Thank you for your comment. To help clarify the point, this process actually did use AI as stated in the video. The justification is that we mimicked a process of a user who manually depth-matched logs. Then we developed a special workflow using extensions to perform the exact same step. We welcome the opportunity to provide more information. Please contact us at [email protected] for more details.

Hello, we thank you for your questions. Here is our response to both: The inversion analysis process only processes the individual CPD’s at well locations. So if there are 10 wells in an inversion project, the inversion analysis processes only 10 CDP’s. This enables QC and parameter experimentation to be performed efficiently before running the whole volume through inversion. Inversion analysis also sets values such as the model-seismic scaling (the model has very small values <1.0 while the seismic usually has values 10,000+ times larger). The last set of parameters used during inversion analysis are remembered by Strata. Apply to volume, then applies the remembered parameters from inversion analysis, to the complete seismic volume. It is important that inversion analysis is run so that scaling and parameters are set, before applying to the whole volume.

We used Greenberg and Castagna (1992) method which proposes that the shear-wave velocity for a brine-saturated rock with mixed mineral components can be given as a Voigt-Reuss-Hill average of the volume components of each mineral. This is an iterative method and was applied only to the samples in the selected reservoir zone which matched zone conditions. For all other samples, Castagna's equation was used.

@@ReservoirExperts thank you. Did you carried out any fluid substitution at the hydrocarbon zone to brine condition to honor the castagna equation to get Vs. ?

Hello Bruno, thank you for your comment. Our reply below: The discontinuities are mainly from the mode mixing problem from EMD, which is triggered from the sifting process to extract each instinct mode function (IMF). Using the injection of controlled zero mean, Gaussian white noise, ensemble empirical mode decomposition (EEMD) and complete ensemble empirical mode decomposition (CEEMD) effectively reduces mode mixing. Adding white Gaussian noise helps perturb the signal and enables the empirical mode decomposition algorithm to visit all possible solutions in the finite neighborhood of the final answer, and it also takes advantage of the zero mean of the noise to cancel aliasing [Wu and Huang, 2009]. There are two ways to reduce the discontinuities. First, increase the ensemble times, which will stabilize the results. Second, instead of digging into each IMF property, HRS extracts more robust attributes, like the peak frequency, to explore more information from seismic data.

dropped the negative value

hi salim ahmadi. if you have hampson and russel software plz plz send me my email is [email protected] thanks in advance

@@gamingwithanin3120 bang, pas saya load seg y nya kok, data sumur saya gk muncul ya penampang seismik stratanya

@@eldwinmanalu boleh minta software HRS 9? kalau berkenan mungkin bisa mengirim lewat email saya [email protected], terimakasih kak

@@eldwinmanalu tolong mas, apakah bole saya meminta softwarenya?🙏🙏

You are correct that HRS AVO does not currently support exporting this last probability volume for further quantitative analysis. However, I’m happy to report that this feature is currently slated for release in HRS 10.3 - coming in 2017. As always, thank you for your interest and continued support of CGG GeoSoftware.