A. Ainsley

Proposal for Probabilistic Risk Assessment in Grounding Systems and Its Application to Transmission Substations

An approach to probabilistic risk assessment of electrical system grounding is proposed. The method uses all significant factors that affect the risk of electrocution at substations and takes into account their probabilistic nature. The approach implements an accurate statistical description of IEC479-1 fibrillation and body impedance data, and it uses detailed computer simulations of the modeled grounding system to provide safety voltage distributions that take into account the individuals presence at a site as a random variable. Variation in the power system fault level is accounted for, and extensive data of actual system fault clearance time are included. It is proposed that the probabilistic risk assessment is utilized as a second stage of the grounding system assessment when the first-stage deterministic analysis requires expensive or impractical mitigation. Implementation of the second stage probabilistic risk assessment yields a measure of individual risk. This is then benchmarked against industry-accepted “as low as reasonably practicable” values to determine whether investment in mitigation is required. To illustrate the applicability of the proposed approach, the probabilistic risk assessment is applied to a practical case study of a transmission substation.

Controlled Large-Scale Tests of Practical Grounding Electrodes—Part I: Test Facility and Measurement of Site Parameters

This paper describes the establishment of a large-area outdoor experimental facility at the Dinorwig pumped-storage power station in North Wales for investigating the performance of practical grounding systems. The facility itself and arrangements of test electrodes and measurement equipment are described, providing a background for subsequent test methodologies. In addition, tests to determine the variation in the water and reservoir bed resistivities have been undertaken, resulting in a survey data set that better informs experimental work and simulations of vertical ground rod and grid electrodes. It is proposed that this facility may also be used to calibrate grounding meters and for validating computational modelling techniques. In a complementary paper, results of low-voltage dc, ac, and impulse tests of varying magnitudes are described, and experimental results are compared with computed values obtained from numerical models.

Controlled Large-Scale Tests of Practical Grounding Electrodes—Part II: Comparison of Analytical and Numerical Predictions With Experimental Results

This paper presents tests carried out at an outdoor experimental facility, described in the companion Part 1 paper, to investigate the performance of practical grounding systems. Here, the results of low-voltage dc, ac, and impulse tests performed on rod and grid electrodes are described, and the measured quantities are compared with computed values obtained from numerical models. Measured ground resistance and impedance at low frequency (including power frequency) showed reasonable agreement with simple standard formulae and computational models, but revealed a significant falloff with current magnitude in the range often used for practical testing of high-voltage grounding systems. This may have implications for the specifications of grounding test equipment and extrapolation of measured ground resistance/impedance at low-voltage/current to values representative of realistic fault currents. A frequency dependence of ground impedance was also measured. Specifically, a fall in impedance over a frequency range up to 100 kHz is not generally accounted for in grounding models.

Probability Surface Distributions for Application in Grounding Safety Assessment

The safety criteria of national and international grounding standards do not compare well due to a variety of adopted values for permissible safety voltages. These differences stem from disparities in assumed worst-case conditions which result from the adoption of discrete probabilistic assumptions of safety parameters embedded within an overall deterministic safety assessment. Harmonization of grounding design standards, however, could be achieved through the adoption of a probabilistic safety assessment which will lead to a quantification of the risk of ventricular fibrillation to an individual. Fundamental principles and criteria for such an approach are developed in this paper, and are based on a statistical analysis of the IEC published curves of ventricular fibrillation. The developed numerical approach also takes into account the effect of current path on fibrillation sensitivity and body impedance. Using this approach, probability surfaces of ventricular fibrillation are then calculated as a function of applied voltage and shock duration. This statistical approach to the characterization of the human body under various electric shock scenarios forms a platform for a full probabilistic approach which can be extended to take into account other statistical parameters, and may lead to a more objective measure for making investment decisions in hazard mitigation.

Experimental investigation into the performance of large-scale earthing electrodes

Analytical and numerical simulation techniques have been developed for the calculation of earth resistance/impedance and to estimate the potential distribution in the vicinity of earth electrodes. However, very little literature is available on experimental validation of these calculation techniques. To address this, a programme of experimental tests on various earth electrodes has been carried out at the lower water reservoir of a hydro pumped-storage power station in North Wales. In this paper, the earthing test facility at Dinorwig power station is described including the details of the experimental set up. The results from experimental tests on a 5mx5m earth grid, immersed in water and energised are under ac, dc and impulse, are presented. The values of measured earth resistance/impedance and water surface potential distributions are compared with those obtained from analytical calculations and detailed numerical computer simulations.

Large-scale earthing test facilities at Dinorwig power station

In this paper, experimental set and initial test results are described of low voltage DC, AC and impulse tests on a rod electrode immersed in a water reservoir of a large hydro power station. A detailed water resistivity survey indicates largely uniform temperature and resistivity throughout the reservoir. Calculations and simulations of the rod electrode under DC and power frequency AC show reasonable agreement with experimental results. Further investigations are required to understand better the behaviour of the test system under high frequency and impulse conditions.

Parametric Analysis of Safety Limit-Curves in Earthing Systems and Comparison of International Standard Recommendations

Recommendations and procedures for the design and safety of earthing systems from three national and international standards for earthing systems, namely IEEE 80, BS 7354 and EA-TS 41-24 have been considered in this investigation. In this paper, we present the results of a comparative parametric evaluation of the different recommendations taking into account the grid size, mesh density, earth resistivity and body resistance. Safety limit-curves were obtained, which classified the various designs according to the different standards. Through investigation of the so-generated safety limit-curves, it is possible not only to clarify for what grid characteristics safe conditions are obtained for the installation but also to develop an optimum design procedure that accounts for all conditions considered within the standards

Current distribution on tower lines under fault and surge conditions

In this paper, an investigation of current distribution on tower lines under lightning and power frequency fault conditions is carried out. An equivalent circuit model of a 400 kV double-circuit transmission line was established using EMTP and two models of differing complexity were developed. The power-frequency earth fault was simulated by short circuiting one phase conductor to the earth wire while lightning conditions were simulated by injecting an impulse-type current onto the earthwire and also onto to one phase conductor. The current and potential distribution through the tower line was examined for different fault locations and the effects of soil resistivity and tower footing resistance variations were quantified.