A method for evaluating reservoirs in fracturing wells

Abstract

The invention discloses a method for evaluating a reservoir of a fractured well. The method comprises the following steps: S1, establishing an evaluation database M of the reservoir, wherein M is composed of a parameter set X, a sample set U and an index set Y; S2, dividing all the parameters into three kinds, and establishing a three-stage evaluation system including an index layer, an element layer and a target layer; S3, calculating the correlation between each parameter and open-flow capacity in the sample set U, and establishing a single-factor evaluation criterion set V; S4, calculating the influence weight W of each parameter to fracture improvement effect and the influence weight K of each kind of parameters to the fracture improvement effect in the sample set U by using a grey correlation method; S5, calculating and evaluating the membership degree R of each parameter in the object X in four different stages in the single-factor evaluation criterion set V by using normal-distribution membership functions; and S6, carrying out blurring operation on the weight W and the membership degree R, and grading quantifiably and comprehensively. The method for evaluating the reservoir of the fractured well, which is disclosed by the invention, has high objectivity and accuracy; a novel decision method is provided for selecting a fractured improvement well.

Description

technical field [0001] The invention belongs to the field of oil and gas field development, and in particular relates to a fracturing well reservoir evaluation method using a fuzzy mathematical method combined with oil field geology and fracturing operation data. Background technique [0002] Since the implementation of fracturing stimulation in the 1940s, oilfield engineers mostly use empirical evaluation methods to determine the quality of oil and gas well reservoirs, qualitatively predict the stimulation effect of single wells after fracturing, and make decisions for oilfield engineers to select wells . Therefore, most evaluation methods of this type are subjective and arbitrary, and the estimated coincidence rate is low. At the beginning of the 21st century, from the perspective of fracturing well energy and recoverable reserves, oilfield engineers used oilfield test data as a means to qualitatively evaluate the fracturing potential of a single well, which mainly includ...

AI Technical Summary

Problems solved by technology

The characteristic of the neural network method is that it does not require a specific mathematical model, and the model training is carried out through weight distribution to obtain the predicted production of single well reconstruction, which greatly improves the coincidence rate of the evaluation results, but the model training is complicated and the convergence is poor. At the same time, new data is added Afterwards, the model must be retrained, and the practicability is weak; fuzzy mathematics can effectively explain multi-parameter decision-making events in oilfield engineering, and overcome the problems of long training time and repeated training of neural networks, but the weight distribution often adopts expert scoring. Human subjective factors often reduce the accuracy of the evaluation results. The above quantitative evaluation methods also mainly focus on method research, and there are few researches on parameters that affect the effect of fracturing stimulation. This is also one of the main reasons for reducing the evaluation accuracy. 1 (Li Hongqi, Guo Haifeng, Guo Haimin, etc. Research on Data Mining Method for Complex Reservoir Logging Evaluation [J], Acta Petroleum Sinica, 2009, 30(4): 542-548)

Images