Mehdi Sadeghi and Peyruz Gasimov (AmeriCo Energy Resources) have analyzed historic data to maximize profitability at a mature Permian basin waterflood by developing a cost-effective strategies to optimize rod pump performance. Discovered in 1946, the Coleman Ranch property now has 48 producers and 20 active injectors. Rod pump failures are a major cost center and cause of production loss. A data pipeline from historical daily reports and emails was analyzed with Python code and regex text parsing and into a well operations activity matrix covering 15 years of data from 88 well and 500k data points. A reliability model was constructed from downhole failure data comprising pump time to failure, time to repair and other KPIs. An ‘average’ reliability model was developed using the Akaike information criterion is used to choose the best fit for reliability model. Wells were then ranked with a profitability index, combining failure history and productivity into a single metric. The study demonstrated the value of chemical treatment. ‘For every dollar spent on chemical treatment we save on average 6 dollars on workovers’. AmeriCo is now investigating the effect of weather, injection parameters and workover cost as a predictors of time to failure.
Ron Peterson extolled the merits of Unico’s suite of sucker rod pump controllers. Unico’s variable speed drive controllers use dynamometer data to adapt pump speed to well conditions in real time. The system avoids ‘over-pumping’ and minimizes energy consumption. An inclinometer or crank sensor compares Dynacard data with pre-calculated rod-string models in real-time, providing continuous inferred surface and downhole states and allowing accurate estimation of pump load and fill. The system is said to generate ‘a comprehensive well report for every pump stroke’. Interaction is via an HMI on a touchscreen in the field, laptop or smartphone, or from Unico’s GMC remote monitoring station. Unico’s system would appear to be a prime example of ‘edge computing’ although it was probably developed before the term appeared!
On the topic of what could be termed ground-truthing the analytics, Mick Harrison (Harrison Wright Oil Co.) argued in favor of ESP ‘DIFA’ (dismantle inspection failure analysis). DIFA is the subject of the API Recommended Practice 11S1*. Harrison believes that pump run-life is often not maximized because ESP failures are not properly identified. DIFA aims at determining the root cause of failure and involves physical inspection of the pump along with an analysis of operating conditions and externalities including amperage data, downhole gauge data, scada and field records. ESP run-life can be extended when more care and attention is given to failure mode and root cause analysis. ‘If you don’t have a DIFA standard then establish one, set a goal and establish a run-life improvement plan’.
More from the LBCG Artificial Lift Conference website.
* API RP for Electrical Submersible Pump Teardown Report 3rd Edition September 1997.
© Oil IT Journal - all rights reserved.