First Break December 2023 - Data Management and Processing by EAGE

VO L U M E 41 I I S S U E 12 I D E C E M B E R 2 0 2 3

SPECIAL TOPIC

Data Management and Processing EAGE NEWS Unmissable field trips for Oslo 2024 TECHNICAL ARTICLE Deep structural imaging in the Vienna basin

TPS

(TUNED PULSED SOURCE)

Innovative Low Frequency Broadband Marine Seismic Source

VISUALIZE COMPLEX GEOLOGIES LIKE NEVER BEFORE TPS provides geophysical clarity by enhancing the quality and value of seismic data and enabling effective imaging of the most complex geologies. It has been successfully deployed in commercial marine acquisition operations. Choose TPS to Optimize Your Marine Seismic Acquisition •Improved imaging •Reduced environmental impact •Enhanced operational efficiency •Improved safety

For more information

Nantes, France sales.nantes@sercel.com Houston, USA sales.houston@sercel.com

www.sercel.com

FIRST BREAK® An EAGE Publication

CHAIR EDITORIAL BOARD Gwenola Michaud (gmichaud@gm-consult.it) EDITOR Damian Arnold (arnolddamian@googlemail.com) MEMBERS, EDITORIAL BOARD • Lodve Berre, Norwegian University of Science and Technology (lodve.berre@ntnu.no) • Philippe Caprioli, SLB (caprioli0@slb.com) • Satinder Chopra, SamiGeo (satinder.chopra@samigeo.com) • Anthony Day, PGS (anthony.day@pgs.com) • Peter Dromgoole, Retired Geophysicist (peterdromgoole@gmail.com) • Kara English, University College Dublin (kara.english@ucd.ie) • Stephen Hallinan, CGG Stephen.Hallinan@CGG.com • Hamidreza Hamdi, University of Calgary (hhamdi@ucalgary.ca) • Clément Kostov, Freelance Geophysicist (cvkostov@icloud.com) • Pamela Tempone, Eni (Pamela.Tempone@eni.com) • Angelika-Maria Wulff, Consultant (gp.awulff@gmail.com) EAGE EDITOR EMERITUS Andrew McBarnet (andrew@andrewmcbarnet.com) MEDIA PRODUCTION Saskia Nota (firstbreakproduction@eage.org) PRODUCTION ASSISTANT Ivana Geurts (firstbreakproduction@eage.org) ADVERTISING INQUIRIES corporaterelations@eage.org EAGE EUROPE OFFICE Kosterijland 48 3981 AJ Bunnik The Netherlands • +31 88 995 5055 • eage@eage.org • www.eage.org EAGE MIDDLE EAST OFFICE EAGE Middle East FZ-LLC Dubai Knowledge Village Block 13 Office F-25 PO Box 501711 Dubai, United Arab Emirates • +971 4 369 3897 • middle_east@eage.org • www.eage.org EAGE ASIA PACIFIC OFFICE UOA Centre Office Suite 19-15-3A No. 19, Jalan Pinang 50450 Kuala Lumpur Malaysia • +60 3 272 201 40 • asiapacific@eage.org • www.eage.org EAGE AMERICAS SAS Av. 19 #114-65 - Office 205 Bogotá, Colombia • +57 310 8610709 • +57 (601) 4232948 • americas@eage.org • www.eage.org EAGE MEMBERS CHANGE OF ADDRESS NOTIFICATION Send to: EAGE Membership Dept at EAGE Office (address above) FIRST BREAK ON THE WEB www.firstbreak.org ISSN 0263-5046 (print) / ISSN 1365-2397 (online)

Streamlining energy and production data management from field to processing.

Editorial Contents 3

EAGE News

Personal Record Interview — Daniella Bordon

Monthly Update

Crosstalk Industry News

Technical Article Deep structural imaging in the Vienna basin Ahmed Mamdouh, Klaus Pelz, Sandor Bezdan, Erika Angerer, Alexandra Oteleanu, Harald Granser, Abdelrahman Abubakr, Alexander Sakharov and Ian F. Jones Applied fault topology: understanding connectivity and uncertainty of fault systems that define and affect commercial and environmental projects Frank Richards, Mark Cowgill and Megan Rayner

Special Topic: Data Management and Processing Automating the data flow — how AI and ML can reimagine subsurface data management Christopher Hanton and Venkatesh Anantharamu Streamlining energy and production data management from field to processing Kristy DeMarco Large-scale industrial deployment of machine learning workflows for seismic data processing Julien Oukili, Jyoti Kumar, Jon Burren, Steve Cochran, Martin Bubner, Denis Nasyrov and Bagher Farmani How the latest SEG-Y revision will improve data management Jill Lewis, Shawn New, Joel Allard and Victor Ancira Data agility: Innovative approaches to subsurface data management Jose Chapela Affordably making the invisible unmissable Neil Hodgson, Karyna Rodriguez, Helen Debenham and Lauren Found Advanced imaging of hybrid acquisition data: Exploring new frontiers Sylvain Masclet, Fang Wang, Guillaume Henin, Loic Janot, Olivier Hermant, Hao Jiang, Nicolas Salaun, David Le Meur and Daniela Donno Calendar

cover: This month we publish papers showing how the field has progressed in the past year.

FIRST

BREAK

VOLUME

DECEMBER

2023

European Association of Geoscientists & Engineers

Board 2023-2024

Near Surface Geoscience Circle Esther Bloem Chair Andreas Aspmo Pfaffhuber Vice-Chair Micki Allen Contact Officer EEGS/North America Adam Booth Committee Member Hongzhu Cai Liaison China Deyan Draganov Technical Programme Officer Wolfram Gödde Liaison First Break Hamdan Ali Hamdan Liaison Middle East Vladimir Ignatev Liaison CIS / North America Musa Manzi Liaison Africa Myrto Papadopoulou Young Professional Liaison Catherine Truffert Industry Liaison Mark Vardy Editor in Chief Near Surface Geophysics Florina Tuluca Committee member

Oil & Gas Geoscience Circle Edward Wiarda President

Laura Valentina Socco Vice-President

Pascal Breton Secretary-Treasurer

Caroline Le Turdu Membership and Cooperation Officer

Peter Rowbotham Publications Officer

Yohaney Gomez Galarza Chair Johannes Wendebourg Vice-Chair Lucy Slater Immediate Past Chair Erica Angerer Member Wiebke Athmer Member Tijmen Jan Moser Editor-in-Chief Geophysical Prospecting Adeline Parent WGE & DET SIC liaison Matteo Ravasi YP Liaison Jonathan Redfern Editor-in-Chief Petroleum Geoscience Aart-Jan van Wijngaarden Technical Programme Officer

Sustainable Energy Circle Carla Martín-Clavé Chair Giovanni Sosio Vice-Chair

SUBSCRIPTIONS First Break is published monthly. It is free to EAGE members. The membership fee of EAGE is € 80.00 a year including First Break, EarthDoc (EAGE’s geoscience database), Learning Geoscience (EAGE’s Education website) and online access to a scientific journal.

Maren Kleemeyer Education Officer

Aart-Jan van Wijngaarden Technical Programme Officer

Esther Bloem Chair Near Surface Geoscience Circle

Companies can subscribe to First Break via an institutional subscription. Every subscription includes a monthly hard copy and online access to the full First Break archive for the requested number of online users. Orders for current subscriptions and back issues should be sent to EAGE Publications BV, Journal Subscriptions, PO Box 59, 3990 DB, Houten, The Netherlands. Tel: +31 (0)88 9955055, E-mail: subscriptions@eage.org, www.firstbreak.org. First Break is published by EAGE Publications BV, The Netherlands. However, responsibility for the opinions given and the statements made rests with the authors. COPYRIGHT & PHOTOCOPYING © 2023 EAGE All rights reserved. First Break or any part thereof may not be reproduced, stored in a retrieval system, or transcribed in any form or by any means, electronically or mechanically, including photocopying and recording, without the prior written permission of the publisher.

Yohaney Gomez Galarza Chair Oil & Gas Geoscience Circle

FIRST

BREAK

PAPER The publisher’s policy is to use acid-free permanent paper (TCF), to the draft standard ISO/DIS/9706, made from sustainable forests using chlorine-free pulp (Nordic-Swan standard).

Carla Martín-Clavé Chair Sustainable Energy Circle

VOLUME

DECEMBER

2023

HIGHLIGHTS

Near surface geoscience shines in Edinburgh

Borehole geophysics back and alive in Milan

Volcanic meeting in Prague

ield trips you won’t want to miss F at 2024 Annual in Oslo Plans for the field trips to be included in the programme for next June’s 85th EAGE Annual Conference and Exhibition in Oslo are already underway. As tradition, we will offer a wide range of interesting and knowledgeable field trips to our delegates as part of the main conference. Looking at all the positive feedback that we got from participants of field trips during the EAGE Annual Conferences in the past, we want to keep expanding our portfolio of field trips in 2024. We offer various topics in which you can choose to participate depending on your own interests. This is a great opportunity for geologists and engineers in the energy industry to have a better insight into the geology of the Oslo region, exchange knowledge and learn from each other. A total of five field trips will be held on Sunday 9 June, Monday 10 June and Friday 14 June 2023. The Geology Cambrian field trip is set to take place prior to the conference on Sunday 9 June. It will be a one-day field trip, starting in the morning from Lillestrøm to Huk-Bygdø to discover some small faults and folds. In the afternoon, the group will continue to visit a seismic scale fault by Oslo graben in Nessoden. On the same day, we will organise the Architecture and Geology field trip, with the support of the Natural History Museum of Oslo. Participants in this special field trip will be offered the opportunity to visit the museum

with a guided tour and an explanatory walk in the Tøyen area. The following day, Monday 10 June, we offer a half-day field trip to Hovedøya. The island of Hovedøya is only 5 mins sailing from the Oslo harbour and the Askerhus castle. Mooring on the northern corner, delegates will walk across the southern part of the island, focusing on three uppermost Ordovician formations: Skogerholmen, Husbergøya and

where the group can get an overview of companies engaged in the energy transition and then end with the tour of the Equinor facility focused on CO2 capture transport and sequestration. Another field trip will take participants to the Shearwater and Norwegian Geotechnical Institute (NGI) technology centres. While NGI facilities will be the opportunity to witness geotechnical testing that can measure behaviour of geomaterials

Field trip to Herøya Industrial Park will conclude the Annual Conference.

Langøyene. Along the walk, the group will pass the ruins of a Cistercian monastery dating back to 1147 which was destroyed after the reformation in Norway in 1536. On the last conference day, to wrap up the programme, together with our event host Equinor, is the CCS tour: Herøya Industrial Park field trip will take place on Friday 14 June 2023. The visit will kick-off with an introduction session about the main facility, followed by a tour of the park FIRST

BREAK

when subjected to the loads that come from planned construction such as buildings, roads, tunnels, platforms, and the like, the Shearwater Technology and Innovation Center will guide you through the development of products to be used in geophysical data acquisition. If you are interested in joining one of our field trips, please visit our website www.eageannual.org for more detailed information. Stay tuned for new updates in the upcoming months. I

VOLUME

DECEMBER

2023

EAGE NEWS

T ime is now to prepare presentations for EAGE Annual in Oslo The 85th EAGE Annual Conference & Exhibition, in Oslo on 10-13 June 2024, is set to be another milestone event for the industry as we explore the contributions of geoscientists and engineers under the theme ‘Technology and talent for a sustainable energy future’. Now is the time to consider submitting an abstract for the main technical programme for which the deadline is due on 15 January 2024. The location in Norway presents a valuable opportunity for the global geoscience audience to connect with a local

community of researchers, operators and service companies leading the transformation of the Norwegian Continental Shelf into a broad energy province. Reflecting on the programme for the upcoming event, Erling Vågnes, SVP subsurface exploration production international, Equinor and chair, EAGE Annual 2024 Local Advisory Committee, says, ‘We look forward to a broad and balanced technical programme built on in-depth discipline knowledge, focused on technical integration for holistic solutions and on cross-fertilisation between oil and

The Technical Programme for the Oslo Annual will be a balanced mix of topics, ranging from Energy Transition (CCS, geothermal energy, and hydrogen storage) to Near Surface (infrastructure planning and natural risk mitigation), along with a focus on O&G production and exploration. Reservoir engineering will also hold a significant spotlight. Our primary objective in crafting this programme is to facilitate knowledge sharing and the cross-fertilisation of ideas among various sectors.

Aart-Jan van Wijngaarden Leader work processes in E&P, Equinor and EAGE Technical Programme Officer

gas, CCS, renewables and infrastructure geosciences. Abstract submissions should reflect the latest technological advances and key case studies demonstrating the contributions of our deep technical disciplines in achieving a more sustainable and secure energy future.’ Technical submission topics include Geophysics, Geology, Reservoir Engineering, Integrated Subsurface, Energy Transition, Mining & Infrastructure, and Data & Computing Science. Presenters will also have the chance to choose between oral and printed poster presentation formats. The poster sessions provide a great space for presenters looking for a high level of interaction in a more relaxed setting. It also provides an opportunity for networking with more extended one-on-one discussions during the hour-long sessions. Our Technical Programme Committee is also working closely with the Local Advisory Committee, the Research Committee and the Technical Communities to prepare high-quality workshops and dedicated sessions. More details will be released soon. You can submit your papers and check out the new developments on the way to the EAGE Annual 2024 at www.eageannual.org.

ADDITIONS THIS MONTH Centered around ‘Unlocking New Energy Resources’, the 4th EAGE Eastern Mediterranean Workshop (4-6 December in Athens, Greece) delves into the intricate regional geology, structural styles, and depositional systems, shedding light on their impact on prospects and future potential in the Eastern Mediterranean. 35 papers will be presented. The first issue of Geoenergy is available. New issues of Near Surface Geophysics and Basin Research will be published in December.

FIRST

BREAK

VOLUME

DECEMBER

2023

Let's play

SPOT THE DIFFERENCE Hint: The power of least-squares imaging with the full wavefield.

1. DUG’S MULTI-PARAMETER FWI IMAGING USING RAW FIELD DATA

2. RTM USING THE SAME RAW FIELD DATA Re ce iv

urc So

t pe ar -c

e arp e -c

ndary t bou

dar un bo

Near-surface amplitude-extraction Answers: The top image has designature, deghosting, demultiple, interpolation, regularisation, superior focussing, superior illumination, true amplitudes, superior vertical and horizontal resolution.

Data courtesy of Shell

Full waveform inversion. Multi-parameter. Superior resolution. Superior imaging. A complete replacement for the conventional processing and imaging workflow.

9001

•

IS O

COMPASS

RVICES SE

•

SURANCE AS

•

dug.com/fwi

SURANCE AS

RVICES SE

COMPASS

Let’s talk — contact info@dug.com O 27001

EAGE NEWS

WGE Community highlights diverse leadership as key driver of energy transition

COMMUNITY

The energy industry is changing and, more than ever, creating bridges across disciplines, people and careers has become necessary to fulfill its demands. In 2023, this premise was at the heart of the EAGE Women in Geoscience and Engineering activities.

Dr Lim, she delivered a ‘thought-provoking keynote that articulated how driving equity is not just a moral imperative but also a strategic necessity for achieving sustainability. Her nuanced insights into the systemic challenges at the workplace, supply chain, and community levels

WGE committed to growing a collaborative network across genders, ages and specialisations.

Dr Anna Lim, WGE Community chair, says that in the context of current global challenges, such as energy transition, diverse leadership can offer unique perspectives and solutions. That’s why the team organised a session on the role of women’s leadership in securing a sustainable future at the 84th EAGE Annual Vienna, ‘for its timely relevance and critical importance to promote diversity and inclusion while spotlighting the role of women in sustainability’. The session featured Mikki Corcoran, VP of sustainability at SLB. According to

offered both a deep understanding of the issues at hand and a call to immediate action.’ The WGE Community also continued exploring ways to contribute towards an inclusive industry by hosting a discussion on ‘Neurodiversity in geosciences’. During this engaging online panel session, attendees were provided with an introduction to the topic, making it accessible to those new to the subject. The four keynote speakers not only shared their personal journeys and experiences but also discussed practical strategies

for navigating neurodivergence while pursuing fulfilling careers. Additionally, the panel addressed the prevalent but potentially misleading notion of a trade-off between accommodating work environments and productive work environments, shedding light on a complex issue within the discussion. Participants share that ‘thanks to these engaging initiatives, they could walk away not only with actionable insights but also with a renewed sense of empowerment.’ In 2024, the WGE community will continue with its mission of growing a collaborative network across genders, ages and specialisations to support equity in the fields of geoscience and engineering. A session on ‘Empowering diverse talent in a tech-driven world’ is in the planning for our upcoming Annual Conference in Oslo. If you would like to connect with the WGE Community, just update your affiliations – with EAGE Circles, Local Chapters, Special Interest and Technical Communities.

Update your EAGE affiliations!

EAGE Online Education Calendar START AT ANY TIME

START AT ANY TIME | VELOCITIES, IMAGING, AND WAVEFORM INVERSION - THE EVOLUTION OF CHARACTERIZING THE EARTH’S SUBSURFACE BY I.F. JONES (ONLINE EET)

SELF PACED COURSE

6 CHAPTERS OF 1 HR

START AT ANY TIME

GEOSTATISTICAL RESERVOIR MODELING BY D. GRANA | SELF PACED COURSE | 8 CHAPTERS OF 1 HR

SELF PACED COURSE

8 CHAPTERS OF 1 HR

START AT ANY TIME

CARBONATE RESERVOIR CHARACTERIZATION BY L. GALLUCIO

SELF PACED COURSE

8 CHAPTERS OF 1 HR

START AT ANY TIME

NEAR SURFACE MODELING FOR STATIC CORRECTIONS BY R. BRIDLE

SELF PACED COURSE

9 CHAPTERS OF 1 HR

10 OCT 19 DEC

GEOLOGICAL CO2 STORAGE BY A. BUSCH, E. MACKAY, F. DOSTER, M. LANDRO, P. RINGROSE

EXTENSIVE ONLINE COURSE

24 HOURS (INCL 7 WEBINARS OF 1-2 HRS EACH)

8 NOV 8 DEC

INTRODUCTION TO MACHINE LEARNING FOR GEOPHYSICAL APPLICATIONS BY JAAP MONDT

EXTENSIVE ONLINE COURSE

14 HOURS (INCL 4 WEBINARS OF 1-3 HRS EACH)

4-7 DEC

GEOPHYSICAL DATA ANALYSIS: CONCEPTS AND EXAMPLES BY R. GODFREY

INTERACTIVE ONLINE SHORT COURSE

4 HRS/DAY, 8 MODULES

* EXTENSIVE SELF PACED MATERIALS AND INTERACTIVE SESSIONS WITH THE INSTRUCTORS: CHECK SCHEDULE OF EACH COURSE FOR DATES AND TIMES OF LIVE SESSIONS

FIRST

BREAK

VOLUME

DECEMBER

2023

EAGE NEWS

First data processing workshop in Cairo

Join us for this highly anticipated first workshop on data processing in Cairo.

The inaugural EAGE Data Processing Workshop planned for 26-28 February 2024 in Cairo, Egypt, will focus on discussions concerning the significance of data acquisition technology and computational capacity. With the increasing demands of energy transition initiatives, such as the development of offshore wind farms, the extraction of critical minerals, decarbonisation efforts, and the shift away from fossil fuels, there is a growing need for high-resolution data processing. Furthermore, satellite imagery data is becoming increasingly prominent as the cornerstone of precise site characterisation.

Participating in this workshop presents a valuable opportunity to share the outcomes of advancements in technology and innovative thinking. It will bring together experts from the industry to engage in discussions and share their ground-breaking concepts with a global audience. Don’t miss this exceptional chance to gain insights into state-of-the-art technologies, connect with industry leaders, and contribute to the ongoing discourse. Register now to take advantage of the early bird registration rate and join us in exploring the future of energy.

Papers invited for Special Issue on multi-scale hydrogeophysics for water sustainability Our Near Surface Geophysics (NSG) journal is calling for submissions of original research work relating to the multi-scale hydrogeophysical studies carried out to achieve water sustainability. Efforts will be made to include geographically well-distributed studies conducted over diverse hydrogeological terrains. The contributions can deal with water sustainability-related issues at global, regional, or local scales. Studies on the development of improved methodology for modelling and those dealing with the application of emerging technologies employing UAVs or ATVs etc. are welcome. The journal’s premise for this Special Issue is that our planet is in the throes of extreme climate variability. Inextricable linkages of climate with the hydrosphere at different scales are playing havoc with the availability of water resources essential for the well-being and survival of humanity. Anomalous climate extremes are causing unexpected floods and long

periods of droughts in different parts of the earth, exacerbating water sustainability challenges. Hydrogeophysical investigations effectively address various related issues at global, regional, and local scales. We intend to highlight the latest applications of hydrogeophysics at various scales to address the problem of water sustainability. Topics for this call for papers include but are not restricted to: Groundwater sustainability in the time of climate extremes; Geophysical studies for groundwater sustainability at different scales and Recent advancements in geosciences to address groundwater sustainability. Guest editors are Saurabh Verma, Subash Chandra and Virendra Tiwari (National Geophysical Research Institute, India), John Lane (USGS, USA) and Tim Munday (CSIRO, Australia). Manuscripts should be prepared according to the author’s guidelines published on the NSG website and submitted using the online submission webpage. FIRST

BREAK

When submitting your manuscript, please answer the question ‘Is this submission for a special issue?’ by selecting the special issue title from the drop-down list. All manuscripts will be peer-reviewed in accordance with the journal’s established policies and procedures. The final selection of papers will be based on the peer review process as well as reviews by guest editors and the editor-in-chief. Submission deadline is 30 June 2024. I

VOLUME

DECEMBER

2023

EAGE NEWS

Our report on this year’s EAGE Near Surface Geoscience Conference and Exhibition (NSG2023) in Edinburgh on 3-7 September 2023 It was thrilling to welcome 560 participants to the Edinburgh event, bringing together a diverse group of academic, research, and commercial practitioners of geosciences, united with the common goal of sharing knowledge and insights.

Networking crowd.

Through the convergence of four parallel conferences: the 29th European Meeting of Environmental and Engineering Geophysics, the 3rd Conference on Geophysics for Infrastructure Planning, Monitoring and BIM, the 2nd Conference on Hydrogeophysics and the 1st Conference on Sub-surface Characterisation for Offshore Wind, the event succeeded in showcasing the synergy between different geophysical techniques. One of the standout aspects of the conference was the growing interest in the integration of geophysical and geotechnical approaches, with a strong emphasis on supporting civil engineering, geotechnical, and infrastructure sectors. This emphasis on understanding the strengths and weaknesses of different methodologies and using the right tools for the job significantly enhances the chances of a successful survey and meaningful results.

The community’s commitment to addressing the challenges society faces today was also evident throughout the event. In a world grappling with geo- and anthropogenic hazards, energy transition, and shifts in human behaviour concerning infrastructure and soil and groundwater use, the need for geoscientists to work together and offer their insights to the public and policymakers is more crucial than ever. The willingness and urgency of policy-makers and the general public to make changes were highlighted, reinforcing the need for geoscientists to bridge the gap between their solutions and the wider world outside. Arre Verweerd, chair of the Local Advisory Committee, said: ‘Looking back on a successful four days of NSG2023 in Edinburgh, I feel we have created deeper understanding and focus as a geoscience community on the issues society is struggling with today.’

Discussions focused on need for an integrated disciplinary approach.

FIRST

BREAK

VOLUME

DECEMBER

2023

NSG2023 also shed light on significant developments in geophysics for offshore wind. With offshore wind farm expansion growing globally to meet climate targets, a deep understanding of the near-shallow subsurface is paramount for successful foundation designs. The first conference on subsurface characterisation for offshore wind held in one of Europe’s offshore wind capitals, Edinburgh, was timely. It showcased a range of new technologies and data processing methods, notably ultra-high-resolution 3D seismic imaging. This was a key highlight, as it is instrumental for wind farm foundation designs. The conference dedicated its last half-day to near-shore investigations for landfalls, an area often underrepresented but equally important for offshore wind projects. Discussions focused on the need for early consultations with various stakeholders to optimise ground modelling approaches and quantify risks effectively. Another series of talks delved into quaternary geology, data integration, and ground modelling at various offshore wind farms across Europe. The discussions provided valuable insights into the challenges related to sedimentary and erosional environments, chalk formations, and the presence of boulders. The glacial history around the UK was a central topic, shedding light on its implications for subsurface characterisation.

