Using retrieved data from malfunctioned components to avoid another costly failure.
In 2015, an incident at a gas separation plant where one of the major gas turbines faced a devastating failure which left many of its internal blades heavily damaged. This was simply the tip of the iceberg. The focal point of this incident centers around the consequence which had halted the oil production by a total of 5 full days, setting the client back by up to THB 400 million in value.
One of many reasons that caused this unplanned shutdown could be attributed to the malfunction of the sensors even though they were embedded in the gas turbine.
The information obtained from The Soothsayer’s software, based on data input from the malfunctioned gas turbine indicated that had the predictive machine-learning software been implemented 3 months prior to the component failure, the incident could have been avoided — including the downtime of 5 days.
When comparing the formulated information from the predictive maintenance module against the actual information, the results were relatively close — suggesting that the information provided by the software is reliable.
“Since 2014, The Soothsayer has effectively saved clients
THB 4,000 Million
and prevented weeks of unplanned shutdown.”