Petrophysical interpretation of well log datausing VFSA-I 2 algorithm

摘要

In the paper a global optimization method solving the nonlinear geophysical well logging inverseproblem is presented. We adopted a Very Fast Simulated Annealing (VFSA) optimization technique,which is advantageous to use because of the quickness of searching the absolute optimum of the objec-tivefunction. At first a point by point inversion method using local response equations is applied toestimate the petrophysical parameters (porosity, water saturations, shale and matrix content) in differ-entdepth-points separately. It results in better estimations compared to the conventional linearizedinversion methods, but the accuracy is still limited because of the marginal overdetermination of theinverse problem. To improve inversion results, we formulate the VFSA interval inversion procedure(VFSA-I 2 ), which uses the well log data set along the overall depth-interval in a joint inversion proce-dure.In the framework of this algorithm the depth-dependent petrophysical parameters are expandedin terms of an appropriately chosen set of functions and the expansion coefficients are considered asunknowns.To test and compare the inversion methods in-situ data measured in a hydrocarbon exploratory bore-holeare inverted. It is shown by means of the model variances that the interval inversion algorithmresults in much more accurate petrophysical parameters than the local point by point inversionmethod. Moreover, it is able to give an estimate for the layer-thicknesses as an additional geologicalunknown by inversion.