QI considerations
« Quantitative Interpretation » versus « Reservoir Characterization »
Namely, the former is often restricted to a complex workflow combining seismic inversion and classification into various facies (e.g. reservoir / non reservoir) whereas the latter stresses the impact of seismic data quality. Those definitions do not oppose one to another, they rather complement each other.
As a matter of fact, seismic inversion does involve seismic data quality control via the key step of wavelet estimation (aka quantitative wavelet estimation). However, when you think about it, it involves a very parcimonious dataset and does not allow verifying the wavelet properties (frequency, phase and amplitude content) stationarity/variability. Well locations are sparse pinheads over a map, especially in an exploration context (not so much in a field development project), even if you are lucky enough to get well penetration over the seismic volume to be qualified.
Combining seismic-to-well ties to areally extensive metrics based on the sole seismic data allows to evaluate the randomness or spatial organisation (likely due to overburden variations).
Wavelet estimation is a highly unstable process. Indeed, this yields a complete deterministic result (only hard data plus some form of numerical stabilization) but sparse. It is always worth complementing with some highly redundant and mappable result depending on some hypothesized geological set-up.( fed by geological regional knowledge, if any. e.g. seabed).
In theory, seismic inversion can be seen as a way to remove the band limited impact of the propagating wavelet. However, due to natural limitations (loss of higher frequencies during propagation and noise corruption) this is only partially effective. As a consequence, thin layers are not resolved and actual elastic properties are not properly recovered, which is a limitation to the classification step that follows. Many authors have described techniques to evaluate reservoir thickness, more precisely net reservoir thickness or net-pay in favorable conditions directly from seismic data of easily derived volumes thereof (e.g. relative, pseudo or coloured inversion).
In addition, traditional inversion parameters (e.g. P-wave, S-wave ipedances or acoustic impedance, Poisson’s ratio or Vp/Vs ratios) are not necessarily the most efficient parameters to capture the targeted reservoir properties.
That is to say that many reservoir characterization objectives can be robustly achieved using techniques dealing directly with seismic data, provided they are duly QCed and conditioned based on those QC results. I believe that the term « Quantitative Interpretation » stresses more effectively the importance of qualifying seismic data for the purpose of « reservoir characterization« . and the potential of direct amplitude interpretation.