Enough engineering to be dangerous (and useful)!

Workflow transformation looks beneath tip of data iceberg. AI identifies workover candidates.

Speaking at the SPE/Reed Exhibitions Intelligent Energy event earlier this year, Steve Cassidy described workflow transformation at Chevron’s San Joaquin Valley operations—involving a 100 year old field with 17,000 wells. Steam injection involves a huge daily workload as most wells see some form of intervention at least once per year.

The complexity and scope of operations led to system unreliability and Chevron is looking to apply more automation, and computer-based advisories. Developing these requires new skill sets—such as folks with an IT background combined with ‘enough reservoir engineering to dangerous (and valuable)!’ Artificial intelligence is used to perform predictive analytics. This is to look beneath the ‘tip of the iceberg’ of the vast amount of data currently acquired of which much is wasted.

All Chevron’s forecasts are currently data-driven (sans simulation). These leverage large numbers of low cost sensors to instrument a few wells in an intensive way. Data techniques such as genetic algorithms for cyclic steam injection and neural nets to select workover candidates are being trialed. The problem is that ‘very few people understand this stuff.’ One answer is the collaborate ‘environment.’ Not necessarily with everyone in same room, maybe just ‘well-connected individuals sharing the same data.’

Click here to comment on this article

Click here to view this article in context on a desktop

© Oil IT Journal - all rights reserved.