--> Abstract: Support Vector Regression to Estimate Sonic Log Distributions and Overpressured Zones, by Cranganu, Constantin; Breaban, Mihaela; #90163 (2013)

Datapages, Inc.Print this page

Support Vector Regression to Estimate Sonic Log Distributions and Overpressured Zones

Cranganu, Constantin; Breaban, Mihaela

In the oil and gas industry, characterization of pore-fluid pressures and rock lithology, along with estimation of porosity, permeability, fluid saturation and other physical properties is of crucial importance for successful exploration and exploitation. Along with other well logging methods, the compressional acoustic (sonic) log (DT) is often used as a predictor because it responds to changes in porosity or compaction and, in turn, DT data are used to estimate formation porosity, to map abnormal pore-fluid pressure, or to perform petrophysical studies. However, despite its intrinsic value, the sonic log is not routinely recorded during well logging. Here we propose the use of a method belonging to the class of supervised machine learning algorithms — Support Vector Regression (SVR) — to synthesize missing DT logs when only common logs (such as natural gamma ray — GR, or deep resistivity —REID) are present. The Support Vector Regression approach can be divided into three steps: (1) supervised training of the model; (2) confirmation and validation of the model by blind-testing the results in wells containing both the predictor (GR, REID) and the target (DT) values used in the supervised training; and (3) applying the predicted model to wells containing the predictor data and obtaining the synthetic (simulated) DT log. SVR methodology offers significant advantages over traditional deterministic methods: strong nonlinear approximation capabilities and good generalization effectiveness. These result from the use of kernel functions and from the structural risk minimization principle behind SVR. Unlike linear regression techniques, SVR does not overpredict mean values and thereby preserves original data variability. SVR also deals greatly with uncertainty associated with the data, the immense size of the data and the diversity of the data type. A case study from the Anadarko Basin, Oklahoma, involving estimating the presence of overpressured zones, is presented. The results are promising and encouraging.


AAPG Search and Discovery Article #90163©2013AAPG 2013 Annual Convention and Exhibition, Pittsburgh, Pennsylvania, May 19-22, 2013