DESIGN AND TESTING OF A NEW MICROPROCESSOR- BASED CURRENT DIFFERENTIAL PROTECTION SCHEME FOR TEED CIRCUITS

Ali  H. HUSSEINI

The advantages of employing microprocessors in Protective relaying for Power Systems are well known and a number of digitally based schemes are now available for protecting EHV plain feeders. Teed circuits are more difficult to protect than plain feeders and although differential protection is generally regarded as the method best suited to such circuits, however it is only recently that it has become feasible to be able to develop new high speed digital techniques for fault detection on such lines using Current Differential Principles. This has been made possible due to two main reasons which are :

         1) The advent of fiber optical communication channels .

         2) The advancements made in high speed digital signal processing (DSP) chips.

This thesis describes in some detail the developments undergone in the design and building of the hardware for an EHV teed feeder protection scheme based on previously developed CAD techniques. The relay operating principle involves an advanced differential scheme based on master and slave principles. It utilizes wide‑band fiber optic link for data transmission and signaling between local and remote ends. Novel techniques of digital processing‑and filtering along with their dedicated hardware explained together with a sophisticated decision process. The relay has been tested in the laboratory using fault simulated current waveforms  which are generated on a local programmable transmission line (PTL).

The relay offers a high performance and has various advantages over existing differential schemes, such as immunity to the presence of feed‑around paths and CT saturation and an increased sensitivity to high resistance faults. Its operating times for a majority of faults are in the range of 4 to 6 ms.

The thesis clearly shows that the relay design based on CAD techniques can be engineered into practical model using DSP technology and this is vividly demonstrated by the very close correspondence between the Differential and Bias signal waveforms attained from the practical and theoretical models.