Speaking in the conditional maintenance track at the 2018 National Instruments' NI Days event in Paris, Lodovico Menozzi described a ‘big shift’ in maintenance from manual to automated online data collection. Although this is not without its disadvantages, ‘walking around a plant can provide awareness of future issues’. But the new approach allows 24x7 analytics and remote diagnosis of captured waveforms by SMEs from ‘anywhere with network access’. NI provides connectivity with OSIsoft PI, OPC UA, Asset 360 (Black & Veatch), Avantis Prism and more. NI tools are used to view trends and ‘displace’ other software for vibration, temperature and motor current signature analysis from the voltage bus. IR thermography and EM signal analysis detect hot spots and insulation probs in transformers. All this is now delivered from a single software tool CompactRIO, deployed at the network edge. One oil and gas use case is the flagship Texmark Chemicals refinery of the future.
Thierry Romanet presented a rejuvenated ThreatScan, BHGE’s real time pipeline impact detection system. Hydrophones along the line use local real time processing spectral analysis to detect signatures from e.g., digging. It is not practical to send all the data to the cloud because of bandwidth limits. Real-world tests with diggers breaking into (empty) pipes give source spectra types. These are convolved with attenuation data on pipe type, fluid and burial conditions. ThreatScan has been around for a while. Back in the day, the system was drowned out with nuisance alarms. Today, improved firmware and neural net algorithms can distinguish noise types from normal operations. The system was completely redesigned in 2017 with Step Automation and the software ported to an NI sbRIO-based computer. Google Map and Beacon also ran. In April 2017 the system detected its first real impact on an offshore pipe. Subsequent ROV and pig surveys found damage 18 meters from the estimated location.
Less digital but of interest is Optel Thevon’s specialized optical mires and cameras for very high speed (1.5MHz) surveillance of rotating equipment. The tools provide real time torsion vibration monitoring at the edge.
Finally, a spectacular video of a future (or futuristic?) railway link from Beijing to London that had champagne drinking passengers travelling on the upper deck with automated cargo handling systems beneath them. China’s CRRC railroad operator plans to do the whole 8,000 km trip at 400kph for a 20 hour journey time. CompactRIO will be doing the onboard edge computing looking for small signals from the wheels with machine learning applied to pattern recognition and prediction of failure. Watch video.
© Oil IT Journal - all rights reserved.