Demonstrating good practice in the safe operation of gas assets with predictive analytics
The oil and gas industry operates large and complex facilities. Technical integrity (and thus licence to operate) must be maintained through routine inspection and maintenance regimes. Corrosion attacks every component at every stage in the life of every oil and gas field or plant (Schlumberger 1994). Globally, corrosion management accounts for US2.5Tr cross-industry spend (NACE International 2016). Risk-based approaches for internal corrosion based on susceptibility of a process item to corrode, have been utilised to assist with identifying appropriate and more cost-effective maintenance and inspection strategies. The aim of such approaches is to protect integrity and not compromise safety; however, they do nothing to minimise regret cost. These approaches use only known physical characteristics of piping equipment and rely on repeat inspection data to calculate corrosion rates and associated maintenance schedules. The present paper will leverage the challenges and shortcomings of using existing risk-based inspection (RBI) approaches and demonstrate how Accenture in collaboration with Woodside and others is utilising predictive analytics to more accurately determine likelihood of corrosion to exist in a more granular resolution, thus managing likelihood and consequence of corrosion to produce an improved risk-based model. The analytics model considers physical, geospatial and external factors for external corrosion. This is a work in progress, with very promising initial results, that leads into the implementation of an improved RBI strategy, enabling Woodside to reduce inspection scope, physical site activity and associated management cost. In addition, it better manages plant risk in conjunction with smart visualisation tools.