Glen Sartain (Teradata) and Olivier Germain (Halliburton) took part in a Hart Energy webinar this month that investigated the potential application of ‘next generation’ database technology in drilling. The thesis is that existing systems are or soon will be overwhelmed by the data volumes streaming into a drilling rig—particularly with the advent of wired drill pipe such as National Oilwell Varco’s IntelliServe1/Intellipipe downhole network streaming ‘up to’ 1,000,000 times more data to the rig. Even today, a rig can create around 4 terabytes of data a day, more than can be analyzed in real time by a database/flatfile combo. Sartain believes we have to change the way we value and view data—moving from a ‘reporting’ standpoint to an environment where data is constantly analyzed and compared with historical data during operations—blending strategic and operational intelligence and moving towards ‘autonomous’ drilling.
Enter Teradata’s ‘shared parallel’ database with its ‘in place’ analytics. Teradata instantiates analytical models inside the database—removing data latency. Germain sees this as a logical progression from the introduction of systems like Pason’s AutoDriller which were met with skepticism when they came to market. Tests comparing humans on the brake against the AutoDriller showed the latter to be much more performant. Artificial intelligence and parallel databases will take such automation to the next level. There is also a feeling that post Macondo, it will become mandatory to capture real time data to a ‘pristine’ environment for root cause analysis inter alia. Webinar replay available on www.oilit.com/links/1103_11.
1 www.oilit.com/links/1103_10.