- Home
- Conferences
- Conference Proceedings
- Conferences
First EAGE Workshop on Optimizing Project Turnaround Performance
- Conference date: 23-24 February 2021
- Published: 23 February 2021
-
-
Seismic Processing - What is required?
Authors T. Martin, Ø. Korsmo and N. CheminguiSummaryIn a turnaround and cost conscious environment, do we really need to apply all the algorithmic processes in a seismic processing sequence? Full wavefield migration may be one way to eliminate certain processes. If we treat the full wavefield migrated image as part of an inverse problem in a least-squares migration, we may exclude more steps. Least-squares full wavefield migration (Lu et al., 2018) uses the raw seismic data, and many processing steps in both the data and image domain can be excluded, potentially reducing turnaround whilst maintaining, or improving, image quality for the entire data record.
-
-
-
Seismic Data Compression Impact on Quality and Near Realtime Data Transfer Impact on Delivery
More LessSummaryWe demonstrate that for modest compressions (below 50:1) the quantization noise in the image domain is low. Such data compressions can facilitate near real time streaming from the acquisition platform and lead to reduce turnaround by overlapping both subsurface model derivation and signal processing with the acquisition. We demonstrate this reduction on turnaround on two projects hampered by onward logistics issues which makes tradition data delivery via physical media much more difficult. We summarise the use cases for different levels of compression where quality constraints may outweigh turnaround reduction.
-
-
-
From Acquisition to First Image with Full Waveform Inversion
More LessSummarySubsalt imaging is an industry level challenge that has existed for many years in Gulf of Mexico. Both correct salt shape and kinematically correct detailed sediment velocity are required to successfully image the reservoir level especially below salt bodies. The conventional method in the industry is to build the model from top down with flooding with sediment and salt velocity then delineate the salt bodies by manual interpretation. This process requires intensive labor and usually creates unoptimized subsalt images in the presence of complex salt geometries. Full-waveform inversion (FWI) can build a high-resolution model including salt and sediment delineation and reflectivity can be easily extracted from this model for early viewing which can be a jump start for the interpreters to browse around especially if the project is large scale. In this paper we demonstrate a quick velocity and reflectivity generation from raw acquired data.
-
-
-
Can Data Analytics Help Reduce Seismic Processing Turnaround
More LessSummarySeismic data volumes are increasing and pressure is growing to accelerate the turnaround time of seismic processing projects. Testing, validation and production administration are time consuming components of a project. Testing is performed to optimize the parameters for each step in the processing sequence. It can be both computer and human resource intensive, as many testing phases require numerous repeat runs and significant interaction with the data. A Proof-of-Concept (POC) example illustrates that parameter testing can be helped by mining for parameters from a database.
-
-
-
Robot Butlers, Flying Cars, Automated Seismic Processing and Imaging – How near are we?
Authors J. Brittan, R. O’Driscoll, J. Walpole and Y. CoboSummaryLike many of the promised technological marvels we may have expected by 2020, automated seismic data processing and imaging remains stubbornly out of reach. The explosion in the widespread use of ML based algorithms has the potential to significantly alter the way seismic processing and imaging projects are undertaken. However, as well as much remaining technological development there are significant non-technical barriers (often economic and psychological) that hinder the widespread adoption of these approaches. To better understand these barriers, we present three examples of recent technological development that are intended to reduce human project time and improve turnaround and discuss the path and landscape required for widespread adoption of technologies of this type.
-
-
-
Cloud for faster, better and increased project confidence
By M. SkinnerSummaryCloud is a broad term including compute, storage and networks, offering a range of solutions that can be considered as bespoke, private or public. Cloud suppliers are varied in their services and the playing field for customers can be complex but navigating through this leads to compelling technology and business models.
Here we consider how and why cloud is being adopted and how cloud can not only help us to go faster but can also improve the way we work with seismic survey data. Gaining higher levels of data confidence and faster knowledge.
