This project is associated with computational sciences activity group under the collaboration framework between Schlumberger, the largest oilfield services company in the world, and IPT (Petroleum Engineering & Applied Geophysics) and IDI (Computer and Information Science) at NTNU. The aim is to delineate (semi)-automatically the boundary of the geometric complex geobody (e.g. salt bodies) within a seismic cube. In the image/geometry processing world this task is equivalent to the extraction of 3D point clouds from a set of non-random images.
Figure 1 showed the inline slice of velocity model from the EAGE salt model data. It seems not too hard for the demo data since the structure is relatively simple and the noise is absent. From the real data (Figure 2) it can be seen that the boundary is not that easy to pick given the complicated structures and low signal to noise ratio, especially around the flank.
Figure 1: Inline slice of EAGE salt model (velocity model)
Figure 2: A real seismic example from the Gulf of Mexico (left). A multi‐Z interpretation and mesh (right)
While delineating salt traditionally requires interpreting multiple single depth value horizon segments, many commercial software like Petrel, Geoframe, Landmark … have started the study of multi-Z interpretation in recent years.
Petrel implemented the Multi-Z Interpretation by (1) Manual interpretation on time slice, INL/XLN/Random line. Both, the original seismic and the attribute were used; or (2) Manual interpretation + horizon autotracking (define the top of the body).
1. Flank auto tracking: the bottom and top of the body can be auto tracked by the horizon auto tracker. While for the flank of the body, it is cumbersome to auto track the data with the high dip angle and poor signal (indistinguishable in some case).
2. Mistie: Body is a 3D object. Even the boundary can be auto tracked or manual picked from inline/cross line/time slice/random line/composite line, the difference from different slices will be a problem.
By having the Multi-Z autotracking module, user will simply need to input a seismic cube and get (semi)-automatically, i.e. with minimal interaction, the geobodies delineated with multi-Z interpretations, that itself will yield 3D point cloud data.
The project is ambitious and could lead to a publication. It will be implemented as a Plugin for Petrel, which is a Schlumberger owned E&P software platform that provides an integrated solution from exploration to production.
[Necessary] Good programming skills in Matlab, C#, TDT4195 (Visual Computing Fundamentals), Image Processing and Analysis.
Dataset to be used: Will be provided by Schlumberger
Advisors: Liyuan Xing (NTNU), Qiang Fu (Schlumberger), Nader Salman (Schlumberger), Victor Aarree (Schlumberger), Theo Theoharis (NTNU)
You can ask us for full description with images.