Making industrial installations safer and more efficient by identifying real and false alarms

Narrative

Large-scale industrial installations - such as oil refineries and power stations - rely on safety-critical systems that employ distributed sensor networks coupled to alarms. A sub-optimal or erroneous alarm configuration at these installations can have catastrophic environmental, health, and economical consequences.
Researchers at The University of Manchester (UoM) and the SME Argent & Waugh Ltd have used new mathematical algorithms and methods to develop an innovative approach to improve alarm configurations that identify redundancies reliably and in real time. The new approach is deployed at industrial sites via the widely-used Sabisu software platform, with more than 6,000 individual license owners worldwide, at companies such as SABIC, Royal Dutch Shell, and Huntsman. For SABIC alone, the savings from using Sabisu are estimated at GBP1,500,000 per annum, resulting from more efficient real-time monitoring of equipment.
Further benefits arise from increased plant safety and reduced operator workload in the plants, and for Argent & Waugh through increased profits and enhanced capability.

Impact date	1 Aug 2013 → 31 Jul 2020
Category of impact	Economic
Impact level	Adoption