EAGE NEWS

Enjoying field excursions.

Maarten Vanneste, co-chair of the 1st Conference on Sub-surface Characterisation for Offshore Wind, said: ‘As the EAGE’s first dedicated offshore wind conference, it received a lot of interest, excellent presentations, and marks a new and exciting part of the EAGE Near Surface Geoscience programme.’ Other talks during NSG2023 explored broadband processing for diffraction imaging, the repurposing of legacy seismic data, and advances in seismic data. Novel geophysical approaches, such as

surface and shear wave methods, were also discussed, including the use of fibre optic-based geophysics, i.e., digital acoustic sensing (DAS). Jonathan Chambers, chair of the 29th European Meeting of Environmental and Engineering Geophysics, said: ‘This is an exciting and rapidly emerging technology that will shape the field of near-surface geophysics for years to come.’ Moreover, the integration of machine learning in geophysics was emphasised, with a focus on using algorithms to assist

in interpretation rather than blindly following computer-generated results. This shift reflects the evolving landscape of geophysics, moving toward more intelligent and informed decision-making. EAGE Near Surface Geoscience Conference and Exhibition will celebrate its 30th edition next year in Helsinki, and it promises to be an extraordinary milestone. Make sure to save the date, 8-12 September 2024, and keep a close watch on www.eagensg.org for the latest updates and exciting developments.

Get ready for everything digital in Paris next March Anticipation is already building for the next EAGE Digital conference to be held in Paris on 25-27 March 2024 where the theme will be ‘Digital: Delivering better energy in a transforming world’. As the energy industry undergoes a significant change, EAGE Digital 2024 promises to be a crucial gathering for industry leaders, innovators and experts seeking to explore and shape the future of energy. The rapidly changing landscape of the energy industry provides exceptional challenges across the value chain. Energy companies are constantly pressed to adapt their business models and practices to stay competitive in the face of pricing fluctuations and shifting demand. Amidst these challenges, the power of digital transformation emerges as a game-changer, offering unprecedented opportunities

for sweeping changes and value creation. Furthermore, collaboration with new-tech industries has become imperative in our collective pursuit of the energy transition. In the inaugural edition of the EAGE Digital conference, we laid the foundation by focusing on the leadership aspects needed to enable digitalization and its impact on our business. The following edition dealt with digitalization as a key enabler to innovation and to the industry transformation needed through energy transition. These previous editions set the stage for the forthcoming one where our focus will be on organising with new tools and workflows to deliver the energy the world needs. We are looking for innovative and disruptive technologies that transform the digital space across different applications

FIRST

BREAK

of geoscience and energy transition. Are you one of the pioneers in the digital transformation of the energy industry? Have you gained insights, knowledge, or valuable lessons from your journey with digital transformation? If so, EAGE Digital 2024 offers you a unique opportunity to share your expertise and contribute to this transformative journey. Be integral to the discussions and presentations that will shape the future of the energy sector. There are still chances to join the line-up of our technical programme speakers. Hurry up and submit an abstract before 10 December 2023 to become a part of the event. Your contributions can make a significant impact on the industry, influencing the path it takes in this era of digital transformation. Visit eagedigital.org for more details.

VOLUME

DECEMBER

2023

EAGE NEWS

Geotech 2024 will cover fibre sensing and reservoir monitoring in parallel workshops

OUR JOURNALS

THIS MONTH

The Third Geoscience Technologies and Applications Conference (EAGE GeoTech) will make a comeback next spring (8-10 April 2024) in The Hague with the bonus of being two workshops packaged into one event.

Near Surface Geophysics (NSG) is an international journal for the publication of research and developments in geophysics applied to the near surface. The emphasis lies on shallow land and marine geophysical investigations addressing challenges in various geoscientific fields. A new special edition (Volume 21, Issue 6) will be published within December.

Basin Research (BR) publishes primary research on the science of geophysics as it applies to the exploration, evaluation and extraction of earth resources. A new edition (Volume 35, Issue 6) will be published in December.

Geoenergy focuses on the publication of timely and topical research in subsurface geoscience, critical for this new era of sustainable energy. The journal considers articles on the themes of: energy storage, subsurface disposal and storage, geothermal energy, hydrogen energy, critical minerals and raw materials, and sustainability. The first issue is still in progress and will close at the end of 2023. Accepted papers are available to read on EarthDoc.

CHECK OUT THE LATEST JOURNALS

NSG

FIRST

BREAK

VOLUME

Geoenergy

DECEMBER

2023

Following the success of the previous editions in 2021 (online) and 2022 (hybrid and London), the third edition of EAGE GeoTech is being adapted to in-person format, offering two technical workshops as part of one high-powered multi-disciplinary programme, namely the 4th EAGE Workshop on Distributed Fibre Optic Sensing and the 4th EAGE Workshop on Practical Reservoir Monitoring. Participants will benefit from exposure to two dedicated technical programmes, providing unique opportunities for interaction and fostering cross-discipline knowledge sharing to tackle some of the major challenges facing the industry with the ongoing energy transition. The 4th EAGE Workshop on Distributed Fibre Optic Sensing will continue with the exploration of the well-established use of distributed optical fibre sensors for diverse energy and industrial applications. The workshop will focus predominantly on geoscience and engineering applications utilising distributed acoustic (DAS), temperature (DTS) and strain (DSS) systems, including new emerging fields of application. The technical programme will contain presentations on advancements in instrumentation including 3C optical sensors, applications, and technology integration. A special focus will also be placed on current challenges, business impacts and evolving best practices. The 4th EAGE Workshop on Practical Reservoir Monitoring aims to show how the use of modern reservoir surveillance practices can be applied to ensure safe injection and drainage, while optimising field production and maximising value. The workshop will continue to investigate the benefits through multi-disciplinary data integration, of geophysical and engineering data, and increased digitisation that allow for improved work processes to enhance reservoir monitoring capabilities. The workshop positions itself as a great platform for showcasing embryonic ideas and technologies as well as future trends. To learn more or submit an abstract for these workshops, please visit the conference website at www.eagegeotech.org. Abstracts submission deadline is 22 January 2024.

EAGE NEWS

Why our learning catalogue continues to grow This year, EAGE offered its members more than 90 short courses in diverse formats, covering topics from geoscience and energy to engineering. Some of our presenters explain the rationale behind their recently introduced programmes. A new self-paced course titled ‘Near surface modelling for static corrections’ by Ralph Bridle is accessible on Learning Geoscience, EAGE’s online learning platform. Bridle says: ‘There are many reasons to study the near-surface and it is important to define the scope of this study. For civil engineering and archaeology, the aim is to create a highly defined image of the near surface. In engineering and exploration applications there is a huge difference in scale of the projects, and required accuracy of the modelling. In this class we are concerned with calculating a time correction for seismic reflection surveys as used in mineral and hydrocarbon exploration. For calculating time corrections for seismic reflection surveys, we are not so much interested in an exact model to image the near surface. Rather, the objective is a time shift to correct near-surface travel times, in order to eliminate false time structures, thereby avoiding the possibility of dry wells. The class introduces the complexities of the near surface. Then it describes the theories, assumptions and shortcomings of some methods in modelling the near-surface’. One of our new Interactive Online Short Courses (IOSCs) is ‘The interpreter’s guide to depth imaging’ by Dr Scott MacKay. It is intended to cater for seismic interpreters looking to incorporate depth imaging into their evaluations and depth-processing imagers aiming to enhance their collaboration with interpreters. ‘Basics of carbon capture and storage’ by Prof Mike Stephenson, part of the EAGE Education Tours (EET), was offered on-site during the EAGE Annual

Conference in Vienna, attracting over a hundred participants. Prof Stephenson explains: ‘The course satisfies a part of the market that is not currently catered for – the wider science, risks, financing, planning and social licence aspects of CCS. These are issues that are as important as the technical issues in the sense that any of them can be a show-stopper for CCS. Many geologists in companies first starting out in CCS will need to have a broad background of the science, technology, risks and policy planning aspects of CCS - also technical civil servants and planners want to know about the new technology of CCS including its risks, environmental aspects and social licence.’ Prof Stephenson also offers an IOSC on ‘Palynology for Geologists’ which, despite its name, is not exclusively for palynologists. It is designed to bridge the gap between palynology specialists and geological specialists working on plays or prospects. He says: ‘This course provides succinct information and insight into palynological data and techniques by instilling an understanding of the main palynomorph groups, their uses, advantages and disadvantages, and what they can

and cannot do - helping the non-specialist geologist get the most out of their palynological data.’ In 2023 ‘Navigating career challenges and oportunities of the energy transition’, also known as the Coaching Programme, was launched as one of the Extensive Online Short Courses (EOSCs). Our coaching organisers explain ‘The EAGE Coaching Programme prepares early career professionals with soft skills and strategies to navigate the challenge of choosing a career among several options. Its purpose is to maximise participants’ personal and professional potential by raising open questions and self-discovery. Reflecting on and dispelling false feelings and beliefs will allow young professionals to plan activities and track progress toward stated career goals. As well as to develop new habits to prioritise their following tasks and gain confidence to overcome obstacles that could arise in their original plan.’ No matter where you are in your professional journey, EAGE is committed to supporting your professional development and learning needs by providing the latest and high-quality learning opportunities.

Making education affordable EAGE strives to make online education accessible to everyone, particularly early career professionals, by offering courses at affordable prices. Those interested in taking multiple courses per year can save up to 50% on total registration fees with Education Packages, which are valid for one year. Long-term EAGE members who are between jobs also have the opportunity to apply for financial assistance through the EAGE Economic Hardship Programme.

EAGE Student Calendar 6 DEC

ASIA PACIFIC EAGE STUDENT CHAPTERS MEETING

ONLINE

31 DEC

DEADLINE TO UPLOAD STUDENT CHAPTERS’ ANNUAL REPORT

ONLINE

FOR MORE INFORMATION AND REGISTRATION PLEASE CHECK THE STUDENT SECTION AT WWW.EAGE.ORG.

FIRST

BREAK

VOLUME

DECEMBER

2023

EAGE NEWS

Prof Koenraad Johan Weber 1934-2023 Koenraad Johan Weber, recipient of the EAGE’s Alfred Wegener Award in 2001, died peacefully on 12 September 2023 at the age of 89. Koen Weber had a remarkable career as an engineer and geologist. He was widely known for his contributions that were key to development geology and reservoir geology being recognised as distinct technical disciplines in the oil and gas industry. His contributions are reflected by the honours he received from his peers in the industry. In addition to the Alfred Wegener Award from the EAGE, Weber received the Van Waterschoot van der Gracht Medal from the Royal Geological and Mining Society of the Netherlands, the Sidney Powers Memorial Award from the AAPG, and the NGMS/Shell Award from the Nigerian Mining and Geosciences Society. He also received an honorary doctorate from Heriot-Watt University in Edinburgh. Weber graduated in 1960 as a mining engineer at Delft University of Technology and joined Shell as a research engineer at Shell’s E&P laboratory in Rijswijk near The Hague. Shortly afterwards, he visited Iran to review the geological model and the production performance of the Gachsaran field. During this assignment he learnt two fundamental lessons that he applied throughout his career. Lesson 1: Never accept the data that you receive at face value. Always check how data are acquired and their accuracy. Lesson 2: When proposing a new method or technique that differs from ‘standard practice’, always clearly outline the expected benefits of your proposal. In 1968 Koen Weber transferred to Shell’s Nigerian operations. There he laid the basis for his main expertise: geologically realistic modelling of oil and gas reservoirs. A well-known paper from this period is: ‘Sedimentological aspects of oil fields in the Niger Delta’ [Geologie & Mijnbouw 50, 559-620]. In the late seventies,

FIRST

BREAK

VOLUME

during a second posting to Nigeria and in spite of increased managerial tasks, Weber still found time for fundamental scientific work. Together with his Nigerian colleague Edmund Daukoru, he wrote the much-cited ‘Petroleum geology of the Niger Delta’. Some 20 years later, Weber’s close collaboration with Daukoru paid off in an unexpected manner. At a symposium organised by the students of the Faculty of Petroleum and Mining Engineering at Delft University of Technology, the Secretary General of OPEC 2006/2007 – Edmund Daukoru – was the keynote speaker. Not bad for the student society in question. In 1985 Shell promoted Weber to senior consultant reservoir geology. At the same time, he was appointed as professor production geology at the Faculty of Petroleum & Mining Engineering at Delft University of Technology. As senior consultant he was excused from most management tasks and could concentrate on technical work and teaching. He combined his part-time professorial role with tours as distinguished lecturer for the AAPG, EAGE and the SPE. In addition, he was associate professor at École Nationale Supérieure du Pétrole et des Moteurs in Paris, external examiner at Heriot-Watt University (Edinburgh) and at Imperial College London. During this period, he published papers such as ‘Framework for constructing clastic reservoir simulation models’ and ‘The role of faults in hydrocarbon migration and trapping in Nigerian growth fault structures’. In 1993 Weber formally retired from Shell at the age of 59 (retirement age was 60, with service in tropical regions counting double). Retirement allowed him to dedicate more time to research and teaching at Delft University of Technology. In total he supervised some 70 MSc and/or PhD projects, and most of his students followed rewarding careers in the oil and gas industry. In 1999 Weber also

DECEMBER

2023

retired as professor at Delft University of Technology, but he did not rest on his laurels. As well as continuing with his geological research he resumed work on research hobbies from his youth: archaeology and history. He also became actively involved again in the ‘Mijnbouwkundige Vereeniging’, the student society at the Faculty of Petroleum & Mining Engineering at his alma mater. His ‘Puzzles for mining engineers’, were regularly published in the student magazine. Koen Weber’s last major public communication project was a documentary about the history of ‘Mijnbouwstraat 120 – Een markant gebouw’. Mijnbouwstraat 120 housed the Faculty of Mining Engineering / Petroleum Engineering for well over 100 years. Weber was a student in this building in the Fifties, returning there some 25 years later as a full professor. In the documentary [www. youtube.com/watch?v=EQCVQzLrqDY] Weber shares memories from his distinguished career. With Koen Weber’s passing we have lost a good friend and our mentor during the early years of our own careers in Royal Dutch / Shell. Contributed by J. Evert van de Graaff and Lucia van Geuns

EAGE NEWS

Exciting line-up of EAGE events in LATAM for 2024 We are excited to unveil an exciting array of events set to take place in Latin America in 2024. These conferences and workshops are designed to bring together leading experts and professionals in the geosciences and engineering fields, focusing on advancing knowledge, fostering collaboration, and promoting innovation. On 6-8 August 2024, Mexico City will take centre stage in the world of geoscientific exploration as EAGE organises its 3rd Conference on Near Surface & Mineral Exploration in Latin America, complemented by the 2nd Workshop on Water Footprint plus the 3rd Workshop on Geothermal Energy in Latin America. This great event promises to explore all the challenges and opportunities presented by the region’s unique geological characteristics. On 12-13 September, Cartagena, Colombia, will be the focal point of a significant event in the geosciences calendar as experts from around the world convene to explore the role of Artificial Intelligence (AI) in Full-Waveform Inversion.

This workshop aims to foster collaboration among experts to unlock the full potential of this advanced recovery technique. Once again, Mexico City will be a focal point on 24-25 October 2024, this time for the 3rd EAGE Workshop on Advanced Seismic Solutions in the Gulf of Mexico showcasing cutting-edge technologies and methodologies to address the unique challenges of this dynamic region. Then on 5-7 November in Trinidad & Tobago, the 1st EAGE Conference on Energy Opportunities in the Caribbean will be held promising to be an event that will hold groundbreaking discussions about new energy opportunities in the region. To cap off the second half of the year, we are thrilled to announce that Rio de Janeiro, Brazil will host the 1st EAGE/ SBGf conference on the roadmap towards low carbon emissions. This collaborative event will be a platform for sharing strategies and initiatives to reduce carbon emissions in the region’s energy sector.

UNLEASH YOUR DATA’S POTENTIALSM

Katalyst Data Management’s Geopost Virtual Data Room Katalyst’s VDR platform allows you to market your assets to a worldwide audience. With industry leading tools for promotion, visualization, and interpretation of your assets, Katalyst can support you with your asset divestiture.

The event will emphasise AI’s potential to revolutionise seismic data analysis. The energy-rich region of Rio de Janeiro, Brazil, will host the 4th EAGE Conference on Pre-Salt Reservoirs on 17-19 September enabling researchers and professionals to share their insights and experiences in exploring and exploiting these valuable reserves. Argentina’s capital, Buenos Aires, will be the hub of discussions on Enhanced Oil Recovery (EOR) on 3-4 October 2024.

We are sure these events will be an invaluable resource for professionals and researchers in the geosciences and engineering fields. With a diverse range of topics and locations across Latin America, we hope to foster knowledge exchange, technological innovation, and collaboration in these critical fields. Mark your calendars for these exciting gatherings which promise to shape the future of geosciences and engineering in the LATAM region. FIRST

BREAK

katalystdm.com ®

VOLUME

Katalyst Data Dec23.indd 1

DECEMBER

2023

09/11/2023 15:57

EAGE NEWS

Borehole geophysics workshop back and alive in Milan

WORKSHOP

REPORT

Technical Committee co-chairs Husain Nassir (Saudi Aramco) and Rafael Guerra (SLB) report on the 7th EAGE Borehole Geophysics Workshop held in Milan in September. Our workshop, entitled Borehole Geophysics ‘Bridging the gap between surface and reservoir’, provided a forum for lively discussion between operators, service companies, consultants and academics. As on previous occasions, the location was chosen to be easily accessible to an international audience, to have a strong local operator and geoscience research presence, and to have points of cultural interest for a good social programme. The Technical Committee, drawn from across the industry reflecting the diversity of the delegates, selected a technical programme of 29 oral and seven poster papers to be presented over three days, interspersed with four keynote presentations by invited speakers. There was an exhibition area for companies to showcase their latest technology and a social programme to provide networking opportunities. The workshop attracted around 60 delegates from 26 different organisations and feedback from delegates on the choice of venue, the Melia Hotel in Milan, was overwhelmingly positive. Keynotes The committee was honoured to receive the opening address from Davide Calcagni, Eni head G&G operations. He highlighted the role of DAS technology in improving the efficiency of Eni’s operations, including CCUS monitoring. This was followed by a keynote presentation from David Hill, CTO of Sintela, on distributed fibre optic sensing technology and its applications in the energy sector and beyond. On the second day, Eric Verschuur (Delft University Seismic Inversion & Imaging) promoted the integration of borehole and surface seismic, including in survey design and imaging. During the joint session, Ahmad Riza Ghazali (PETRONAS chief scientist) gave a high-level overview of seabed seismic and DAS 3DVSP technologies from a company perspective. 14

FIRST

BREAK

VOLUME

Shujaat Ali (SLB borehole seismic expert) reported on advances in cross-well seismic, including the use of anisotropic eFWI to obtain high quality images from a CCUS project in the USA, and showed how to overcome the challenges of DAS 3DVSP noise in producing wells. Technical programme The technical papers reflected the broad scope of borehole geophysics and were divided into eight sessions: Borehole data acquisition & greener operations, Conventional VSP applications and robust well ties, Advanced processing, anisotropy and inversion (two parts), Microseismic monitoring, bridging the scale gap of acoustic measurements, Time lapse VSP Monitoring, and a joint session between the Seabed Seismic and Borehole Geophysics workshops. At the end of each session, presenters were invited on stage for a panel discussion to further explore the studies presented. The authors of each poster submission described their work in a short introduction, before the delegates were invited to view the posters in the break-out area. The complete list of workshop abstracts is available at EarthDoc with some highlights mentioned here. Liborio (Eni) presented the geological interpretation of high-resolution DAS 3DVSPs recorded in four producing wells which supported the drilling of successful development wells. Alfataierge (Aramco) showed a high-density DAS walkaway VSP recorded in two deep wells with fibres strapped to tubing, with very good data quality. Lesnikov (TotalEnergies) obtained very good geophone walkaway results, including AVO and anisotropy estimation, from a project offshore Guyana. His colleague, H. Klemm, went a step further and inverted the walkaway gathers at the well location for Vp and Vs logs. Mizuno (SLB) discussed the importance of velocity model

DECEMBER

2023

calibration in microseismic monitoring in event detection (less of a problem for migration-based methods) and in event location. Rufino (SLB) showed a case study of high quality LWD seismic VSP imaging from offshore. Martinez (SLB) presented DAS offset VSPs recorded in very challenging completions with multi-mode fibres inside coiled tubing clamped to tubing in cased hole, the data quality was good and better than expected. Shashkin (Curtin University) showed the correct formula relating DAS strain amplitudes to rock density and velocity. His colleague, Pevzner, analysed the non-linear effects on the vibroseis signals recorded by downhole DAS. Soulas (ASL) showed comparisons between different DAS manufacturers and geophone tools, demonstrating that all DAS interrogators are different. Guerra (SLB) showed a DAS VSP recorded in minutes offshore UK using extra-strength hybrid logging cable technology. This allowed the cancellation of conventional geophones, saving hours of rig time and tonnes of CO2. Nassir (Aramco) demonstrated how to attenuate vibroseis harmonic cross-correlation artefacts using template matching. Belleza (OGS) presented an interesting DAS VSP project recorded in Türkiye in shallow but very hot wells (2500C), using engineered fibres and a miniature electric vibroseis source, but the results had relatively low SNR. During the joint session with the seabed workshop, Yu (BGP) presented a world record DAS 3DVSP survey recorded for ADNOC in 13 flowing wells offshore UAE, during OBN acquisition with two vessels shooting simultaneously, with deblending and some advanced processing applied. Haacke (CGG) obtained good DAS 3DVSP results in a survey recorded for BP in the Caspian Sea in a producing well with fibres clamped to tubing, but no increase in

EAGE NEWS

DAS resolution compared to OBN data was achieved due to the complex velocity model and high flow noise. Moore (CGG) presented a nice DAS 3DVSP test in one of Equinor’s Johan Sverdrup water injection wells with fibres clamped to tubing; 4D modelling was performed to assess repeatability and detectability of CO2 injection. Verschuur (Delft University) discussed the design of sparse geometry surveys, including the use of migrated image quality to automatically update source and receiver locations. Haldorsen (MagiQ) showed the audience how to locate acoustic sources in the sea using a multi-component array with DAS fibre and point sensors. Finally, Gupta (SLB) presented good S-DAS PrP and PrSv

results with dark fibres buried under the seafloor over a distance of 80 km. Short course Following the workshop, a one-day short course on distributed fibre-optic sensing technology was given by Dr David Hill (Sintela CTO). The course explained the key concepts of distributed fibre optics in DAS, DTS, DSS and was well received by the participants who were generally more familiar with seismic methods than with optics. There were many questions and clear explanations followed. Closing remarks The committee voted on the best technical contributions to the programme and presented the best paper prize to

Carolina Liborio, Eni geoscientist, for her thorough geological integration of DAS 3DVSP and well data. The resulting improved reservoir model has helped in the successful drilling of development wells. The best delegate paper was presented to Eric Verschuur for his constructive engagement and insights shared during the workshop. The proceedings were brought to a close by co-chairs Husain Nassir and Rafael Guerra, who thanked everyone involved in the success of this workshop and the event’s invaluable Platinum level sponsors: Saudi Aramco, BGP and Eni, as well as the sponsors: SLB, VSProwess and Avalon Sciences. Planning is already underway for the next workshop in 2025!

