A weathered crust karst fracture-cavity type oil reservoir oil well water breakthrough time prediction method based on machine learning

Abstract

The invention discloses a weathered crust karst fracture-cavity type oil reservoir oil well water breakthrough time prediction method based on machine learning. The method comprises the following steps of S01 determining the karst background of a fractured-vuggy oil reservoir to be weathered crust karst, and determining the reservoir type of a reservoir where an oil well is located; S02 identifying the oil pressure change type of the water-seen high-yield oil well by using a knn algorithm; S03 establishing a characteristic corresponding relation between the fracture-cavity reservoir driving stage and the oil well oil pressure stage; S04 establishing a quantitative plate of each reservoir type of the fracture-cavity oil reservoir; S05 counting water breakthrough influence factors and waterbreakthrough time of the fracture-cavity oil reservoir oil well; S06 determining an initial clustering center k of the fracture-cavity reservoir oil well; S07 establishing a mean value clustering prediction model of the water breakthrough time of the fracture-cavity type oil reservoir oil well. According to the present invention, the water breakthrough early warning quantitative chart and water breakthrough time k-means clustering prediction method has the advantages of being wide in application range, being easy to operate and the like, and being able to quickly guide adjustment of on-site measures.

Description

technical field [0001] The invention relates to a method for predicting the water breakthrough time of an oil well in a fracture-cavity reservoir, in particular to a machine learning-based quantitative prediction method for the water breakthrough time of an oil well in a fracture-cavity reservoir. Background technique [0002] Carbonate oil and gas reservoirs rank first in the world's oil and gas reserves and production, and are highly concerned research objects at home and abroad. There are large areas of carbonate formations distributed in my country, and there is a huge room for innovation in the geological research of carbonate reservoirs. Fracture-cavity carbonate reservoirs, as a special type, occupy a large proportion of oil and gas resources at home and abroad. The Ordovician oil reservoir in Tahe Oilfield of Sinopec Northwest Oilfield Company is a typical fracture-cavity carbonate reservoir. After more than 20 years of exploration and development, it has built the ...

AI Technical Summary

Problems solved by technology

The theory of this method is strong, but it is affected by many factors in practical application, the measurement accuracy required for the parameters is high, and the multi-solution is serious.

②Indoor macroscopic 3D physical simulation method: This method takes the 3D seismic sculpted body of the fractured-vuggy reservoir as the prototype model, prepares the macroscopic 3D physical model of the fractured-vuggy unit according to the similarity theory, and studies the water drive characteristics and bottom water cone during elastic development through laboratory experiments. The bottom water coning stage and the main controlling factors are analyzed, so as to predict the water breakthrough time of fractured-vuggy reservoirs. Limited by factors such as scale and model dimensions, it is impossible to reproduce the three-dimensional flow characteristics of real reservoir pressure and temperature conditions. The experimental results have certain guiding significance for the understanding of the mechanism of bottom water coning in fracture-vuggy reservoirs, but they cannot guide the actual oil well

[0004] It can be seen that the above-mentioned methods for predicting the water breakthrough time of oil wells in fractured-cavity reservoirs have the problem that their practical application is greatly affected by external factors, and their scope of application is limited, resulting in poor prediction results.

Images