Invention Grant
US09229127B2 Methods program code, computer readable media, and apparatus for predicting matrix permeability by optimization and variance correction of K-nearest neighbors
有权
方法程序代码,计算机可读介质和用于通过K-最近邻的优化和方差校正预测矩阵磁导率的装置
- Patent Title: Methods program code, computer readable media, and apparatus for predicting matrix permeability by optimization and variance correction of K-nearest neighbors
- Patent Title (中): 方法程序代码,计算机可读介质和用于通过K-最近邻的优化和方差校正预测矩阵磁导率的装置
-
Application No.: US13773231Application Date: 2013-02-21
-
Publication No.: US09229127B2Publication Date: 2016-01-05
- Inventor: Nicolas Leseur
- Applicant: Saudi Arabian Oil Company
- Applicant Address: SA Dhahran
- Assignee: Saudi Arabian Oil Company
- Current Assignee: Saudi Arabian Oil Company
- Current Assignee Address: SA Dhahran
- Agency: Bracewell & Giuliani LLP
- Agent Constance Rhebergen; Christopher L. Drymalla
- Main IPC: G06F17/18
- IPC: G06F17/18 ; G01V9/00 ; G01V1/50 ; G01V99/00

Abstract:
Methods, program code, computer readable medium, and apparatus estimating permeability at unsampled (uncored) but logged interval locations in a reservoir based upon a similarity with log signatures from cored intervals, are provided. An example of a method can include performing a core plug-by-core plug blind test, defining an objective function which qualifies the quality of the permeability prediction to minimize discrepancies between measured and predicted permeability, optimizing calculated parameters, and estimating permeability values of uncored logged intervals. The method can also include applying a debiasing/variance restoration to compensate for inherent averaging artifacts.
Public/Granted literature
Information query