to three-phase and last for a duration of 862 ms, with values of 23.7/23.7 kA, which means that the insulation of these phases has failed under the thermal effect of the arc, generalising the three-phase fault; the current I₂ is 16.5 kA for 1.02 s. At the appearance of the arc, a highly dynamic arc pressure is also observed, exceeding the 1.5 bar peak in the first 20 ms, measured at the side wall of the compartment. In its external appearance, the R.M.U.-type panel shows no deformations, detached parts or thermally induced holes, but the indicators burned, as shown in Figure 11b, demonstrating an inappropriate result for the R.M.U.-type panel.

3.5. Three-Phase Experiment with Arc Ignition in the Fuse Compartment

This experiment was carried out on new panels, as shown in Figure 2, using the circuit shown in Figure 3. The occurrence of this fault situation implies a malfunction of the fuses, such as when they interrupt a short-circuit current and the porcelain tube breaks, unable to isolate the fault.

Generally, R.M.U.-type panels have a very small fuse compartment volume and the energy developed in this space creates enormous pressures with extremely dangerous mechanical effects.

To carry out this experiment, a 0.5 mm² diameter fuse wire is fitted along each of the three fuses for arc ignition, as shown in Figure 13.

In general, the volume of the switchgear and connection compartment of an R.M.U.-type panel is on average 0.5 m³ ÷ 1 m³, and the fuse compartment is about 10 times smaller.

The arc energy released in both situations is the same, resulting in a pressure to the order of tens of bars in the fuse compartment, and it is difficult to find a solution to this, especially as these compartments do not have pressure-relief valves.

The mechanical strength of these compartments must therefore be such that they can absorb the energy released or transfer it to the switching compartment, which is even larger and equipped with a pressure-relief valve. In this situation, the destruction of the R.M.U.-type panel is certain, but the body of the operator in front of the panel is protected. Photos taken during the experiment are shown in Figure 14.

The oscillogram obtained in this experiment is shown in Figure 15.

Figure 15 shows the oscillograms obtained from the three-phase experiment, in which the R.M.U.-type panel was subjected to an internal arcing test in its fuse compartment, initiated with a ϕ 0.5 mm² copper wire. The power supply was three-phase, with all switching devices in the R.M.U. power-on position and all doors closed and secured. The R.M.U.-type panel was fully equipped with its original equipment. The parameters obtained are applied voltage values between phases of 6.2/6.2/6.2 kV; peak current values: 32.8/33.4/−42.8 kA; r.m.s. values over the whole short-circuit current range: 16/16.1/16.1 kA; between-voltage drop values: 0.573/0.6/0.59 kV; and duration: 1 s.

The behaviour of the R.M.U.-type panel during the experiment is totally inadequate: the panel disintegrated, the indicators were destroyed, as shown in Figure 14, and the result, for the operator, would have been tragic.

Inside the panel, the effects of the arc were as expected because of the extremely high pressure in the fuse compartment, but the mechanical strength of the front part should have prevented the spread of flames and pieces of melted metal to the outside area.

4. Conclusions

R.M.U.-type panels are evolving towards a robust, cyber-secure, green and digital design, safe in operation and adapted to the evolution of smart grids, with a high level of management and the integration of distributed energy resources. R.M.U.-type panels are a mix of cutting-edge technologies, combining circuit breakers with a vacuum extinguishing medium, a technology preferred by users; compressed air as an insulating medium, an environmentally friendly solution; cloud-embedded sensors and algorithms for preventive maintenance and the integration of metering for network management; and high-mechanical-strength swich disconnectors for frequent network reconfiguration.

This article presents various situations in which arcing can occur, in which compartments it can develop and how it progresses, based on real experiments with the most severe operating conditions. This article has been produced to assist users in selecting the correct equipment by properly considering the possibility of faults leading to internal arcing. Understanding these phenomena can reduce risk to a tolerable level, where risk is a combination of the probability of a failure occurring and its severity; therefore, the selection of an R.M.U.-type panel should be based on documented analyses to achieve a tolerable risk.

The effects of internal arcing inside an R.M.U.-type panel can be devastating, as highlighted in this article, and it is therefore recommended that measures are taken to limit its external effects, such as fast fault clearance times, initiated by pressure or heat sensors or differential busbar protection; the use of appropriate fuses in combination with switching devices to limit the current flow and the duration of the fault; rapid arc clearance by diverting the arc to a metal short circuit using rapid detection and closing devices; and moving a retractable part to or from the operating position only, with the front door closed.

The purpose of the tests described in this article was to verify the effectiveness of the R.M.U.-type panel in protecting people in the event of internal arcing under normal switchgear operating conditions. The tests carried out do not assess behaviour under all conditions, e.g., when the low-voltage compartment is open or dismantled. In addition, for indoor installations, internal arcing can cause a surge in the room where the equipment is installed.

The tests were carried out at a voltage such that the arc did not extinguish prematurely in any of the phases in which it was initiated, which was successfully achieved even though the standards allow temporary single-phase quenching, provided that the cumulative duration of the no-current intervals does not exceed 2% of the duration of the test. The timing of the closing of the circuit has been chosen so that the peak current occurs in one of the outer phases and also creates a large loop in the other outer phase. As the experiments were conducted at 5o Hz, the assumed peak current was set at 2.5 times the r.m.s. value. Since the voltage obtained in the tests was lower than the nominal voltage, the peak value of the assumed current is irrelevant, but the peak value of the short-circuit current must not be less than 90% of the nominal peak value successfully obtained in the tests. In the bi-phase experiments, the closing time was chosen to ensure a maximum DC component.

In order for the panel to be considered compliant, tests were carried out to ensure that there would be no panel fragmentation and no ejection of fragments with a mass greater than 60 g and that the indicators would not ignite under the influence of hot gases or flaming liquids. In the experiments where the indicators ignited, the situation was assessed and the phenomenon was caused by glowing particles rather than hot gases.

The tests carried out are an important input for designers of R.M.U.-type panels and, even if the equipment did not respond adequately in every situation, it is very important to know how this equipment behaves when this type of fault occurs.

After carefully observing the behaviour, during internal arcing, of R.M.U. panels, we have determined the possible causes of internal arc faults and possible preventive measures in terms of the location they are most likely to occur in the connection compartment. The possible causes of internal arc faults are inadequate choices of cable, faulty installation and failures of insulation. Possible preventive measures include increasing the dimensions of the compartment, the use of higher quality materials, the avoidance of crossed cable connections and the execution of dielectric tests on site. The operation of disconnectors and switches is the cause of maloperation and can be prevented with interlocks and training. Bolted connections and contacts caused by corrosion can be prevented by encapsulation and supplemental heating to prevent condensation. Instrument transformers are caused by ferro-resonance and short circuits on the low-voltage side of VTs. These can be avoided through the suitable design of the circuit and protection covers. Circuit breakers caused by insufficient maintenance and can be prevented through regular maintenance and training.