Join our latest marine acquisition technologies event Technical Committee chair Martin Widmaier extends an invitation to join the geophysical acquisition community in Oslo. The 4th EAGE Workshop on Marine Acquisition will take place once again in Oslo on the 2-4 September 2024. This popular workshop will be a forum for operators, contractors, manufacturers, and academia to discuss latest geophysical and technical developments and innovations as well as applications. We will take a closer look at recent advances in marine seismic equipment, operations, and survey design. Recently, the energy trilemma has forced the industry to review strategies and the corresponding technology demand. Challenges and change can drive innovation in the industry. The workshop will also cover energy transition related applications such as CCS and offshore wind. While marine seismic methods will again be the key focus of the workshop, contributions related to other relevant marine geophysical methods are very welcome. The workshop aims to provide a comprehensive overview on the latest advances in marine acquisition covering

Seismic survey over the Northern Endurance CCS project (photo courtesy of Northern Endurance Partnership and PGS).

seismic source and sensor technologies, novel acquisition geometries and survey design solutions, as well as operational aspects. We would like to share recent experiences and lessons learned from case studies and explore future visions. The workshop scope covers a wide range of applications from hydrocarbon explora-

FIRST

BREAK

tion and reservoir monitoring to energy transition-related topics such as high resolution near surface methods for offshore wind, CCS development surveys and monitoring, nuclear waste management as well as marine mineral exploration. Abstracts can be submitted until 31 March 2024. For more details visit the Calendar of Events at www.eage.org.

VOLUME

DECEMBER

2023

EAGE NEWS

A volcanic meeting in Prague Our Local Chapter Czech Republic chose to organise their autumn meeting on an explosive topic. Jeroným Lešner (Geotechnik CZ) was invited to lead a lecture on volcanism including his recent expedition to the Litli-Hrútur volcano, Iceland.

The talk was about eruptions.

The participants enjoyed wonderful images, stories and videos of the spectacular eruption of Litli-Hrútur (312 m), which is part of the Fagradalsfjall, an area that has already seen three eruptions since 2021. Lešner described the work and the experiments carried out with the flowing lava on-site. We saw unique videos capturing little air twisters forming over the lava flows and discussion on formation of Pele’s hairs and tears. We tried holding the long shovel designed to extract fresh lava samples, saw how rock transforms when exposed to fresh lava and were instructed on how to

breathe through air filters to keep safe from volcanic fumes. The talk attracted the attention of scientists from the Geophysical Institute which is mapping seismic activity in this region of Iceland through their own seismic network as well. Originally scheduled for one hour, the discussion extended well beyond that and provided an excellent opportunity for networking. Local Chapter Czech Republic is also the organiser of the Vlastislav Červený Student Prize for the best Master and Bachelor thesis in applied geophysics, which will be awarded later in the year.

LC Stavanger celebrates first in-person meeting On 2 November, EAGE Local Chapter Stavanger celebrated a Grand Opening – its first in-person gathering! LC Stavanger was established two years ago by Surender Manral and a group of local volunteers, amid the Covid-19 pandemic. From its inception, the Chapter has operated primarily online, driven by the mission to promote geoscience for the next generation, facilitate the exchange of knowledge, and champion diversity within the local geoscience community. The latest event marked a pivotal shift, representing the Chapter’s first endeavour into in-person gatherings and a promising new phase in the Chapter’s journey, solidifying its vision for the future. The event was held at the iconic Norsk Oljemuseum in Stavanger, an apt choice allowing members to mingle and network amidst the captivating artifacts that illustrate the evolution of oil and gas offshore operations as Norway’s most important industry. The event was attended by around 60 enthusiastic geoscientists who came together to make it a resounding success. One of the

FIRST

BREAK

VOLUME

evening’s highlights was the outstanding and diverse presentations delivered by keynote speakers Carlos Alonso Gil Figueroa and Kjetil Westeng, igniting compelling conversations about the future of industry, whether in terms of energy transition or digitalization. Gil, vice-president of subsurface and area development at Equinor, discussed the critical role that subsurface will play in energy transition, with geoscientists positioned at the core of the transformation of the Norwegian Continental Shelf from an oil and gas province into a broad energy province. He illustrated this through real case examples that Equinor has undertaken. Kjetil Westeng, advanced petrophysicist at Aker BP Norway, presented his award-winning project on next-gen subsurface solutions. The project was awarded the Exploration Innovation award at the NCS Exploration conference held in Oslo earlier this year. He demonstrated how Aker BP has ventured into a journey towards the integration of automated solutions and digitization in subsurface interpretation.

DECEMBER

2023

The success of the event was the result of tireless efforts and meticulous planning by the board members and volunteers of LC Stavanger including Anastasiia Jacobsen, Alexandra Zaputlyaeva, Victor Aarre Madsen, Maria Josefina Cuello, Khushboo Havelia, Olesya Zimina, Cristiano Camatel and Hilde Grude Borgos.

Photo op at Stavanger meeting.

The feedback from attendees resonated with a shared sense of satisfaction and anticipation. They were pleased with the Chapter’s transition to in-person gatherings and voiced their expectation for it to continue providing an avenue for the community to convene and build meaningful connections. To stay updated, you can follow the EAGE LC Stavanger on LinkedIn.

EAGE NEWS

CONFERENCE

REPORT

IMOG 2023 maintains record for new science initiatives

The 31st International Meeting of Organic Geochemistry (IMOG) took place in Montpellier, France from 9 September to 15 September 2023. Convened by the European Association of Organic Geochemistry (EAOG) and EAGE, the meeting again featured the latest science of organic geochemistry for environment, paleoclimate, and petroleum industry applications. EAOG consists of over 500 members and is one of the few communities whose participants cross broad applications using geochemical tools. Knowledge sharing and community engagement are both intrinsic to the value of IMOG which always showcases new discoveries and advances.

Posters were popular.

At the 2023 conference, a new analytical approach emerged as a significant theme in the scientifical programme. The orbitrap mass spectrometer provides intramolecular stable isotope analysis for a range of compounds. Researchers are using this instrument to advance understanding on the formation and fate of important molecules. Mass spectrometer imaging, another emerging technique, was featured in the opening presentation of the Congress, by the Geoff Eglinton Lecture awardee Janina Grongina from the University of Bremen. Industry scientists presented work featuring the ways in which energy companies use organic geochemistry to understand the origin and production of naturally occurring petroleum, with study areas from Mozambique, the Middle East, China, and throughout North and South America. For example, researchers showed new approaches to the analysis and interpretation of diamondoids, as well as new tools that provide geochemical interpretation early to guide drilling decisions. Company scientists also presented the value organic geochemistry brings to the energy transition topics such as carbon capture and storage, hydrogen, and reducing stray gas emissions.

Impressive turnout for event.

FIRST

BREAK

Mingling at Montpelier venue.

Doctoral students are always important participants in the scientific programme. IMOG 2023 featured 57 student presenters, including plenary and session talks and posters. The IMOG Scientific and EAOG Awards committees collectively judged and selected the winners of the best student presentations awards based on scientific and presentation quality. The winners were Katrin Haettig, Luke Brosnan, Ilya Kutuzov, and Alice Fradet. Important awards were also conferred on both early career and senior scientists. Gordon Inglis from the University of Bristol received the EAOG’s Pieter Schenck Award. Stuart Wakeham received the prestigious career Treibs Medal, awarded by the Geochemical Society, for his lifetime contributions to understanding the carbon cycle, especially in anoxic basins. During the general assembly, EAOG elected new Board members, and paid tribute to the exiting members. As Courtney Turich takes on the chairperson role, she notes that the EAOG Board has reached historic gender parity for the first time in its 60-year history. IMOG 2025 planning is already in full swing, and we look forward to welcoming scientific content and enthusiastic participation in Porto, Portugal in September 2025. Organic geochemistry is important in many different aspects of our science and energy transition. IMOG will remain the conference where cutting-edge research in organic geochemistry is introduced and further developed. I

VOLUME

DECEMBER

2023

EAGE NEWS

EAGE Student Fund needs your support in 2024 In 2023, the EAGE Student Fund played a pivotal role in empowering the next generation of geoscientists and engineers to foster innovation in our industry. Thanks to the generosity of you, EAGE members, we could support EAGE Student Chapters from around the world and promote the learning of geosciences among new scholars.

The EAGE Student Fund supported the Laurie Dake Challenge finalist teams to attend the Annual Conference in Vienna. In the photo: EAGE Student Chapter Universidad Nacional de Colombia, Bogotá.

In June, for example, the Student Chapters from the University of Miskolc (Hungary), Universidad Nacional de Colombia (UNAL Bogotá), and the Federal University of Bahia (Brazil) were able to participate in the 84th EAGE Annual Conference and Exhibition in Vienna thanks to the support of the Fund. The EAGE Student Chapter Universidad Nacional de Colombia stated: ‘We had a wonderful experience in Vienna, where we participated in the Laurie Dake Challenge, the EAGE Annual, and the Global Geoquiz. We worked on a realworld problem related to geothermal ener-

gy and presented our solution to a panel of experts. We also networked with professionals and researchers from different countries and disciplines and learned about the latest developments and innovations in geosciences. We are very grateful to the EAGE Student Fund for supporting us and giving us this unforgettable experience.’ The EAGE Student Fund also recognised and supported the exceptional performance of the University of Strasbourg, at the Global GeoQuiz 2023, and those of The University of Manchester and University M’hamed Bougara Boumerdes in their winning of the 2023 editions of the

Laurie Dake Challenge and Minus CO2 Challenge respectively. In 2023, the Student Fund provided EAGE Student Chapters with a platform for networking, knowledge exchange, and collaborative activities under the EAGE umbrella. In this regard, the Fund donated 15 student memberships for each of our 38 EAGE Student Chapters, provided 300 complimentary memberships for first time student members, and financially supported early careers dedicated events. Marie Gärtner, Jan-Phillip Föst, and Rune Helk told us: ‘Thanks to the generous support of the EAGE Student Fund, we successfully hosted the Geophysikalische Aktionsprogramm (GAP) in Karlsruhe, Germany from 11-14 May 2023. This year marked a significant turning point for GAP, as we welcomed 75 participants and 30 volunteers. Our participants came from 11 different universities, and we organised various activities, including excursions, talks, and networking opportunities. We owe a special thanks to our sponsors, including the EAGE Student Fund, for making this event possible.’ To all members, please continue to provide students with your support enabling a lasting impact in 2024 in nurturing the next generation of geoscientists and engineers. Donations can be made while renewing your 2024 membership, or via donate.eagestudentfund.org.

The EAGE Student Fund supports student activities that help students bridge the gap between university and professional environments. This is only possible with the support from the EAGE community. If you want to support the next generation of geoscientists and engineers, go to donate.eagestudentfund.org or simply scan the QR code. Many thanks for your donation in advance!

D O N AT E T O DAY ! 18

FIRST

BREAK

VOLUME

DECEMBER

2023

PERSONAL RECORD INTERVIEW

Daniella Bordon

Personal Record Interview

Environmentalist in the oil business With a few twists and turns Daniella Bordon has pursued her early fascination with the marine environment in a career that has now evolved into global ESG manager, BGP Offshore based in Rio de Janeiro. Much of her work in the last 16 years has involved seabed seismic technology companies working offshore Brazil including early adopters such as RXT, Oceangeo and ION Geophysical.

All started with biology My parents came from low-income families in Brazil and worked very hard to put me and my siblings into good schools. I have always been a good student and in the first year of high school I had a list of possible careers in the back of my notebook. I loved science and my mother wanted me to go to med school. But, influenced by a great biology teacher who lived in Australia and worked with coral reefs, I decided to become a marine biologist. Many adventures came with this choice - living in amazing places doing internships and working with marine mammals. Learning English in US On my last day at university, I was given a brochure on my way out the gate for a work abroad programme. I did not speak English back then and it sounded like the next important step in my career plan. My parents did not approve of my choice. Anyway I sold my car and paid for the programme. From my first day of work at a hotel breakfast buffet, I was forced to learn English quickly (often making a fool of myself in the process). After three months in cold Pennsylvania, I went to Hawaii to fulfil one more dream. What was meant to be a three-week trip became the next chapter of my life. I got a job providing spray on tattoos, then selling cookies at a swap meet and later at a store in the mall. My main goal was to work at the Sea Life Park aquarium, but when I did get a job there I actually felt bad for the captive animals. I wanted

to be back at sea working with wild animals, and so that’s what I did. Experience offshore After two years abroad I returned home but planning grad school back in the USA. However, at the time, there were lots of geophysical surveys happening in Brazil, and I was offered work as a marine mammal observer. It proved an incredible experience working with a multi-national crew, learning new things and getting five weeks off to travel some more which is a passion of mine. First seabed seismic job A company called RXT was coming to Brazil to work on a two-year project doing ocean bottom cable surveys for Petrobras, technology then in its infancy. They needed someone to manage the environmental aspects of the survey and liaise with local authorities. I was referred to them by a friend and later interviewed at the international airport by my future boss. Career progress I embarked on several courses from certificates in environmental management to a master’s degree in development practices while working with various companies in the seabed seismic business. Luckily I still work with many of the same amazing people from when I started, but these days focused on environment, social and governance (ESG), my real passion. My responsibilities range from managing risks at project level to developing BGP Offshore’s enviFIRST

BREAK

ronmental and social strategies at corporate level. I have visited 40 countries and seen many interesting cultures, living the challenge and delights of being a woman in a male-dominated industry. Community projects My company started implementing at least one community outreach programme per project. For example we engaged with a local foundation in Sao Tome and Principe to support their school with supplies and their local orchestra with uniforms. In Suriname we donated wheelchairs and industry-recognised courses with our partners TGS and CGG. Now we are launching our first small grants programme focused on the recovery of our oceans. On the dark side It did take me almost 10 years to feel good about what I do. As a biologist, I felt like I was working for the dark side! What changed was understanding how energy access is essential to pull people out of extreme poverty and feeling like I was able to make a real change. Today, I believe my company is very much onboard with the idea of leaving a positive legacy for the environment and people. Away from work When not working I try and be on the beach and nature as much as possible with my children, and twice a week, religiously (when not travelling on business), I play beach volleyball … what a cliché for a Brazilian! I

VOLUME

DECEMBER

2023

Make sure you’re in the know

EAGE MONTHLY UPDATE Show how your research can significantly impact the advancement of geoscience and engineering knowledge!

S UBMI T Y OUR A B S T R A C T S B Y 15 J A NU A R Y 2 0 24 EAGEANNUAL.ORG

Geoenergy

RENEW YOUR EAGE

MEMBERSHIP FOR 2024

EARTHDOC

• Digitally enabled geoscience workflows: unlocking the power of our data • Sustainable geological disposal and containment of radioactive waste • The energy-critical metals for a low carbon transition

Submission deadline: 31 December 2023 to ie s wa y rt u n it g a te ppo Yo u r le d o ra ll e unpa

R E N E W AT E AG E . O R G / W E L C O M E

CONTRIBUTE TO THE TECHNICAL PROGRAMME! Submission deadline: 22 January 2024

FIRST

BREAK

VOLUME

Submit your abstracts

by 1 February 2024!

W W W. E AG E G E OT E C H .O R G

2-5 SEP TEMBER 2024 O S L O , N O R WAY

DECEMBER

2023

10 December 2023 Fourth EAGE Digitalization Conference & Exhibition

U P CO M I N G D E AD LI N E S

3 Special Issues – Call for Papers FIND OUT MORE ON

Abstract Submissions

14 December 2023 Fourth EAGE/AAPG Hydrocarbon Seals Workshop 31 December 2023 6th Asia Pacific Meeting on Near Surface Geoscience and Engineering

Registration 30 December 2023 Regular fee EAGE/AAPG Workshop on New Discoveries in Mature Basins 10 January 2024 Early Bird fee Fourth EAGE Digitalization Conference & Exhibition 19 January 2024 Early Bird fee EAGE Sub-Saharan Africa Energy Forum 26 January 2024 Early Bird fee First EAGE Data Processing Workshop

Our innovative airborne geophysical technologies allow for efficient exploration reducing environmental impact

X Abu Dhabi • X Astana • X Bogota • X Geneva • X Houston • X Hyderabad X Kinshasa • X Luxembourg • X Madrid • X Ottawa • X Perth • X Pretoria • X Rio de Janeiro • X Riyadh • X Toronto

CROSSTALK BY AN D R E W M c BAR N E T

•

Unsettling times for science In his 2015 State of the Union address President Obama was checked out online a doctor’s diagnosis, indeed who hasn’t used perhaps rash to state unequivocally that the ‘debate was settled. the internet for a bit of self-diagnosis of some ailment or other. Climate change is a fact …’ It provided deniers an immediate The point is that we can all become instant experts, and opening to counter that science is never settled, which of course instant critics in the science sphere. The evolution of social media is indisputable. Scientific conclusions can only be based on the has magnified the trend enabling everyone to express their views with no filter on their validity. Hence we have a much-discussed available evidence, balance of probabilities, etc and are most convincing when mainly or overwhelmingly pointing in one information crisis with everyone challenged to know whom to direction. trust. Traditional media sources based on at least some semblance At issue is what determines when we as a society can trust of journalistic research and verification of the facts are in serious science to offer an acceptable certainty upon which we can base decline as a guide. This is not just because of the sheer volume decisions. For example, apart from some weird minority sects, of competing news and opinion feeds out there – Twitter (X), we all go along with the research and practical experience that emails, Instagram, Facebook and so on. Rapid dissemination of shows anti-polio vaccinations are a worthwhile protection, but huge amounts of material also takes its toll on a clear narrative. Covid-19 not so much. A significant swathe of people in many Most significantly this loose structure is open to obvious countries proved unpersuaded by the medical evidence. The manipulation by vested interests, e.g., perpetrating fake news, flooding media with false/bias opinions, distorting the consensus, current parliamentary inquiry into the UK government’s handling of the Covid crisis under Boris Johnson has revealed a chaotic etc. We may often suspect governments and business interests are response. Medical counsel was weighed against alternative culpable, but often it requires a whistleblower to reveal the extent of involvement. explanations of the virus as well as political and other self-serving considerations. In the US, the president of the time at one point Needless to say, geoscience is not immune from this world advocated swallowing disinfectant as an antidote to Covid. of disinformation. Objections to recent proposals by Danish Such goings-on illustrate the vulnerability of science, and company Ørsted to install two offshore wind turbine developthe trust we are prepared to place in it. Overcoming doubt ments off the coast of New Jersey because of potential danger about scientific hypotheses through research, to whales provide a classic example of how experiments, tests, pilot studies, etc are all part ‘The process is literally science can be questioned in spite of seemof building scientific knowledge. In practice ingly incontrovertible evidence. According to the process is literally not a perfect science. not a perfect science’ The Guardian newspaper, ex-president Trump Findings are challenged, mistakes are made. at a 2024 presidential campaign rally in South In the medical field, the dispensing of Thalidomide medicaCarolina on 24 September, had clearly judged the issue. He tion to pregnant women to counter nausea caused terrible birth attributed a wave of recent whale strandings on the US East Coast defects. It took a major newspaper campaign, among other things, to ‘windmills’ driving the mammals ‘crazy’ and ‘a little batty’. He to expose the tragedy. That was in the pre-internet days of the noted ‘You wouldn’t see that once a year – now they are coming Sixties and Seventies. up on a weekly basis’. Today public discussion of science topics has changed draTrump was correct in his observation that stranding of matically. The internet makes an incredible amount of information whales has become a more common phenomenon in recent on just about every subject you choose instantly accessible, from years. He would have known that last December, 30 New Jersey the authoritative to the bogus. Sticking with medicine, who hasn’t coastal mayors called for a moratorium on offshore wind activity,

FIRST

BREAK

VOLUME

DECEMBER

2023

CROSSTALK

prompted by the ninth dead whale to be washed ashore in the sound’ and expressed her fears for the area’s fishing business state that month. and those who depend on it. Carlson noted ‘Whales use sound to navigate. So why wouldn’t that be affecting or killing them? For those familiar with the marine seismic industry’s challenges to comply with increasingly strict mammal protection You’re applying the principles of science,’ adding that offshore regulations to enable surveys in offshore regions around the world, wind was the ‘DDT of our times’ referring to the toxic pesticide that was banned in 1972. this is well travelled territory. It requires minimal reference to know that the US Bureau of Ocean Energy Management and predecessor As a spokesperson for the fisheries business, Lapp has since authorities have played an important role in establishing the repeated the gist of her argument at a packed public hearing in Wildwood, New Jersey, in March and in various subsequent interguidelines for seismic vessels working in US territorial waters, views, also echoing concerns over the practicality of fishing around particularly in the Gulf of Mexico. wind turbines. Locals have also complained that the turbines will In the New Jersey case, the National Oceanic and Atmospheric spoil views from the New Jersey coast. Administration (NOAA) made clear in a statement nearly a year It is no surprise that the controversy has been overtaken by ago that any surveys for offshore wind companies have been politics. Local Republican Party representatives and organisations focused on collecting data for future projects and ‘produce much such as the Texas Public Policy Foundation (TPPF), said by opposmaller impact zones (compared with seismic) because in general nents to be fossil-fuel friendly, have weighed in with support. They they have low noise, higher frequency and narrow beam-width’. have spotted an opportunity to attack President Biden’s renewable A noted marine environmental researcher Dr Douglas Nowacek energy policy which envisages substantial of Duke University, currently engaged in a expansion of windpower, rolling out 30GW of five-year study on offshore wind energy and ‘An opportunity to offshore wind capacity by 2030. its impact on wildlife, was quoted by Factchek. attack President The emotional charge of whales dying on org as stating that ‘there’s basically zero chance the beach has turned out to be an achilles heel that those surveys have caused any mortality.’ Biden’s renewable for those who want to stick to the rational, According to the NOAA, explanation for energy policy’ science-based explanation. It is impossible to the death or stranding of an increasing number adequately deny the coincidence of marine of different species of whales (humpbacks, surveys and the increased strandings, and it is pretty easy to drum North American right whales and minke whales) can most probup public sympathy. ably be put down to vessel strikes and entanglement with fishing Coincidence or not, Ørsted, the biggest company in the gear. The conclusion was based on the evidence of some 40% business worldwide, last month cancelled its plan for the two of carcasses it was possible to examine. This fits with a number big Ocean Wind I and II developments off the New Jersey coast of factors that may be in play, some predating the relatively few incurring an estimated impairment cost of more than $3 billion. offshore turbines so far installed off the US east coast. High inflation, rising interest rates and supply chain bottlenecks First, climate change is said to be warming the oceans and were given as reasons for the cancellation. The company is now altering the distribution of available nutrition in the ocean. Whales faced with sorting out generous state subsidy arrangements agreed are apparently moving out of protected areas and closer to shores with New Jersey Democrat Governor Phil Murphy, who had been which make them more vulnerable to vessel strikes and fishing a strong advocate of the projects and the opportunity for renewable gear entanglement. Secondly, New York and New Jersey ports energy investment. For the record, Ørsted last month also quit have reported significant growth in shipping activity meaning more a consortium with Fred Olsen Renewables and Hafslund Eco vessels and potential collisions with whales. Thirdly, the whale developing windpower offshore Norway, citing ‘a prioritization of population in the case of humpbacks in the mid-Atlantic has grown, investments’ in its portfolio. thereby increasingly the likelihood of mishaps. It is tempting to dismiss the New Jersey fracas and turning With such a clear rebuttal of the claims made against offshore science evidence into a political football as a peculiarly American turbine-related surveys available, the Trump allegations need some phenomenon born out of hopelessly polarised politics. Yet a similar context from which, depending on your appetite for conspiratorial debate over the impact of seismic on marine life is underway explanations, conclusions can be drawn. However, let’s not forget in southwest Victoria, Australia. A local environmental group is that the whole episode points to a serious disregard for scientific protesting against a ConocoPhillips survey plan in the Otway findings which seemingly goes unbridled. Basin, currently at the consultancy stage, based on the noise factor, Trump may well have taken his cue from a Fox News despite decades of regulated seismic surveys having been carried interview earlier this year between the now dismissed news comout offshore Australia. mentator Tucker Carlson and Meghan Lapp, fisheries liaison for The problem is that the science is never totally settled, and that Seafreeze, the largest producer and trader of sea-frozen seafood applies to the much bigger issues such as climate change mitigation on the US East Coast. Lapp referred to offshore wind developers and protection against pandemics. as ‘essentially carpet bombing the ocean floor with intense Views expressed in Crosstalk are solely those of the author, who can be contacted at andrew@andrewmcbarnet.com.

