[Sgems-users] Problem in COSGSIM results

[Bernardo L. <ber...@gm...>]

Hello.

I am trying to create porosity scenarios from a 3D grid with uncertainty,
modeled as missing points in the grid, and using the grid's permeability as
a secondary variable with no uncertainty.
The porosity grid has values that range from 0  to 1 inclusive, and the
permeability grid has values equal or greater than 0. When performing
simulations on COSGSIM, the resulting porosity scenarios possess values
outside the original range, with negative and greater than 1 values. Am I
doing something wrong or is there a way to constrain the realization values
to a specified range?

Here are the paremeters set for the simulation:

Simulation grid: A newly created cartesian grid containing the original
permiability grid
Kriging type: Simple
Cokriging option: Markov Model 1

Primary data: Porosity data from a point set with missing points
(uncertainty)
Ranges: 15 | 15 | 15
Angles: 0 | 0 | 0
Secondary data: Permeability data from the simulation grid (complete grid
with all values)

Primary variable: Variogram model constructed from the original porosity
values (no uncertainty)
Secondary variable: Coefficient of correlation between the original
porosity and permeability grids


Thank you in advance,
Bernardo Lindgren