Summary
Fault location on highly branched networks is particularly problematic, especially with a limited number of observations from the network. This paper focuses on two emerging fault location algorithms. The first is based on a multi-ended travelling wave approach relying on the timestamp of the fault induced transient. The second is based on the principle of time reversal and back injection. The methods are compared on a large, rural 11 kV distribution network. The simulations are performed using JMarti frequency dependent line models in the ATP-EMTP. The performance of the techniques are evaluated with particular emphasis on the number of observation points required for a fault to be located anywhere on the network. It has been found that using only a single observation point does not always unambiguously locate a fault, but performance improves as more observation points are added. With observation points on every branch termination, both the time reversal and timestamp methods are shown to be able to locate a fault at multiple locations on a branched network.
Additional informations
Publication type | ISH Collection |
---|---|
Reference | ISH2015_552 |
Publication year | 2015 |
Publisher | ISH |
File size | 466 KB |
Price for non member | Free |
Price for member | Free |
Authors
Jiafu Wang, WU Lin, Zhang Chaohai, LIU Yi-yong