FIRST

BREAK

VOLUME

DECEMBER

2023

HIGHLIGHTS

INDUSTRY NEWS

PGS shoots CCS survey offshore Norway

Renewables meeting half of new energy demand

TGS wins GoM OBN contract

UK offers 27 oil and gas licences in the North Sea The UK has offered 27 licences in the North Sea in areas prioritised because they have the potential to go into production more quickly than others. Shell, Total Energies, Equinor, Eni, Anasuria Hibiscus, Athena Exploration, Bridge Exploration, DNO, Ithaca Energy, Neo Energy, Ping Petroleum, Tailwind Energy and Tangram Energy are among the 14 operators to have won contracts. Norwegian operator DNO which, with JV partner Aker BP, landed blocks 9/9f, 9/10c, 9/14c and 9/15d, which are contiguous to the Norwegian maritime boundary and just west of the Aker BP-operated Alvheim hub offshore Norway. The offshore UK area also comprises the Agar discovery from 2018, in which DNO held a 25% interest until it was relinquished in 2020. DNO said it will acquire additional 3D seismic data and potentially reprocess the data to reduce risk and volume uncertainty. Meanwhile, the UK government has introduced the Offshore Petroleum Licensing Bill to require annual oil and gas licensing rounds subject to stringent new emissions and imports tests. In addition, six more blocks, which were also ready to be offered, have been merged into five existing licences. The 33rd Oil and Gas Licensing Round was launched on 7 October 2022 with 931 blocks and part-blocks made available for application. In total, the UK North Sea Transition Authority (NSTA) received 115

applications from 76 companies for 258 blocks/part-blocks when the application window closed on 12 January 2023. This was the highest participation since the introduction of the Innovate Licences in 29th Round in 2016/17. There are currently 284 offshore fields in production in the UK North Sea and an estimated 5.25bn boe in total projected production to 2050, said the UK Transition Authority. Oil and gas currently contribute around three quarters of domestic energy needs and official forecasts show that, during the energy, they will continue to play a role in our energy mix for decades to come, it added. Stuart Payne, NSTA chief executive, said: ‘Ensuring that the UK has broad options for energy security is at the heart of our work and these licences were awarded in the expectation that the licensees will get down to work immediately. The NSTA will work with the licensees to make sure that where production can be achieved it happens as quickly as possible. A recommendation for the remaining 203 blocks will be made once the Habitat Regulation Assessment Further Appropriate Assessment process has been completed. Companies winning a share of blocks include Aker BP, Triangle Energy, BP and Dana Petroleum. ‘The introduction of regular licensing for exploration will increase certainty, investor confidence and make the UK FIRST

BREAK

Oil platform in the North Sea.

more energy independent,’ said the NSTA in a statement. ‘This new regime will be subject to two key tests being met: that the UK is projected to remain a net importer of both oil and gas; and that the carbon emissions associated with the production of UK gas must be lower than the average of equivalent emissions from imported liquefied natural gas.’ Supporting continued production in the UK will also reduce reliance on higher-emission imports – with domestic gas production having around one-quarter of the carbon footprint of imported liquefied natural gas, said the NSTA. Secretary of state for energy security and net zero Claire Coutinho said: ‘As energy markets become more unstable it’s just common sense to make the most of our own homegrown advantages and use the oil, gas, wind and hydrogen on our doorstep in the North Sea. Rather than importing dirtier fuels from abroad, we want to give industry the certainty to invest in jobs here and unlock billions of pounds for our own transition to clean energy. I

VOLUME

DECEMBER

2023

INDUSTRY NEWS

CGG Joins Phase 3 of Northern Lights CCS seismic monitoring project CGG has joined Phase 3 of a project to monitor background seismicity at the Northern Lights CO2 storage project offshore Norway, the world’s first large-scale open-source infrastructure for receiving and storing CO2 from multiple sources and industries. As a partner in the HNET Horda Platform Region Project, CGG has made available the 3D velocity model from its 44,000 km2 Northern Viking Graben multi-client seismic data set. Phases 1 and 2 of HNET investigated onshore and offshore seismic monitoring instrumentation solutions, by implementing a

new onshore array (HNAR) on Holsnøy island north of Bergen and assessing the integration of offshore nodes which form part of existing permanent reservoir monitoring (PRM) sensor arrays in the area. Phase 3, lasting three years, aims to understand natural seismicity in the planned CO2 injection site. With a background seismicity data set the operator will be able to assess the nature of tectonic seismic activity prior to CO2 injection underground and more accurately assess any induced seismicity during the injection period. For this, the HNAR

Holnsøy island where a monitoring array has been put in place. (image courtesy of CGG).

onshore monitoring array and a selection of offshore PRM nodes will continue to be used to refine and improve detection accuracy.

PGS shoots 3D survey for CCS scheme offshore Norway PGS has completed a 3D seismic survey over the Poseidon CCS licence area (licence EXL005) in the Norwegian North Sea for clients Aker BP and OMV. The 500 km2 survey, 100 km off the Norwegian coast, was carried out by the vessel Ramform Atlas. The campaign aimed to generate high-resolution imaging of the CO2 storage complex and to provide a baseline for monitoring of the storage integrity. In March 2023, Aker BP ASA and OMV (Norge) AS were awarded the Poseidon licence (licence number EXL005) in accordance with the CO2 Storage Regulations on the Norwegian

PGS Ramform Titan.

FIRST

BREAK

VOLUME

Continental Shelf. The licence has a work programme which includes a 3D seismic acquisition and related studies, followed by a drill-or-drop decision by May 2025 at the latest. Poseidon is a 50/50 partnership, operated by Aker BP. The Poseidon 3D seismic survey was safely executed within schedule and budget, said Aker BP and OMV. Meanwhile, PGS has won a 3D exploration contract in the Mediterranean. The vessel Ramform Titan is scheduled to mobilise for the survey this month and the contract has a total duration of approx. 160 days. ‘The Mediterranean is a hotspot serving the European gas market and we see potential for more work in this prolific region,’ said Rune Olav Pedersen PGS, president and CEO. PGS, in partnership with SNPC, has also completed a fourth phase of reprocessing of the Congo Vision project, adding more than 3900 km2 rejuvenated 3D seismic data to create a regional dataset, and covering open blocks Marine XIX and Marine XXX. The company has increased regional coverage of its Congo

DECEMBER

2023

Vision dataset by targeting deepwater open acreage. The latest phase expands Congo Vision to the south, along strike from significant Angolan fields in the core of the Congo Fan depocentre. Congo Vision now comprises a contiguous volume of nearly 13,000 km2 broadband processed KPSDM data across the shelf and deep water, offshore Congo to create a seamless 3D seismic dataset that reveals prospectivity in both pre- and post-salt stratigraphy. Acreage included in the fourth phase of Congo Vision has previously yielded Upper to Lower Miocene discoveries, and the rejuvenated data images inboard Albian and pre-salt fields. Congo Vision provides the means to further evaluate exploration opportunities with confidence, at both play and prospect level. Finally, PGS has won a 4D acquisition contract offshore West Africa. Mobilisation is scheduled for Q4 and the contract has a total duration of approx. 60 days. ‘We experience continued high activity in our core West African markets,’ said PGS president and CEO, Rune Olav Pedersen.

INDUSTRY NEWS

iDROP signs deal to provide ocean bottom node system iDROP has signed an agreement with ExxonMobil, Hess Corporation and Woodside Energy Technologies to develop the Oceanid system – an autonomous OBN (ocean bottom node). The system comprises a self-navigating drone – Oceanid, that by gravity and ballast shift propels itself autonomously to a pre-plot location on the seabed. A complementary and automated handling system is under development for unmanned deployment, and subsequent recovery of nodes from the ocean surface.

‘The agreement involves both technical and financial support, opening up exciting field test options and is a commercial breakthrough for iDROP and its Oceanid system,’ said Kyrre J. Tjøm chief executive officer and founder. The Oceanid development project is funded by the European Union, the Eureka global R&D and innovation funding network, Innovation Norway and the Norwegian Research Council, with industry support from several European major energy companies.

Land Seismic Noise Specialists Our Full-Wave Correction (FWCTM) Technology Can Address Surface Scattering and Improve Your Challenging Seismic Data

Realtimeseismic processes vintage UK data for geothermal project Realtimeseismic has agreed a multi-year collaboration with Star Energy Group to carry out seismic processing covering more than 3000 km of multi-vintage 2D seismic data, acquired over UK urban areas from 1970s up to the present day. The processing campaign is split into five discrete projects, each supporting a hydro-geothermal development from the Star portfolio. ‘Realtimeseismic, with its pro-

prietary near-surface, denoising and velocity-modelling solutions, is ideally positioned to address all the data challenges and project objectives,’ said Realtimeseismic in a statement. Star’s processing campaign for geothermal applications is the 75th seismic processing commitment undertaken by Realtimesseismic in the ‘New Energy’ sector, and the 24th onshore UK over the past five years.

TGS and PGS merger moves closer TGS and PGS’ respective boards of directors have unanimously approved a definitive merger agreement in line with the terms previously announced. The merger is to be structured as a statutory, triangular merger between TGS NewCo AS, a newly established wholly owned subsidiary of TGS designated for such purpose, TGS and PGS in accordance with Chapter 13 of the Norwegian Companies Act. TGS NewCo will be the surviving entity and PGS shareholders will receive 0.06829 ordinary shares of TGS for each PGS share held. FIRST

BREAK

After the completion of the merger, TGS shareholders will own around a two thirds of the company and PGS shareholders will own the remaining third. Completion of the merger is subject to approval by extraordinary general meetings at TGS and PGS, expected to be called for shortly and held within one month thereafter. The merger is expected to generate synergies worth $100 million, including up to $70 million in operating costs (above a previous indication of $50 million). Vessel utilisation will be 2-3% higher (a worth $15-20 million). I

VOLUME

DECEMBER

2023

N��

INDUSTRY NEWS

Renewables are meeting more than half of new global energy demand, says DNV outlook Over the last five years fossil fuels have met only half of the new demand for energy globally, according to DNV’s Energy Transition Outlook. The report found that between 2017 and 2022 renewables met 51% of new energy demand, while the remaining demand was supplied by fossil fuels. Renewables are still just meeting increased demand rather than replacing fossil fuels and in absolute terms fossil fuel supply is still growing, said DNV. To reach the goals of the Paris Agreement, CO2 emissions would need to halve by 2030, but DNV forecasts that this will not even happen by 2050. CO2 emissions will be only 4% lower than today in 2030 and 46% lower by mid-century. Energy-related CO2 emissions are still hitting record highs and are only likely to peak in 2024, which is effectively the point at which the global energy transition begins. ‘Globally, the energy transition has not started, if, by transition, we mean that clean energy replaces fossil energy in absolute terms,’ said Remi Eriksen, Group president and CEO of DNV. ‘Clearly, the energy transition has begun at a sector, national, and community level, but globally, record emissions from fossil energy are on course to move even higher next year.’ Energy security has strengthened as a driver of energy policy due to changes in the geopolitical landscape. Governments are willing to pay a premium for locally

Energy Transition timeline shows how renewables are ramping up.

sourced energy, which has had a notable impact on the Outlook’s results. For example, the Indian Subcontinent is now forecast to transition slower with more coal in the energy mix. In Europe the transition is accelerating with the alignment of climate, industrial and energy security objectives. Even if the transition is yet to get out of the starting blocks, once it starts renewables will outsprint fossil fuels. From now, most energy additions are wind and solar, which grow 9-fold and 13-fold respectively between 2022 and 2050. Electricity production will more than double between now and 2050, bringing efficiencies to the energy system. The fossil to non-fossil split of the energy mix is currently 80/20 but this

will move to a 48/52 split by mid-century. Solar installations reached a record 250 GW in 2022. Wind power will deliver 7% of global grid-connected electricity and installed capacity will double by 2030, despite inflationary and supply chain headwinds. ‘There are short term set-backs due to increasing interest rates, supply chain challenges, and energy trade shifts due to the war in Ukraine, but the long-term trend for the energy transition remains clear: the world energy system will move from an energy mix that is 80% fossil fuel-based to one that is about 50% non-fossil fuelbased in the space of a single generation. This is fast, but not fast enough to meet the Paris goals,’ said Eriksen.

CGG signs deal with AI company for HPC solutions CGG and AI company LightOn have joined forces under a new contract to leverage CGG’s industrial high-performance computing (HPC) solutions. This will enable LightOn to optimally evaluate and test Large Language Models (LLMs) to support the industrial deployment of AI.

FIRST

BREAK

VOLUME

LightOn can now use CGG’s HPC and AI Centre of Excellence to benchmark their LLMs, exploring innovative hardware configurations hosted by CGG to identify the most optimal solution. ‘LightOn is at the forefront of innovation in AI, creating from scratch and delivering no less than 12 LLMs,’ said Laurent

DECEMBER

2023

Daudet, Co-CEO, LightOn. ‘We’re on a mission to transform business productivity for our clients and revolutionise how companies perceive and use generative AI. Having access to CGG’s latest HPC technology will accelerate our evaluation and testing, and ultimately our overall business strategy.’

INDUSTRY NEWS

PGS gets $30 million bonanza after wining fee dispute A British arbitration tribunal has ruled in PGS’s favour in a dispute over transfer fees under several data licensing agreements. In 2022, PGS recorded revenues of approximately $30 million of transfer fees relating to a change of control event under several data licensing agreements, where the amount was still not agreed with the client. Given the inability to conclude the negotiations, PGS initiated two separate arbitration proceedings under the dispute resolution provisions in the agreements. The tribunal in the first arbitration proceeding has now issued a decision in PGS’s favour. The award includes late payment interest and reasonable legal costs. The amount due to PGS, net of any revenue share to third parties, is estimated to be approximately $40 million. The second arbitration proceeding is expected to conclude during the second half of 2024, unless a settlement between the parties is reached earlier. As a result of the proceedings, PGS expects to recognise extra revenues in Q3 2023 of $15 million – in addition to the estimated $168 million Q3 2023 revenues

reported in the company’s pre-announcement to the market October 10, 2023. PGS will in addition recognise ‘a significant amount’ as interest income. Meanwhile, PGS will soon release results from its 2D feasibility study earlier this year in Danish waters to support screening and characterisation of potential CCS sites. The study for the Geological Survey of Denmark and Greenland (Geus) and industry partners is phase 1 of a regional 2D rejuvenation project in the Danish

North Sea to be launched at the beginning of 2024. Work is focused on offshore sites identified by GEUS for maturation. It includes seven seismic sections from three surveys using state-of-the-art 2D processing including pre-processing, tomography and final VTI Kirchhoff PSDM (alternatively PSTM). ‘We aim to demonstrate the uplift achievable by reprocessing the data and to prepare the processing sequence for the larger project planned in 2024,’ said PGS.

Prestack depth migration generates more accurate images (left) reducing exploration risk and improving reservoir definition, said PGS.

TGS wins OBN contract offshore Gulf of Mexico

Kristian Johansen, TGS CEO.

TGS has won a three-month proprietary ocean bottom node (OBN) data acquisition contract in the Gulf of Mexico (GoM) for a repeat customer. Acquisition will start in Q4 2023 and complete during Q1 2024. Kristian Johansen, CEO at TGS, said: ‘This project further highlights the integral role OBN acquisition has in providing our clients improved seismic data quality and help them to make better reservoir development and management decisions.’ Meanwhile, TGS, in partnership with SLB and Petrobangla, has com-

FIRST

BREAK

pleted the 12,636 km first phase of a 2D seismic acquisition program in April 2023, offshore Bangladesh. The project covers the Bengal Fan, one of the world’s largest deep-water fans with good evidence of working petroleum systems. It is considered one of the most extensive unexplored frontier areas, said TGS. Finally, TGS has won a four-year extension to a proprietary reservoir monitoring and source acquisition contract in Norway, with a further option to extend by another two years.

VOLUME

DECEMBER

2023

INDUSTRY NEWS

BRIEFS SeaBird Exploration has reported third quarter revenues of $6.5 million with EBITDA of $4.2 million and cash flow of $3.2 million. Net-interest bearing debt was $12.6 million. Vessel utilisation was 96%. Equinor is preparing to sell a 20% stake in its Rosebank oil development in the British North Sea, which could fetch about $1.5 billion, two industry sources said. Equinor has agreed with Libya’s National Oil Corporation to study and evaluate the oil and gas potential offshore Libya. Eleven blocks are being offered offshore Suriname in shallow offshore areas, in water depths up to 150 m, located south of the deepwater discoveries and north of the onshore producing oilfields. The data room will open from 18 December with bids required by 31 May 2024. Australia is inviting bids on eight onshore areas for petroleum exploration in Western Australia. The blocks are in the Canning, Northern Carnarvon, Amadeus, and Perth basins, and vary in size from 400 to 7070 km2. Hydraulic fracturing will not be authorised within any of the exploration permits. The application period will close on 19 January 2024. The US Bureau of Ocean Energy Management is postponing Lease Sale 261, originally scheduled for September 27, 2023, and later scheduled for November 8, 2023, in response to judicial orders. Until the court rules, BOEM cannot be certain of which areas or stipulations may be included in the sale notice. The US Bureau of Ocean Energy Management (BOEM) has finalised four new Wind Energy Areas (WEAs) in the Gulf of Mexico. Option J is 495,567 acres located 47.2 miles off the coast of Texas. Option K is 119,635 acres 61.5 miles off the coast of Texas. Option L is 91,157 acres 52.9 miles off the coast of Texas. Option N: 56,978 acres located 82 miles off the coast of Louisiana.

FIRST

BREAK

VOLUME

TGS third quarter revenues approach $300 million TGS reported a net profit of $16 million on revenues of $293 million compared with a net loss of $2 million on revenues of $119 million in Q3 2022. POC operating profit of $68 million was up from a loss of $7 million in Q3 2022. Multi-client revenues were $178 million during Q3 2023 compared to $140 million in Q3 2022. Early sales of $88 million compared to $39 million in Q3 2022. Late sales of $72 million, compared to $65 million in Q3 2023, were ‘below expectations’. Total POC order backlog, including acquisition, is $475 million. Full-year multi-client investment guidance has been increased to $400 million with a minimum early sales rate of 75%. ‘We are pleased to present a Q3 2023 report that is strong on all parameters. POC revenues are up 34% y/y (pro-forma), driven by robust performance in all business areas. Moreover, excellent operating performance and tight cost control leads to a healthy operating margin of 23%. Finally, cash flow was solid during the quarter, with a free cash flow of $45 million, further enhancing the balance sheet ahead of the announced

acquisition of PGS. With leading exposure across the energy data value chain, TGS is well positioned to continue benefiting from the expected market growth going forward,’ said Kristian Johansen, CEO of TGS. Cash flow from operations of $117.6 million, compared to $177.9 million in Q3 2022 In the quarter the company focused on multi-client programmes in Europe, Brazil, Egypt and Malaysia. ‘Prefunding revenues of $101 million reflect strong interest for ongoing acquisition projects and significant sales from surveys in the processing phase, resulting in a prefunding level of 144%,’ said TGS. The company’s order book increased 28% sequentially. A major part of the increase is contract work with pricing for the winter season at similar levels to the summer. TGS has 60% of 3D vessel capacity booked for the first half of next year. Meanwhile, TGS has won a four-year extension to a proprietary reservoir monitoring and source acquisition contract in Norway, with a further option to extend by another two years.

PGS reports net loss of $7 million on falling revenues PGS has reported a net loss of $7 million on revenues of $185 million compared to a net profit of $3 million on revenues of $216 million. Operating profit (excluding impairments) of $12 million compared with $34 million in Q3 2022. Rune Olav Pedersen, president and chief executive officer said: ‘As the global energy transition evolves, PGS expects global energy consumption to continue to increase over the longer term with oil and gas remaining an important part of the energy mix. Offshore reserves will be vital for future energy supply and support demand for marine seismic services. The seismic market is improving on the back of increased focus on energy security, several years of low investment in new oil and gas supplies, and high oil and gas prices. DECEMBER

2023

Offshore investments in oil and gas exploration and production are increasing in 2023. The seismic acquisition market is benefiting from the higher exploration and production spending, and a limited supply of seismic vessels. PGS expects full-year 2023 gross cash costs to be below $550 million. Multi-client cash investments are expected to be approximately $190 million. Capital expenditures for 2023 are expected to be below $100 million. Some 45% of 2023 active 3D vessel time is expected to be allocated to contract work. The order book amounted to $437 million on September 30, 2023. On June 30, 2023, and September 30, 2022, the order book was $341 million and $253 million, respectively.

TECHNICAL ARTICLE

Deep structural imaging in the Vienna basin Ahmed Mamdouh1, Klaus Pelz 2, Sandor Bezdan2, Erika Angerer 2, Alexandra Oteleanu2, Harald Granser 2, Abdelrahman Abubakr1, Alexander Sakharov3 and Ian F. Jones1*

Abstract Reprocessing of a recent wide azimuth vibroseis survey in the Vienna basin, followed by detailed anisotropic velocity model building and pre-stack depth migration, has led to an improved understanding of the subsurface structure, helping to de-risk future drilling decisions. Several processing and imaging routes were investigated, including common reflection surface stacking, 5D regularisation, Kirchhoff and common reflection angle migrations, using an interpretation-constrained velocity model. Results for the different routes are compared, with an assessment of the interpretational uplift obtained with the different approaches.

Introduction The Vienna basin is a mature and well established gas-producing region, with current production from the relatively simple, shallow flat-lying Neogene section and the more structurally complex pre-Neogene section, at depths greater than 3 km. Previous studies, combining gravity and seismic information (Granser, 1987; Rossi et al., 1998; Vesnaver et al., 2000; Spitzer and Gierse, 2008; Pfeiler et al., 2011; Salcher et al., 2012; Pelz et al., 2018; Garden and Zühlsdorff, 2019; Šamajová et al., 2019; Harzhauser et al., 2022) have outlined the overall structural picture of the Vienna basin, but obtaining the detail required for accurate deep well placement remains problematic. In addition, the recent interest in carbon capture and storage and CO2

enhanced oil recovery has led to further investigation to identify structures suitable for these purposes (e.g. Mikunda et al., 2020). Figure 1 shows a tectonic overview (Garden and Zühlsdorff, 2019 as adapted from Wessely, 1992): as can be inferred, whereas the Neogene shallow section (extending to a depth of approximately 3 km), is structurally simple, the deeper pre-Neogene section is extremely complex, combining regional thrust faults with highly distorted overturned beds. The data used for this study were acquired between 2017 and 2019 as part of a large wide azimuth survey over the Schönkirchen area using a Vibroseis source, emphasising the low frequencies to help with deep imaging. For most of the survey, the sweep range used was 2 Hz-90 Hz, with a 64 s sweep producing 9 s

Figure 1 Tectonic overview. Geological section of the Vienna Basin. Wells drilled by the time the section was produced are indicated. Most of the wells target shallow, thin reservoirs within the Neogene sands (yellow). Green indicates oil reservoirs, red indicates gas reservoirs, known or expected. The deep carbonate reservoirs (hashed light blue) are of exploration interest.

Brightskies Geoscience | 2 OMV, Vienna | 3 SeisEdge, Stavanger

Corresponding author, E-mail: Ian.Jones@bsgeoscience.com

DOI: 10.3997/1365-2397.fb2023097

FIRST

BREAK

VOLUME

DECEMBER

2023

TECHNICAL ARTICLE

Figure 2 Survey area, showing the location relative to Vienna (a), and indicating different phases of the acquisition (b). In 2b, the denser red area had a finer 200 m receiver line spacing, compared to the 400 m spacing over the rest of the area.

