Techlog K.mod

Objectively reconstruct your missing data. Use log data and the powerful neural network capabilities to predict non-recorded parameters e.g. to reconstruct poor or missing data and to manage the shift from core to reservoir scale.

Support Contact Us Related Resources

Overview

Interactive, intuitive, and fast

The K.mod module is based on complex technology, but it remains easy to use. It is a straightforward, efficient tool that provides simple interpretation and accurate reservoir characterization.

Network design diagrams can be created via an interactive learning process.

Applications

Supervised neural networks

Parameters can be reconstructed or modeled directly from log data via an interactive learning process. The original variability in the data can be retained using the powerful, nonlinear Multilayer Perceptron modeling tool.

Benefits

Fully quantified uncertainties

Full control of input parameters is retained, while clear feedback on log and model quality is provided; the K.mod module is not a “black box” tool. Uncertainties can be managed on input (back propagation method to check the contribution of each input) and on output (self-organized map categorizes training and validation data for their effectiveness in modeling the target data). Inputs can also be weighted, which allows the forcing extreme values. If required, output and learning data distributions can be standardized to match dynamic ranges.

Features

Quantitative parameter modeling

The Techlog K.mod module extracts essential information from log data to:

Predict nonrecorded parameters (Ф, K).
Reconstruct missing or poor quality measurements to compensate for bad hole conditions, environmental effects, acquisition problems, etc.
Control scale shift management from core to reservoir scale.
Compare well log and core data to reduce the need for coring and plug analysis of subsequent appraisal wells.

Training