Research of a transformer fault diagnose method using multi-source recording waves

Power transformer is one of the most important and expensive devices in power system. Fault in a transformer has a large influence so that the operation statuses of it directly affect the safety of power system. However, it is difficult to diagnose fault location because transformer faults may be caused by many factors and occur in many places for its complex structure. This paper proposed a transformer fault diagnose method by using voltage and current from fault recorder and protection devices of all voltage levels, as well as sequence components of them. Characteristics of current from all voltage levels are recognized by analyzing sequence components of currents derived from special TAs and bushing TAs under several fault situations. Then, a correspondence relationship between the transformer fault location and fault record wave is established to determine fault location quickly. Finally, the paper analyzed a real case of 220kV transformer fault to verify the effectiveness and usefulness of the method. Transformer fault can be quickly and accurately located using the proposed method, which makes the outage time shortened, thus improving the stability and security of power grid.