Prioritizing Non-Rig Well Work Candidates Using Data Science

According to Wikipedia, "Data Science is an interdisciplinary field that uses scientific methods, processes, algorithms and systems to extract knowledge and insights from data in various forms, both structured and unstructured, similar to data mining." The oil and gas industry is increasingly expanding its activities by moving into the Data Science and analytics space to increase efficiency, reduce costs, make better decisions and improve quality of technical products and services. Through the extraction of knowledge and insights from historical data, oil and gas companies can systematically process the huge data available to them using scientific methods and algorithms to identify trends for problem identification and optimization opportunities. The data processing can also be used to perform analytics to provide Descriptive, Diagnostic, predictive or Prescriptive solutions for value creation. For Chevron offshore and onshore non-rig wellwork, the existing methodology of planning and scheduling Non-Rig Workovers (NRWOs) for execution is a spreadsheet or a Project typically run on Microsoft applications or software. This process does not incorporate numerous factors that affect the value realization through executing the NRWO such as historical Data Analytics, predictions and several extreme constraints. The value in building a prioritized candidate selection schedule is allowing the business to shift to a data-driven model based from a method of simple basic programs with limited options and typically biased by human input. Historical data from various sources is being collected to provide an encompassing view of the NRWO prioritization, planning and scheduling environment. The scope of this study involves utilizing Data Science to generate solutions comprising of prioritized scheduled workovers that are optimized by various constraints to rank these workovers such as individual well Non-Rig workover cost per barrel. The approach can be replicated using other operational and well related constraints to generate alternative optimized rigless well prioritization solutions. The resulting wells will be gauged against established business drivers to develop an optimal prioritized solution which is then applied at the start of the business plan year to provide an optimized wellwork schedule for the planning year. Data Science applied to this project utilizes the various systems of records within the offshore and onshore fields such as Wellwork candidate listings and categorization database, project maturation database, cost schedules, possibility of success, reserves, production profiles, etc. The systems of records are then integrated through Data Science and prioritized by ranking the various parameters through automation based on constraints specified by customers. The long-term project will reduce NPT by 2-3% annually, save well work maturation recycle time, and increase efficiency in executing wellwork through an optimized schedule. Equivalent cost savings of between $650,000 and $1m was estimated for the initial pilot simulation run for the business planning cycle evaluated. The methodology applied in this study provides a multidiscipline and integrated approach to bridge the conventional optimization void of Data Science and the big data approach to make quicker non-rig well scheduling decisions.