Our products have a number of proven applications in the engineering sector created by a common desire to provide early fault detection for a range of assets. In many cases, it is too difficult or expensive to provide an accurate simulation model of an asset and so understanding asset performance from identified patterns in the performance data provides a valuable means of identifying potential problems.
With the availability of wireless technology and affordable computer storage, it is possible to monitor many hundreds or even thousands of sensors installed to inform on asset performance. Over time, such intensive monitoring creates substantial, large data sets and when spread over many assets, this data becomes too large to analyse so this valuable resource tends to be ignored. The SDE range of products have been designed to allow users to create models based on features in this data, detect similar patterns across a fleet of assets and monitor current performance using these models. These models can detect abnormal or unexpected events across the asset (using AURAalert), assist with a diagnosis of an unknown event (‘have I seen this before’) or assist with incident investigation.
Two typical examples of the use of SDE are:
- During transient situations when plant and machines are being started up or closed down or there is a sudden change in load. This is where many malfunctions and faults start to become evident and SDE is able to distinguish between even small changes in behaviour.
- Detecting patterns in vibration data where modelling of real world behaviour is difficult to achieve.
As many fixed assets age yet continue to be used often way beyond their design life or capacity, it has become even more vital to monitor their condition. The functioning of bridges, tunnels, rail track, overhead power systems, pipelines and buildings are critical but opportunities for failure are many, including deteriorating materials, failure of design or of construction and improper load ratings. Consequently, there is an increasing trend towards continuous monitoring of structural integrity of these assets. It is vital to detect early signs of failure rather than wait for a catastrophic event; SDE’s pattern matching approach is ideal to find random or unpredictable events across a range of assets where manual inspection of the data due to its size and complexity is clearly not feasible.
The high economic cost of unexpected down-times makes performance monitoring of assets a priority for power generation companies. A typical turbine system including steam generation and boiler system may collect data from thousands of sensors once every second. Monitoring of performance is essential and SDE’s role in understanding and modeling asset behavior using the extensive archived data available is proving to be a valuable tool for this industry. Typical example use cases include evaluation of a multi-stage gas-powered turbine and a super-critical steam generation system.
This sub-sector of the power generation industry provides an excellent application for the SDE tool range. Wind turbine farms can comprise of hundreds or even thousands of individual turbines and with the performance of these assets still under assessment and the current high failure rates impacting on the economics, there is a need for analytical tools to model the performance of the different sub-systems of the wind turbine. Simulation based modeling has limitations in a complex, changing environment and so SDE is used to understand farm performance in different wind conditions and to categorize failure modes with individual turbines.
The aerospace industry focuses on extremely high safety standards, rigorous testing of assets prior to putting them into service complimented by a comprehensive maintenance regime. Consequently, these assets exhibit high reliability and unexpected failures occur relatively infrequently. Nevertheless, the cost of delays and cancellations in the civil aviation industry is expensive and damages reputation whilst lack of availability of defense aircraft cannot be an option. Consequently, monitoring of asset performance is becoming increasingly important both during routine testing and during flights and with significant data volumes across large numbers of assets, SDE Professional and SDE Premium are valuable tools for diagnostic engineers. In particular, SDE has been used to assess complex vibration data where it is not possible to develop simulation based models which accurately reflect real-world behaviour. Incident /accident investigation requires comprehensive analysis of historical performance data and SDE’s visualization and browsing features give engineers the ability to rapidly assimilate large amounts of signal data. Events that are discovered in the data leading to the incident can be evaluated quickly answering questions such as ‘have I seen this event in this asset before’ or ‘ has this event been seen in other assets in the fleet?’.
Cars, trains, ships and other vehicles are now fully instrumented so that their performance can be assessed allowing early detection of fault conditions and pre-planning of maintenance. Using SDE Professional or SDE Premium, an engineer can inspect large data sets, identify anomalies and develop routine models which can be used to assess performance in real-time.
Typically, process industries are interested in running their plant 24/7 and any asset failure, even minor assets such as pumps and filters, which result in idle time is extremely expensive. Given the range of assets involved, SDE’s AURAalert has the advantage of providing an inexpensive way of modeling a wider range of assets to detect abnormal performance whilst SDE’s pattern matching allows further assessment of an abnormality.