Figure 3 Vintage preSDM image from the 2020 processing, showing the full section. Although the shallow pre-Neogene section (down to approximately 3 km) is well imaged, the deeper structural image is still ambiguous.

records (in urban areas, a 12 Hz low-end of the sweep had to be used). These data were initially processed between 2018 and 2020 as part of a project that also incorporated some earlier vintage data (acquired in 2016) covering an area of about 1500 km2. In the study presented here, only a 500 km2 subset of the recent wide azimuth Vibroseis data were used. Figure 2 shows a location map, indicating the different phases of acquisition. Most of the area was shot with 400 m receiver line spacing with 40 m receiver interval, and 400 m shot line spacing with 20 m shot interval, but the denser red area in Figure 2b has denser 200 m receiver line spacing. Methodology The anisotropic velocity model for the overburden Neogene sedimentary sequence had been adequately constructed in the previous 2020 processing project using a combination of first-arrival refraction tomography and subsequent reflection tomography. In this phase of the study we concentrated on the more problematic, structurally complex pre-Neogene sequence. At these depths (typically greater than 3 km) are found the highly deformed nappes of the Northern Calcareous Alps (NCA) formed during an active tectonic period in the Late Cretaceous period. In the earlier processing project that focused on the shallower geological section, the processing parameters were optimised 32

FIRST

BREAK

VOLUME

DECEMBER

2023

for high resolution shallower imaging, and not for preservation of the deeper structure. Hence, in the work considered here, the pre-processing was repeated optimising parameters for the deeper, pre-Neogene section. Overall, the parameterisation of the many and various noise suppression methods was selected to be milder than in the original processing in order to better preserve the signal content of the pre-Neogene section. Figure 3 shows a final image from the 2020 project showing good shallow imaging of the mostly flat-lying Neogene structures. Both 5D regularisation and common reflection surface processing (CRS) were assessed, and the CRS data selected as input for anisotropic Kirchhoff migration, as it preserved the deeper signal more than the 5D approach. Azimuth information was preserved during OVT processing to assess any azimuthal velocity variation. An initial pass of (ray-based) anisotropic tomographic inversion was performed, to optimise gather flatness using the available offsets (a maximum of ~8 km, but more generally about 6 km), and to refine the anisotropy parameters. The delta values were updated by calibrating against the available well-control, and epsilon adjusted during the tomographic inversion, resulting in delta values of about 4% down to 1 km, 9% from 1 km - 2.5 km, and decreasing to around 2% in the pre-Neogene section, with epsilon ~1.5*delta. For the three main wells in the target area, the resulting misties at around 3600 m depth were: -20 m, 0 m, and +10m. Thereafter a migration-scanning approach followed by automated identification of the most coherent elements of the scanned images was used to update the velocity model. Two passes of scanning were performed initially perturbing the tomographic velocity field from 91% to 109% in steps of 3%, with a second pass starting from the output model from pass #1, perturbing that model between 96% and 106% in steps of 2%. Following this, a reflectivity-based geobody insertion method was used to adjust the velocities in the regions where high velocity dolomite rafts were believed to exist. This method relied on the instantaneous phase of high reflectivity events, constrained with well-control to determine the geobody velocities (Valler et al., 2017). Results Figure 4 outlines the evolution of the velocity model starting from the 2020 legacy anisotropic model which can be seen to be lacking in any structural detail in the deeper section. After an initial

TECHNICAL ARTICLE

pass of tomographic inversion updating only velocity, the model was recalibrated against the well-database to facilitate an update of delta, followed by recalibration of the epsilon parameter to maintain gather flatness (higher order moveout was not visible on the gathers in the deeper section). In migrations using the final anisotropic velocity model, the depth mis-ties at the pre-Neogene

level (depth ~3600 m) in the three wells in the target area were 0 m, 10 m, and -20 m. Kirchhoff 3D depth migration with an intermediate velocity model on the relatively raw field data already showed better preservation of the target events (Figure 5). The dipping event outlined in the yellow ellipse is known to be a real event

Figure 4 a) legacy anisotropic velocity model; b) anisotropic velocity field following initial tomographic update and recalibration to wells, c) velocity after scanning and automated update; d) final velocity field after high contrast geobody insertion for the dolomite rafts.

Figure 5 Zoom on the target area, comparing an inline from the final vintage 2020 processing a) and an early stage image from the new processing before denoise and final model update. The target event (outlined in the yellow ellipse) is more clearly visible in the new (but still noisy) image. The vintage image has good resolution of the shallow section and is ‘clean’ but has degraded the integrity of the deeper target structure.

Figure 6 Crossline comparing a) vintage result from 2020 processing; b) new denoised processing with final model; c) 5D regularisation with final model; d) CRS input with final model. The 5D regularisation has slightly degraded some of the very deepest structures, hence the CRS result was preferred. In the CRS result (d) it can clearly be seen that the steep dip fault planes and other structural elements are much better preserved than in the vintage data (a).

FIRST

BREAK

VOLUME

DECEMBER

2023

TECHNICAL ARTICLE

Figure 7 Zoom on the deeper target structure comparing a) the 2020 vintage preSDM; b) preSDM from the current project using the reprocessed input data and final geobody velocity model with structurally orientated smoothing applied to the resulting image; c) and the reprocessed input data with additional CRS processing. This is the line for which the velocities were shown in Figure 4.

Figure 8 Zoom on the deeper target structure comparing a) Kirchhoff preSDM inline from the current project using the reprocessed input data with additional CRS processing, b) CRAM preSDM inline, c) Kirchhoff CRS crossline, d) CRAM crossline. For the most-part, the CRAM result clearly improves the steeply dipping events and fault planes, as well as suppressing noise.

(from well penetrations) and is better preserved in the new data processing, even before application of additional denoise procedures. Figure 6 shows a central crossline comparing the newly processed data migrated with the final geobody model for three denoise and interpolation strategies and contrasts these results with the corresponding vintage 2020 image. Figure 6a is the preSDM of the vintage data, Figure 6b is the migration of the basic denoise data, Figure 6c is the migration after 5D regularisation/interpolation, and Figure 6d is the migration of CRS processed data. The 5D regularisation has slightly degraded some of the very deepest structures, hence the CRS result was preferred for interpretation. Figure 7 shows an inline through the target area comparing the vintage result with Kirchhoff preSDM images produced with the newly denoised data and the CRS processed data. Again, the vintage result is suboptimal in the deep section. In addition to the Kirchhoff migration, a common reflection angle migration (CRAM) was performed. CRAM performs its ray-tracing directly from the subsurface and with respect to the subsurface dip field, so is usually better able to suppress migration noise than a conventional Kirchhoff scheme. Overall, this gave a significant uplift in terms of signal to noise, with improved imaging of most of the steeply dipping events (Figure 8). 34

FIRST

BREAK

VOLUME

DECEMBER

2023

Discussion The success of OMV’s upcoming drilling programme in the Vienna Basin depends to a large extent on a reliable interpretation of the deep pre-Neogene structure from seismic images. Seismic data from this area are known to be of notoriously poor quality in terms of the signal-to-noise ratio of the deeper section. Hence, here we strove to be extremely cautious in the application of various de-noise procedures so as not to compromise any weak underlying signal: the emphasis being on retaining as much signal as possible, guided by the knowledge of dips and locations of various structural elements obtained from well logs. Being able to better understand both the basin architecture and the faulting and compartmentalisation of the reservoir units was central to de-risking the prospectivity of the area. Based on the images obtained during this reworking of the recent data, we have been able to advance our knowledge with sufficient confidence to guide the infill drilling programme. OMV’s recent natural gas discovery in this area (the largest in Austria for 40 years), confirmed with the successful Wittau Tief-2a exploration well, points to the importance of this prospect, and the new imaging presented in this study should help with future appraisal well drilling.

TECHNICAL ARTICLE

Acknowledgements We would like to thank Oliver Langton and Paul Keane for helpful suggestions on improvements to the paper, AspenTech for provision of the Common Reflection Angle Migration software, and to OMV E&P GmbH for permission to publish this work. We would also like to thank Gwenola Michaud the other reviewers for their constructive and helpful suggestions.

Pfeiler, S., Fuchsluger, M., Stotter, C., Chwatal, W. and Brückl, E. [2011]. 3D Model Based Acquisition Design for Imaging the Deep Vienna Basin. Eage Annual Conference. Rossi, G., Böhm, G., Madrussani, G., Vesnaver, A., and Granser, H. [1998]. 3D adaptive tomography and imaging in the Vienna Basin, SEG annual. Salcher, B.C., Meurers, B., Smit, J., Decker, K., Hölzel, M. and Wagreich, M. [2012]. Strike-slip tectonics and Quaternary

References

basin formation along the Vienna Basin fault system inferred

Garden, M. and Zühlsdorff, L. [2019]. Acquisition of a modern seismic

from Bouguer gravity derivatives, Tectonics , 31, TC3004,

survey in the Vienna Basin, SEG annual.

doi:10.1029/2011TC002979.

Granser, H. [1987], Three-dimensional interpretation of gravity data from

Šamajová, L., Hák, J., Csibri, T., Bielik, M., Teták, F., Brixová, B., Sliva,

sedimentary basins using an exponential density-depth function,

L. and Šály, B. [2019]. Geophysical and geological interpretation of

Geophys. Prospect., 35, 1030–1041.

the Vienna Basin pre-Neogene basement (Slovak part of the Vienna

Harzhauser, M., Kranner, M., Mandic, O., ´Cori´c, S. and Siedl, W.,

Basin). Geologica Carpathia, 70(5), 418-431.

[2022]. High resolution Late Miocene sediment accommodation

Spitzer, R. and Gierse, G. [2008]. Target-oriented Seismic Data Process-

rates and subsidence history in the Austrian part of the Vienna Basin,

ing in the Vienna Basin – A Key to Improve Imaging Using Vintage

Marine and Petroleum Geology, 145 (2022) 105872.

Data, EAGE Annual Conference.

Mikunda, T., Franců, J., Pereszlényi, M., Hladík, V., Kolejka, V., Kulich,

Valler, V., Nathan Payne, N., Hallett, T., Kobylarski, M., Venkatraman,

J., Götzl, G., Kollbotn, L. and Jankulár, M. [2020]. Report: Towards a

G., Rappke, J. and Fairclough, D. [2017]. A holistic approach to

strategic development plan for CO2-EOR in the Vienna Basin. Euro-

model-building in and around injectites: a case-study offshore

pean Union Horizon 2020 research and innovation programme under grant agreement No 653718, ENOS, http://www.enos-project.eu.

Norway, Eage Annual Conference. Vesnaver, A., Böhm, G., Madrussani, G., Rossi, G. and Granser, H.

Pelz, K., Granado, P., Strauss, P., Roca, E., König, M., Thöny, W., Peresson, H. and Muñoz, J.A. [2018]. Overturned to recumbent thrust

[2000]. Depth imaging and velocity calibration by 3D adaptive tomography. First Break, 18, 303-312.

sheets in the Northern Calcareous Alps (NCA)—The role of inflated

Wessely, G. [1992]. The Calcareous Alps below the Vienna Basin in

salt on fold-and-thrust belt structural styles: EGU General Assembly

Austria and their structural and facial development in the Alpine-Car-

Conference Abstracts 20, 7676.

pathian border zone: Geologica Carpathica, 43, 347-353.

INTRODUCING

Providing faster access to subsurface images for quicker decisionmaking by processing seismic data right where it was acquired

Compared with conventional processing methods, STRYDE Lens can unlock significant data processing time savings on both 2D and 3D data:

80% 60% up to

up to

Reduction on field style processing time

Reduction on conventional style processing time

DISCOVER MORE

FIRST

BREAK

VOLUME

DECEMBER

2023

30th

5th

4th

European Meeting of Environmental and Engineering Geophysics

Conference on Geophysics for Mineral Exploration and Mining

Conference on Airborne, Drone and Robotic Geophysics

ABSTRACT SUBMISSION

IS NOW OPEN! LEARN MORE AT EAGENSG.ORG

8-12 SEPTEMBER 2024 I HELSINKI, FINLAND

# N S G 2 0 24

TECHNICAL ARTICLE

Applied fault topology: understanding connectivity and uncertainty of fault systems that define and affect commercial and environmental projects Frank Richards1*, Mark Cowgill1 and Megan Rayner1.

Abstract For subsurface commercial ventures or environmental projects that rely on maps of faulted horizons, accurate maps are fundamental. Fault topology provides an ideal tool for analysis of connectivity of fault systems. The data required to undertake the analysis is straightforward to extract from fault maps and can readily be compared to analogue data. In this paper we introduce the concept of fault topology and present existing and new analogue data. To get the most from applying topology the analysis must be coupled with knowledge of the structural history. This includes, the magnitude of faulting, the number of phases of activity and the angle of intersection of successive faulting events. We present a series of case studies that firstly illustrate how topology can capture and define variations in connectivity of fault systems and, secondly, demonstrate how fault topology can be used to identify potential anomalies.

Introduction Subsurface maps of horizons and the faults that affect them are of fundamental importance to the energy industry. The size, geometry, connected elements and the position of potential barriers is equally important for emerging energies and carbon storage as it is to petroleum accumulations (Stober et al 2017, Mulrooney et al 2020). The application of topology to better understand connectivity in fracture networks has been discussed by Ortega & Marrett (2000) and Sanderson & Nixon (2015). More recently, topology has been applied to fault systems from a variety of data types; outcrop, horizontal seismic sections,

seismically mapped horizons, analogue models and remote sensed data (Morley & Nixon 2016, Duffy et al 2017, Morley & Binazirnejad 2020, Mendes et al 2022, Kania & Szczęch 2020, Osagiede et al 2023). In this paper we will briefly review fault topology, but the main objective is to demonstrate how topology can be used to clearly quantify the connectivity of a fault system at a particular level (mapped horizon or unconformity). Also, how topology can be used to spot significant changes in connectivity and how by integrating knowledge of the tectonic setting and structural history it is possible to identify potential errors in the interpretation.

Figure 1 (a) (i) Map showing a population of normal faults (ii) Node types (iii) Branch types (b) Node types with the number of branch connections (c) Fault node topology graph (d) Fault branch topology graph. Both graphs show the CB ratio as defined by Sanderson and Nixon (2015).

CGG

Corresponding author, E-mail: francis.richards@CGG.com

DOI: 10.3997/1365-2397.fb2023098

FIRST

BREAK

VOLUME

DECEMBER

2023

TECHNICAL ARTICLE

Fault topology The first step of undertaking topological analysis requires the identification of how the fault elements connect; the size, magnitude of displacement and geometry of the faults are immaterial. Fault terminations (tips) are defined as isolated i-nodes, fault connections as y-nodes and crosscutting faults as x-nodes (Figure 1). Fault branches are the portions of the fault located between nodes. There are three types of fault branch: ‘i to i’ (isolated), ‘i to c’ (connected at one end) and ‘c to c’ (connected at both ends) (Figure 1). The number of branches (NB) is given by: NB = ½(Ni + 3Ny + 4Nx) The level of connectivity of a fault system can be expressed in terms of the average number of connections per branch (CB): CB = (3NY + 4NX)/NB The CB ratio is a dimensionless number between 0 and 2: completely isolated to completely connected (Sanderson and Nixon, 2015). Topology data plotted onto ternary graphs, shows the relative proportions of the different node types (i, y or x) or branch types (i-i, i-c, c-c). In the node topology graph CB is expressed as a series of contours, in the branch topology graph the ratio plots as a curved line with increasing CB (Sanderson and Nixon 2015). In both plots, the degree of connectivity increases towards the base of the graph. In this paper we focus on using fault branch topology, as fault populations – unlike fracture populations – tend to be more binary (i and y nodes) as cross cutting faults evolve with displacement from x nodes to discrete pairs of y nodes and thus cluster along the i-y axis of the node ternary plot (Duffy et al 2017). Figure 2 shows a set of analogue data derived predominantly from maps of surfaces generated from good quality 3D seismic data. Also included are data from outcrop images and from the

Figure 2 Fault branch topology plot of analogue data.

FIRST

BREAK

VOLUME

DECEMBER

2023

results of analogue modelling (Clifton et al 2000, Henza et al 2011). New data presented in this study is also summarised in Appendix 1. Seismic quality is infinitely variable and can vary greatly even within a single data set. However, to capture as reliable data as possible, we endeavor to use only what we consider to be ‘good quality’ seismic data. We define ‘good quality’ seismic as: i. data with a high signal to noise ratio, i.e., clean data where seismic reflections are clear and faults, for the most part, can be readily identified particularly at the level of interest. ii. Where faulted, multiple reflectors of the relative upthrown and downthrown seismic sequences can be correlated with a high level of certainty, or in the case of syn-depositional faulting, a growth package can be clearly identified. For published maps, where there is no (or limited) seismic data available, we include maps that show clear, unambiguous faults and show no anomalous faults with the application of basic structural tests (such as fault displacement – length relationships). Published data from Duffy et al (2017) has also been included in the ternary plot in Figure 2. Despite the inherent uncertainty in using such a mix of analogue data, the data shows a good fit to the CB trend line (Figure 2). Analogues from areas subject to just a single phase of extension display low levels of connectivity (e.g., Top Miocene, Bonaparte Basin, Figure 2) and areas where two sets of intersecting faults are present show high levels of connectivity (e.g., Gullfaks field map, Figure 2). These observations are consistent with published studies where topology has been applied (Morley & Nixon 2016, Duffy et al 2017, Morley & Binazirnejad 2020). In essence we have the makings of a consistent theoretical and analogue-based framework which can be used to evaluate maps which, for example, were generated from poor quality data, where there is limited coverage or where there may be a viable alternate interpretation.

TECHNICAL ARTICLE

Figure 3 Fault branch topology data derived from physical models (a) Open red squares; data from Duffy et al (2017) (based on Henza et al 2011). Multi-phase rifting; i to iii = increasing single phase strain, iv-vi = increasing superimposed oblique strain (small inset maps show the detail of the fault patterns at selected stages). Black squares; derived from Clifton et al (2000). α is the angle between a pre-existing structural trend and the angle of the imposed extension in each model. (b) Adapted finite fault maps from Clifton et al (2000) showing the angle α and the position of faults for seven experimental models.

Duffy et al (2017) generated topology data from the analogue clay models of Henza et al (2011) that investigated the nature of faulting associated with two superimposed, oblique extensional faulting events. One of the principal conclusions was that increasing connectivity (CB) could be correlated with increasing strain along a single fault trend for low ratios (CB = 0 – 0.8) and that higher ratios (CB = 0.8 – 2) could only be achieved by having progressively greater movement during a second phase of crosscutting faults (Figure 3). The magnitude of the two events is important, if one of the events is weak, fault connectivity will be limited. The above enhances the potential insights that fault topology may provide. In that, knowledge and understanding of the geological history of a region can be cross-referenced with the graphical representations on the ternary plot. Figure 3 also shows topology data generated from a suite of physical models undertaken by Clifton et al (2000). In these experiments a layer of clay was subject to a single phase of extension orientated at variable angles with respect to an underlying rift trend (angle α). The modelling and finite extension was very similar to that of Henza et al (2011). The geometry of the rift axis was controlled by overlapping steel plates overlain by a latex sheet. The results showed that when extension was orthogonal (α = 900) or at a high angle (down to α = 600) to the direction of the rift, the resulting connectivity was broadly similar to a high strain single phase rift in the experiments of Henza et al (2011). As the angle decreased the connectivity increased significantly to a maximum CB ratio of 1.25 when α was equal to 300 (Figure 3). At very low angles (α = 150 and 00) the connectivity rapidly decreased. It should be noted that two data points plot some distance from the CB ratio line in the ternary plot (α = 00 and 900), the overall trend is, however, clear. The significance of this is a discussion point later in the paper.

These results suggest that some caution is required in the transition zone between single and multiphase rifts (around CB 0.8 to 1.25) and we can’t automatically assume two phases of faulting. It’s possible that a single trend is being affected by, and interacting with, an underlying pre-existing trend. These data also suggest that where there are two phases of extension the angle between the phases is an important factor and needs to be considered. Case studies In the next section we will look at a series of contrasting case studies to see how topology can be applied to aid the understanding of connectivity in subsurface maps. Case Study 1: Niger delta extensional fault system

This relatively simple case study looks at faults that have been mapped on high-resolution 3D seismic data (Figure 4a). The maximum displacement on these faults is low (tens of metres) and the faults shown were derived from a mapped surface. To the southwest of the map a single trend is observed. Despite a high proportion of isolated faults there are areas where some larger faults are well connected. The curved nature of these faults has enhanced the connectivity (CB ratio around 0.9) (Figure 4b). To the northeast, the presence of a second crosscutting fault trend can be linked to a significant jump in the connectivity with a CB ratio of around 1.8 (Figure 4b). This example is a good analogue, it shows how fault topology can be used to quantify changes in fault connectivity on a subregional scale. This is important if we want to apply this methodology, say, on variation that may occur across a producing oil field. The resolution of the data and identifying changes in the fault population (additional faults and changing patterns of strain) are the key factors in this analytical method. FIRST

BREAK

VOLUME

DECEMBER

2023

TECHNICAL ARTICLE

Case study 2: Passive margin extensional fault system

This case study represents the opposite end of the spectrum with regards to data quality; the quality in this case was very poor. The tectonic setting was well understood and along strike analogues, based on better-quality seismic lines, show crustal scale normal faults with kilometre scale displacements. The presence of largescale transfer zones has also been recognised (linked to pre-existing structural trends) on a regional scale (Figure 5a). The mapped fault population contains a very high proportion of unconnected faults and plots at the top of the ternary diagram (CB = 0.15, Figure 5b). Populations that plot in this zone are associated with very low strain and with a single set of faults (Figure 3). As we are in a position where we know something about the regional setting, we can feed our understanding of the fault topology into the interpretation: we would expect a high strain rift and the presence of transfer zones to possess a much higher CB ratio (Figures 2 and 3) (Duffy et al 2017). The ratio of branches to connected nodes, in

this case, is far too low. Of course, this approach does not inform us as to which faults may need to be modified, but it does give licence for the interpreter to be much more assertive with the mapping and structural model. In studies with challenging data sets, such feedback is extremely valuable and can significantly improve the quality of the interpretation. It may also prevent a region from being written off and make it easier to push ahead and acquire new data. Case study 3: Multiple models

This case study considers multiple alternative interpretations of an identical data set. It highlights the inherent uncertainty associated with interpreting when there are data issues. In this instance, it was data coverage rather than seismic quality that was the source of the problem. The data was comprised of good quality 2D seismic. Faults imaged on individual lines were clear and consistently identified by the different teams working on the

Figure 4 Case Study 1: (a) Fault branch topology map of faulting affecting a shallow surface in the Niger Delta. (b) Fault branch topology plot showing two structural domains. Plot includes data from Figure 2 (grey points).

Figure 5 Case Study 2: (a) Fault branch topology map of faults interpreted in a mature passive margin (a surface towards the top of the pre-rift sequence) (b) Fault branch topology graph of the faults in map 5(a). Plot includes data from Figure 2 (grey points).

FIRST

BREAK

VOLUME

DECEMBER

2023

TECHNICAL ARTICLE

Figure 6 Case Study 3: (a) Fault topology plot based on data collected from eight alternate interpretations of the same 2D seismic data set. Plot includes data from Figure 2 (grey points). (b) Fault branch topology maps.

