Fault and the induction of the distributed energy

Fault Detection Methods for Interconnected Power Distribution NetworksA.M LiaqatDepartment of Electrical Engineering, North China Electric Power University Beijing P.

O. 102206, PRCEmail: [email protected]: As per recent advancements and the induction of the distributed energy resources to existing power systems its really very demanding to analyze the systematic parameters for the maintainability, reliability, safety of engineering application products and emerging load demands.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

Distribution networks are getting complex day by day with the interconnection of various energy generating resources. Prediction of faults and their locations in such situations is extremely important for utilities in hassle free running of system. Load growth and high penetration of regional power makes the smart grids more itching to utility personals. Various theoretical and practical methods are there as part of the global research activities in said concern. This research is in order to make researchers aware of the alternate method of software implementation in lieu of Micro-PMU in system with the ability to obtain, communicate and store the real time data for calculations of systematic parameters. Its capability to determine the phasors in time domain is an advantage over the existing devices. IntroductionPower systems are always prone in terms of fault conditions as they are the source of quite deep interest regarding functionality, integration, up-gradation and modification.

Fault usually interrupts the systems normal functioning leading to breakdowns and total failure in extreme cases. To reduce damage and isolation of fault to small area is always a common interest for various researchers. Literature regarding said methodologies involves the following methods:Relay based methodsNeural/ knowledge based MethodsModel based MethodsSynchrophasor based MethodsMicro-PMU Relay based methods are quite conventional and often very familiar in power industry due to less technicality involved in obtaining the data from various nodes under fault condition and corresponds them with normal data to judge the nearest point of interest. System variation in terms of voltage, current, frequency and power are the relative parameters to predict the relay operation under critical conditions. Majority of the operational relays belong to family of Earth leakage relays, over-current relays, Impedance Relays and Phase failure relays. In case of High impedance faults concerning line to ground the voltage and current values are not stable which usually lead to new advance methods to involve for relative fault diagnosis. V & I sampled are not accurate and leading towards system damage in case of functional delay. Accuracy of system is compromised.

Knowledge based methods involves the pre-state and post-state analysis of system parameters for feeding the neural system in order to be aware of the fault conditions in abnormalities. Artificial neural network, fuzzy neural network, adaptive neural network (AANN). Many of the advanced methods for fault calculation and detection are using the above said methods due to their capability of predicting nature and accuracy of 91%- 97.3%. Problem pertaining to these methods are also the same as of relays that the sampled signals are large hence the responsive time is slower leading to delay operation of protective equipments installed at the vicinity. Model based methods of fault locating strategies are very common in the field as they require a residual signal to be generated as calculations for the aforementioned purpose. Its importance can be found by the fact that industry and academician are using these methodologies since last four decades.

Difference of the output and the estimated output is equating the residual signal. Fault is diagnosed whenever the residual signal deviates from zero value. Transients, noises, transformer tapping and disturbances lead to no signal deviation as they are adjusted in the threshold level of residual signal. Fault alarm turns on when system meet the below mentioned condition,Fault Alarm ? Threshold levelThis method is an easy approach toward fault dependent models of power systems and the designing of fault detection filters based on observatory mechanism.