CIRED
20th International Conference on Electricity Distribution
Prague, 8-11 June 2009 Paper 0642
EFFICIENT FAULT MANAGEMENT USING REMOTE FAULT INDICATORS Eilert BJERKAN Nortroll AS – Norway
[email protected]
Inflexibility of products designed to order for specific applications and networks. Incorrect installation of the FPI Poor product quality.
ABSTRACT The paper contains a brief historical review of fault passage indicators (FPIs also called FCIs: Faulted Circuit Indicators), both in terms of development and also technology behind the detection. Limitations and advantages are elucidated both technically and economically. Local vs. remote indication (with communications) is discussed. The paper concludes that FPIs is a cost-efficient measure in order to increase operational efficiency during Fault Management in MV distribution systems.
INTRODUCTION Requirements to QoS (Quality of Supply) are increasing in all industrialized and developing countries. This enables the use of FPIs as cost-reducing measures in all MV distribution networks since it reduce duration of outages and thus increase QoS. Remote FPIs combined with a system of remote controlled sectionalizing switches opens up for a very efficient system for localising and limiting faults rapidly. Electro-mechanical FPIs have been used for more than 60 years [1] with limited possibilities of adjusting the units to the specific networks. The first fully electronic FPI was released in 1976 [2]. This was a major breakthrough in fault management since the device was programmable and led to a big reduction of the time spent for fault location. The first units were line-mounted and suitable for detecting over-current in most overhead line, high voltage distribution systems. Similar FPIs for underground cable networks followed shortly after. During the 80’s, new electronic FPIs enabled detection of earth faults in both overhead line and underground cable systems with isolated and resistancegrounded neutral increasing the operational value of FPIs during fault management. Non-directional threshold-sensing FPIs have been widely used during the last decades to track down faults as fast and efficient as possible, reducing usage of man-hours and number of trial connections. Trial connections often lead to switching surges that may lead to damage of customer equipment and installations. While FPIs have been used for many years, utilities have in certain cases been disappointed, due to: Incorrect settings, resulting in false or missed tripping. CIRED2009 Session 3
By investing in state-of-the art microprocessor based FPIs, utilities can avoid many of these problems and then dramatically reduce average outage time. Distributed Generation (DG) is being implemented on a large scale in some countries. DGs connected to the MV grid support the fault current and directional FPIs will therefore be needed in near future. Technology exists for directional detection of both earth faults and short circuits. The directionality is usually added by adding voltage measurements or comparative analysis (several feeders) to traditional FPIs, but other methods do also exist. EFFICIENT FAULT MANAGEMENT The Nordic countries were among the first in the world to deregulate the power markets and to develop penaltyschemes for the utilities depending on severity and duration of outages. These regulations were enforced to encourage new investments to reduce the number of faults. Remote control systems integrated with fault locating equipment such as fault indicators with communication have proved to be a cost-efficient and time-saving solution for all utilities operating under such penalty-schemes. The trend is that most industrialized countries adopt these regulatory frameworks, since customer requirements to quality of supply are increasing rapidly. The penalty-scheme in Norway is based on CENS (Cost of Energy Not Supplied) with differentiated costs for different customers (Industry, households etc.) and affects the maximum allowed network income cap. Many studies have been performed to investigate outagecosts in different types of networks, and most of these studies conclude with the fact that investments in remote control equipment and fault locating equipment such as fault indicators have very short pay-back time (in some cases less than one year [3], [4]).
