Study On The Methods To Improve The Reservoir Geological Modeling Accuracy Of Horizontal Well Platform Area In Fuyu Oilfield

In eastern parts of our country,new oilfields would be increasingly hard to found,strengthening the depth of investigation of old blocks,increasing more recoverable reserves of the remaining oil,improving the recovery factor is becoming more and more urgent.Fuyu oilfield is a typical middle-later periods of the development of oilfield,the water content is 90.5%,well spacing density is 43.73 well/km2.To establish reservoir geological model which could accurately represent the underground reservoir characteristics,is the key to predict the remaining oil distribution and the horizontal well design.because the traditional geological modeling existing the problem,such as:determining the accurate reserve parameters,calculating exact surface variation function,establishing the different scale reservoir framework,geological model accuracy is not good enough,lead to the low horizontal well drilling rate.Given these problems,this paper takes platform 12 well block of Fuyu oilfield as an example,through the integration of comprehensive study,by using multiple direction horizontal well,seismic and vertical wells,analyzing the delta plain distributary channel sand body,using different materials work for different levels of variation function,introducing new methods to obtain accurate physical parameters at the same time.constantly improving the understanding skeleton sand body of underground reservoir,reducing the uncertainty of reservoir recognition,this paper draws the following results:1.Proposed research method that using KRR to predict the reservoir porosity based on detection of nonlinearity,in order to improve the single well porosity interpretation accuracy.Comparing the results of a variety of methods,the root mean square error of KRR is 1.6132,far less than others.It can conclude that KRR model can extract the nonlinear mapping relationship between reservoir porosity and logging information asmuch as possible,solve nonlinear problem well.KRR method predicts the reservoir porosity has better prediction effect.2.Proposed research method that uses modified K-C equation to interpret permeability in different flow units,in order to improve the single well permeability interpretation accuracy.Based on the cementation prediction,calculate the FZI and identify seven kinds of flow unit,and establish the interpretation model,permeability correlation coefficient between modified K-C equation and core analysis is 0.9831.Results shows that: cementation index has great influence to the flow unit index,modified K-C equation can improve the accuracy of reservoir flow unit identification,then permeability interpretation accuracy can be improved.3.Discusses the research method that using different data for planar variation function.The major range calculated by vertical well logging is from 133.22 m to256.62 m,the minor range is 80 m to 112.81 m.The major range calculated by seismic wave impedance is from 471 m to 632 m,the minor range is 354 m to 561 m.The major range calculated by horizontal well logging data is from 22 m to 63.7m,the minor range is 10.2m to 16.8m.Results shows that:variation function range which get from different data has big difference.It reflects sand body variation characteristics of different scales,can provide random simulation of different levels geologic body with appropriate variation function.4.Explore a set of methods that establish model in different technical process and different hierarchical.Analyzing the results of modeling with different scales of variation function,the restriction of seismic compound sand body model is set up.And with the restriction of single sand body configuration interface,single sand body model is established.