 |
P A D |  |
Arcing earth faults in cable networksAn increased attention has been directed towards the problem with arcing earth faults in cable networks. This type of fault has become more frequent as the earth fault currents have been reduced. When a single-pole earth fault occurs, the "neutral point" of the network is transferred to the fault location, while the neutral point is charged with phase-to-neutral voltage. When the fault location voltage is close to zero, with limited current, the tendency of self extinction is high. In that way, deficiency in the insulation may cause repetitive flashovers that quickly self extinguish.These repeated earth contacts of short duration do not result in a stable 50/60-cycle current that can be used by the protections for detection. However, a more specific "neutral point" voltage is obtained which, at every ignition, gets closer to the phase-to-neutral voltage and thereafter slowly falls towards zero with the time constant of the network. When the arcing earth faults are being repeated at a high frequency - which is normally the case - the neutral point voltage protection (NUS) will trip the entire substation after a preset time if none of the line protections be able to identify the misconducting line. During the past decade, the suppliers have tried to avoid these difficulties by giving the earth fault protections functions that count zero sequence current peaks during a specific period of time or similar functions with varying outcome. One supplier is now about to release such an improved algorithm, also including the longer time constant of the voltage in its measuring criteria. A completely new measuring method has been developed by Protrol. This method focuses on the phase currents (2 Iph+3Io or 3 Iph) which contain all information regarding both current peaks as well as rechargers. Furthermore, there is a characteristic difference in these currents when comparing the misconducting and the correct line. The "normal" procedure of using the residual current and its characteristic current peaks when dealing with earth faults is not entirely undisputed, as peaks occur even in healthy lines. The network design, the earth capacitance of the line/-s as well as where in the network the fault occurs and the possible existence of varying fault resistance, even if it is very low, decide to a large extent the appearance of these peaks. The "residual current peak method" may therefore be difficult to adjust to achieve a reasonable balance between reliability and stability. This is due to the influence on the "peaks" by the entire network, the protected line, where on the protected line the fault occurs and if any fault resistance exists. The measuring method developed by Protrol (Phase Asymmetry Detection, PAD) is not sensitive to the amplitude of the current peaks, nor to the properties of the network, the degree of tuning or the phase displacement of transformers. Therefore, the measuring criteria is much more precise than that of other known methods. Apart from higher reliability in detection, this also leads to an easier way of adjusting the protections and independence of the common power system appearance. All protections can be set accordingly in the same way. The protections therefore maintain their characteristics at switches and changes in the network configuration. Protrol also introduces a new protection (Arcing Earthfault Protection) PPC to be used for retrofit in existing substations. These protections should be set with a time delay of 0.5 sec. shorter than the neutral point voltage protection thereby making it possible to disconnect the faulty feeder and avoiding the disconnection of the entire primary substation. In this way, the life of existing protections can be extended and at the same time be given "state of the art in protection", without any real change in existing protection scheme. |
|
|