Figure 7 Case study 4: (a) Fault Branch topology map of the Buchan Field, the thick continuous red line separates the uncompartmentalised west from the highly compartmentalised east (adapted from Marshall & Hewett (2003) (b) Fault branch topology plot. Plot includes data from Figure 2 (grey points).

data (Richards et al 2015). How each team chose to link together the faults led to the production of a suite of very different maps (Figure 6). Figure 6a shows the fault branch topology of eight maps included in the study. The alternate maps show a surprising amount of variation with regards to connectivity, with CB ratios ranging from as low as 0.7 up to 1.8. Three of the maps are shown in Figure 6b, Maps 8 and 6 represent the end members in terms of their fault connectivity; Map 6 was compiled without a single isolated fault, and although Map 8 displays faults with similar trends, 46% of the faults are isolated (Figure 6a). Regardless of the intended use of these two maps, be it defining containers or predicting the movement of fluids, these two maps exhibit radically different scenarios. For example, it’s possible to navigate in a north-south direction through the centre of Map 8 without encountering a single fault barrier (Figure 6b). Can we do more? Can we look at these results and use topology to suggest that some maps are better representations of the subsurface? If we consider the tectonic setting, some of the interpretations appear less likely. The extensional strain at this level is

relatively low and although the faults are curved, they generally strike in a north-south direction. There is also evidence of fault reactivation but no change in the direction of extension. Based on this, one could argue that maps that display connectivity close to 2 are unlikely (this would affect two of the eight maps). Although it has yet to be established whether a significant departure from the dashed CB line in the ternary branch topology graph represents an anomalous map, two of the maps sit significantly outboard of the analogue data (see discussion). So, with comparatively little effort and expenditure we can seriously question four of the eight maps. At the very least these interpretations should be rigorously checked, and additional analysis undertaken (Richards et al 2015). This case study illustrates the human element of the interpretation process and shows it’s possible to generate results covering greater than 50% of the total range of connectivity (CB 0.7 to 1.8) from identical data. A further conclusion is that anyone using 2D data to define fault patterns for faulted horizons should be very cautious and probably review the implications of alternate scenarios. FIRST

BREAK

VOLUME

DECEMBER

2023

TECHNICAL ARTICLE

Case Study 4: The Buchan Field

The Buchan Field produces from Upper Devonian to Lower Carboniferous reservoirs and is located in the UK North Sea. Early maps, based on 2D data, were comparatively simple and showed the field to be comprised of four fault-bounded compartments (Edwards 1991). For this study an analysis of a more recent map based on 3D data was undertaken (Marshall & Hewett 2003, Figure 7a). It should be noted that the quality of the 3D seismic was comparatively poor (Marshall & Hewett 2003). A topology assessment revealed a potential anomaly in the fault interpretation. The eastern portion of the field is heavily compartmentalised by faults (18 compartments). In stark contrast, a similar-sized area to the west is comprised of a single irregular compartment (Figure 7, a red line defines the boundary between the compartmentalised east and open west). The fault topology from the western side has a CB ratio of 0.9. The connectivity of the eastern half of the field is predictably much higher (CB ratio around 1.7, Figure 7b). Case study 1 (Figure 4) showed an example of a rapid variation in connectivity where a new fault trend locally develops and crosscuts an existing trend. Could this be happening across the Buchan Field and was it missed during the interpretation of the 2D data? The regional structural trend is dominated by sub east-west trending faults and intersecting north-south faults have been described by Zanella and Coward (2003). So, it’s feasible that north-south faults may be affecting the eastern half of the field but not the west. Two observations suggest this is not the case: the irregular geometry of the transition and the fact that the nature and trends of faults across the whole map, for the most part, are similar. Seemingly, the only difference is that the eastern half is just more connected. As the seismic data is relatively poor, we would argue that the origin of the variation is more likely to be related to how the data was interpreted. Is it possible we are dealing with a merged interpretation? Regardless of the origin of the variation, the analysis strongly suggests that the field needs to be remapped and the connectivity made more consistent. Case Study 5: The Scott Field

The final case study is based on the fault topology of the Scott Field, which is also located in the UK North Sea. The 3D seismic

data that covers the field is generally of good quality. The structure varies in complexity: to the southeast, a single set of northwest-southeast rotated fault blocks are observed, to the west they merge and intersect with a second trend running roughly northeast-southwest (Brook et al 2010). Where the two fault sets are present, the structure is more complex. Numerous intra block faults, with variable orientations are also observed. The strain on all the major faults is high with throws of more than 1000 m. It should be noted that in areas where the complexity is high, the quality of the data is somewhat reduced. The fault topology analysis shown in Figure 8a is based on faults from within the structural zones defined by master faults (zones A-E, Figure 8b, n.b. the map upon which these data were derived is not shown here), it includes faults that splay or intersect with the larger, block-defining, faults. The connectivity is highly variable but a consistent pattern is evident. Blocks A and B are located where a single fault trend dominates and accordingly displsys lower values of connectivity (CB 0.5 – 0.8). Blocks C and D, in the more complex area to the NW show higher values of CB (1.4 to 1.8). The topology of Block E is anomalous: despite lying in the more structurally complex part of the field the connectivity is relatively low with a CB of around 0.75 (Figure 8a). Fault topology in this case study was used to recommend revisiting this zone and reassessing the fault interpretation. Conclusions When generating subsurface fault maps, with the exception of simple structures illuminated completely by good quality data, uncertainty is a critical issue. The degree of fault connectivity is one of the key uncertainties. The application of fault topology allows us to quantify connectivity and critically, enables us to compare the areas of economic importance to analogues. Case Study 1 represents a good analogue, the quality of the data is very good and there is a high certainty as to how the faults connect. It demonstrates how there can be a rapid transition of fault connectivity over a relatively short distance (Figure 4). The second case study (Figure 5) lies at the opposite end of the

Figure 8 Case study 5: (a) Fault branch topology of intra-block faults of the Scott Field. Points A to E represent data from structural subregions shown in the inset map, n.b. the data shown on the ternary plot is based on proprietary data that is not shown here. Plot includes data from Figure 2 (grey points). (b) Simplified map of the Scott Field showing the position of the major faults and intra block subregions A to E. Grey Circles show the position of production wells (adapted from Brook et al 2010).

FIRST

BREAK

VOLUME

DECEMBER

2023

TECHNICAL ARTICLE

Figure 9 Fault branch topology plot showing an amalgamation of analogue data (in grey), case study data and data from the models of Clifton et al (2000). The plot also shows a ‘buffer zone’ which runs parallel to the CB ratio line and bounds the analogue data from Figure 2.

spectrum with regards to mapping uncertainty being based on very poor quality seismic. In this case fault topology can act as a guide, given the understanding of the tectonic setting (a high strain rift margin, with transfer zones) faults were under connected. This gives licence to the interpreter to be more assertive and introduce more connected faults into the interpretation. Case study 3 is a good reminder of the amount of variation inherent in making interpretations. It’s also a reminder that the origin of variation is not necessarily about seismic quality, but also about data coverage and the limits of using 2D data (Richards et al 2015). Fault topology was able to identify outliers based on a basic understanding of the regional geology and by comparison with analogue data. In case studies 4 and 5, fault topology was able to quantify the connectivity of structural domains and, by integrating an understanding of the structural geology, help to identify potential anomalies. Several methods of integrated structural analysis have been available for a number of decades. Commonly applied techniques include cumulative frequency of fault length (and fault displacement) and the ratio of fault displacement with respect to length (Walsh & Watterson 1988, Walsh & Watterson 1992). Freeman et al (2010) demonstrated how the distribution of strain along faults could be used to identify structural anomalies. Other relationships have also been identified, for example the relationship between the magnitude of displacement and width of relay ramps to help predict the probability of ramp breaching (Imber et al 2004). Fault topology offers something new, as it deals directly with fault connectivity. It’s also possible for it to be applied directly to lineaments observed on horizontal slices through seismic data (Morley & Nixon 2016), so it can be applied before, or during a mapping exercise. Discussion: Topology as a QC tool? The case studies described in this paper show how topology can effectively be deployed to quantify variations in connectivity of fault populations observed on maps. From a QC perspective this alone makes it a useful tool. When integrated with structural geology it is much more powerful. Duffy et al (2017) showed that increasing connectivity (CB ratio) could be related to increasing strain and multi-phase fault evolution. However, there is still

considerable work required to fully understand the evolution of connectivity. For example, data generated from the analogue models of Clifton et al (2000) show how pre-existing trends may enhance connectivity during a single phase of extension. The implication is that it may be difficult in some settings and at certain ratios (around CB 0.8 to 1.25) to distinguish between single and multi-phase rifting from topology alone. The CB ratio line in fault branch ternary plots (Sanderson and Nixon 2015) is mathematically defined ‘from randomly assigned node types to branches, weighted by probabilities of occurrence’ (Morley and Binazirnejad 2020). Fault topology analogue data presented here (Figure 2) displayed a clear tendency to cluster close to the CB line (Figure 9). This tendency is also evident in the published data of Duffy et al (2017), Mendes et al (2022) and Osagiede et al (2023). Some data, however, plots away from the line (Figure 9). Some of the published data from Morley and Nixon (2016) and Morley and Binazirnejad (2020) also shows this tendency. It’s possible this may be related to geology, it’s also possible that it may be linked to the quality of the interpretations, or indeed, some combination of the two. Clearly not all the maps in case study 3 (Figure 6) represent accurate representations of the subsurface. Maps 1 and 5 lie some distance off the CB ratio line; we propose that this is used as a possible quality flag. We define a buffer zone that bounds the outer limit of the analogue data in Figure 2 and treat data located outside the zone as ‘potentially poor’ (Figure 9). As more analogue data becomes available and these relationships become better understood, this approach will almost certainly need to be adapted. With time we predict that the use of fault topology will become increasingly routine and form an important part in any structural QC or review process. Acknowledgements We would like to thank the anonymous reviewer for feedback that greatly improved the direction and quality of this paper. Thank you CNOOC for allowing us to use data derived from their map of the Scott Field. We would also like to thank Nick Richardson and Marguerite Fleming for the usual feedback and encouragement. FIRST

BREAK

VOLUME

DECEMBER

2023

TECHNICAL ARTICLE

Map

Data type / quality

Short structural description

Reference / Figure used

Niger Delta A

3D seismic / good

Two intersecting fault trends

N/A Propriety seismic data

Niger Delta B

3D seismic / good

Single Fault Trend

N/A Propriety seismic data

Beatrice Field

3D seismic / good to fare

Single fault trend 2002

Husmo et al (), Figure 10.31a

Gullfaks Field, Rannoch Fm.

3D seismic / good to fare

Single event high strain extensional faults with oblique intra-block faulting

Fossen & Hesthammer (1998), Figure 3

Gullfaks Field, Statfjord Fm.

3D seismic / good to fare

Single event high strain extensional faults with oblique intra-block faulting

Fossen & Hesthammer (1998), Figure 4

Bishop, USA

Satellite image / good outcrop

Single fault trend, low strain

Iny County, California, USA (GR 37.458804, _118.448992)

East Africa Rift, Kenya

Satellite image / good outcrop

Dominant N-S trend affected by localized oblique faulting

Esonorua, Kenya (GR -1.35751, 36.339892)

Groningen Field, Holland

3D seismic / good to fare

Multiple intersecting fault trends

Jager & Visser (2017), Figure 5

Base Zechstein, Southern North Sea

3D seismic / ant tracking, Good

Multiple intersecting fault trends

Preiss and Adam (2021), Figure 7A

Base Zechstein, Southern North Sea 2

3D seismic / ant tracking, good

Multiple intersecting fault trends

Preiss and Adam (2021), Figure 7B

Top Eocene, Bonaparte Basin Australia

3D seismic, ant tracking / good

Single fault trend

Frankowicz, E. and McClay (2010), Figure 8d

Top Miocene, Bonaparte Basin, Australia

3D seismic, ant tracking / good

Single fault trend

Frankowicz, E. and McClay (2010), Figure 8d

Valanginian Unconformity, Bonaparte Basin, Australia

3D seismic / good

Two fault trends

Frankowicz, E. and McClay (2010), Figure 8d

Faults from physical models

Clay analogues models / good image quality

Variable depending on the model; single trend to two crosscutting trends

Clifton et al (2000), Figure 3

Appendix 1 Details of data used to populate the fault branch topology plot in Figure 2.

References

Frankowicz, E. and McClay, K.R. [2010]. Extensional fault segmentation

Brook, G.R., Wardell, J.R., Flanagan, S.F. and Regan, T.P. [2010]. The Scott Field: revitalization of a mature field. Geological Society, London, Petroleum Geology Conference Series. 7, 387-403.

and linkages, Bonaparte Basin, outer North West Shelf, Australia. AAPG Bulletin, 94, No.7 977-1010. Freeman, B., Boult, P.J., Yielding, G. and Menpes, S. [2010]. Using

Clifton, A.E., Schlische, R.W., Withjack, M.O. and Ackermann, R.V.

empirical geological rules to reduce structural uncertainty in seismic

[2000]. Influence of rift obliquity on fault-population systematics:

interpretation of faults. Journal of Structural Geology, 32, 1668-

results of experimental clay models. Journal of Structural Geology, 22(10), 1491-1509.

1676. Husmo, T., Hamar, G.P., Høiland, O., Johannessen, E.P., Rømuld, A.,

Cowie, P.A. and Scholz, C.H. [1992]. Displacement-length scaling

Spencer, A.M. and Titterton, R. [2003]. Lower and middle Jurassic.

relationship for faults: data synthesis and discussion. Journal of

The Millennium Atlas: Petroleum Geology of the Central and

Structural Geology, 14(10), 1149-1156.

Northern North Sea. Geol. Soc. London, 129-156.

De Jager, J. and Visser, C. [2017]. Geology of the Groningen field–an

Henza, A.A., Withjack, M.O. and Schlische, R.W. [2011]. How do the

overview. Netherlands Journal of Geosciences, 96(5), s3-s15.

properties of a pre-existing normal-fault population influence fault

Duffy O.B., Nixon C.W., Bell R.E., Jackson C A.L, Gawthorpe R.L.,

development during a subsequent phase of extension? Journal of

Sanderson D.J. and Whipp P.S. [2017]. The topology of evolving rift fault networks: Single-phase vs multi-phase rifts. Journal of Structural Geology, 96, 192-202.

Structural Geology, 33(9). Imber, J., Tuckwell, G.W., Childs, C., Walsh, J.J., Manzocchi, T., Heath, A.E., Bonson and Strand, C.G J. [2004]. Three-dimensional distinct

Edwards, C.W. [1991]. The Buchan Field, Blocks 20/5a and 21/1a, UK North Sea. Geological Society, London, Memoirs. 14, 253-259.

element modelling of relay growth and breaching along normal faults. Journal of Structural Geology, 26(10). 1897-1911.

Fossen, H. and Hesthammer, J. [1998]. Structural geology of the Gullfaks

Kania, M. and Szczęch, M. [2020]. Geometry and topology of tectonolin-

field, northern North Sea. Geological Society, London, Special

eaments in the Gorce Mts. (Outer Carpathians) in Poland, Journal of

Publications, 127(1), 231-261.

Structural Geology 141. 104186.

FIRST

BREAK

VOLUME

DECEMBER

2023

TECHNICAL ARTICLE

Manzocchi, T. [2002]. The connectivity of two-dimensional networks of

Osagiede, E.E., Nixon, C.W., Gawthorpe, R., Rotevatn, A., Fossen, H.,

spatially correlated fractures. Water Resources Research 38, 1162.

A‐L. Jackson, C., and Tillmans, F. [2023]. Topological character-

Marshall, J.E.A. and Hewett, A.J. [2003]. Devonian. The Millennium

isation of fault network along the northern North Sea rift margin.

Atlas: Petroleum Geology of the Central and North Sea. Geol. Soc.

Tectonics, e2023TC007841. Preiss, A.D. and Adam, J. [2021]. Basement fault trends in the Southern

London. Mendes, L de C., Correia U.M.C., Cunha O.R., Oliveira F.M. and Vidal,

North Sea Basin. Journal of Structural Geology, 153, 104449.

A.C. [2022]. Topological analysis of fault network in naturally frac-

Richards, F.L., Richardson, N.J., Bond, C.E. and Cowgill, M. [2015].

tured reservoirs: A case study from the pre-salt section of the Santos

Interpretational variability of structural traps: implications for explo-

Basin, Brazil, Journal of Structural Geology, 159. 103930

ration risk and volume uncertainty. Geol. Soc. of London, Special

Morley C.K. and Binazirnejad H. [2020]. Investigating polygonal fault topological variability: Structural causes vs image resolution, Journal

Publications 421, 7-27. Sanderson, D.J. and Nixon, C.W. [2015]. The use of topology in fracture network characterization. Journal of Structural Geology, 72, 55-66.

of Structural Geology, 130. 103930. Morley, C.K. and Nixon, C.W. [2016]. Topological characteristics of

Stober, I., Fritzer, T., Obst, K., Agemar, T. and Schulz, R. [2017]. Deep

simple and complex normal fault networks, Journal of Structural

Geothermal Energy: Principles and Application Possibilities in

Geology, 84. 68-84.

Germany. Federal Ministry for Economic Affairs and Industry.

Mulrooney, M.J., Osmond, J.L., Skurtveit, E., Faleide, J.I. and Braathen,

Walsh, J.J. and Watterson, J. [1988]. Analysis of the relationship between

A. [2020]. Structural analysis of the Smeaheia fault block, a potential

displacements and dimensions of faults. Journal of Structural

CO2 storage site, northern Horda Platform, North Sea. Marine and

Petroleum Geology 121.

Nicol, J., Watterson, J., Walsh, J.J. and Childs, C [1996]. The shapes, major axis orientations and displacement patters of fault surfaces, Journal of Structural Geology, 18, 235-248.

Geology, 10(3), 239-247. Walsh, J.J. and Watterson, J. [1992]. Populations of faults and fault displacements and their effects on estimates of fault-related regional extension. Journal of Structural Geology, 14(6), 701-71. Zanella, E. and Coward, M.P. [2003]. Structural framework. Chapter 4,

Ortega, O. and Marrett, R. [2000]. Prediction of microfracture properties using microfracture information Mesaverde Group sandstones, San

The Millennium Atlas: Petroleum Geology of the Central and North Sea. Geol. Soc. London.

Juan Basin, New Mexico. Journal of Structural Geology, 22(5), 571-588.

th Annual Meeting 16th & Conference Courses

13 13 -- 16 16 May May 2024 2024 Shangri-La Shangri-La Hotel Hotel Qingdao, Qingdao, China China Conference Conference Courses Courses 12 12 & & 17 17 May May InterPore.indd 1

FIRST

BREAK

VOLUME

03/10/2023 08:38

DECEMBER

2023

Special Topic

DATA MANAGEMENT AND PROCESSING The industry has innovated to offer more and better-quality data. Geoscience companies are competing with each other to offer improved data processing and management packages for new acquisition and reprocessing of vintage data, often using the latest innovations in machine learning and artificial intelligence. Ever more powerful computing capability is aiding geoscientists’ application of complex algorithms and integration with different data types to give a more complete picture of the subsurface – both for oil and gas reservoirs and increasingly for renewable energy projects needed for the Energy Transition. Christopher Hanton et al review the limitations of traditional subsurface data management tools and look at how machine Learning and artificial intelligence offer the capability to redesign how we manage data in our industry. Kristy DeMarco addresses key challenges in data management and provides useful solutions for data aggregation, transportation and activation Julien Oukili et al discuss the benefits of implementing deep neural networks for certain steps of seismic data processing on data examples from around the world. Jill Lewis et al explain how the SEGY update to SEG-Y_r2.1 is updating historic data and enabling auto-access to historic data. Jose Chapela explores the challenges associated with data management and review solutions to overcome them. Neil Hodgson et al demonstrate how reprocessing vintage data can bring new prospectivity to basins in Oman and East Coast India. Sylvain Masclet et al illustrate how the leveraging of sparse node data through an interferometry approach and the use of elastic FWI can enhance streamer seismic imaging.

Submit an article

Special Topic overview January

Land Seismic

First Break Special Topics are covered by a mix of original articles dealing with case studies and the latest technology. Contributions to a Special Topic in First Break can be sent directly to the editorial office (firstbreak@eage.org). Submissions will be considered for publication by the editor.

February

Digitalization / Machine Learning

March

Reservoir Monitoring

April

Unconventionals and Passive Seismic

May

Global Exploration Hotspots

June

Securing a Sustainable Future Together

It is also possible to submit a Technical Article to First Break. Technical Articles are subject to a peer review process and should be submitted via EAGE’s ScholarOne website: http://mc.manuscriptcentral.com/fb

July

Modelling / Interpretation

August

Near Surface Geo & Mining

September

Reservoir Engineering & Geoscience

October

Energy Transition

November

Marine Acquisition

December

Data Management and Processing

You can find the First Break author guidelines online at www.firstbreak.org/guidelines.

More Special Topics may be added during the course of the year.

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Streamlining energy and production data management from field to processing Kristy DeMarco1* addresses key challenges in data management and provides useful solutions for data aggregation, transportation and activation. Abstract Exploration and production companies continuously strive to find and develop new recoverable hydrocarbon reserves. For every segment, activating critical data faster can give companies a competitive advantage. From limited infrastructure, barriers to scale, limited transmission bandwidth and network capacity, and improved clock time from data acquisition to processing, as well as the difficulty of aggregating, transporting, and activating large data sets, presents a clear challenge to exploration and production companies’ turnaround time. As the evolution of data acquisition technologies has outpaced approaches to data management in the field, this problem has only compounded over time. This paper will address the key challenges to energy industry data management and provide useful solutions for data aggregation, transportation and activation. As data becomes more and more of a significant strategic resource for exploration and production enterprises, it will become vital that service providers discover a new approach to acquiring and processing data in the field. This solution will need to be a cost-effective, efficient, and scalable physical data transfer solution that is built both for mass-capacity storage in the field and frictionless physical transfer to any cloud service. I. Introduction Exploration and production companies continuously strive to find and develop new recoverable hydrocarbon reserves. For every segment—upstream, midstream, and downstream—activating critical data, faster, offers operators and service providers alike a critical advantage over their competitors. However, before you can get to the decision making and delivery stage, intense data workflows starting at the field level have always presented significant roadblocks. As data becomes a more significant competitive advantage for exploration and production companies, it will become vital that service providers discover a new approach to acquiring and processing data in the field. This physical data transfer solution will need to be cost-effective, efficient, and scalable, built both for mass-capacity storage in the field and frictionless physical transfer to any cloud service.

Seagate Technology

Corresponding author, E-mail: kristy.demarco@seagate.com

The importance of data to energy industry operations

Whatever the energy project, when it comes to data time is of the essence. Whether used to assess subsurface geologic features, changes in dynamic physical conditions of a well, or to identify potential risks that may exist downhole or in surface facilities, it is crucial to capture and process large digital information datasets with the highest possible speed and integrity. Knowledge-based data analysis enables improved understanding of physical subsurface processes and physical conditions to help drive important decisions in well operations, future field development designs, and field development execution.

Figure 1 PXGEO 2 14-streamer seismic vessel.

Figure 2 Ocean bottom node survey.

DOI: 10.3997/1365-2397.fb2023100

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Jurick added that this challenge will only further complicate in-field operational decision making as more and more sensor types collect larger and larger sets of data, increasing the strain on data storage systems and rendering bandwidth limitations and network bottlenecks a limiting factor in some regions where infield communications infrastructure is still in development. ‘This will become very difficult to manage in the long term,’ Jurick emphasised. ‘I also worry about the integrity of data on individual hard drives under tough environmental conditions of heat, humidity, dust, and dirt, not to mention driving data drives out of the field through some very rough terrain and then travelling long distances back to the central processing centre.’ To keep up with these ever-evolving data management challenges, the industry will need to compress the overall time from data acquisition to customer delivery without sacrificing scalability and data security at the edge.

Figure 3 PXGEO’s ocean bottom node survey.

II. Key challenges to data workflows • In 2020, the global energy industry generated 5 exabytes of data every two days, which is equivalent to the total amount of data created by humans until 2003. – Department of Industrial and Systems Engineering, Hong Kong Polytechnic University, Hong Kong • The big data market in the energy sector is poised to grow at a CAGR of more than 19% between 2021 and 2025 – Research from Technavio Figure 4 Lyve Mobile Array.

