Method and device for predicting total organic carbon content of hydrocarbon source rock

A technology of total organic carbon content and prediction method, applied in the field of petroleum exploration, to achieve the effect of improving generalization ability

CN113219553AInactive Publication Date: 2021-08-06CHINA UNIV OF PETROLEUM (BEIJING)
4 Cites 3 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Publication Date
2021-08-06
Estimated Expiration
Not applicable · inactive patent

Smart Images

  • Figure 1
    Figure 1
  • Figure 2
    Figure 2
  • Figure 3
    Figure 3
Patent Text Reader

Abstract

The invention provides a method and device for predicting the total organic carbon content of hydrocarbon source rock which are applied to a hydrocarbon source rock total organic carbon content prediction system. TOC prediction is carried out through a Bayesian neural network method of principal component analysis. Model training is carried out according to different salinity, the prediction precision of TOC is improved through optimization of the method, and the problem of hydrocarbon source rock prediction under different salinization conditions is solved. Meanwhile, when a logging curve is missing, a delta logR model optimized by a supplementary method is used for prediction, so that the requirement of a well location area with insufficient data on TOC prediction precision is met.
Need to check novelty before this filing date? Find Prior Art

Description

Technical field

[0001] The present invention relates to the field of petroleum exploration techniques, and in particular, there is a prediction method and apparatus for predicting a total organic carbon content of hydrocarbon source rocks. Background technique

[0002] The total organic carbon content of the hydrocarbon source is an important basis for identifying an effective hydrocarbon source rock. The hydrocarbon organic carbon abundance is an important position in the evaluation of the ease of oil and gas reserves. In practical applications in oilfield, the impact of cost and technology is difficult to continuously obtain the total organic carbon content, so generally use continuous logging parameters to predict the total organic carbon content prediction.

[0003] At present, there are more commonly used two hydrocarbon source rocks to log prediction methods, multi-regression, ΔLogR, artificial neural network law, and support vector machines and nuclear magnetic resonance w...

Examples

Embodiment

[0065] See figure 1 However, the total organic carbon content prediction method of hydrocarbon source rock provided in the present invention is applied to the total organic carbon content prediction system of hydrocarbon source rocks, including:

[0066] Step S101, acquire sample TOC data and log curves for training to be tested;

[0067] Because the neural network is subsequent, it is necessary to support a large amount of data, so the sample TOC data and log curves in the area to be tested first. These data are also also used to select logging parameters with high correlation with TOC data. By obtaining a single logging parameter curve of the hydrocarbon source rock and the correlation of the sample TOC data, it is determined which log parameters are higher and the TOC data is higher, and the selection is selected. Due to differences in sample TOC data and log curves in different to-be tested, there may be different regions, and the parameters and quantities selected by logging ...