Questions of application of attributes in processing seismic data.
In the last some years use of various attributes at processing seismic data finds the increasing application.
The major factor constraining their use is the ineradicable belief in a miracle. That is the belief that after calculation of new attribute, it will show us where is oil and where is not present.
As a rule, the miracle does not happen, that is quite natural. And the trust to attributes falls up to their full disbelief. As a matter of fact, the question of trust becomes a question of competence of the geologist supervising the project. The qualified geologist understands that any geophysical method can be used only in a complex with other geological and geophysical information. That the same complex can give well result in one geological conditions and appear useless in others.
Whether use of spectrum seismic attributes for searches of oil is possible? Certainly. But thus there should be not analytical, but an empirical approach. The attribute should be tested on the investigated area and to give a positive effect confirmed by statistics. At a stage of searches, at absence of productive wells, it is possible to be based on positive results in similar geological conditions. Like - the alphabet. In practice these approaches are ignored pretty often. Works which somewhere have given good result are ordered. Often, the client gives a material, hiding the information accessible to him on productive horizons. Such approach at once reduces opportunities of processing, excepting from process the empirical approach. It opens road to the different voodoo, ready to find oil anywhere on the basis of ostensibly new analytics.
The academic circles very much do not love the empirical approach. They do not consider it as the serious approach. Clearly, the analytics to a science is closer. But let's recollect that the empirical approach has given the human bronze, gunpowder, the field-glass. Yes what there to speak, at the analytical approach the first fire would be about 200 years ago, not earlier. Analytics - a thing fine. Let's recollect the equation of a pendulum. Beautifully, simply. One trouble - does not work. For in the nature there is no dot weight and not extensible strings.
And what in seismic prospecting? Let's remember, that the seismic signal is defined not by lithology, he is defined by acoustic border. In infinite thickness of sandstone (as well as any other rock) at all will not be any signal. Thus is value of any spectra attribute will depend on such chasm of factors what to try to solve a problem of his corresponding to lithology or to presence of a product analytically - pure voodoo.
