Introduction: Condition monitoring plays an increasingly important role in asset management and diagnostics for high-value equipment. New technology and advances in sensing capabilities enable us to understand more about the asset and thus make optimal maintenance decisions (e.g. maintain on condition). Minimising the requirements for installation and maintenance of these sensors, and removing the need for cables and batteries are the key aspects of the desirable fit and forget functionality.

Existing approaches to substation diagnostics typically involve mains-tethered instrumentation for data acquisition. It is prohibitively expensive to roll out this type of scheme widely due to cost and cabling constraints, which inevitably leaves gaps in condition monitoring coverage that should ideally be filled. In addition, diagnostic systems have become significant assets in themselves, requiring trained personnel to operate them. This approach adds additional complexity to the task of a monitoring engineer, whose primary concern must be the operational state of plant rather than the intricacies of a diagnostic system. Therefore, a non-obtrusive, integrated approach to diagnostics should be followed.

Recent developments in miniaturisation of digital electronic devices have fuelled the development of wireless sensor network technology. These networks are made up of a number of discrete sensor nodes, which integrate processing, sampling, storage and communications capabilities. By taking advantage of this technology, wireless diagnostic sensors have the potential to increase condition monitoring coverage without the need for cross-site cabling, simplifying deployment and reducing costs.

Through identifying general requirements for wireless condition monitoring systems, a modular approach could be defined for a multitude of sensors to be attached to the same underlying platform (for instance: RF, ultrasonic and thermal). In addition to sensing, wireless sensors such as this with suitable analytical capabilities can also support a level of on-board defect diagnosis. By diagnosing defects on-sensor, the volume of monitoring data can be drastically reduced at source so that only pertinent defect information is transmitted to monitoring engineers. This reduces the burden of transmitting data back to corporate networks, increasing system scalability and minimising the requirement for wideband communications links.

An initial laboratory study into this type of approach, targeted at Partial Discharge (PD) monitoring, has resulted in a promising new diagnostic technique built upon wireless sensor technology. This method has demonstrated detection and basic classification capabilities and, based on the knowledge gained from this study, implementing the UHF technique on a wireless sensor node has been recognised as feasible. Based upon this prior work, a wireless condition monitoring platform technology demonstrator could be created using partial discharge detection and diagnosis as a reference application.

Objectives: The aim of the project is to create:

• A low-cost, readily distributed diagnostic system architecture suitable for operating wirelessly within a substation

• A report detailing the feasibility and expected functionality of fully autonomous wireless sensors deployed in a range of environments when integrated with energy harvesting devices.

• Capability for integrating the technology within the Smart Asset Management System (SAM) to provide real-time diagnostics such as PD, to monitoring engineers

• A technology demonstrator based upon a low-power partial wireless discharge (PD) detector and diagnostics package that can be used for PD identification.