With the advent of cloud computing, Internet of Things (IoT), heavy duty sensor technology, and high resolution downhole communication, predictive analytics gained popularity in the energy sector.
In upstream and oil field services in particular, the complexity of the subsurface environment and the vast amount of data collected from disparate sources—such as seismic, core floods, production history, and well logs—make the case for application of predictive analytics.
Predictive models have proved effective in estimating the remaining life of an asset while exposing the parts with the highest likelihood of failure. Leading companies focus on predicting failures of components that will most impact asset operations. They overlay the asset usage schedule and patterns with its remaining life forecast to identify which components put operations at risk. The key benefit to this approach is in targeting only repairs with the potential to significantly impact the operation. This enables efficiency improvements without substantially increasing the maintenance frequency, repair cost, and parts replacement cost.
Predictive analytics should not be applied across the board to every asset within an organization. Different types of analytics techniques may be required to address different business requirement and asset classes. Based on the asset segmentation shown in figure B, one can identify assets best suited for various applications of predictive analytics.
Principal, PwC US
Director, Analytics, PwC US
US Energy Advisory, Managing Director, PwC US