ReLoc

ReLoc – is a software product that involves obtaining predictive maps of any reservoir parameter using neural networks. ReLoc has three modes of operation, depending on which parameter needs to be predicted.

ReLoc – is the mode of operation of the ReLoc program used to predict net reservoir thicknesses.

One of the main questions for an oil and gas company: where to drill? A common problem in seismic net reservoir thickness prediction is the low degree of their correlation with seismic attribute maps. Neural networks can solve this problem by finding non-linear relationships between net reservoir thicknesses and sets of different seismic maps.

The ReLoc program allows you to create, configure, train neural networks and obtain predictive maps of net reservoir thicknesses. The effectiveness of this method can improve the efficiency of production drilling. Use Machine Learning to predict net reservoir thicknesses.

PetroLoc – is the mode of operation of the ReLoc program, used for the forecast of the injection water front and the map of remaining reserves (LTRO).

Input maps are used to predict waterflooding zones: net pay thickness, average permeability, reservoir pressure. Training is carried out on point well data with the values ​​of the current water cut in production wells and the accumulated water injection in injection wells.

In order to get free trial version of ReLoc software please proceed to this page.

Load seismic maps and well data

Analyse data and get correlation matrix

Tune machine learning

Water cut map based on the results of neural network forecast

Our software is powered up by advenced machine learning frameworks TensorFlow and Keras