Font Size: a A A

Applied Research Of Reservoir Prediction In The Area Of Hongtai

Posted on:2013-10-03Degree:MasterType:Thesis
Country:ChinaCandidate:W B GaoFull Text:PDF
GTID:2230330374476556Subject:Earth Exploration and Information Technology
Abstract/Summary:PDF Full Text Request
At present, in the most of big oil and gas fields in this world, especially the oil field in the east part of China, the newly found oil and gas in place are predominantly from subtle reservoirs, which are very hard and costly to be developed. The existing problems are mainly due to the shortage of effective methods and technologies to predict and describe this kind of reservoir, especially the poor resolution being far from what is required.The Geophysicists establish a common long-term purpose at inversion accuracy of lithological parameters. With the operation of Oilfield by seismic, either reservoir prediction or oil and gas delineation, reservoir description, impedance inversion is a very important and reality method. It is a deterministic primarily method to solve lithological and reservoir problem with impedance inversion technology, which has clear physical meaning and intuitive expression comparing with statistical methods. According to logs feature of region of interest, the article takes reservoir inversion of generating a new acoustic curve. Using acoustic curve mixes effectively natural potential, gamma ray and special resistance of reflecting layer lithological alternation, and translating a new acoustic curve with dimension of acoustic curve, so it can reflect alternation of formation velocity and impedance, and also can reflect hairline of formation lithology. Finally, we process inverse impedance utilizing Jason inversion software. By means of comparison of inversion results before and after processing acoustic curve, we obtain that the processed results can show the divergence feature of reservoir impedance preferably.The seismic attribute is another important mean of horizontal reservoir prediction. Seismic attributes are specific measurements of geometric, kinematic, dynamic, or statistical features derived from seismic data. Seismic attributes indicate all information subsets in initial seismic data and the relation of all kinds of attributes is very complicated. Firstly, Seismic attributes are more profoundly understood through studying signification, classification, extraction methods and influence factors of seismic attributes. This dissertation expounded the extracting method and classifying method. At the same time, it expounded the principle of seismic attribution analysis techniques and the problems, then discussing the relationship between seismic attributes and oil-gas possibility.Neural network is a non-linear dynamic systematic, establishing memory through studying samples to determine closest memory in unknown mode. It could solve problems with complex information, unclear background knowledge and uncertain rules, besides allowing samples with larger loss and distortion. Geophysicists get more important to this advantage. Appling neural network to petroleum-gas prediction plays an important role in improving the accuracy of reservoir prediction and further improving the success rate of drilling, will also have greater economic and social benefits. Therefore, the research of artificial neural network has great prospects and potential. BP neural network use gradient descent to decrease error function, and to study a large number of seismic parameters. This article uses BP neural network to dealing with and analysis, with seismic attributes and testing data as samples, forecast the oil-gas possibility, and gets some results.The paper take S2sand ground of Hongtai area as an example, combining previous results, applies BP neural network technology, seismic attribution technology and impedance inversion technology, then researching oil and gas favorable areas, and submits two wells deployment to guide further development.
Keywords/Search Tags:Impedance inversion, curve reconstruction, Seismic attribute, BP neural networks, Reservoir prediction
PDF Full Text Request
Related items