This year’s International Meeting for Applied Geoscience and Energy, beginning August 28 in Houston, doesn’t have a single, overarching theme. But one concept keeps popping up again and again: Innovation.
Interested in carbon capture and storage? IMAGE has the technical panel session, “Innovative Global Technologies Applied to CCS – Current and Future Emerging Technologies.”
Want to know more about the role of innovation in oil and gas? Try the IMAGE panel session, “Innovation, the Role it Played and its Potential Impact on the Industry.”
Other sessions will explore commercializing innovation, assessing emerging technologies and innovating change in the dynamics of energy use. The meeting includes an Aug. 29-30 “U-Pitch New Technology Showcase” to promote technical innovation and new products.
AAPG and the Society of Exploration Geophysicists, in conjunction with SEPM, are hosts for the second annual IMAGE meeting. According to a statement from the event’s co-chairs, “near-surface engineering and environmental sciences, digitalization, carbon management and global opportunities will be in the limelight.”
It’s designed as a broad-interest event, with 30 key topic areas ranging – alphabetically – from acquisition and survey design to structure, tectonics and geomechanics. As an AAPG/SEG-sponsored event, IMAGE will include several sessions related to geological analysis, imaging and interpretation.
Offerings in Geophysical Imaging
Joël Le Calvez is geophysics adviser and fiber optic program manager for Schlumberger in Houston, and a member of the IMAGE discussion panel, “Imaging Acquisition and Interpretive Technologies: Surface and Subsurface.”
Le Calvez said there are several fronts to consider in geophysical imaging innovation, including acquisition, processing and interpretation.
Acquisition: “Fiber and other tools can be deployed cheaply as a distributed measurement tool rather than punctually/sensor point.”
Processing: “Techniques have evolved with more data to look at and consider, while signal-to-noise reduction techniques are improving.”
Interpretation: “Based on newly acquired data only, based on integrating old and new data, integrating from various domains now that computers are more powerful and software has evolved.”
Data techniques: “The ubiquitous artificial intelligence, machine learning, etc., which call for training datasets with the usual notions of bias.”
Heather Bedle, assistant professor in the school of geosciences at the University of Oklahoma, will be a member of the same discussion panel.
“I would say that the greatest innovation emerging will be in the visualization of multidimensional data, combining products from seismic, well logs and production, in ways that the interpreter can quickly interpret patterns from these different data sources,” Bedle observed.
“The other aspects of innovation are along the Big Data avenue – improved tools to combine, correlate and assess,” she added.
New Acquisitions Technologies
In the imaging area, attendees at IMAGE ’22 “should be on the lookout for new acquisitions technologies that will help them see faster and better for less money, with faster turnaround times,” Le Calvez advised.
As an example, he cited the use of fiber optic data acquisition for vertical seismic profiles. Zero-offset VSPs that once took weeks to produce “now can be a matter of less than an hour,” he said.
Innovative approaches to data processing include computer-based, automated analysis and application of very large datasets. However, quality control is a critically important part of utilizing large datasets, and human judgment remains essential to the evaluation, according to Bedle.
“Merging large datasets can still be unwieldy, and to a certain point, we do not want to automate everything. There is essential value in QCing the data and incorporating the geologic knowledge of the interpreter,” she noted.
“In this line, automated workflow in software analysis packages is not ideal, but rather we need programs to guide the interpreter through the steps, which may include QC work and parameterization of the algorithms,” she said.
In a similar way, Le Calvez emphasized the importance of ensuring data quality as new and innovative data-capture technologies and interpretation techniques emerge.
“It’s like a good meal. If you get good quality food, unless your cook is a crook, you’re going to get a good meal,” he said.
Starting with poor quality food guarantees a poor meal and the same is true of poor data, Le Calvez observed. Bad data leads to bad output no matter how advanced the computing and interpretation techniques.
IMAGE ’22 also will delve into computing and data applications in energy, sometimes combining the innovation/energy/data themes in sessions like “Leveraging Software and Infrastructure to Drive Innovation through Data” and “Innovative Solutions and Data Science Applications in Unconventional Resource Evaluation.”
“On the interpretation side, more and more people are looking at – the buzzword of the day – machine learning,” Le Calvez noted.
He’s not as big a fan of the application of AI/ML as some data scientists, viewing them as works-in-progress in the computing world. Le Calvez noted the importance of reliable data, good training datasets and effective interpretation algorithms in machine learning.
“We don’t have as yet, in my opinion, enough properly-labeled data points or datasets,” he said.
Bedle said she frequently uses both AI and machine learning as helpful tools, where “the caveat is that users need to know what they are doing and how it works, and not use it without awareness of its limitations and assumptions.”
Nonetheless, Le Calvez observed, automation in data applications does offer a number of significant advantages, including the ability to automate and quickly accomplish a wide variety of routine tasks.
“The good thing is that we can free people to do tasks that can’t be simplified,” he said.