Method and system for intelligent identification and early warning of drilling risk

Abstract

The invention discloses a method and system for intelligent identification and early warning of drilling risks. The method comprises the steps: collecting real-time logging data of historical wells and drilling risk record data, and constructing a drilling risk case database and a normal drilling logging database; based on the drilling risk case database and the logging database, constructing and training by using a random forest machine learning algorithm to obtain a plurality of drilling risk identification early warning models; and inputting the real-time logging data of the target well into each drilling risk identification and early warning model according to the input parameter requirements of different drilling risk identification and early warning models so as to execute real-time identification and early warning processing of the drilling risk of the target well. The method can meet the requirement of safe drilling, reduces the false alarm rate and the missing report rate, achieves the real-time and effective recognition and early warning of the drilling risk, and achieves the purpose of improving the risk early warning accuracy.

Description

technical field [0001] The invention belongs to the field of petroleum engineering, and in particular relates to a method and system for real-time identification and early warning of drilling risks. Background technique [0002] With the continuous deepening of exploration and development, the difficulty of oil and gas exploration and development is getting higher and higher, the geological conditions are becoming more and more complex, the buried depth of reservoirs is increasing, and drilling engineering is facing more and more complex situations, resulting in higher and higher costs for dealing with drilling risks and accidents. The realization of safe drilling is the primary goal of the drilling industry. Although theoretically relying on calculation models of key parameters such as formation pressure, wellbore pressure, and frictional torque, early warning and analysis of drilling downhole risks can be carried out, but because these calculation models have many assumpti...

AI Technical Summary

Problems solved by technology

Although theoretically relying on calculation models of key parameters such as formation pressure, wellbore pressure, and frictional torque, early warning and analysis of drilling downhole risks can be carried out, but because these calculation models have many assumptions at the beginning of their establishment, there are deviations from the real downhole environment. Affected the accuracy of risk warning results

At the same time, factors such as downhole high-frequency vibration, high temperature and pressure, and drilling fluid flow directly restrict the use of sensors to obtain downhole conditions for early warning of risks.

[0003] At present, the drilling early warning methods provided in patents and literatures are mainly divided into two categories: one is through the calculation of key parameters such as wellbore circulation pressure during drilling, wellbore fluctuating pressure during tripping, and drill string friction torque. However, there is a problem of insufficient applicability of the model, which limits the accuracy of its discrimination and early warning. This is mainly reflected in the fact that there are many assumptions in the model, the empirical parameters are related to the drilling area, and some model data are actual. It is difficult to obtain other aspects during application; the other is to use gray correlation, decision tree and other methods to analyze the changes of relevant parameters based on the logging data obtained by ground sensors on the drilling site for risk identification, but the parameters set based on expert experience and cases There are obvious regional differences in data such as thresholds and weights, and poor applicability leads to a high false positive rate

Generally speaking, the accuracy rate of early warning of drilling kick risk by the current technical solutions is still difficult to meet the needs of safe drilling, and the false alarm rate and missed alarm rate remain high in actual application, and the accuracy rate is not high

Images