Dana Jurick, EVP and GM of Neubrex Energy Services (US), which serves the Americas energy and power industries with advanced fibre optics-based monitoring and surveillance technology, explained the importance of reliable data acquisition and delivery to professionals like petroleum engineers, geothermal energy engineers, and carbon capture and sequestration (CCS) engineers and geoscientists. These professionals rely on many data types to make information-driven assessments that help them to recommend and make decisions to improve energy production, execute safe and environmentally sound operations, and run operations more efficiently and effectively. ‘The industry collects a lot of raw data in the field - hundreds of terabytes, sometimes up to a petabyte of different digital data types on a single project throughout its life cycle. Our engineers and data acquisition systems will stay in the field for weeks and months at a time, collecting data. It’s a very high-volume, data-intensive acquisition process,’ Jurick said. ‘Our customers require full analysis of these large sets of data, and they need it in a very timely manner. But it’s a challenge to manage such large data sets when we’re that far out on the edge.’ The difficulty of moving very large raw data sets to a processing centre where they can be signal processed and analysed presents a clear challenge to turnaround time. It can take weeks for a data acquisition project to be fully completed in remote field locations, where network bandwidth is often very limited. Without the ability to physically transfer ever-increasing data sets back to a central processing centre, customers cannot reap actionable insights from that data. In effect, as data ages, it loses value. 52

FIRST

BREAK

VOLUME

DECEMBER

2023

The biggest impediments across data workflows come down to scalability, mobility, and affordability. For service providers relying on traditional data management strategies, such as magnetic tapes, to store and transport the data they generate in the field, it can take weeks for a data acquisition project to be fully completed in remote field locations. Because those datasets have grown from terabytes to petabytes, over the years—and quadruple in size during processing and interpretation—many companies have realised the need to reimagine the way their months-long field projects aggregated, transported, and activated data from the field to final delivery. Field data generally arrives unstructured (e.g. raw data or written reports) or semi-structured (e.g. models and simulations) and is typically collected in raw formats, using industry standards (SEGD, in the case of field seismic). This data, however, is of little use until it is promptly extracted; processed using high-performance computing (HPC), in order to produce various products for geophysical and geological interpretation; delivered; and refined. Processing in the field has now advanced to a point that various stacking and fast-track products can be produced before the raw data reaches an offsite data centre. However, the bulk of pre-stack migration, angle stacks, and depth-converted products are still produced in data centres before shipping to the end client. This necessitates moving data from the field to the core, efficiently and securely, as it isn’t until then that product data (typically SEGY and SEGY gathers) — often in the range of 20-30 product types — can be delivered to an operator or service provider partner for reformatting for use within the operator’s preferred software suite.

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

scalability at the edge and frictionless physical data transfer from edge to data centre to cloud, where insights from these data sets can only continue to create value for their users. Because time to insight for analytics is such a crucial part of success in the energy sector — simple, secure, and efficient data mobility is more important now than ever. New strategies for overcoming data logistics roadblocks

Figure 5 Simple, Efficient and Secure DTaaS.

Before energy companies can transform data into a deliverable, they must first establish a physical data storage strategy that guarantees effortless and affordable scalability in the field, frictionless physical data transfer from field to data centre to cloud, as well as coordinated and efficient data management along the way. Data management pain points

When looking across upstream use cases and data workflows today, it’s clear that data capture, transfer and analysis are hindered. Put simply, the critical data challenges across the industry include: • Inability to scale and reduce footprint of data storage infrastructure in the field • Raw data sets too large to be transmitted over satellite or 5G • Difficulty offloading and transferring data from subsurface imaging technologies in remote locations and harsh environments to a processing centre or near site location • Delays when ingesting data to the cloud for faster access for analysis, visualisation and business decision making A glaring pain point across the industry use cases is limited physical edge infrastructure that’s unable to accommodate growing data needs. The rapid deployment of internet of things (IoT) devices and other sensors across energy operations generates massive datasets. The reality is that on-premise edge storage devices and data centres have reached their limits. Due to limited storage space onsite, many operations have had to purge data to make room for new datasets, leaving valuable insights behind. Quickly scaling up or down in storage capacity to match current energy operations is also difficult, as IT teams often budget and make storage infrastructure investments designed to last for several years. These challenges are pushing energy companies to look for data-storage alternatives. III. Imagining a path forward Research indicates that, in order to employ intelligent field development and management solutions in remote settings, the industry must move with the times to future proof its edge operations. It is vital that energy project operations embrace a truly disruptive data storage and transfer strategy, one which bypasses bandwidth limitations and guarantees effortless and affordable

The question is: How do operators and digital oilfield service providers overcome these data challenges? To accomplish this, many organisations are incorporating Data Transfer as a Service (DTaaS)-based data management strategies into their field-tocloud workflow, which not only offer them access to the tools they need for fast ingestion of mass data sets for storage, back-up, and archive, but leaning on a data storage provider to import that data to the multi-cloud gives teams time to focus on the work ahead of them, all while making sure they only pay for the storage devices they need, when they need them. Unlock the benefits of DTaaS After all, the cornerstone of a good data management workflow is data flexibility — this means having your data in the right place, at the right time, and with the right accessibility. In a world where exploration and production efforts are producing more data than ever, the issue isn’t whether we have enough space to store all of it — it’s getting that data moving to where it’s needed fast enough. Not only does an on-demand, consumption-as-a-service model simplify device management, preventing unnecessary

Figure 6 Mobile data centre in oil field.

Figure 7 Mobile data centre in oil field.

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

on-set IT costs, it also gives teams in the field the flexibility they need to alter the numbers of devices they deploy to complete certain projects, as storage needs change. By circumnavigating the headaches that come with owning data storage infrastructure outright, like maintenance fees and technology upgrades, DTaaS-based strategies give users the freedom to expand production, accelerate timelines, and even reallocate budgets without having to worry about data. Each project requires an entirely unique data storage infrastructure, both for data consolidation and back-up in the field and field-to-cloud data transportation. Deploying DTaaS-based

Figure 8 Lyve mobile Pelican cases in the field.

storage strategies gives teams the ability to consolidate data on location quickly and securely, using a right-sized, direct-attached, on-prem storage system. The benefits of deploying high-capacity storage arrays When it comes to managing data in the field, instead of recording seismic, survey, and surveillance data on hundreds of tapes, service providers can significantly reduce their hardware footprint by recording data to storage arrays, substantially increasing disk space while reducing IT support requirements and maximising field data storage resiliency. By leveraging high capacity, ruggedised, and securely encrypted storage arrays to create secure copies of raw data so that, before data is manipulated or moved, one copy of the raw data is securely archived, users can keep up with simultaneous recording, back-up, and processing workflows at the same time that they maintain a small data storage infrastructure footprint, making physically shuttling data from field to ingest logistically straightforward. Additionally, using securely encrypted drives that implement a user key management service (KMS) layer protects data both in the field and in transit for robust data security against both cyber and physical threats. In the end, bypassing limited field infrastructure, bandwidth limitations, and data management silos by organising all data consolidation, transportation, and processing using one device not only improves time to data but ensures upstream companies have a competitive advantage over their competitors, delivering essential information to stakeholders quickly and maintaining data infrastructure scalability, accessibility, and security throughout the process. The benefits of embracing a DTaaS strategy for cloud import Additionally, these operations should take advantage of a subscription-based data transport service to simplify data management logistics for physical transfers and ingest to the cloud. By utilising a cost-effective, Data-Transfer-as-a-Service (DTaaS) model, operators benefit from just-in-time device delivery to and from any location, unburdening data management logistical overhead and maximising existing budgets by ensuring individual projects only pay for the hardware they need. Leaning on the additional strengths of high-capacity, portable devices to aggregate mass datasets in the field, operators can expect optimised performance in remote environments, robust data security, as well as scalable and efficient in-field storage for business insights.

Figure 9 Lyve mobile Pelican cases in the field.

Impact of deploying a DTaaS Strategy – PXGEO

As an innovative marine geophysical service provider, PXGEO provides solutions that bring together the latest in seismic data acquisition techniques to collect superior quality data in challenging environments. However, as PXGEO developed its turnkey technology over the years, it found that the amount of data its devices generated exceeded their long-term storage capabilities. To ensure their teams in the field were able to aggregate, transport, and activate data, PXGEO looked to incorporate a DTaaS-based data management strategy into its field-to-cloud workflow.

Figure 10 Lyve mobile Pelican cases in the field.

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

As a result, PXGEO aggregated, stored, processed, and mobilised more data. For the geoscience team in the field whose job it is to quality control and package the data prior to delivery, this meant they were able to quickly pass data to its specialised partners, accelerating the delivery time by 3-4 times. After the data was fully processed and interpreted, PXGEO used a cloud import service to quickly transfer that data to a designated Amazon S3 bucket for client delivery, which meant their client was able to receive the final project data five weeks before deadline. Because PXGEO embarked on a massive workflow shift away from tapes and towards a digital-first workflow, it was able to modernise its entire process, unlocking the data logistics challenges that slowed them down before and significantly accelerating their data delivery schedule. Impact of deploying a DTaaS strategy – Neubrex Energy Services

Neubrex uses distributed fibre-optic sensing technology to make detailed measurements of 2D and 3D physical features and dynamic processes associated with underground well operations. As a critical part of the energy development and carbon management market, it collects and processes this data in near real time at the field sites, allowing it to deliver insights and answers fast. In the past, Neubrex used individual hard drives and small RAID systems to store on-site data. This made initial processing and filtering of top-line data difficult. To accelerate the data collection and hand-off process, it turned to a data storage provider to offer integrated, high-capacity data transfer solutions for its edge-to-cloud workflows. As a result, Neubrex is getting data to the cloud up to 10 times faster and reducing processing time by 30-40%.

Not only does this solution allow Neubrex’s field engineers to quickly write data from their proprietary data acquisition systems onto high-capacity data storage arrays, offering them an on-location data centre that is no longer at the mercy of network bottlenecks and dependencies, it also makes it easier (and safer) for them to physically move data by eliminating the risks and confusion that come with using individual hard drives. Factor in the unit’s secure, ruggedised transport case, and teams can seamlessly move data by vehicle to wherever it needs to go next. Key takeaways Each of the above case studies prove that it is best practice to partner with a data storage expert that offers a high-capacity physical data transfer solution, designed to accommodate mass data physical transfers from field to core, and scale them alongside growing data needs. To optimise large-scale data movement, such as in seismic mapping applications, upstream exploration and production companies should select devices that lift data out of the field and rapidly transport it to a data centre or headquarters for redundancy and back-up without loss, corruption, or breach. From a hardware perspective, investing in cost-optimised data storage hardware, software, and services enables users to move petabytes of data from one location to the next for immediate processing and analysis, as well as long-term data archival, accessibility, and security. Working with a data transfer partner whose services help teams quickly and easily transfer their data from any endpoint, field, or core location to the cloud of their choice also helps operators to digitalise and thereby optimise their legacy data — saving costs on the front end and better organising data archives on the back end.

February 29 + March 1 EXHIBITION CENTER OFFENBURG www.geotherm - offenburg.de/en

GeoTHERM.indd 1

FIRST

BREAK

VOLUME

07/11/2023 16:29

DECEMBER

2023

25-27 MARCH 2024 I PARIS I FR ANCE

DELIVERING BETTER ENERGY IN A TRANSFORMING WORLD

D R I V E T H E D I G I TA L T R A N S FO R M AT I O N

OF THE ENERGY SECTOR

DIGITALIZATION & INSIGHT GENERATION

APPLIED ANALYTICS & MACHINE LEARNING

THE DIGITAL SUBSURFACE

DIGITALIZATION AS A LEVER FOR ENERGY TRANSITION

SUBMIT YOUR

ABSTRACTS! DEADLINE: 10 DECEMBER 2023

HOSTED BY

W W W. E A G E D I G I TA L . O R G

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Large-scale industrial deployment of machine learning workflows for seismic data processing Julien Oukili1*, Jyoti Kumar1, Jon Burren1, Steve Cochran1, Martin Bubner1, Denis Nasyrov1 and Bagher Farmani1 discuss the benefits of implementing deep neural networks for certain steps of seismic data processing on data examples from around the world. Introduction Seismic data processing is often thought of as a non-deterministic journey where signals are incrementally separated from noise, with the inherent challenge of the noise being, in a lot of cases, distinctly different and separable from the signal but in some cases remarkably similar and difficult to distinguish. At the numerous steps of data preparation, migration and post-processing, the opportunities to improve quality are many, and so are the risks of harming the desired signal. Even evaluating the results remains a challenge, especially in the early steps of an imaging flow. Machine learning (ML) applications have caught the interest of many, with the desire for faster and more thorough quality control (QC), more reliable processes, or simply to automate some of the more mundane and highly repetitive tasks of the geophysicists working on increasingly larger amounts of data. At the same time, the accelerated energy transition has put increased pressure on geophysicists to get the most out of each seismic dataset which are often used for multi-purpose subsurface investigations (e.g., both oil and gas exploration and carbon storage screening). In recent years many case studies have demonstrated the potential of ML methods for processing, QC and interpretation of seismic datasets. However, often these have come with caveats about the use of the results, especially if the actual data being processed significantly differs from those used for training the ML algorithm. In this paper, we look at two different types of machine learning use cases where the neural network workflows have been deployed to many marine seismic processing projects, at large scale, and with special attention given to the challenge of highly varying data characteristics, geological and geographical environments. The application examples are focused on problems where signal and noise separation were critical, either for robustness of subsequent processing steps or for the quality of direct interpretation efforts. Conventional data processing flows broadly follow a strategy of 1) parameter testing, often on a very limited subset of data, 2) production set up on the full dataset, and 3) QC, which potentially reveals the needs to change or even rerun current processing steps with different parameterisation. Often the final

PGS

Corresponding author, E-mail: julien.oukili@pgs.com

parametrisation choice represents a compromise that is reached after a considerable amount of time has been spent on trying to optimise the said processing steps on a limited assessment of the data. Using ML methods has allowed us to speed-up specific processing steps and has allowed the geophysicists to concentrate on improving the resulting data quality rather than spending time on optimising processes and parameters. We are sharing some of our main learnings from routinely using ML methods in a number of specific applications scenarios over the last two years. Use case 1 – Noise removal in raw data prior to wavelet processing The first use case focuses on the very early stage of seismic data processing: denoising of raw recorded data. Applying sufficient denoise prior to the first multi-channel processes is crucial to avoid spreading noise or enhancing it, making noise more apparent to desired signal and difficult to remove. Noise characteristics can also vary greatly between seismic surveys, as well as within surveys, sail lines, individual shot records and between the different types of recording sensors in the case of multi-sensor acquisition systems. Pressure and particle motion data from multi-sensor streamer acquisition can be combined to separate the wavefield into up- and down-going components (Carlson et al., 2007). To guarantee the generation of high-quality up-going and downgoing wavefields, the noise from both records must be attenuated before the data are combined. ML tools have increasingly become the method of choice in a drive to increase automation and improve output consistency. Farmani et al. (2023) presented specific workflows for pressure and particle motion data that employ deep learning to suppress the noise in the records efficiently. Real image denoising network (RIDNet), a convolutional neural network, sits at the core of the proposed workflows. The method, presented here, uses a single RIDNet model with a specific structure for both pressure and particle motion recordings, as opposed to machine learning-based workflows presented by Farmani and Pedersen (2020a; 2020b; and 2022), that attenuate incoherent noise in

DOI: 10.3997/1365-2397.fb2023101

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 1 Shot gather from deep-water Eastern Mediterranean, offshore Egypt: RIDNet application on hydrophone data (top), particle motion sensor data (middle) and wavefield separation step (bottom). The raw hydrophone data are contaminated with towing noise (linear events at near offsets, left side) and turn noise at far offset (right side). The particle motion data have a 10 Hz low cut filter and initially shows visible noise towards the far offsets. The denoise was very stable in high- and low signal-to-noise areas. Furthermore, the difference of the particle motion data shows noise being removed at near offsets where signal and noise seem equally strong in amplitudes as well.

the bandwidth where most of the noise exists. As a result, the core components of the processes are greatly streamlined and strikingly comparable for the two types of sensor recordings. A network built on the RIDNet architecture is used to attenuate incoherent noise on pressure and particle motion records within the bandwidth of interest. Additional processing techniques can be used to attenuate noise that is present outside of the RIDNet application bandwidth. The approach presented has been successfully applied to numerous datasets worldwide with high consistency and higher efficiency and has enabled us to produce denoised data for both pressure and particle motion sensors very quickly after the data have been acquired offshore. Successful noise attenuation using RIDNet applications The first field data example is from a 2023 seismic survey, part of multi-client campaign in the Eastern Mediterranean Sea, offshore Egypt. The data were acquired using a triple-source configuration and 12 10,000m-long multi-sensor streamers separated by 150 m. Although the sea was generally calm during 58

FIRST

BREAK

VOLUME

DECEMBER

2023

the acquisition, some sail lines contained significant noise. Shot gathers from one of the noisy sail lines are shown in Figure 1. The top image displays the results of applying machine learning denoise to the pressure dataset using the RIDNet methodology. It is evident that noise has been efficiently attenuated, and the difference plots demonstrate that there has been no primary leakage during the process. The outcome of the particle motion dataset is shown in the middle image. Finally, the bottom row shows the upgoing wavefield generated through the wavefield separation process using pressure and particle motion input after ML denoising, as expected, without receiver ghost interference and without artifacts. A second example is shown in Figure 2. This dataset was acquired in 2023 as part of multi-client campaign offshore Malaysia. Again, a triple-source configuration was used and 12 9000m-long multi-sensor streamers separated by 112.5 m. As observed in the previous example, the machine learning denoise workflow effectively attenuates the noise in pressure as well as particle motion datasets, which in turn manage to produce a high-quality upgoing wavefield. Even if minor residual noise can still be observed, it has not deteriorated through

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

the wavefield separation process and hence can be dealt with effectively at a later stage. The final example is from a dataset acquired in 2015 using also multi-sensor streamers in the Faroes Shetland Basin, offshore UK. This survey was acquired using a dualsource configuration and 10 7050m-long cables separated by 100 m. The data were reprocessed in 2023 as part of a regional multi-client rejuvenation project. Machine learning denoise once more worked admirably for both the pressure and particle motion datasets (Figure 3), demonstrating that even older datasets can benefit from the new denoising technology. The machine learning denoise workflow described in these use-case examples has shown good and robust performance in attenuating common noise on datasets from all around the world and enabled generation of high-quality separated wavefields. And since the presented method using RIDNet is largely automated it only took a few days on each project to set up the respective wavefield separation workflow and apply it to several thousand square kilometres reliably despite the high variations in input data characteristics.

A similar RIDNet workflow based on a lighter network, more focused on signal and noise classifications, is currently being trialled to identify residual noise which might require further attention and investigation. Design work is continuing that should allow us to graphically visualise the results of this new residual noise detection workflow so that geophysicists can make quick and informed decisions. Use case 2 — Noise attenuation in the postmigration image domain In the following section we shift our attention from raw data denoise to the late stages of a standard seismic data processing flow, post-migration image denoise, where the starting assumption is fundamentally different from the previous application since the desired signal is expected to be already in focus (i.e., adequately imaged). Seismic images are often contaminated by migration noise, sometimes referred to as migration smiles, swings, artifacts or defects. This noise is generated when assumptions made by the migration algorithm begin to break down, e.g., the midpoint, offset and azimuth sampling requirements of the input data are

Figure 2 Shot gather from shallow water Sarawak, offshore Malaysia: RIDNet application on hydrophone data (top), particle motion sensor data (middle) and wavefield separation (bottom). Compared to the examples on Figure 1, the raw hydrophone data show more towing noise at far offsets (right side) and spurious noisy traces. The particle motion data show more noise in the form of vertical stripes, likely to be caused by more active birds (instruments which steer the streamers).

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 3 Shot gather from Faroe Shetland Basin, offshore United Kingdom: RIDNet application on hydrophone data (top), particle motion sensor data (middle) and wavefield separation (bottom). Very strong noise is present in the raw hydrophone data, as well as weaker linear noise, both well attenuated through the denoise, which reveals a lot of signal. The amplitudes of the noise on the particle motion data are also higher than in the examples of Figure 1 and 2.

not adequate (Long et al., 2006). The assumptions may break locally when the seismic wavefield propagates through complex media or is exposed to strong amplitude (reflectivity) contrasts. Data regularisation and filtering of aliased energy prior to migration can mitigate effects of the sampling challenges (e.g., Schonewille 2000; Chemingui and Biondi, 2002). The effects of under-sampled field data can be further compounded by subsurface complexity. Preserving the amplitudes accurately is a key objective of any imaging exercise and care needs to be taken not to alter amplitudes in any noise removal process especially at the post-migration stage. Traditionally, applying any noise removal post migration has taken a lot of time and effort. Klochikhina et al. (2021) described a machine learning method for tackling this problem. In their paper they demonstrate how a convolutional neural network (CNN) with U-net architecture was trained using synthetic data examples and the results demonstrated on some field data examples. This ML-based denoise technique has now been widely adopted and applied to a wide range of different datasets as part of commercial imaging projects. It has proved very effective in 60

FIRST

BREAK

VOLUME

DECEMBER

2023

removing or reducing migration noise, suggesting the training data were sufficiently representative. Application scenario 1 — Structural Imaging and post-migration clean-up A field data example from offshore Newfoundland demonstrates the ability of the neural network to attenuate noise on the migrated section, making the image more easily interpretable and improving the structural picture (Figure 4). A second field data example from offshore Norway (Figure 5) represents a more challenging scenario where some of the migration noise, especially from the Top Chalk reflection, does interfere with shallower structures of similar dips in heavily faulted formations. In this application example the CNN noise removal had to be limited to a horizon-bound interval and was complemented by structurally conformable filtering. This resulted in a considerable improvement in signal-to-noise ratio without detrimental image distortion. We postulate that more traditional denoise tools would have failed in preserving the smaller details if they had been designed to achieve a similar noise reduction.

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Application scenario 2 — Supporting quantitative interpretation (QI) work In a further example (Figure 6) we illustrate how the CNN denoising capabilities positively impact on QI workflow. The signal-to-noise ratio of the data influences the stability of any attributes derived from the data, so a typical QI workflow necessitates the inclusion of steps to precondition the pre-stack and stack-domain data to address unwanted noise. This can be a time-consuming process, depending on the quality of the input data. Including the neural network in the data preparation workflow provides the required noise suppression, to the benefit of the QI attribute derivation. Figure 6 shows the impact on the AVA gradient attribute; the level of noise is clearly reduced while the signal continuity is maintained. Considerable uplift is also seen on the relative P-wave impedance attribute computed from the same data.

Figure 4 Migrated section before (top) and after (bottom) application of the neural network. The migration artifacts are more prominent above strong interfaces and could be in the worst case interpreted as faults.

Application scenario 3 — High-end imaging using least squares migration workflows Incorporating the same neural network methodology into highend imaging workflows has helped overcome challenges with the migration noise frequently observed in least-squares migration (LSM) results. LSM solutions can be divided into two main categories: data-domain and image-domain solutions. Both approaches are powerful techniques for overcoming challenges with subsurface illumination and image blurring and ensure more reliable amplitude preservation in the imaging process,

Figure 5 Migrated cross-section (left column) and time slice (right column), before (top row) and after (middle row) targeted image denoise and difference (bottom row). Contrary to the example of Figure 4, the migration artifacts above the Top Chalk (blue arrow) are not easily distinguished from the shallower complex geology. However, the artifacts have a more distinct lineation pattern on time slices which is taken just above the Top Chalk. Features which are parallel to the noise pattern on time slices have been well preserved.

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 6 Left side: AVA gradient attribute calculated on data before (top) and after (below) migration noise attenuation applied to angle-stacks, using the neural network. Right side: relative Ip attribute calculated on data before (top) and after (below) migration noise attenuation applied to angle-stacks, using the neural network.

important for AVA studies. Data-domain solutions have a lot in common with full waveform inversion (FWI), involving forward modelling, followed by comparison between modelled and observed data and the back-propagation of residuals, resulting in the estimation of reflectivity instead of velocity. This approach to LSM can prove computationally expensive, especially for higher frequencies. Image-domain LSM solutions derive and apply corrections to a conventional migration. If the signal-to-noise ratio of the conventional migration is poor, the LSM result will be noisy,

Figure 7 Raw migration (top) showing clear evidence of migration noise. The output of the least-squares migration workflow (bottom), incorporating the neural network, shows improved horizontal and vertical resolution with minimal contamination from the migration noise on the raw migration.