We can consider benefits from cloud for processing compute capacity, remote operations, collaboration, transparency, data access and commercially.
-
-
-
‘Race-track’ Depth Imaging; A case study from the Campos Basin
Authors P. Karalis and R. BartlettSummaryLicensing rounds require novel acquisition, processing and project management solutions in order to deliver high quality data to prospective clients in demanding timescales. During 2019, the latest phase of a large 3D exploration-scale seismic acquisition programme was completed in the Campos Basin, offshore Brazil, targeting data delivery for the 16th Brazil licensing round.
This abstract uses the Campos 3D case study to highlight how modern acquisition design can work hand-in-hand with fast-track processing and depth imaging. Onboard compute power, cloud resources and data compression algorithms were harnessed to achieve an efficient collaboration between the seismic vessel and onshore multi-disciplinary support and client teams. This resulted in a successful delivery of a high quality pre-stack depth migration volume and added value in decision making in this time-critical licensing round.
-
-
-
Deep Learning Techniques Revolutionize E&P – Two practical applications
Authors C. Jaikla, E. Alkan, C. Sutton, Y. Cai, S. Mannava, A. Gala, P. Devarakota, D. Knott, L. Chernis, H. Badat, G. Madiba, D. Segonds and D. HohlSummaryMajority of the deep learning techniques in seismic image analysis focus on solving one task at a time and ignore the richness of presence of many other structures in the vicinity and their correlation with the task of interest at hand. These approaches work best in solving the identification of simple structures in the shallow areas of the survey where the signal-to-noise ratio is high and struggle in deeper areas as the signal becomes weaker. In addition, it is a challenge to acquire the right data and quality labels to train the deep learning models for some of the fundamental challenges in geoscience. In this paper, we present two recent applications of deep learning in seismic processing that are targeted to reduce the cycle time of seismic processing projects. In the first example, we present the concept of learning multiple related tasks and demonstrate on a use case that the multi-task learning learns common representations of related tasks (salt body and salt boundary) and improves the accuracy of identifying and segmenting salt boundary structures even in low signal-to-noise areas. In the second example, we reconstruct the regularly and irregularly missing in narrow-azimuth data using an encoder-decoder style convolutional neural network.
-
-
-
Keynote: From Acquisition to Final PSDM Volume in Days via Full-bandwidth FWI
Authors M. Warner, T. Kalinicheva, F. Mancini and H. DebensSummaryFWI run to full bandwidth can remove multiples and ghosts from raw field data. Differentiation of the resultant high-resolution velocity model then allows the generation of a full-bandwidth PSDM reflectivity image without conventional processing or migration. We have run FWI to 100 Hz on raw field data from a marine towed-streamer dataset. We show that the results are similar to, and are broader bandwidth than, conventionally processed PSDM images. When run on the public cloud, this approach allows the production of final full-bandwidth PSDM volumes within in a few days of completion of data acquisition.
-
-
-
Accurate Velocities and Reduced Cycle Times from Cloud-enabled Full Waveform Inversion Using XWI (AWI and RWI)
More LessSummaryThe central objective of advanced Full Waveform Inversion is to enable rapid turnaround of accurate velocities directly from raw seismic data. AWI with its convolutional filter-based residual represents a fundamental change in the way FWI is normally run, as an ‘add on’ to time-consuming velocity- model building performed on pre-processed data, where its role is to finesse a tomography starting model.
Here we show the combined solution of AWI and RWI known collectively as XWI serves as a predictor for unseen drilling logs. The inversion is run on raw data from NW Australia (6 sailline validation test) demonstrating convergence to essentially the same result from two simple 1D starting models. It is able to predict deviations in a sonic log from a starting position over 1000 m/s away from the measured value.
The cloud environment where XWI ingests traces directly from blob storage consists of a dynamic pool of interruptible compute instances. This allows for cost-effective frequency sweeps through the iterations and scalable hyperparameter scans. It also is configured for interfacing XWI with interactive processing software for seamless trace preparation and project start up.
-