It wasn’t so long ago that geologists and geophysicists each labored in their own separate universe, so to speak, with little or no direct interaction.
In the mid-to-late 1990s, 3-D seismic grew to prominence as a kind of end-all, be-all in the E&P realm, soon creating a synergy between these professions that is considered to be routine today.
It was a significant turning point in the industry.
Geophysical data have proved invaluable to the geologist in myriad ways, particularly as a means to visualize aspects of the subsurface over large areas.
Think 3-D geological models.
“In its basic form, a 3-D model communicates the same information as a geological map,” said Mark Jessell, WA Fellow/Winthrop professor at University of Western Australia. “It’s a visualization of a geologist’s view of the distribution and structural relationships between rock units.”
In fact, the model essentially serves as the foundation for further investigation.
A number of different modeling schemes have been developed over the course of the last 30 years to enable geologists to build 3-D geological models, according to Jessell.
They vary in their employment of primary observations and geological knowledge to constrain the 3-D model geometry.
He emphasized that existing 3-D geological modeling systems are well-adapted to environments rich with data, such as basins where 3-D seismic provides stratigraphic constraints. Yet they are poorly adapted to regional geological problems.
“There are three areas where improvements in the workflow need to be made,” Jessell said, pointing to:
- Handling of uncertainty.
- The actual model building algorithms.
- Interface with geophysical inversion.
For the novice, geophysical inversion is a mathematical process enabling explorers to obtain added knowledge from geophysical data by converting geophysical measurements into subsurface 3-D images. These images can then be integrated with other geologic information, both subsurface and above ground.
Noting that all 3-D models are under-constrained, Jessell cautioned that the practice of creating just a single model ignores the enormous uncertainties underlying model construction processes. This hinders the relay of meaningful information to the end user about the elementary risk entailed in using the model to solve geological problems.
“Future studies need to recognize this and focus on the characterization of model uncertainty, spatially and in terms of geological features,” he said, “and produce plausible model suites instead of single models with unknown validity.”
Implicit Algorithms
Jessell, who will present the paper “Next Generation 3-D Geological Modeling and Inversion” at the SEG annual meeting at the end of October, noted that the most promising systems for understanding uncertainty use implicit algorithms given that they allow the inclusion of certain geological insight, such as relative ages of faults and onlap-offlap relationships.
However, these existing implicit algorithms lack inclusion of normal structural criteria, such as lineations and poly-deformation recognition, owing to their origin at the mine or basin scale. Such criteria are basic implements for a field geologist who is working to map the geology in a structurally complex area.
As a result, the modeling workflow requires manual intervention.
“One area of future research will be to establish generalized structural geological rules that can be built into the modeling process,” he said.
The most formidable challenge, according to Jessell, is the need for geological meaning to be maintained during the model building processes.
Currently, complex 3-D geological models incorporate geological and geophysical data along with the prior experience of the modeler, by means of the interpretation choices.
“These inputs are used to create a geometric model, which is then transformed into a petrophysical model prior to geophysical inversion,” he said. “All of the underlying geological rules are then ignored during the geophysical inversion process.
“Examples exist that demonstrate that the increased use of uncertainty characteristics in the workflow can at least partially overcome the loss of geological meaning between geological and geophysical modeling.
“The use of uncertainty metrics provides several potential pathways for the improved integration of geological, petrophysical and geophysical data during inversion,” Jessell emphasized.