Search for a command to run...
Abstract Projecting production fluid streams in the oil and gas business during the production phase is very important as it impacts daily operations, depletion planning, reservoir management strategies, and other planning activities. Conventional projection tools such as decline curve analysis and reservoir simulation are either too simplistic or too expensive (computationally) for required details. Thus, a ‘fit-for-purpose’ approach was developed which balances the accuracy and run-time of the said projection based on the requirement. The new approach combines forecasting techniques to project the variation in the produced fluid's phase fraction (water cut and gas oil ratio) over production and optimization via linear programming over a set of time-based constraints to identify the most optimal route of well operations satisfying the facility and reservoir constraints such as water handling, gas handling and voidage replacement ration for wells. It also identifies cases where projected constraints for facility and reservoir would make wells non-operable. A comprehensive visualization and case handling database is also provided for maintaining projection history and perform iterative case building. This data-driven approach provides optimal rates of producing and injecting wells over a span of time while adhering to facility-level total production-injection constraints and subsea voidage replacement ratio constraints placed on wells to maximize oil production. It uses rate-based estimation to capture only the required operating complexities compared to a simulation model, which is an overkill, while being more sophisticated than decline curve analysis, which is too simple. Medium-term variation profile of fluid phase fraction (water cut and gas oil ratio) can be added to capture longer term forecasts and look for investment opportunities (equipment repair and replacement, initialize development of infill wells). Time based constraint changes are added to emulate operational scenarios (such as planned pigging, well testing campaign, pressure data acquisition and infill well start-up) and help evaluate optimal timing of conducting the activities while minimizing downtime. It also allows optimization for multi-facility operations connected through a common reservoir, primary for pressure management. This approach offers a fast and computationally inexpensive solution to complex operation planning problems while trading precision for speed (as compared to full field numerical simulation for quick scenario planning), assisting production and reservoir engineers in making better decisions faster. A combination of data-driven forecasting and numerical optimization (linear programming) is introduced which can enhance an engineer's capability to identify optimal projection of the fluid stream for different operating scenarios on the field.