AUTOMATIC STRAY CURRENT MONITORING IN THE POWER SUPPLY CONTROL SYSTEM OF A DC RAILWAY SECTION: A COMPUTATIONAL ASSESSMENT

Abstract

On DC-electrified railway sections, part of the traction current leaks from the running rails into the ground and causes electrocorrosion of buried metal structures, so the power supply control system of a section should monitor the condition of the track insulation. The aim of the study is to substantiate the set of measurements and the action rules of an automatic stray current monitoring system using a computational model. A two-layer rail–earth model with distributed parameters was used; the longitudinal rail resistance (15–20 mOhm/km) and the rail-to-earth resistance (from 100 to 0.4 Ohm·km) were taken from EN 50122-2 and published studies. The rail potential and leakage current were calculated for single-end and double-end feeding and feeding zones of 1–20 km, and a Monte Carlo simulation was used to test the automatic estimation of the rail-to-earth resistance from the difference of rail currents at two points with sensor errors of 0.1–1 %. For a train current of 1000 A and a 10 km zone, a decrease in the rail-to-earth resistance from 100 to 0.4 Ohm·km reduces the maximum rail potential only from 99.8 to 72.2 V, whereas the earth current grows from 2.5 to 409 A; hence the rail potential cannot serve as the main indicator of insulation degradation. An average leakage current of 2.5 A/km is reached at about 2.0 Ohm·km for a 2 km zone and about 9.8 Ohm·km for a 10 km zone; double-end feeding reduces the leakage current about fourfold. With 0.5 % sensor error, the current-difference estimate detects the limiting insulation state in a 10 km zone, while zones of about 2 km require more accurate sensors or averaging. A structure of the automatic monitoring loop is proposed. The results are computational and require verification by measurements on an operating section.