Data can provide great value to maximize the efficiency of industrial systems, thanks to the technological revolution centred on the information management and automatic analysis. Among the countless methods of obtaining value from data, one of the most explored is that which focuses on the early identification of deviations from normal operation or performance, the so-called predictive monitoring.
The main objective of predictive monitoring is the early detection of events (for example, malfunctions, degradation, etc.), with the aim of improving the reliability, security and efficiency of the monitored assets.
Timely identification of deteriorating performance levels makes possible to maximize the time to act, thus alleviating any possible damage caused. Likewise, maintenance costs are also reduced thanks to allow maintenance tasks planning and replacement of components.
From plant life data, including operational and environmental conditions, the predictive monitoring techniques identify incipient deteriorations or malfunctions in the equipment. From the historical records of the normal state of the equipment in different function modes and environmental conditions, different models can be developed which are able to estimate, in real time, the value of each of the significant variables of the process for given operating and boundary conditions. These estimated values are compared with the actual values in order to identify the deviation between the estimated and real value, as precursors to any abnormalities in the operation or condition of the equipment.
Predictive monitoring can be used in a wide variety of systems and equipment, including not only dynamic equipment such as turbines, pumps, compressors, but also static equipment such as condensers, heat exchangers, reactors, etc. This technology, therefore, provides valuable results for power generation plants (nuclear, combined cycle, wind, hydraulic, solar thermal, etc.), oil and gas facilities, aviation and rail sectors, etc., undoubtedly a wide range of industries.
It is generally used on components and systems with a defined operation pattern, which have operational historical data and are sufficiently sensorised to provide information on their status. Its deployment is quick and flexible because it is done using readily and available instrumentation.
Tecnatom is currently applying predictive monitoring as a relevant tool at the Monitoring Center. The result from the monitoring are combined with our experts knowledge of the process to add maximum value when optimising the operations and maintenance of power plants and other industrial facilities.