A Fault in power system can occurs due to many factors such as lightning strike, crane or tree encroachment, human error, and many other factors. Conventional fault location technique requires the workers to go to the substation after a fault occurrence to extract the fault information from protection devices. This process is time consuming and will make the faulted line to be reenergized at a delayed time. This research developed the wide-area fault detection and location algorithms for transmission network where the fault information can be gained directly from a central location. Besides that, the algorithms also have been tested for charging current effect and applicability on hybrid non-homogeneous circuit. For wide-area applications, the measurements are taken from strategically located phasor measurement units (PMUs) in the network. This research developed a PMU placement strategy using the proposed five rules to determine the optimal number and strategic locations of PMU. Based on the results of PMU locations, the measurement sources for all lines in the networks were determined. For buses without PMU, the connected line currents and bus voltages were determined using Kirchhoff’s current law (KCL) and Kirchhoff’s voltage law (KVL), respectively. There are two methods been developed to detect the fault occurrence and determine the actual faulted line in the network. The first method is the comparison of total positive-sequence fault current between pre-fault and fault conditions. When a fault occurs at a line, the total combination of currents from both ends of the line will be very high compared with its value during pre-fault time. The second method is the negative-sequence bus voltage checking method. This method was used in conjunction with the first method to determine the actual faulted line among several lines which share the same measurement sources. Following next is the fault location. The wide-area fault location algorithm was developed based two-terminal impedance-based fault location method using the negative-sequence values as the inputs. Both fault detection and location algorithms have been tested on IEEE 5 and 39-bus networks for various fault locations, fault resistances and fault types. The results showed that the algorithms successfully detected the faults at all lines and accurately located all the faults with average estimation error less than 1 km and 1 %. The accuracy of fault location also is not affected by charging current. Finally, the modified algorithms also correctly detected the actual faulted sections and accurately estimated the fault locations for hybrid circuit.
