Organization
School of Petroleum Engineering, Changzhou University, Changzhou 213164, China
摘要
常规油田中后期开发往往存在油井压力不足、油井供液能力与抽汲能力不匹配等问题，现有间歇采油、变频控制等方法难以充分考虑多种油井监测数据的动态变化。为此，文中提出一种基于数据驱动的油井生产参数智能调控方法，结合采油工程理论知识，筛选了示功图、动液面和产液速度3种指标用于指导油井生产参数调控。根据示功图，建立基于卷积神经网络的油井供液程度智能识别模型，将示功图量化为供液程度；对筛选出的3种时序数据开展分析，建立油井生产参数调控单因素决策因子计算方法；在此基础上，考虑多种因素对调控决策的综合影响，建立了基于层次分析法的综合决策因子计算方法，实现了油井生产参数调控多因素量化决策。现场应用表明，使用油井生产参数智能调控方法后，油井供液程度平均提升13.6百分点，动液面下降137.903 m，产液速度上升0.510 t/d，供液稳定性明显提升，有效实现了油井生产参数与供液情况的自适应匹配，采油效率明显提升。
Abstract
In the middle and late development of conventional oilfields, there are many problems such as insufficient oil well pressure and mismatch of oil supply capacity and swabbing capacity. Existing methods such as intermittent oil production and frequency conversion control are difficult to fully consider the dynamic changes of various oil well monitoring data. To this end, a data-driven intelligent control method of oil well production parameters is proposed. Combining with the theoretical knowledge of oil production engineering, three indicators of dynamometer diagram, dynamic liquid level and liquid production rate are screened to guide the control of oil well production parameters. According to the dynamometer diagram, an intelligent recognition model of oil well liquid supply degree based on convolutional neural network is established, and the dynamometer diagram is quantified as the liquid supply degree; the three kinds of time series data selected are analyzed, and the decision factor calculation method of single factor of oil well production parameter control is established; on this basis, considering the comprehensive influence of various factors on control decision-making, a comprehensive decision-making factor calculation method based on analytic hierarchy process is established, which realizes multi-factor quantitative decision-making for oil well production parameter control. Field application shows that after using the intelligent control method of oil well production parameters, the oil well liquid supply degree increases by 13.6% on average, the dynamic liquid level decreases by 137.903 m, the liquid production rate increases by 0.510 t/d, and the liquid supply stability is significantly improved, which effectively realizes adaptive matching of oil well production parameters and liquid supply conditions and significantly improves oil production efficiency.
论文详情