基于最小二乘支持向量机的二氧化碳驱油沥青质沉积对储层伤害的动态分析

油田开发过程中注CO2驱替原油常会发生沥青质沉积现象,沥青质沉积由于吸附和桥塞作用会伤害储层,会降低储层的孔渗性。本文利用数值模拟方法研究沥青质沉积过程,及对油田实际生产的影响。首先通过LSSVM机器学习算法拟合了沥青质沉积与气体浓度及压力之间的非线性关系；之后运用渗透率伤害GPT模型来建立砂岩储层沥青质沉积量与渗透率降低率的伤害模板；最后建立了五点井网下的地质模型,模拟相应工作制度下沥青质伤害前后的注入井注入能力、生产井生产能力及剩余油分布状况。模拟结果表明,沥青质沉积主要集中在近井地带,对注入井的注入能力影响较大,同时由于沥青质对储层的伤害会导致渗流阻力的增大影响气水波及效率,从而引起生产井产量的下降及最终采收率的降低。

Dynamic analysis of reservoir damage caused by carbon dioxide displacement asphaltene deposition based onleast square support vector machine

Asphaltene deposition often occurs in CO2-injected crude oil during oilfield development. Asphaltene deposition will damage the reservoir due to adsorption and bridging action and will reduce the porosity and permeability of the reservoir. In this paper, the process of asphaltene deposition and its effect on oilfield production are studied by numerical simulation. First, LSSVM machine learning algorithm was used to fit the non-linear relationship between asphaltene deposition and gas concentration and pressure. Then the damage template of the amount of asphaltene deposition and the permeability reduction rate of sandstone reservoir was established by using the permeability damage GPT model. Finally, a geological model under the five-point well network was established to simulate the injection well injection capacity, production well production capacity and residual oil distribution before and after asphaltene damage under the corresponding working system. The simulation results show that the asphaltene deposition is mainly concentrated in the near-well zone, which has a great influence on the injection ability of the injection well. Meanwhile, due to the damage of asphaltene to the reservoir, the increase of seepage resistance will affect the gas-water conformance efficiency, resulting in the decrease of production well output and the final recovery rate.