Skip to content

DIGITAL TECHNOLOGY IN AID OF SAFETY

Because risks are part and parcel of the energy industry, safety is one of our top-priority values. First, because we consider safety as a vital issue. Second, because safety guarantees our activities will stand the test of time. So, improving our capacity to prevent major accidents is an absolute must.

With this in mind, we are entering the era of digital prevention. Developed to optimize the management of risks associated with a gas leak on one of our industrial installations, TADI (Total Anomaly Detection Initiative) represents an innovative approach that combines:

  • Early anomaly detection (equipment failure, gas leak) owing to new-generation sensors able to rapidly detect the presence of gas, identify it, quantify its concentration, locate the source, provide the leak rate, and in the most critical cases, visualize in 3D and in real time the hazardous area related to the formation of a gas cloud.
  • A system to acquire/process/valorize big data in real time, in order to transform raw monitoring data into relevant information on the anomaly in progress.
  • The transfer of information to the operator in the control room, in the form of a diagnosis and recommendations optimized by artificial intelligence to help the decision-making process.
TADI, toward digital prevention of major accidents

A UNIQUE HSE TESTING INFRASTRUCTURE

In 2018, the TADI project acquired a site to test and qualify innovative gas detection and quantification technologies which is located on the Lacq Pilot Platform (PPL) at the PERL. Its equipment, taken from the dismantled installations of our old plant (pipes, valves, tanks, columns, wellheads, flares, etc.), makes it a realistic and modular industrial theater where we can re-enact, in safe conditions, a large number of the accidental scenarios obtained from feedback from the field.

Designed with the teams of the PERL (Platform for Experimental Research in Lacq) as part of the “Major accident prevention” R&D project , this lasting R&D infrastructure is one of a kind. First of all, owing to its location in a SEVESO 3 industrial environment. But also, because it is currently the only site that can reproduce such a wide range of leak rates, from 0.5 g/s to 300 g/s.

It will enable us to continue the sensor qualification program, that until now was carried out on temporary installations, in much better conditions. Two test campaigns have already been conducted:

  • The first, in June 2018, to test low-cost acoustic technologies for monitoring the safety of installations.
  • The second, in October 2018, to evaluate several optical and acoustic technologies in different forms (ground, airborne, on a robot, drone, etc.) to remotely detect/locate/quantify gas for emergency situations and safety surveillance.

Take a look at the CNBC article detailing innovations developed at the PERL to reduce costs and the ecological footprint.

 

GIVING A VOICE TO DATA IN A "VIRTUAL TADI"

These and future test campaigns will generate large amounts of data that will be transferred, stored and classified in a cloud alongside context data specific to each test. What is the objective? To extract pertinent information, using big data and artificial intelligence, so it can be passed on in real time in the right form and to the right target (operator, crisis cell, maintenance service, etc.) via TADI safety apps, as user-friendly as those of our smartphones, to deliver diagnoses and recommended actions.

The development of these applications is based on an “operations of the future” room located a few hundred meters away from the test infrastructure on the premises of the PERL. An operator will automatically receive alerts and diagnoses on a big screen, in the form of visual messages. By navigating through a virtual TADI, a digital twin of the test site, the operator will be able to access, in just a few clicks, all the context data necessary to understand, at a distance, what is happening on site, confirm the diagnosis and implement an action plan adapted to the anomaly detected.

The next step, due for 2021, will be the delivery of a demonstrator validated on a number of simple and multiple leak scenarios played out on TADI.