Touch voltage in MV systems: elements to identify a Global Earthing System

The concept of Global Earthing System (GES), with reference to MV distribution systems, was introduced by the CENELEC Harmonization Document HD 637 S1 (published in 1999) and, later, by the International Standards EN 50522 and IEC EN 61936-1 (published in 2010-2011). It is defined as: "an equivalent earthing system created by the interconnection of local earthing systems that ensures, by the proximity of the earthing systems, that there are no dangerous touch voltages". The identification and official classification of GES areas would lead to a simplification of the design and verification procedures of MV/LV substations grounding systems, with associated economical savings for both Distribution System Operators (DSOs) and MV users. However, no practical guidelines are provided by the Standards. Although many works in literature have tried to solve this issue, until now, none of the available methodologies has been massively adopted by the Italian DSOs. In fact, in Italy, just few cases of GES are certified. This happens mostly because the identification methods that are actually available lack in rigorous scientific strength or are money and time consuming. This thesis is based on the work conducted in the Meterglob Project. A pool of 6 partners (Enel Distribuzione S.p.A., Istituto Italiano del Marchio di Qualità IMQ, La Sapienza Università di Roma, Politecnico di Torino, Politecnico di Bari, Università di Palermo) has studied the phenomena linked to GESs. The goal of the project consists in providing a set of guidelines for GES identification. The main aim of this thesis is to contribute in this research field. In this thesis, first, definitions regarding earthing and the conditions to be satisfied for the design of safe MV/LV substation earthing systems (ESs) with regard to touch voltages are shown. Moreover, the maintainability procedures that ensure a safety usage of the electrical installations are introduced. Then, the GES definition is discussed in depth. Some of the main concepts related to GES (e.g., interconnection, proximity and quasi-equipotentiality), which are only cited in the Standards, are instead here fully formalized. The methods available in literature to identify a GES are described and tested in a real urban MV network. Moreover, the main phenomena resulting from the presence of a GES are presented; according to the GES definition, the close interconnection of the grounding systems to each other and to utility installations (water/gas pipelines, railway tracks, etc.) sets up an overall low resistance grounding system and has mainly two effects: a distribution of the fault current between grounding electrodes (of the faulty substation and of the neighboring ones) and MV cables sheaths; a smoothing of the earth surface potential profile, reducing the hazardous voltage gradients. With reference to the first effect, a dedicated simulation tool is presented. A parametric analysis is carried out in order to evaluate which of the influence factors are the most significant. The setup of field measurements and their results are reported: a real single line to ground fault was produced in a MV network and the currents flowing into the ES and MV cable shields of 5 consecutive substations were measured. Both the field measurements and the model results suggest that the current injected in the faulted MV/LV substation is just a small portion of the fault current. A large part flows back in the feeding cable's shields. According to the simulation results, the main factors that influence the fault current distribution (and therefore the reduction factor, defined as the ratio between the current to earth and the total fault current) are the presence of bare buried conductors, the presence of LV neutral conductors, the per unit length resistance of the cables sheaths and the number of interconnected MV/LV substations. The effect of bare buried conductors deserves a special comment: in fact, they reduce the current to earth to less than 1% of the total fault current and make the substations mutual distance irrelevant. A similar effect is provided by the LV neutral conductors, which drastically reduce the importance of the substations mutual distance. Then, the second main effect of a GES, which is the smoothing of the earth surface potential, is also examined in this thesis. A model based on the Maxwell’s sub-areas method is shortly presented and the comparison with field measurement results carried out for its validation is shown. After the model validation, a simulation of the ground potential profile in a real urban scenario affected by a MV fault is implemented. The main goal is to evaluate how the electric potential profile on the soil surface is modified by the presence of buried metallic parts and distributed LV neutral conductors, together with their “reinforcement grounding rods”. In addition, an analysis about the Transferred Potentials (TPs) on floating metallic parts is carried out in order to have a more complete point of view of the phenomenon. As far as the first contribution is concerned (buried metallic parts), the simulation results show that when the additional buried metallic structures are not connected to the ESs, their influence on the ground potential distribution is not significant. The main changes in the ground potentials in fact take place only at a certain distance from the active ES, where the ground potential values are already low. In addition to this, the results show that the potential gradients are worsened by buried metallic structures, in particular near the active ES: this means that the presence of floating metallic parts can introduce localized problems that otherwise are not present. Moreover, the simulations show that dangerous potentials are transferred to LV ESs in the proximity of the MV/LV fault substation. Considering that the fault can last for a certain time before being cleared, TPs can expose people to dangerous touch voltages. Adding floating metallic parts does not significantly modify the TPs. Vice-versa, as far as the second contribution is concerned (distributed LV neutral conductors, together with their “reinforcement grounding rods”), simulation results show an Earth Potential Rise (EPR) reduction of 77% in the faulted substation with reference to the scenario where the faulted substation is called on to disperse the entire fault current. The interconnection between grounding electrodes determines a significant modification in the soil surface voltage distribution. Even if a voltage increase can be detected in areas far away from the faulted substation, a reduction of the maximum touch voltage and earth potential profile gradients detectable in the area under investigation can be observed. For this reason, the descripted LV distributive scheme surely improves GES efficiency and its realization should be recommended. However, considering the results of the studies reported in this thesis, which have investigated the GES effects, both the fault current distribution and the smoothing of the earth potential profile, it can be concluded that the most significant is the first one; for this reason, the fault current distribution is set as the basis of the methodology to identify a GES herein proposed. A general guideline shall meet two opposite goals: first, it should be based on a method than is able to faithfully represent a complex reality; secondly, it should be simple to use. Since the current distribution is the basis of the method proposed to identify a GES, a practical and fast instrument to compute the portion of the fault current injected into the ground in the fault substation is needed. According to this, a simplified formulation of the reduction factor based on the main influence parameters is developed. Finally, a suggestion for practical guidelines to identify the presence of a GES is presented.