Data-Driven Injection/Production Optimization for Horizontal Well Pattern in a Complex Carbonate Oilfield

Abstract

M oilfield is complex carbonate reservoirs in the Middle East, with strong heterogeneity, high permeability zones, local dissolution fracture area, high viscosity oil area and asphalt layer, etc. Strong heterogeneity leads to early water-out, rapid water cut rise and large production decline for horizontal wells, slow reservoir pressure restoring by water injection and inefficient utilization of horizontal section. Because of great difference in the production performance of single well and unclear development law, it is difficult to achieve multiple goals and good waterflooding effect. In this paper, big data-driven strategy module, and Capacitance Resistance Modeling(CRM), multi-objective optimization modelling are used to establish a technical process and platform for real-time waterflooding optimization on the complex reservoir, which hasn't been put forward in previous research for horizontal well pattern and already successfully applied in M oilfield. Big data driven analysis was adopted to quickly process the geological characteristics and production dynamic data from database set, used for cluster analysis based on neural networks to describe the distribution of dominant water flowing channels and residual oil distribution, evaluated waterflooding law and optimized rational production-injection strategies for its main controlling factor areas. CRM were established through simple geological data, PVT data and prodcution history data, which was an equivalent simplified model to caculate injection allocation factors matched with liquid rates. Real-time connection network has been established to determine injection allocation factors from injectors to producers for large number of horizontal wells. Multi-objective optimization modelling was established to solve the realization conditions for super-achieveing the lowest water cut rising, the slowest production decline, the most reasonable pressure restoring, the highest cummulative oil production and the balanced Voidage Replacement Ratio(VRR) for each main controlling factor area. Integrated continuous, dynamic and quantitative adjustment will be output and implemented during weekly and monthly cycle, and comprehensive monitoring, timely warning and accurate diagnosis are realized for the oilfield. M oilfield has been adjusted about 634 wells to rational performance, and then water cut was controlled from 67.1% to 64.7%, water cut rising rate was decreased from 7.9% to −13.84%, yearly production decline rate was reduced from 25% to 7%, reservoir pressure was built up by 158 psi, and total incremental oil is 5.48 million barrels, which indicated that the waterflooding performance has been greatly improved. This novel methodology and platform provide important reference significance for the waterflooding optimization in Middle East. It can rapidly realize waterflooding optimization in balancing reservoir pressure, controlling water cut rise, slowing down production decline and so on, and obtain better incremental oil and economic benefit under low operation cost.