Drilling Analysis Using Big Data has been Misused and Abused

Abstract

Abstract The authors have been contacted by many people in the industry lately that are incorrectly utilizing big data to produce correlations that attempt to identify operational "sweet spots". This paper will show examples and address the need to add several steps to big data before any meaningful correlation results can be obtained, mainly understanding (and this is not a comprehensive list): - The sensors involved and their limitations; - The errors in the placement of these sensors (e.g. hook load sensor on the deadline); - The frequency of the data and how this impacts the analysis (some companies provide 10-second data); - The quality of the data itself; - The appropriate filtering of data to ensure apples-to-apples comparisons; - The rig state must be known - Understanding of the physics involved.