BibTeX
@MISC{Demmenie95forwardmodelling,
author = {P. Demmenie},
title = {Forward Modelling of Resistivity Logs using Neural Networks},
year = {1995}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
In the oil industry, one uses the difference in electrical conductivity between hydrocarbons (oil and gas) and water to determine the hydrocarbon saturation of rocks. To this end, a measuring device is lowered in a well bore to record a so-called resistivity log. Due to environmental effects this resistivity log differs from the resistivity of the formation, the true resistivity. To invert the measured log to the true resistivity one uses an iterative forward modelling process, involving the numerical solution of differential equations. Although the current modelling algorithms have significantly improved in speed comparing to a few years ago, they are still not fast enough for inversion on a well site. Therefore, we have investigated the feasibility of using neural networks to perform the forward modelling process. Once trained, neural networks are very fast in producing output to certain input. The one-way mapping between the earth models (true resistivity model and environmental co...
Keyphrases
neural network forward modelling resistivity log true resistivity iterative forward earth model true resistivity model numerical solution differential equation certain input measured log environmental co current modelling algorithm hydrocarbon saturation oil industry environmental effect so-called resistivity log well site electrical conductivity resistivity log differs speed comparing one-way mapping measuring device
