The implementation of a fracturing measure becomes a common yield increasing measure for high-water-content later-stage
oil field development and exploitation, and on-
site selection of a fracturing construction well before fracturing also depends on an empirical method. The method has the characteristics of
blindness and subjectivity, and the final fracturing effect is often not satisfactory. Along with the proposal of an
oil field digitization strategy of a
petroleum system,
oil field production pays more and more attention to collection and accumulation of data. A large amount of production data are also accumulated in oil field fracturing production, the data are actually a set of information related to fracturing production activities, and when the data are accumulated to a certain degree, the data inevitably contain available rules. According to the patent, a model for judging whether fracturing is effective or not is established by using a
support vector machine data mining algorithm on the basis of oil field reservoir
physical property factors such as oil layer mid-depth, thickness,
porosity, permeability, shale content and
water saturation. And furthermore, a new fracturing operation is expected to be predicted by utilizing the established model for judging whether fracturing is effective or not, the new fracturing operation is guided, the future production is predicted to lower the development risk, and the economic benefit of an enterprise is improved.