DETECTION METHODS FPIs use detection methods similar to protective relay equipment but often with simplified signal processing or reduced capabilities. Fault types that are possible to detect using FPIs are : OC : Phase to phase faults (OverCurrent) EF : Phase to ground faults (Earth Faults) DBC : Downed or Broken Conductor CCF : Cross-Country Faults (Double Earth Faults)
Paper No 0642
CIRED
20th International Conference on Electricity Distribution
Prague, 8-11 June 2009 Paper 0642
Suitable methods are chosen with respect to the grounding method of each specific distribution network. The usual grounding methods are : Directly grounded neutral Resistor grounded neutral Isolated neutral Compensated neutral Depending on the fault type and grounding method, different detection methods may be applied in order to detect the faults : Ammetric detection (current measurement) o Absolute treshold o dI/dt (rapid increase relative to load current) o Proportional treshold (relative to load current) Transient detection (zero sequence transients) [5]. Wattmetric detection (zero sequence 50Hz) [6]. Phase Asymmetry Detection [7]. Negative sequence (downed/broken conductor) [8]. Differential methods (combinations of several) Overcurrent faults such as short circuits and cross-country faults are easy to detect using ammetric detection (nondirectional treshold-based current detection). In most cases this is sufficient even in networks with distributed generation as long as CB tripping occurs as normal. In some cases a directional detection would be desirable, and in those cases a voltage measurement is necessary to calculate direction of the fault current flow.
development of a new generation of fault indicators that use the fault-initiated transients for directional detection was started during the late 90’s [5]. By utilizing the fault-initiated transients for directional indication, new possibilities open up compared to traditional non-directional detection. Closed loop networks or networks containing DG are possible. Even intermittent earth-faults may be detected since little or no stationary current is necessary. DETECTION OF DOWNED AND BROKEN CONDUCTORS Downed and broken conductors are not always detected and disconnected by the principal relay-protection. This faulttype leads to very dangerous situations for human beings, animals and assets. Recently, distributed systems are able to detect all kinds of broken and downed conductors, including back-feed faults, broken loop and even blown HV-fuse. Such systems have been tested several times with success in real MV distribution systems [8]. When conductor breaks and falls down on the load side rather than the supply side, a back-fed fault situation occurs. The faulty phase is energized via the distribution transformer depending on the connection group of the transformer and the degree of secondary loading of the transformer. MV/LV station Activated detector Inactive detector
The fault current during phase to ground faults are fairly low for isolated and compensated neutral. In some cases directionality must be added to pinpoint the fault location. This could either be done by using a transient based earthfault detection [5], a wattmetric detection using a 50Hz phase comparison of the zero sequence voltage and current [6], or other new methods such as Phase Asymmetry Detection [7], which require no voltage measurement but can be applied in for earthfault-detection in compensated networks. The latter together with the transient method is also applicable for intermittent earthfaults. In long overhead networks the short circuit power will vary greatly with the distance from the feeding substation. In these cases it is beneficial to use dI/dt method for overcurrent detection, since this gives less configuration of FPIs. With treshold sensing one must calculate the setting of each unit. With a proportional treshold a far-end fault will not be detected. dI/dt is sensitive to load pickup. EARTH FAULT DETECTION IN COILCOMPENSATED NETWORKS During the last 20 years, many utilities have converted to coil-compensated networks leading to lack of methods for detecting earth-faults by FPIs. The Peterson-coil is mainly used to compensate for the capacitive fault current during a phase to ground fault in order to suppress the arc and leave the customers undisturbed during temporary faults. The traditional threshold sensing fault indicators can only be used for detection of earth faults on relatively short braches since the steady state fault current is very low. The CIRED2009 Session 3
HV MV
Broken conductor
Figure 1: Downed conductor detection by distributed LV measurements Equivalent fault resistance (in series with the distribution transformer) can easily reach values above 100 kOhms, making such faults undetectable by the relay-protection, but for units measuring on the low voltage side of a set of distribution transformers in the network as shown in figure 1, the fault will clearly be visible for units outside the fault side due to the voltage imbalance. This makes dangerous and difficult fault types detectable. CROSS-COUNTRY FAULTS (Double Earth Faults) During cross-country faults there is often a mix of OC and EF detection and tripping by feeder relays. In many cases a relay mis-operation is the final explanation because the -
Paper No 0642
CIRED
20th International Conference on Electricity Distribution
Prague, 8-11 June 2009 Paper 0642
country fault may disappear during the long reclosing-cycle (typical for aged surge arrester problems combined with a single pole EF). FPIs may help interpreting such a complex fault situation in cases both when there is cross-country fault within the feeder or shared between two feeders. In case of isolated or resonant grounded neutral, the FPIs may be able to differentiate between a phase to phase overcurrent (OC: I
