Drill bit wear quantitative evaluation method suitable for machine learning

Abstract

The invention discloses a drill bit wear quantitative evaluation method suitable for machine learning. According to the invention, analysis is completed on the basis of statistical analysis of actual construction data of a completed well, and the analysis is actually statistical analysis; the change characteristics and rules of historical well data are analyzed, and the characteristics and rules of data reaction are actual reactions of actual engineering; according to the invention, the discrete drill bit wear degree representation is converted into the continuous numerical variable, so that the mathematical difficulty for a machine learning theory is solved. A drill bit abrasion degree calculation model is established in a data driving mode, and drill bit abrasion degree calculation quantitative analysis is achieved; through main influence factor univariate analysis, the method can achieve the effects that (1) the factor which has the largest influence on the abrasion degree of the drill bit is determined, drilling parameters are optimized, serious abrasion of the drill bit is avoided, and the service life of the drill bit is prolonged; and 2) the wear degree of the drill bit is quantitatively analyzed, and thus making a decision reference for drill bit replacement in tripping.

Description

technical field [0001] The invention belongs to the field of drilling engineering, and in particular relates to a quantitative evaluation method for drill bit wear applicable to machine learning. Background technique [0002] The degree of wear of the drill bit has always been a hot topic in research. In engineering practice, mastering the degree of wear of the drill bit is often the key point for the next decision-making. However, due to the concealment characteristics of the drilling project, there are few monitoring methods, and the actual wear degree of the drill bit is difficult to grasp. In practice, ROP is used as the main reference index, and other engineering parameters such as drilling pressure, torque, and rotational speed are used to comprehensively analyze the degree of drill bit wear. Whether the ROP is normal is often judged by regional experience and expert experience, which brings many uncertain factors to the project. The bit wear monitoring based on the m...

AI Technical Summary

Problems solved by technology

The bit wear monitoring based on the mechanical specific energy (MSE) theory can be used to calculate the bit mechanical specific energy quantitatively through the drilling parameters or qualitatively analyze the bit wear trend by drawing a trend line. In practical applications, the bit torque and sliding friction coefficient are the same The two key parameters are difficult to obtain through direct measurement, which directly affects the calculation accuracy

[0003] It is also difficult to use the traditional ROP equation to evaluate the degree of drill bit wear. The theoretical ROP is not easy to obtain. There is no reference. The bit wear coefficient, formation drillability coefficient, and wellbore cleaning coefficient required by the ROP equation are not easy to obtain. Cause its use value is not high

The method of calculating rock mechanical strength and rock drillability by using logging data to optimize the drill bit has lag, especially in the case of exploration wells, it is difficult to directly analyze rock drillability and drill bit wear degree through real-time logging data

Images