用修正学习曲线测算井段的钻井时间

Forecasting Drilling Time Section by Section by Modified Learning Curve

将学习曲线与传统的分段方法结合起来，既克服了分段法中不考虑学习效果的缺陷，又克报了学习曲线中不能灵活调整的缺陷，形成两种方法的优点叠加．用学习曲线测算井段的钻井时间，其前提条件是：各井在该井段的钻头选型基本相同，井段直径相同，所穿过的地层相同，所属的井型相同．对钻井参数（钻压、转速、泵压、排量）、泥浆密度与性能、喷嘴直径与个数等时变性因素，不单独列出，而作为在钻井过程中通过逐步学习加以改善的对象．用塔北某地区１号构造上所钻井的二开井段（４４４．５ｍｍ）实际资料测算，得到的结果明显优于钻井监督的计划时间，测算的最大相对误差１１．８％，平均为８．８％，而监督组计划误差为４１．９％，平均误差率降低了一半以上．

The authors put forward an approach of forecasting drilling time section by section by modified learning curve. This approach can be used under following conditions: (1) in all sections, hole diameter, drilled stratum, well type must be the same, and the types of used drilling bits should be roughly like. (2) every section must be straight. (3) bit weight, drilling speed, pumping pressure, discharge capacity, mud density, the number of nozzles and their diameters, etc. are regarded as the parameters to be modified step by step by learning while drilling. According to the data of a two section well in a structure of Tabei Oil Field, the maxinum error of drilling time predicted by this approach is 11.8%, but the error predicted by superintendent is 41.9%. This shows that, this approach is effective.