基于聚类分析法的斜井岩屑运移经验-半经验模型优选

Optimal selection of empirical or semi-empirical models of cuttings transport in deviated wells based on clustering analysis method

陈修平 邹德永

中国石油大学（华东）石油工程学院，山东 青岛 266580

College of Petroleum Engineering, China University of Petroleum, Qingdao 266580, China

斜井钻井过程中，岩屑得不到及时有效运移会造成严重的井下事故。及时了解井眼清洁状况，对于钻进参数的调整具有重要意义。相对于理论模型，利用岩屑运移经验￣半经验模型预测井眼清洁状况的方法方便快捷，但对于同一作业条件，不同模型的计算结果相差很大，为此提出了一种基于聚类分析法的斜井岩屑运移经验￣半经验模型优选方法。首先通过文献调研，确定影响岩屑运移的8个主要因素，同时建立斜井岩屑运移经验￣半经验模型数据库，模型库中的各模型分别作为一个样本，8个主要影响因素作为样本指标；然后通过计算、比较实际作业条件样本与各模型样本之间的距离，对计算结果进行从小到大排序，得到优选结果；最后通过实例计算，验证了该优选方法的可靠性。

The ineffective cuttings transport will result in severe trouble in the downhole during deviated hole drilling, so it is of great importance to monitor the degree of hole cleaning for adjusting drilling parameters. Compared to theoretical model, it is convenient to calculate the hole cleaning by using the empirical or semi-empirical models. However, for a certain operational condition, the results obtained from different models differ greatly. So an optimized model based on clustering analysis method is developed. Firstly, the database for empirical and semi-empirical models is built by literature research, and the eight dominant factors are determined to be the indexes of each sample. The optimized result is obtained by the comparison and ranking of distances between models and actual operational condition. The validity of this optimizing method is verified by the case study at last.

10.6056/dkyqt201402024