Demonstration and Trial of an Adaptive Protection System

This document reports on an extension of the work covered by ACARP project C25069, published in November 2017, and describes and analyses a demonstration and trial of adaptive protection. The aims of the adaptive protection system originally proposed were:

• To continually calculate, and (eventually), to automatically alter the required electrical protection settings in order to insure compliance with the touch voltage requirements of the relevant standards.
• At the same time, maximise trip margins, based on actual statistical data for each load in the system, in order to minimise nuisance tripping and maximise productivity.
• Provide a useful tool for mine electrical engineers to help them minimize the burden of managing these protection systems, reduce the probability of human errors, and receive data logs and reports that demonstrate compliance.

After producing and testing the system software, the trial was carried out at a modern, high productivity, underground coal mine and operated for a period of several months. The equipment monitored was for development systems using miners and shuttle cars that were connected to DCBs and powered by 1000 V substations.

The protection relay used in all of the substation and DCB outlets was the Ampcontrol IPX. This was attractive to the trial because these relays gather extensive information and could deliver it to the surface at fairly high sample rates without requiring the mine to undertake any significant modifications to their standard operating and monitoring environment. In some ways, the use of these relays was also a disadvantage for the project because they appear to have superior noise immunity on the earth continuity system, compared to older protection relays used by the bulk of the mining industry. This means that the data gathered on trip margins may not be representative of the situation for much of the industry. On the other hand, determining these margins was not an aim of the trial. The aim was purely to show how these margins could be determined for any particular intelligent protection relay, cable system and load if communication was adequate.

The web interface for the system does not seem to have been used much by the mine so we did not gain much feedback on the experience of configuring the system when it was altered. We do not view this as much of a setback because any user experience should only be judged on a system interface that was more like a “well polished” commercial software product. In hindsight, the limited budget requested for the project did not allow the software to be developed and tested to achieve this level of “polish”. The important aim was to capture enough data to show that the critical functions we envisaged were indeed feasible on a real working system and we believe that this aim was met.

Data gathered during the trial clearly demonstrated that:

• Detecting a cable change should be a relatively easy and reliable task for an adaptive protection system (at least on the equipment used in the trial system). The length of cable added can also be estimated automatically by the system in most cases.
• As data is gathered, reliable trip margins can be determined for the one or two (more usually) earth continuity protection circuits (substation + DCB) whose trip settings help determine the maximum touch voltage possible on each load. These margins will then be based on actual data for those circuits rather than gut feelings or general experience. As time goes by these trip margin estimates will improve and will track changes in the system as the mine develops. For the trial equipment, logged data indicated that earth continuity trip margins were always positive (i.e. noise errors lead only to overestimates of the pilot-earth loop resistance). This may not apply in general and it is likely that some protection relays could underestimate this resistance due to the effect of noise. The minimum and maximum resistance estimates are used in two different parts of a calculation - one for compliance checks and one for reliability checks.

The Electrical engineering manager at the trial site was supportive of the aims of the system and stated that he believed the system had the potential to be a useful tool if implemented well. He also suggested the possibility of extending the functions of the system to include over current protection. This was taken to primarily apply to short circuit protection. While not within the scope of this extension project, we have briefly examined the potential for expanding the range of adaptive protection functions for which an adaptive system could be used.

It is a valid question to ask whether increasing trip settings to minimise the chance of a nuisance trip is acceptable under the WHS act requirement of doing all that is reasonably practicable to ensure safety (ALARP principle). We have undertaken a numerical analysis of this issue, as applied to the touch potential risk, and also compared it to other situations that are commonly encountered in mining electrical protection systems. We believe that we have made a strong argument that, if an electrical protection system complies with the requirements of the Lp curve then, for the range of reliable protection settings, there is no identifiable safety gain from a further reduction of trip settings while there may be considerable costs due to lost production. This appears to be completely in line with the requirements of the WHS act.