Earth surface potential induced by ground-return current of HVDC links — Part I: Measurement and estimation of the earth resistivity

During the commissioning stage or pole outage of high-voltage direct current (HVDC) links, ground-return currents (GRCs) are injected into dc grounding electrodes. Therefore, earth surface potential (ESP) is induced by GRC. Moreover, nonuniform distribution of ESP in wide areas usually leads to half-cycle saturation of power transformers and interference of metallic underground infrastructures, especially pipelines and cables in long distance. As GRCs flow into the earth by dc grounding electrodes, ESP is determined by the resistivity of the earth. However, the heterogeneity of the earth brings difficulty in simulating ESP. In this paper, measurement and estimation of the earth resistivity is studied in detail. First, measurement technique for earth resistivity is introduced, including both the four-point method (FPM) and the magnetotelluric method (MT). A group of apparent resistivity can be obtained by changing the interval of FPM, which indicates the shallow earth resistivity. By using MT, another group of apparent resistivity representing the deep earth is scanned by tuning the receiving frequency. Second, both FPM and MT are proposed to measure the earth resistivity of wide areas. Measurement data of both methods are used as raw data for earth parameter estimation (EPE). EPE is an optimal problem of finding a good enough earth model to interpret the measured data. Third, this paper introduces an overall objective function (OOF) for EPE. The horizontal layered soil (HLS) is chosen as the ideal earth type. Then based on this assumption, HOF is set to a linear combination of the objective functions of FPM and MT. Finally, a hybrid optimization method named LevenbergMarquardt based artificial bee colony (LMABC) algorithm is presented to minimize OOF. Other optimization method such as genetic algorithm (GA), Levenberg-Marquardt (LM), and artificial bee colony (ABC) are also used to compare the performance of EPE. Results shown that LMABC is the most sophisticated method of EPE, which is capable of interpreting the measured data of FPM and MT precisely. This paper helps converting the field-test data into an HLS model, which enable the computation of ESP of wide areas. The numerical analysis of ESP and its applications will be introduced in the second part of this paper.