Page 5 - CE_Industral_Journal_Jan2008

A Novel Approach for Ground Fault Detection

The significance of these hard to detect faults is that A conventional high impedance fault detection tech-
they represent a serious public safety hazard as well as nique that monitors and evaluates the high frequency
a risk of arcing ignition of fires. As such, high imped- components of each cycle of an alternating current is
ance fault detection has been a major concern of pro- described in [3]. The occurrence of a high impedance
tective relaying for a long time. Relaying engineers and fault is determined based on a significant increase in
researchers are challenged for a long time to develop magnitude of the high frequency components for a
a suitable technique that will detect high impedance prescribed period of time and of a prescribed pattern.
faults with a reasonable degree of reliability while be- However, this patent only considers a very specific
ing secure against false detection. feature of high impedance fault currents and does not
Protective relays are usually designed to protect equip- consider high impedance fault currents in a multi-
ment (line, transformer, etc.) from damage by isolating resolution framework. Reference [4] describes a con-
the equipment during high current conditions. High ventional detection technique that compares energy
impedance faults, are typically found on distribution values of load currents to preset thresholds in order to
circuits, results in very little, if any, current. High im- detect an arcing fault on a power line. However, this
pedance faults do not pose a threat to equipment and patent is very basic in its nature and does not consider
by their nature they cannot be detected with conven- or use other features of load currents to detect a fault.
tional overcurrent devices. Nonetheless, the dangers Another high impedance fault detection scheme de-
of a downed conductor are obvious to all. Possibility vised a neural network detection system that is trained
of fire, property damage, and someone coming into to identify high impedance fault is described in [5].
contact with the live conductor are some of the major Although, this study was conducted in the early days
concerns. of neural network development it did highlight the
Also, legal issues in promising a comprehensive solu- potential of adaptive learning systems in high
tion may be very costly for manufacturers (e.g. liability impedance fault detection. That pilot study did not
issues). For example, tripping the breaker follow- involve any experimentally generated data but was
ing high impedance fault detection is not a clear-cut developed and tested using simulated high impedance
choice. While the high impedance fault is a danger, fault data generated using electromagnetic transients
tripping the feeder unnecessarily will create new program (EMTP) software. Furthermore, the neural
hazards by de-energizing homes, traffic signals, of- network used in that study had a relatively simple
fices, etc. The effects of incorrectly determining a high architecture.
impedance fault will have economic and legal ramifi- Reference [6] describes a conventional high imped-
cations. ance fault detection technique that uses a neural
The utility must always have the safety of the public as network to detect high impedance faults which rely
a top priority. However, high impedance fault detection on derivatives of the maxima and minima of the load
has not been possible in the past and realistic detection current. Also, zero crossings are used to detect a fault.
algorithms are not anticipated that can detect 100% of The results of long term high impedance fault detection
all downed conductors, while having 100% security tests conducted with the collaboration of five utilities
against misoperation. The utilities need an economic and including staged faults, naturally occurring faults,
solution and a system that can reliably detect high im- and normal system operation were published in [7].
pedance faults and are also secure in that they do not That study described in great detail some of the practi-
falsely detect a high impedance faults. cal aspects of high impedance fault detection, but it
Some of the earliest high impedance fault detection does show that a 75% high impedance fault detection
schemes involved arcing fault detection techniques that was the best achievable performance using their ran-
used low frequency current components and are de- domness based algorithm.
scribed in [2]. This conventional arcing fault detection BC Hydro and Powertech Labs Inc. tested three high
technique examined low frequency currents, namely impedance fault detection systems including that of
180 and 210 Hz, to determine whether an arcing fault [8]. The most significant result was that the higher
had occurred. A hierarchical detection scheme was frequencies of high impedance fault signatures played
developed that is based on signature energy and using an important role in high impedance fault detection
a dynamic threshold. Although the algorithm was field and in distinguishing high impedance fault from other
tested, its performance limited its practical applicabil- types of faults or normal arcing operations. In another
ity.
Industry Journal 4