466 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 22, NO. 1, JANUARY 2007 Concentrated Discharges and Dry Bands on Polluted Outdoor Insulators Krystian Leonard Chrzan and Federico Moro Abstract—This paper describes concentrated discharges on pol- luted insulators in the laboratory and in the field. The authors have observed the concentrated discharges at the high-voltage laborato- ries in Stuttgart, Zittau, and Mannheim, Germany, and Wroclaw, Poland. The concentrated discharges were also documented under natural conditions. These discharges are very dangerous especially for high-voltage apparatus, bushings, and polymer insulators. It has been shown that due to uneven voltage distribution at very low surface conductivity, the concentrated discharges can ignite even under uniformly polluted and uniformly wetted insulators. Index Terms—Leakage current, pollution, surface conductivity, surface discharges. I. INTRODUCTION O UTDOOR insulation is contaminated by natural or in- dustrial pollution (sea salt, salt sands, industrial dust). Additional sources of pollution are rain (e.g., acid rains) and gases—especially sulphuric oxide and nitric oxide (SO , NO ). With heavily polluted areas, the surface conductivity on outdoor insulators can exceed the value of 100 S, which leads to arcing development and eventually to flashover at continuous operating voltage. Due to large environmental destruction in Middle Europe during the 1950s and 1960s, the faults caused by pollution flashover became an important factor in reliable power de- livery. After the introduction of better insulators, the outage rate was limited to the acceptance level. During the 1980s and 1990s, the dust and gases emission in Middle Europe decreased considerably due to the production limitation of heavy industry and the introduction of new clean technologies. As a result, pollution flashover seems to no longer occur. Despite this comfortable situation, the stress caused by pollution should not be ignored. This paper describes the phenomenon that can decrease the flashover voltage which is very dangerous for high-voltage ap- paratus and for polymer insulators [1]. II. ELECTRICAL DISCHARGES DURING THE POLLUTION TEST IN THE LABORATORY During pollution testing, the real conditions are modeled by a simplified procedure and during a limited time. The method Manuscript received June 20, 2005; revised February 7, 2006. This work was supported by the German Foundation DAAD. Paper no. TPWRD-00355-2005. K. L. Chrzan is with the Wroclaw University of Technology, Wroclaw 50-370, Poland (e-mail: krystian.chrzan@pwr.wroc.pl). F. Moro is with the Dipartimento di Ingegneria Elettrica, Università di Padova, Padova 35131, Italy (e-mail: federico.moro@di.unipd.it). Color versions of Figs. 1–8 are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TPWRD.2006.887093 called “flow on” is characterized by a very simple procedure and short test duration. According to this procedure, the insulator is first covered by a contamination suspension, and then the com- pletely wet insulator is stressed by the voltage whose value is close to the 50% flashover voltage [2]. During the voltage ap- plication, the insulator is not wetted. The pollution layer dries quickly due to the large current value at the test beginning and intensive discharges. As a result, flashover occurs usually in a time interval from a few tens of seconds up to 2 min after the voltage application and the current reaches a value of order 1 A. The test duration of other procedures is longer and insulators are continuously wetted (e.g., the standardized “solid layer” or “salt fog” methods). Since the aim of the tests is the evaluation of 50% flashover voltage, the electrical discharges are very intense. The test voltage is usually higher than the continuous operating voltage and the degree of contamination is very high. As a result, the flashover probability is close to 50%. On the other hand, under natural conditions during dangerous events, such as rain, drizzle, or fog, the flashover probability is usually very low, close to zero. On lightly polluted insulators, the leakage current can reach a value of about 1 mA and, thus, the weak micro discharges are not visible, especially in sunlight. With heavily polluted insulators, the discharges can be intense, visible during the day and leakage current can reach a value of tens of milliamperes. Intense discharges, with currents in the range of a few tens of milliamperes, are not stable. Their length can exceed the shed distance and often result in bridging phenomena, thus decreasing the leakage path length. The electrodynamic and thermal forces result not only in arc elongation but also in them moving around the insulator axis. The typical discharges during pollution testing using the “flow on” method are shown in Fig. 1(a). The characteristic mode of these discharges is a very uniform distribution along the leakage path and long total length of the arcs. Immediately before flashover, the total arc length reaches about 60% of the insulator height. In practice, the arcs burn on each rod between the sheds. In tests carried out according to the “salt fog” or “solid layer” methods, the discharges often concentrate at two places or even at one place on the insulator surface. This discharge type was documented probably for the first time by Gerhard Reverey from Studiengeselschaft fuer Hochspannungsanlagen e.V. in Nellingen, Germany [Fig. 1(b)] [3]. The weak discharges are not easy to observe due to dense fog or steam inside the pollution chamber. When the test stops and the fog is quickly removed, the dry bands become visible [gray areas in Fig. 1(c)]. The remaining wet-insulator part is dark. 0885-8977/$20.00 © 2006 IEEE