“Accelerating the Field Development Planning Process with Data-Driven Technologies”

   A data-driven technology for diagnosing reservoir conditions and generating field development plans for mature assets is presented, based on machine learning and automation of standard petroleum engineering workflows, without requiring a full field simulation model. Ideally suited for large, complex oil fields with large data sets (e.g. thousands of wells producing over many decades), these field development plans are prepared within three weeks using this technology to serve as operating plans and reserves basis.

• Phase I (1 week): Organizing all well data, including logs, production/injection, completions, etc.

• Phase II (2 hours): Automated or machine-assisted workflows for geological mapping, estimating flow contribution by stratigraphic flow unit, decline curve analysis, material balance history matching, fractional flow modeling, drainage area estimation for all wells by flow unit, identifying pay behind pipe, identifying unswept/bypassed oil pockets, and waterflood optimization.

• Phase III (2 weeks): Technical validation by senior engineers, during which those opportunities that are technically valid with high confidence proceed to economic modeling.

  Recent applications of this technology have resulted in field development plans including recompletions targeting pay behind pipe, artificial lift upgrades and optimizations, offline well reactivations, new well locations, and waterflood plans.