用修正学习曲线测算钻井周期

PREDICTION OF WELL DRILLING PERIOD BY RECTIFIED LEARNING CURVES

在测算钻井周期时，常用分段计算和汇总相加的办法。由于其中未考虑学习效果，易产生较大的误差。用普通学习曲线测算可使差异减小，但其应用条件是要求各井的情况相同，这在一般情况下难以满足。文章提出了在井深不同的条件下，对普通学习曲线进行修正的方法，并给出了相应的修正公式。用实例对该方法进行了验证，得到了满意的结果。这说明该方法正确可行，从而扩大了钻井学习曲线的运用范围，为钻井周期的测算提供了新的手段。

In the process of forcasting well drilling period, the well drilling time is firstly predicted by sectional calculation, and then summing the time up. Thus the learning effect has not been considered, the results are often inaccurate. Predicting method by the use of the common learning curve can reduce the difference between the actual time and the measured time, but the required conditions of the common learning curve is too harsh to meet for most wells. The method requires that all wells should have the same conditions, such as depth, well structure, well type (exploration, appraisal or production), destination formation, and structure zone. Usually the depth of each well varies to some extent, even the latter four conditions are the same. The common learning curve is not suitable for this situation. This presentation tries to solve the problem by considering “exponential time increase” when drilling deep well sections, and combines the exponential curve with the common learning curve to form a rectified learning curve. Two actual cases are presented, and the calculation results are satisfactory.