Learnings

Outcomes

The project has developed and delivered RALF radiometric arc fault locating devices which offer potential to locate arc activity associated with faults on the 11kV overhead line network. 

Trials at the PNDC test centre indicatyed the units can detect faults from up to 5km. However the network trials have been unable to validated the device performance. The trials have shown that battery life is the biggest issue to deployment on the SPEN network. 

Lessons Learnt

Initial System design –

The design should try to minimize both the weight of the unit on the pole, and the power dissipation within the electronics to maximize the battery runtime. The use of a tunable, narrowband rf input input stage is essential.

 
Development phase –

This stage of the project has dealt with the implementation of hardware and software to achieve the specification of the RALF system resulting from the earlier stage. Excepting the normal developmental issues expected in a project of this nature, the main learning points are:

• A greater degree of testing than originally anticipated was needed to debug the errors in the system. Although extensive factory testing was conducted on individual FDUs, it was only when a cluster of 4 geographically remote FDUs were deployed in a full end-end test with the server that many of the errors could be identified.
• The performance of the 4G network is not always ideal and provision within the server has been made for situations where data from one of the FDUs is unduly delayed

Evaluation phase –

This stage of the project has dealt with the further development of the system to perform correctly under field service conditions.  The main learning points are:

• The project has demonstrated that arcs generated by utility switching can be located to an accuracy of 64 m when the FDU is located 5 km from the arcing source.
• In order to allow the system to work at distances of ~5 km it is necessary to increase the sensitivity of the FDU, which results in a large number of triggers being recorded.
• The PNDC testing demonstrated that recording a large of number of triggers does not result in any false source locations: the occurrence of matching signals from the FDUs only occurred during PNDC switching events.
• Large numbers of triggers can prevent the system from running in real-time since, on occasions, the 4G network could not keep up with the trigger uploads to the server, leading to a ‘backlog’ of data.
• The backlog issue can be addressed by either
  -Desensitising the FDUs – this could potentially lead to a reduction in operational range (i.e. less than 5 km). This approach is presently being trialed.
  -Uploading the data on demand – in this case the server would request the FDUs to upload the relevant data. This approach requires the time of the fault to be known
• Field trials undertaken with the RALF devices fixed to wooden poles have been undertaken. The devices were deployed for several months however no faults occured in that duration.
• Operational staff identified that battery replacement for the devices required weekly visits.