Group Optimization and Modeling of Mechanized Wells Operating in Intermittent Mode

Abstract Intermittent mode is one of the most efficient methods of managing low-rate ESP wells. However, this approach implies defining optimal working parameters of the wells. So, there is a demand for complex algorithms that can perform simultaneous joint optimization of power consumption and modeling tasks for the groups of wells operating in periodic regime. Besides, it must consider work/idle cycles synchronization to minimize effect of the pressure fluctuations occurring in the oil gathering chain. The mathematical model of the «tubing–annulus-reservoir" system is used, with boundary conditions defined at the submersible equipment intake by mass conservation equations.The simultaneous joint optimization of power consumption is based on existing approaches for steady-mode wells modified to suit intermittent working regimes. Special tools are developed to consider intracycle parameter dynamics and synchronize the on/off cycles, minimizing negative impacts on the infrastructure.Optimal parameters are determined using the adaptation results of complete nonstationary processes model, while potential scenarios are calculated using simplified methods. Finally, the result is refined by recalculating it using the most complex model.As a result, the algorithm was developed that provides functionality for simultaneous joint optimization of ESP well groups operating in both intermittent and steady modes. It can determine the optimal parameters of the system considering geological, technological and regulatory constraints. Besides, it supports two types of optimization scenarios aimed at either maximizing production or decreasing lifting costs. The optimization is built around existing model of non-stationary flow in the well which can calculate dynamics of production, hydraulic and electrical parameters. On top of that, to bring the performance of new algorithms close to those used for steady modes, a set of simplified models was created based on the full physically driven mathematical model. This technique was the key for achieving sufficient algorithm speeds necessary for increasing the scenarios update frequency by using sequential calculations with model complexity enhancement.In addition, the results of numerous calculations are presented. They were compared with actual data from the production fields to verify their predictive ability. This analysis was performed for all types of models, as well as their combinations and covered both group and individual optimization cases. The developed algorithms were tested on pilot fields and the results were confirmed by process engineers.