Overflow risk collaborative monitoring method and system for oil and gas well drilling process

Abstract

The invention provides an overflow risk collaborative monitoring method and system for an oil and gas well drilling process, and the scheme comprises the steps: obtaining ground and underground overflow monitoring parameters, and carrying out the data preprocessing; respectively constructing an expert system risk identification model and an intelligent risk identification model; and on the basis of the preprocessed data, achieving overflow risk monitoring in the oil and gas well drilling process by adopting a collaborative judgment mode combining the expert system risk identification model and the risk intelligent identification model. According to the scheme, the expert system and the deep learning technology are combined, the advantages of the two risk monitoring methods are complementary, and the overflow risk can be rapidly, accurately and intelligently judged.

Description

technical field [0001] The disclosure belongs to the technical field of well control safety in oil and gas drilling engineering, and in particular relates to a method and system for collaborative monitoring of leakage risks in the drilling process of oil and gas wells. Background technique [0002] The statements in this section merely provide background information related to the present disclosure and do not necessarily constitute prior art. [0003] Overflow and lost circulation are two prone risks in the drilling process, which will not only cause serious damage to the reservoir, increase development costs, and reduce development efficiency, but once the control is not effective, it will also induce drilling such as stuck pipe, well collapse, and blowout The accident occurred, causing heavy casualties and economic losses. Therefore, real-time monitoring and early warning of early overflow and lost circulation during drilling is of great significance to realize safe and ...

AI Technical Summary

Problems solved by technology

However, the effective application of deep learning technology is supported by rich monitoring parameters and a large number of risk samples, and it is impossible to establish an accurate identification model when the number of risk samples is small

[0006] On the other hand, most of the existing leakage risk monitoring methods use a single monitoring method, and the comprehensiveness and accuracy of risk discrimination are not enough; the expert system relies too much on expert experience when using it, and it is necessary to manually design a feature extractor to extract drilling parameters. Due to the lack of risk samples for training in the initial application of intelligent identification methods, the accuracy of risk identification is not high

Images