Asset Management - Monitoring Challenges

Monitoring challenges

Global efforts to drastically decarbonise and decentralise our energy systems require more real-time visibility and control at more locations – particularly for assets which are remote or normally inaccessible. However, constraints with conventional monitoring technologies block or limit the insights which can be gained – networks of PMUs are very expensive, and the data accuracy and frequency of conventional SCADA systems is low.

This is unfortunate because there are many opportunities for significant operational efficiencies and cost savings. Transmission and distribution systems contain many decades-old circuit breakers, transformers, and other assets which are especially susceptible to failure, and such unplanned outages can be very costly and disruptive. For offshore wind farms, outage and repair costs are the single biggest opex challenge, with average lost income of £24m per cable failure.

A modern solution

Our distributed sensing hardware, combined with our Synthesis™ software tools, overcomes the challenges with conventional monitoring schemes and will ready the industry for a new approach to asset condition monitoring. This is achieved through the combination of three key unique features:

1. Continuous Point on Wave (CPOW) measurement for sample-by-sample capture of all signals at rates of up to 35 kHz
2. Synchronous sampling of diverse physical quantities including voltage, current, temperature, vibration, and strain, to enable combined analyses
3. Ability to cost-effectively reach and instrument remote generators and other assets, yet providing granular, no-compromises analytics

This provides a low-risk proposition to overhaul the quality of asset condition monitoring practices. Furthermore, the same infrastructure can be used to provide real-time monitoring, protection, and control functions, resulting in further cost efficiencies.

Simplified analysis

Synthesis™ makes it easy to get instant feedback and understanding data from different locations and over different date or time ranges. A significant practical barrier to analysing large-scale monitoring is that data can have gaps or be of limited use due to lack of dependable time synchronisation. Dealing with such “bad data” can involve significant pre-processing. Synaptec’s monitoring approach mitigates these problems.

Grid benefits

Power systems are complex electro-mechanical “machines”. There is continuous interaction and feedback between electrical and mechanical aspects. Only a unified holistic approach can comprehend the total system and evaluate performance – and any degradation. We monitor and correlate in space and time, using electrical and mechanical measurements, to better understand interdependencies. For example, our system monitors thermal behaviours of lines, transformers, and other assets, and readily correlates this with electrical phenomena.

Our sensing hardware will automatically provide fault-recorder level of data and insight for every sensor location. It provides a framework to allow other functions to be added, and allow simple “what if” scenarios to be prototyped quickly and easily – to enable cross-correlation of events for interdependencies. For example, events in one circuit or part of a circuit may create impacts elsewhere that are not currently known. With access to sufficient data, it may be possible to better characterise the “signature” of the fault or other event and therefore improve detection models to capture future events. For example, power oscillations can have disastrous consequences in the grid, as was experienced during the partial Great Britain blackout on 9^th August 2019, and we are performing R&D in this area. Our system can also act as a data historian for critical post-event analysis.

Benefits for wind farms and other renewables

Our sensing platform automatically performs power quality analysis to the 100^th harmonic (or higher, if required). This enables discrete tracking of PQ at every asset – applied to instrumentation at individual turbines, the offshore substation (if present), and at the grid connection point. We can determine trends in harmonic profiles over time and perform anomaly detection i.e. by comparing harmonic profiles of similar assets for differences, or by detecting unusual behaviours that emerge over time. Analysis of this rich PQ data can be used to identify impending component failures e.g. in turbine generators, gearboxes, transformers, power converters, and filters.

Therefore, monitoring any divergence over time, leads to the following benefits:

• Early warning of impending mechanical failure via anomalous harmonics or mechanical readings
• We can predict repair/replace decisions with greater statistical confidence, thereby enabling targeted, efficient maintenance
• Avoid unnecessary downtime
• Extend life of cables and WTG
• Opportunity to correlate electrical PQ data with DTS and DAS