【作者】 张平；

【导师】 管志川；

【作者基本信息】 中国石油大学 ， 油气井工程， 2009， 硕士

【摘要】 深井、超深井由于地质条件或压力系统复杂、可参考邻井资料少,无法对地层压力信息做出较为精确的描述。本文通过研究地层压力预测中的不确定性因素,利用概率分析方法确定其预测模型中参数的概率分布状态,通过直接理论计算或蒙特卡洛方法得出含有可信度的地层压力剖面;利用压力约束准则得出具有可信度的安全泥浆密度窗口,并提出了依据含有可信度压力剖面的自下而上和自上而下的井身结构设计方法,由此方法得出连续的套管层次及下深范围。该方法避免了传统井身结构设计方法对关井井涌和压差卡钻深度进行试算和验证的过程,可直接在满足上述约束条件下进行连续的设计。在实际设计过程中可以根据套管层次和下深的范围结合地质地层情况综合确定最终的井身结构。本文利用所建立的方法对川东北一口深井(普光4井)进行了实例设计。同时本文根据含有可信度的安全泥浆密度窗口,构建出了套管层次及下深的风险概率模型,包括井漏模型、井涌模型、井塌模型及压差卡套模型。利用此模型可对一口所设计出的井的不同套管层次及下深方案进行定量风险评价,得出每一种方案不同深度处的井漏、井涌、井塌及压差卡钻风险,并根据此结果对各种不同的套管层次及下深方案进行优选。更多还原

【Abstract】 The formation pressure information in deep or ultra-deep wells is hard to be described accurately for the complex formation or pressure conditions and lack of offset references. In this thesis, the distribution of parameters in the pressure prediction model are determined via probability analysis method by investigating the uncertain factors in formation pressure prediction. The formation pressure profiles with credibility are acquired by direct theoretical caculation or Monte Carlo method. The safe mud density window with credibility are obtained by using pressure constraint criteria, and then the bottom-up and top-down casing program design methods according to the pressure profiles with credibility are presented. After that, the continuous casing point selection can be conducted. This method avoids the pilot caculation and varification process aiming at shut-in kicking and differential pressure sticking depths in the conventional design method. The continuous design can be conducted under the constraints with no need to determine casing setting depth by pilot caculation. The results acquired by this method is a continuous range. In the practical design, the final casing program can be determined by casing point selection range combining fomation geological information. A deep well in noutheast of Sichuan Province (Well Puguang 4 ) is used as a design example, and by comparison the design results show agreements with the real information. Meanwhile, The thesis presents a probabilistic risk analysis method applied for analyzing quantitatively the risk of the casing program design. The probabilistic risk analysis model for casing point selection had been established. The model generally includes four sub-models and each of them can make quantitative risk analysis on lost circulation, kicking, collapse and sticking respectively. This model can be used to make quantitative risk analysis on different casing point selection designs of the same well. The analysis results can provide valuable information for designers.更多还原