Interactive Petro-Seismic modelling

The Indexed Self Organising Map (I.S.O.M.) technology in Techsis is used to condition the seismic data volumes to the wellbore curves. This step involves sorting and classifying mini-traces (length of contiguous seismic response data) from each of the presented attribute volumes simultaneously into a Self-Organized Map.

INPUT DATA N Seismic attributes

Delimitation of studied area as 3D closed object	Upscaled Log data

The map is then indexed against chosen upscaled log data, for instance Vsh, so that seismic data patterns can be directly associated with the volume of shale signatures recorded at the wellbores. The Indexed Map is then used to back-predict the log reference data at the well locations, to compare to the original curves to verify the robustness of the characterization.
It may be necessary to iterate through the indexation process at this stage before a robust result is achieved.
To assist the comparison and indexation process, Techsis includes an automated method to find the best seismic calibration model based on I.S.O.M.

A solver has been developed to find the best solution in terms of:

the most relevant input blocks for the petrophysical prediction,
the best size of the I.S.O.M.,
the best mini trace size,
the best combination of wells for the indexation phase taking into account both location and log data uncertainties.

The method thus saves the analyst considerable time in studies involving reservoir characterization from seismic multi-attribute data.

This method gives the best model available from the multi-attribute volumes using four main steps in the interpretation workflow:
The creation of all possible maps,
The optimal calibration of the maps sorted according to a confidence criterion,
The application of the best map to create new petrophysical attribute blocks,
The extraction of thematic bodies.

The uncertainty with which the prediction of the petrophysical properties is made is related to the dispersion of seismic patterns within the nodes of the I.S.O.M.
This information is available as an Error Map; and all predicted volumes, slices, logs etc. can be displayed simultaneously with their associated error.