Factor analysis prediction of effective stress from measurable rock attributes and calibration data

We show how to apply the statistical factor analysis method to effective stress prediction from seismic attributes. Two examples are discussed, one based on a laboratory dataset and relevant to the unloading mechanisms of abnormal pore pressure generation, and the other based on well data and a seismic section, relevant to the compaction disequilibrium mechanism. Arguably, this method of seismic interpretation is superior to the existing deterministic and statistical approaches as it reduces the multitude of available seismic attributes to a small number of factors most relevant to the effective stress and,in addition, provides a mathematically systematic way of arriving at most probable effective stress values as well as the confidence range.