Cognite Data Fusion enables Cloud-based Data-driven AI modelling of production performance

• OMV, AkerBP and Cognite jointly developed BestDay, a product that systematically and automatically detects production underperformance on all levels of production systems
• Using data-driven AI modeling to reach the maximum production capacity with an estimated production gain of 0.5% per year
• Information sharing at a fingertip and close collaboration between Field Supervisors and Production Engineering teams at OMV was never easier before

Challenge

The Austrian integrated oil and gas producer OMV operates several fields worldwide. OMV Austria’s mature asset with 800 active wells is already a top performer at 93 % production efficiency and more than seven years of mean time between failure for its artificial lift systems. Yet, with BestDay, a brand-new jointly developed application, OMV aims to take its operations to the next level in terms of production efficiency, starting with the Matzen field in Austria.
Detection and quantification of production underperformance can be a time-consuming process. The challenge is to repeat it consistently and to compare production performance against a reference production capacity on all levels
of a Production System.

The maximum available production capacity is an interplay of various elements of a production system and requires frequent updates to serve as benchmark for production performance analysis – a so far cumbersome and data-intensive process. Hence optimizing production efficiency and detection of underperformance was carried out rather on demand than as part of the daily routine. As a result, performance evaluation was narrowed down to monthly average ranges, omitting pro-active investigation and interconnection of high granularity timeseries.

With the introduction of the DigitUP program in 2017 a clear vision to leverage the access to data in day-to-day operations was formulated aiming to remove data extraction and manual number-crunching with limited accessibility to the team. Still, one crucial ingredient that satisfies the requirement for automated detection of underperformance was missing – a continuously updated set of targets.

Solution

Begin of this year OMV and Cognite as technology provider started the roll-out a jointly developed AI-driven application called BestDay. Beforehand, all necessary relevant real-time and periodic production data from wells and Production Systems were liberated from various sources (including Oracle and OSIsoft PI), routed through HELIUS, OMV’s integrated information ecosystem, an ingested into Cognite Data Fusion (CDF). Apart from an instant and systematic data access, BestDay helps the oil and gas operations reach their maximum production capacity by making use of data-driven AI modeling and reveal what its name suggests: the best possible production day based on historic data. 

The performance of each production system is automatically evaluated within a given timeframe, and a custom set of boundary conditions defined as disqualifying criteria. A data-driven AI maximum capacity model calculation runs updates daily.

Thanks to the built-in deviation detection logic, combined with a tailor-made interactive graphical user interface (GUI) and stunning visuals that convey information streams from several IT & OT data systems to the eye of the user, any detected change of production behavior is immediately brought to the center of attention. The user can also drill down by selecting specified Subsystems or wells to review deviations, associate forecasted and logged deferment events, and comment and collaborate online.
 

Impact

Today, key production indicators from various sources are accessible throughout one tool across the branch office. OMV field supervisors and engineers can make effective use of production data for root-cause analysis, to plan corrective measures, or as predictive cursor. Underperformance can be identified in the blink of an eye and does not remain unflagged, or simply overlooked. Best practice can be shared among all OMV Upstream operations by identifying and amplifying performance enhancing measures.

BestDay creates real-time organizational visibility using an advisory system on all aspects of production optimization, thereby enabling a team of experts to reduce deferment events through improved response time and value of response, thus, increasing production efficiency and ultimately increase or maintain the maximum production capacity possible. The resulting production gain for OMV’s first asset where the application was rolled out is estimated at 0.3-0.5 % per year above current production profile. Additionally, with the BestDay application good or bad actors on performance can be revealed more easily, fostering closer collaboration between field supervisors and engineering teams globally, rendering the exchange of knowledge and best practices no longer a localized phenomenon.
Learn more about the development of BestDay and follow the link to this video clip Cognite BestDay | Reach and Expand Your Production Potential or visit the BestDay website.