الفهرس 
Chapter 1: IntroductionOnly 14 pages are availabe for public view
 |  | from | 16 |  |  |
| | | from | 16 | | |
Abstract
In this thesis, a novel fault management scheme in smart distribution system is
presented. In this scheme, the fault indicators located at secondary substations are
replaced with the fault locators at the lateral substations. This is accomplished via
locating a panel substation at each lateral and the faulted section is identified via
communication between these panels. The isolation of the faulted section can be
implemented through control signals between lateral panel substation and the faulted
section isolators. The service restoration is achieved by a direct communication
between the lateral panel substation and the feeder primary substation. The proposed
scheme is characterized by lower number of communication hops, associated with
faulted section identification, than those exist in the reported control methods. This
reduction is mainly due to the dispensing with information from the substations which
are designated as secondary substations. In order to implement this fault management
scheme, the protective devices must be coordinated and the fault location should be
determined with acceptable accuracy.
The main protection available for distribution networks is overcurrent protection
which includes directional as well as non directional overcurrent relays. One of the
challenging applications is inserting distributed generation (DG) to cascaded parallelradial
distribution feeders. This configuration may yield mis-coordination of the
overcurrent protection. In order to overcome this mis-coordination, a proposed
adaptive overcurrent protection approach for cascaded parallel- radial distribution
feeders in conjunction with the DG is introduced. In this adaptive approach, the miscoordination
and insensitivity of the protective relaying due to the DG insertion are
avoided, where the overcurrent relay settings are adjusted according to the direction of
fault current and the status concerning a single feeder or two parallel feeders in
service. This is accomplished via communicating the relay installed on one feeder with
the circuit breakers associated with the other feeder and vice versa whether the DG is
interconnected to or disconnected from the distribution network. The relay settings are
calculated in all possible current directions using the genetic optimization algorithm. The Performance of the proposed approach is examined via a detailed simulation of a
real field medium voltage distribution feeder.
Other challenges of fault location determination in the distribution system are
encountered due to the presence of the DG and the distribution transformer connection
type. These configurations will cause various loops of fault current which cannot be
measured by the relaying points. This will yield additional errors in the fault location
calculations. In order to overcome these challenges, a new algorithm for fault location
determination is introduced. In the proposed algorithm, these sources of error are
overcome. This algorithm is independent of the fault impedance, earthing state, and
load transformer connections type. Also this algorithm is suitable for distribution
system in conjunction with DG and is applied for parallel feeders. The Performance of
the proposed algorithm is examined via simulating both of a real feeder from the
Egyptian distribution system and the IEEE-33 bus system. Also, the proposed
algorithm is compared with two other existing methods. All applied test results have
corroborated the accuracy of the proposed algorithm for locating faults in distribution
systems.