Optimizing industrial risk management with STORM
- Developed by the R&D researchers at the CSTJF, STORM is a monitoring program for the early detection of gas leaks.
- It analyzes the data from sensors in real time and transforms it into visual information that is easy to interpret.
- After being tested in Lacq, STORM is now part of a pilot on the Tempa Rossa hydrocarbon production site in Italy.
"The development of the STORM application – standing for Safety Tool For Operators and Remote Monitoring – started in early with the company CS Group following a feasibility demonstration," explains Laetitia Urfels, Manager of the Safety R&D theme at TotalEnergies. Its objective is to step up the safety of personnel and facilities and prevent major accidents, particularly those that might cause flammable gas leaks. The story of STORM is inextricably linked to that of the TADI (TotalEnergies Anomalies Detection Initiatives) test site created in 2015, an area of 2,000 m² on the Lacq Pilot Platform equipped to reproduce different risk scenarios in controlled conditions and without any risks (leak rates of 0.1 to 300 g/s). "Thanks to TADI, we have tested, compared, and developed new detection means," explains Xavier Watremez, Manager of the Safety Monitoring and Crisis project. "Several test campaigns have been conducted to qualify innovative tools to “see”, “hear” and “feel” the gas emissions; these include acoustic sensors, infrared cameras, gas detectors, etc." Combining hearing, sight and smell improves the chances of detecting anomalies on time and of saving lives.
STORM gives a voice to data
"As early as 2020, we started to think about how we could use the data that is continuously transferred by the sensors installed on the TADI site," adds Stéphane Collas, in charge of developing STORM as part of the Real Time Risk and Impact Monitoring project. The first step was to store the data in the same place to make it easily accessible. "We opted for a dual storage solution – local with a copy in the cloud – so that information would be available on site, in the affiliates, and at headquarters." The second step was to give a voice to the different types of data, which meant connecting them to one another and making them "visual", a bit like on a Google Map screen: geolocalization of the sensors and state (alarm, routine operation, etc.), coordinates of the leak point, video streams from the cameras, propagation of the gas cloud, weather data, and so on. This allows rapid interpretation by the operators in the control room, who can therefore react faster, sending in robots and/or drones, and limit personnel exposure. "Our objective is to have as faithful a representation as possible of the gas cloud while integrating effective simulation-propagation simulation tools." The Tempa Rossa pilot will serve to validate the latest advances made by STORM on a larger scale. The site compression area (18,000 m²) has been fitted with 24 acoustic sensors, which constantly "listen" to what is happening and are trained by artificial intelligence to identify the sound of a gas leak, and with three infrared cameras. There's a very fine line between safety and the environment, which STORM could very soon cross by integrating environmental sensors. The application could also eventually be used to visualize fugitive emissions of greenhouse gases (AUSEA POC), hydrogen, etc.
Discover STORM in video, a smart diagnostic tool