Abstract
The minimization of the incoherence (the discrepancy between measured and calculated logs) is a principle of complex well log interpretation. If lithological models with different numbers of unknowns (components) occur, however, the incoherences can be compared directly.
The probability, α, of validity of rock component models is introduced as the basis of comparison. The effect of restrictions on the mathematically evaluated mineral volume fractions is discussed. The theoretical Chi-distribution is modified and the limits of its application are examined in case of restrictions.
The model probability, α, concept is applied for practical illustration in the interpretation of a well in a sandstone formation.