A COMPUTATIONAL METHOD FOR POSITIVE CORONA INCEPTION IN THE COAXIAL CYLINDRICAL ELECTRODE ARRANGEMENT IN AIR UNDER VARIABLE ATMOSPHERIC CONDITIONS P.N. Mikropoulos * and V.N. Zagkanas High Voltage Laboratory, School of Electrical & Computer Engineering, Faculty of Engineering, Aristotle University of Thessaloniki, Building D, Egnatia Str., 54124 Thessaloniki, Greece *Email: pnm@eng.auth.gr Abstract: Corona discharge has many practical applications, thus it has been studied extensively experimentally as well through modelling in many electrode arrangements. In the present study a computational method for the estimation of the positive corona inception field strength in the coaxial cylindrical electrode arrangement in air is presented. It is based on streamer theory and involves Hartmann’s expression for the field dependent effective ionization coefficient and the known distribution of the geometric electric field. A very good agreement with literature experimental data referring to wire-cylinder air gaps has been observed for an avalanche number of 10 4 (ionization integral ~9.2) in a wide range of wire radii and under variable atmospheric conditions. A simple absolute humidity correction factor has been introduced in Peek’s formula, allowing for an accurate estimation of the corona inception field strength under variable humidity. 1. INTRODUCTION Corona discharge is of paramount importance in high voltage technology as it has many practical applications including electrostatic precipitators and printing, ozone production, spray coating, biological and chemical surface treatments and the treatment of flue gases. In power transmission systems corona activity results in power losses and interference in communication systems. Moreover, surge voltages propagating on transmission lines are attenuated and distorted by losses due to corona discharges; these corona effects are important in overvoltage protection and insulation coordination of power systems. Therefore, the better understanding of the physical processes involved in corona discharges is important, besides fundamental, from an engineering point of view. Because of its important effects, corona discharge has been studied extensively experimentally as well through modelling in many electrode arrangements. The coaxial cylindrical geometry has drawn a great attention as, besides simulating many practical applications, corona develops symmetrically around the inner electrode and the accurate knowledge of the geometric electric field distribution between the electrodes allows for computations for the estimation of the basic corona characteristics. Thus, corona discharge in a coaxial cylindrical arrangement has been systematically studied since the beginnings of last century [1-18]. The Peek’s formula [1], experimentally derived, is still commonly used for the estimation of the corona inception voltage in many electrode configurations. It is well established that corona characteristics depend on the electric field distribution between the electrodes and upon atmospheric conditions. In the present study a computational method for the estimation of the positive corona inception field strength in the coaxial cylindrical electrode arrangement in air is presented. A comparison is made with literature experimental data referring to wire-cylinder air gaps in a wide range of wire radii and under variable atmospheric conditions. Peek’s formula is modified to consider, besides air density, absolute humidity variation. 2. LITERATURE EXPERIMENTAL RESULTS A great amount of experimental results on corona inception voltage obtained in the coaxial cylindrical electrode arrangement of wire-cylinder in air has been reported in literature [1-12]. Most commonly, estimation of the corona inception field strength at the wire surface is achieved by using the Peek’s empirical formulation [1], which can be written as: 0 1 i C B E A r ⎛ ⎞ = ⋅ + ⎜ ⎟ ⎝ ⎠ (1) where r 0 is the wire radius and constants A, B and C are given in Table 1, as experimentally derived [1-4, 6, 7, 9, 10] or through a theoretical approach [14, 18]. Table 1: Constants A, B and C to be used in (1). Constants Author A B C Peek [1] from [17] 31.53 0.305 0.5 Farwell [2] 35.00 0.241 0.5 Whitehead and Brown [3] 33.70 0.283 0.5 Whitehead and Lee [4] 39.80 0.212 0.5 Stockmeyer [6] 31.00 0.379 0.5 Zalesski [7] 24.50 0.613 0.4 Waters and Stark [9] 23.80 0.670 0.4 Hilgarth [10] 30.00 0.330 0.5 Hartmann [14] 25.94 0.127 0.4346 Lowke and D’Alessandro [18] 25.00 0.400 0.5