FIRST

BREAK

VOLUME

DECEMBER

2023

therefore appropriate data preconditioning and regularisation (as part of the inversion) is critical for this methodology. The example from the Campos Basin, Brazil (Figure 7) shows the results of an image-domain LSM. The inclusion of the neural network-based denoise in the least-squares workflow has ensured that the enhanced image is not contaminated by migration noise. Some takeaways for post-processing of seismic images The application examples above illustrate the potential of an uplift in image quality that benefits the subsequent processing and use of the data. An equally important consideration is the time needed to address the migration noise; using the neural network significantly reduces the preparation time compared to more traditional denoise methods. The user is considerably less encumbered with testing parameters and tuning workflows using the neural network, freeing time which can be spent giving greater attention to more advanced geophysical processes and analysis of QI attributes extracted from the data. Use case 3 — Diffraction event detection In the pre-migration domain, diffraction events, sometimes referred to as ‘tails’, exhibit similar characteristics to the ‘migration smile’ artifacts discussed earlier, albeit in a downward dipping fashion. Building on the similarity of these two artifact types, the same CNN U-net architecture as featured in use case 2 above was used but with an inverted time axis (i.e., free surface pointing downward on seismic traces). Contrary to the earlier example, here the diffraction ‘noise’ is actually desired signal, which can be re-inserted into the migration process to produce a sharper image of small-scale heterogeneities. In Figure 8, a near-offset volume is taken through migration before and after the diffraction identification and separation. The same migration algorithm and velocity models were used, so the images are fully consistent, although one may argue that focusing diffraction tails would require further attention. As expected, the separation is not perfect, as reflection leakage can be observed

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 8 Near offset volume from shallow water offshore Newfoundland and Labrador, Canada: conventional imaging (top) versus diffraction imaging (bottom). The workflow successfully identified both large- and small-scale diffraction patterns, though some weak reflection leakage can be seen. The separation is most impressive towards the shallow (top of the images) where strong reflections were masking weaker diffraction tails.

in the diffraction products. However, the features highlighted by the new image now appear much stronger than the background geology, especially towards the top of the section where highly reflective horizon events were masking smaller details. While the section views are very useful for understanding large fault patterns, we find the diffraction images to be rather chaotic and still noisy for small-scale features. Looking at time slices of the same migration volume (Figure 9) reveals

Figure 9 Slices through the diffraction image volume: seabed (top) and 1368 ms under seabed (bottom). Glacial features are dominating the seabed image except towards the left side of the image: the water bottom goes rapidly deeper and is therefore practically free of iceberg marks. The deeper image shows both a large network of sub-parallel faults as well as small scale polygonal faulting in the lower left side of the image.

patterns and occasionally isolated objects. The benefits of such an approach are significant when the diffraction generators are located close to reflections of similar or higher amplitudes. Furthermore, the separation, although not perfect, does not seem to suffer from amplitude footprint effects and can resolve details down to the scale of the imaging bin size. The method presented here is now being tested on various datasets to aid detailed interpretation for both deep and shallow targets, and using conventional as well as high-resolution and ultra high-resolution 3D seismic data (with approximately 1 m x 1m bin size and 0.25 ms temporal sampling). It is still early days, but so far the machine learning CNN U-net-based workflow has robustly worked for any type of 3D seismic data and any type of frequency bandwidth that it was applied to. Conclusions We have discussed several use cases of ML methods in data processing steps which are always required on any type of seismic data nowadays. The examples shown have proven our implementations to be reliable for a large variety of datasets, nearly irrespective of the region of origin and of the geophysical and geological settings. The ML workflows have either replaced complete steps or been integrated with other tools to achieve at the very least the same quality as before, but most often better. Running machine learning workflows can be compute intensive. However, access to cloud computing has all but removed any computational limits and enabled large-scale deployment of ML technology. The main benefit of employing machine learning technology is that it has allowed us to free up valuable time of the project geophysicists by shifting more of the denoising effort to the computer. This has more broadly allowed the processors to spend time on QC and quality improvements, leading to better project outcomes. The methods described in this paper incorporate a significant amount of automation. However, the role of geophysicist remains crucial, in selecting the appropriate flows, defining the FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

application range, adding adequate complementary processes, or simply discarding the results in parts of a dataset. It is important that using ML technology does not amount to black box processing and that the application user remains in full control. A natural question to ask is how far we can progress towards a fully automated data processing workflow and how quickly and whether experienced geophysicists will be replaced by machines (Brittan et al., 2021). We believe that we are still a long way away from fully automated seismic data processing and that with the continued changes to the type and characteristics of the data we record and the constant evolution of high-fidelity final imaging products, geophysicists will continue to play a critical role in executing successful imaging projects. We have demonstrated in this paper that in specific areas of data processing, fully industrial and dependable implementations of ML technology can ultimately benefit both data providers and data recipients.

are we? First EAGE Workshop on Optimizing Project Turnaround Performance, February 2021, Volume 2021, p.1-5. Carlson, D., Long, A., Söllner, W., Tabti, H., Tenghamn, R. and Lunde N. [2007]. Increased resolution and penetration from a towed dual-sensor streamer, First Break, 25, 71-77. Chemingui, N., and Biondi, B. [2002]. Seismic data reconstruction by inversion to common offset. Geophysics, 67, 1575-1585, doi: https://doi.org/10.1190/1.1512803 Farmani, B. and Pedersen, M.W. [2022]. Stepping Towards Automated Multisensor Noise Attenuation Guided by Deep Learning. 83rd EAGE Annual Conference & Exhibition, Jun 2022, Volume 2022, p.1-5. Farmani, B., Pal, Y., Pedersen, M.W. and Hodges, E. [2023]. Motion sensor noise attenuation using deep learning. First Break, 41(2), 45-51. Farmani, B., Lenses, M. and Pal, Y. [2023]. Multisensor Noise Attenuation with RIDNet. 84th EAGE Annual Conference & Exhibition, Jun 2023, Volume 2023, p.1-5 Klochikhina, E., Crawley, S., Frolov, S., Chemingui, N. and Martin T. [2020]. Leveraging Deep Learning For Seismic Image Denoising.

Acknowledgement The authors would like to thank PGS for permission to publish this paper.

Klochikhina, E., Crawley, S., and Chemingui, N. [2021]. Seismic image

References

Long, A., Fromyr, E., Page, C., Pramik, W., Laurain, R., and Buchan, I.

First Break, 38(7), 41-48. denoising with convolution neural network., First International Meeting for Applied Geoscience & Energy

Bekara, M., and Day, A. [2019]. Automatic QC of denoise processing

[2006]. Multi-Azimuth and Wide-Azimuth lessons for better seismic

using a machine learning classification. First Break, 37 (9), 51-58.

imaging in complex settings. SEG 76th Annual Meeting, New Orleans. Schonewille, M.A. [2000]. Fourier reconstruction of irregularly sampled

Brittan, J., O’Driscoll, R., Walpole, J. and Cobo Y. [2021]. Root Butlers,

seismic data: Ph.D. dissertation, Delft University of Technology

Flying Cars, Automated Seismic Processing and Imaging – How near

Benefit from an additional journal subscription! NEAR SURFACE GEOPHYSICS

FROM SEISMIC MECHATRONICS

Near Surface Geophysics

issn 1569-4445 (print) issn 1873-0604 (online)

Volume 19 · Number 1 · February 2021

Basin Research

Geophysical Prospecting

S E C O N D E A G E D I G I TA L I Z AT I O N CONFERENCE AND E XHIBITION

Volume 33 . Number 1 . February 2021 . http://www.wileyonlinelibrary.com/journal/bre

Published in conjunction with the European Association of Geoscientists & Engineers and the International Association of Sedimentologists

Electromechanic vibrating exploration seismic source for P and S wave excitation. www.geomatrix.co.uk/land-products/seismic/seismic-exploration-systems/lightning

3 Editorial P. Tsourlos 7 A geostatistical Markov chain Monte Carlo inversion algorithm for electrical resistivity tomography M. Aleardi, A. Vinciguerra and A. Hojat

WE ARE THE ONLY EUROPEAN RENTAL SUPPLIER

Volume 69 Number 2 February 2021 pp 267–490

27 Determination of the resistivity distribution along underground pipes in urban contexts using galvanic and capacitive methods S. Rejkjær, C. Finco, C. Schamper, F. Rejiba, A. Tabbagh, J. König and T. Dahlin 43 Cross-borehole geoelectrical time-lapse monitoring of in situ chemical oxidation and permeability estimation through induced polarization T. Bording, A.K. Kühl, G. Fiandaca, J.F. Christensen, A.V. Christiansen and E. Auken

THE ENERGY INDUSTRY I S G O I N G D I G I TA L

59 Transient electromagnetic inversion based on particle swarm optimization and differential evolution algorithm R. Li, N. Yu, R. Li, Q. Zhuang and H. Zhang

AND SO ARE WE!

73 Estimation of earthquake local site effects using microtremor observations for the Garhwal–Kumaun Himalaya, India R. Pudi, P. Roy, T.R. Martha and K.V. Kumar 95 Detection of urban underground cavities using seismic scattered waves: a case study along the Xuzhou Metro Line 1 in China J. Zhang, S. Liu, C. Yang, X. Liu and B. Wang

MARINE PRODUCTS

SOFTWARE

Marine Acquisition

Visit us online for survey planning tools; equipment training videos; magnetic forward modelling program and comprehensive rental pool of engineering geophysical instruments – worldwide shipment.

Pearl changes the rules of

EAGE NEWS EAGE 2023 technical programme shapes up

Ocean Bottom Seismic, oﬀering

TECHNICAL ARTICLE Machine learning-assisted interpretation

better survey designs for exploration,

INDUSTRY NEWS UK launches big licensing round

UK OFFICE 20 Eden Way, Pages Industrial Park, Leighton Buzzard, Beds LU7 4TZ. UK.

appraisal, development, production optimisation and monitoring.

T: +44 (0)1525 383438 E: sales@geomatrix.co.uk

VOLUME 19 · NUMBER 1 · FEBRUARY 2021

SPECIAL TOPIC LAND PRODUCTS

NETHERLANDS OFFICE NOW OPEN • Weena-zuid 130, 3012NC Rotterdam • Fully managed European shipping

shearwatergeo.com

109 A numerical dispersion-suppressed method for shallow seismic migration Y. Liu, Z. Li, J. Wang, M. Sun and Q. Liu

SUBMIT YOUR

ABSTRACT!

WSA1025261 Geomatrix-CMKY-280x210.indd 1

ISSN 1365-2478 (Online)

Scan the QR code for more info!

269 Kirchhoff pre-stack depth scalar migration in a simple triclinic velocity model for three-component P, S1, S2 and converted waves V. Bucha 289 Order-controlled closed-loop focal beams and resolution comparison of primary and multiple reflections for seismic acquisition geometries W. Wei and G. Blacquière 307 Seismic data interpolation using deep learning with generative adversarial networks H. Kaur, N. Pham and S. Fomel 327 Unified elimination of 1D acoustic multiple reflection E. Slob and L. Zhang 349 Elastic wave-mode separation in 2D transversely isotropic media using optical flow X. Wang, Y. Lu and J. Zhang 372 Stress prediction and evaluation approach based on azimuthal amplitude-versus-offset inversion of unconventional reservoirs B. Du, G. Zhang, J. Zhang, J. Gao, X. Yong, W. Yang, R. Jiang, S. Wang, H. Li and E. Wang 388 The impact of seismic inversion methodology on rock property prediction M. Martinho, T. Tylor-Jones and L. Azevedo 404 Elastic properties of a reservoir sandstone: a broadband interlaboratory benchmarking exercise A. Ògúnsàmì, I. Jackson, J.V.M. Borgomano, J. Fortin, H. Sidi, A. Gerhardt, B. Gurevich, V. Mikhaltsevitch and M. Lebedev 419 Deep learning assisted well log inversion for fracture identification M. Tian, B. Li, H. Xu, D. Yan, Y. Gao and X. Lang 434 Subsurface structures of the Xiaorequanzi deposit, NW China: new insights from gravity, magnetic and electromagnetic data A. Shaole, Z. Zhixin, Z. Kefa and W. Jinlin 448 Denoising of magnetotelluric data using K-SVD dictionary training J. Li, Y. Peng, J. Tang and Y. Li 474 Improved controlled source audio-frequency magnetotelluric method apparent resistivity pseudo-sections based on the frequency and frequency–spatial gradients of electromagnetic fields M. Zhang, C.G. Farquharson and C. Liu

Petroleum Geoscience

Geoenergy

EAGE EUROPEAN ASSOCIATION OF GEOSCIENTISTS & ENGINEERS

1 8 -2 1 O C TO B E R 2 0 2 1 I V I E N N A I AU S T R I A Editors: Atle Rotevatn (University of Bergen) Kerry Gallagher (University of Rennes) Peter Burgess (University of Liverpool) Cari Johnson (University of Utah) Craig Magee (University of Leeds) Nadine McQuarrie (University of Pittsburgh)

www.nearsurfacegeophysics.org

www.geomatrix.co.uk

water1122.indd -FB22 November_COVER.indd 1 1-3

69 (2) (2021) ISSN 0016-8025 (Print)

http://www.geophysicalprospecting.org

Volume 69 Number 2 February 2021 Geophysical Prospecting

F I R S T B R E A K V O L U M E 4 0 I I S S U E 11 I N O V E M B E R 2 0 2 2

LIGHTNING: SEISMIC VIBRATOR VO L U M E 4 0 I I S S U E 11 I N OV E M B E R 2 0 2 2

W W W. E A G E D I G I TA L . O R G

05/01/2021 17:15

06/10/2022 18:09

14/10/2022 10:43

DIGI21 GPR 69.2 V2F (February .indd 1 2021) cover.indd 1

04/01/2021 15:06 15:05

As an EAGE member, you have free online access to a journal of your own choice. But did you know that it is also possible to subscribe to other EAGE scientific journals for an additional fee? FOR MORE INFORMATION PLEASE VISIT EAGE.ORG/MEMBERSHIP

Journal Scientific journals.indd 1 FIRST

BREAK

VOLUME

DECEMBER

2023

09/11/2023 09:41

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

How the latest SEG-Y revision will improve data management Jill Lewis1*, Shawn New 2, Joel Allard3 and Victor Ancira1 explain how the SEG-Y update to SEG-Y_r2.1 is enabling auto-access to historic data. Introduction The SEG Technical Standards Committee has issued a new SEG-Y revision: SEG-Y_r2.1 is going through ratification and will enable the application of artificial intelligence (AI) and machine learning (ML). The data format revision now provides random access and has been designed with cloud requirements in mind. This update has been designed to capture the knowledge of the data managers and will enable the ability to auto-read the data. Over the years, the Technical Standards Committee of the SEG has strived to provide new versions of their highly successful formats, especially SEG-D and SEG-Y. These updates are necessary to ensure that the format keeps up with both developments in the industry and changes in IT. The SEG formats have been in use since 1975 and represent one of the only computer-based scientific formats still in use from that era. As an industry, we are almost certainly unique in being able to read mass and multiple volumes of data from almost 50 years ago straight into our modern computing systems. Although this is admirable from a point of longevity, it is very poor from a performance point of view. There is also the issue that in 1975 there was very little computer disk or memory, marine surveys were a rarity and there was no 3D data. One can therefore see that over the years the industry and expectations have changed and the SEG has made enormous efforts to respond to that, but the industry has mainly rebuffed these efforts with very poor format adoption. At the same time the industry complains that the formats are not fit for purpose, but this is mainly due to there being a lack of adoption of the latest versions. The current standards can be found here, with SEG-Y_r2.1 to be added upon ratification: https://library.seg.org/seg-technical-standards In 1999 Jill Lewis of Troika International Ltd approached the UK-based PESGB to form a sub-committee to study the updating of the SEG-Y format. On the back of this work, the SEG-Y_r1 revision was ratified and published in 2002. These meetings, mainly held in London, were attended by upward of 200 interested parties with the whole industry wanting to update their format use. A few highlights of the update were the provision of designated positions for key trace header entries such as easting and northing, the ability to add additional textual headers to the

Troika International | 2 NextEraEnergy | 3 Kosmos Energy

Corresponding author, E-mail: jill@troika-int.com

data volume for further meta-data and an additional binary header flag to provide random access for data on disk. There was also enormous effort at this time to co-ordinate with the International Oil and Gas Producers Association (IOGP) and the Energistics group. Initially known as ‘standards within standards’, this was a successful drive to bring various standards bodies together to ensure we were using the same standards such as Units of Measure and Coordinate Reference Systems. In 1999 we were shown just how important standards are when working with other scientific groups and disciplines when the Mars probe crashed on Mars as Lockheed and Nasa had a discrepancy between the desired and actual orbit intersection altitude. As we all know, the probe crashed at quite an expense. In 2015 the SEG Technical Standards Committee once again targeted the SEG-Y format to create SEG-Y_r2.0 to provide updates required by technology advances. This update provides the ability to write data with a finer sample interval, greatly increased traces per line and ensemble, support for additional data sample formats including IEEE (64-bit) floating-point, little-endian data and much more. The industry would gain so much by updating to these format versions and utilising the work that has been carried out, but it has always been a problem regarding historic data. When we have lost what every trace represents and can only work out the key values for onward processing and interpretation, it is impossible to move the values around to the correct positions within the new revision without fear of overwriting existing values. The solution is SEG-Y_r2.1 and a method of capturing the knowledge of the user via an xml file. When historic data is used, the user will have to identify key trace headers, according to data type, to use the information successfully. SEG-Y_r2.1 provides the ability to capture that knowledge and write it into a simple xml file which is written between the binary header and the trace data. Within the binary header there are flags for format version, data type and whether the traces in the file are all the same length. This information, coupled with the xml file containing the key trace header entries, enables a computer to auto-read the dataset. The dataset at this point would have been checked for quality, binary header read, and xml file written, and therefore these files would be suitable candidates for input to artificial intelligence (AI) and machine learning (ML) processes.

DOI: 10.3997/1365-2397.fb2023102

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 1 Sample Layout file for a standard SEG-Y Revision 2 file

It is necessary to flag in the binary header that the data is SEG-Y_r2.1, that the traces are all the same length and data type. If possible, the complete mapping of the header can be put into the xml file, and this is an extract from the proposed standard Table 1 is an extract from the appendix in the proposed SEG-Y_r2.1 document which, when expanded, describes a schema for an XML file that defines the layout of the trace header entries in a SEG-Y file. With SEG-Y Revision 2, this XML file can be embedded in a SEG-Y file as Extended Textual Header stanza ((SEG:Layout)). In this context, the XML file is an alternative to the method described in r2.0 Appendix D-10: ‘Stanza for Trace Header Mapping’, and the old method may be removed from future versions of the standard. The XML file can also exist as a separate text file that sits alongside a SEG-Y file. In fact, this can be particularly useful for legacy SEG-Y data (typically SEG-Y Revision 0), to define the locations of key entries which did not have a customary location in the original standard, e.g. inline and crossline. This information would typically have been described in the EBCDIC reel header, but the XML file provides a more structured method for capturing this information. Each entry in a trace header layout is given a name, which defines the usage of the entry. Table 3 (“Standard Trace Header”) in the SEG-Y document shows the names and usage of the stand66

FIRST

BREAK

VOLUME

DECEMBER

2023

ard trace header entries with a detailed description. Generally, these are the only names that should appear in a layout, at least when it is being used for exchange purposes. If a name appears in more than one entry definition in a layout, the last entry will supersede the earlier definitions. This is of particular use when the layout contains definitions of entries in Trace Header Extension 1, where an entry can supersede an entry with the same name in the first 240-byte trace header. Note that the XML file does not define the byte ordering of the entries in the SEG-Y file. For SEG-Y Revision 2, the byte ordering used in the file can be determined from the value in bytes 3297-3300 of the file header. SEG-Y Revision 0 and Revision 1 always use big-endian byte ordering. This is a very easy to programme and low-cost method of enabling auto-read of seismic data. The big question is why have these updates not been implemented and why were they not used extensively from 2002 onwards? The SEG-D_r3.0 and SEG-D_r3.1 formats have been adopted by many companies, but the SEG-Y_r2.0 adoption is very low. The oil and gas companies we have spoken to would like to have these formats in use. Many governments around the world are requesting that data be delivered in specific format versions and still the community does not implement the updates. The Norwegian Petroleum Directorate has indicated that it may

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Table 1 XML Elements

<?xml version=”1.0” encoding=”utf-8”?> <SEG-Y-layout name=”rev2”> <desc>SEG-Y Rev 2 standard entries</desc> <entry name=”linetrc” byte=”1” type=”linetrc”/> <entry name=”reeltrc” byte=”5” type=”reeltrc”/> <entry name=”ffid” byte=”9” type=”int4”/> <entry name=”chan” byte=”13” type=”int4”/> <entry name=”espnum” byte=”17” type=”int4”/> <entry name=”cdp” byte=”21” type=”int4”/> <entry name=”cdptrc” byte=”25” type=”int4”/> <entry name=”trctype” byte=”29” type=”int2”/> <entry name=”vstack” byte=”31” type=”int2”/> <entry name=”fold” byte=”33” type=”int2”/> <entry name=”rectype” byte=”35” type=”int2”/> <entry name=”offset” byte=”37” type=”int4”/> <entry name=”relev” byte=”41” type=”elev4”/> <entry name=”selev” byte=”45” type=”elev4”/> <entry name=”sdepth” byte=”49” type=”elev4”/> <entry name=”rdatum” byte=”53” type=”elev4”/> <entry name=”sdatum” byte=”57” type=”elev4”/> <entry name=”wdepthso” byte=”61” type=”elev4”/> <entry name=”wdepthrc” byte=”65” type=”elev4”/> <entry name=”ed_scal” byte=”69” type=”int2”/> <entry name=”co_scal” byte=”71” type=”int2”/> <entry name=”sht_x” byte=”73” type=”coor4”/> <entry name=”sht_y” byte=”77” type=”coor4”/> <entry name=”rec_x” byte=”81” type=”coor4”/> <entry name=”rec_y” byte=”85” type=”coor4”/> <entry name=”coorunit” byte=”89” type=”int2”/> <entry name=”wvel” byte=”91” type=”int2”/> <entry name=”subwvel” byte=”93” type=”int2”/> <entry name=”shuphole” byte=”95” type=”time2”/> <entry name=”rcuphole” byte=”97” type=”time2”/> <entry name=”shstat” byte=”99” type=”time2”/> <entry name=”rcstat” byte=”101” type=”time2”/> <entry name=”stapply” byte=”1t03” type=”time2”/> <entry name=”lagtimea” byte=”105” type=”time2”/> <entry name=”lagtimeb” byte=”107” type=”time2”/> <entry name=”delay” byte=”109” type=”time2”/> <entry name=”mutestrt” byte=”111” type=”time2”/> <entry name=”muteend” byte=”113” type=”time2”/> <entry name=”nsamps” byte=”115” type=”int2”/> <entry name=”dt” byte=”117” type=”int2”/> <entry name=”gaintype” byte=”119” type=”int2”/> <entry name=”ingconst” byte=”121” type=”int2”/> <entry name=”initgain” byte=”123” type=”int2”/> <entry name=”corrflag” byte=”125” type=”int2”/> <entry name=”sweepsrt” byte=”127” type=”int2”/> <entry name=”sweepend” byte=”129” type=”int2”/> <entry name=”sweeplng” byte=”131” type=”int2”/> <entry name=”sweeptyp” byte=”133” type=”int2”/> <entry name=”sweepstp” byte=”135” type=”int2”/> <entry name=”sweepetp” byte=”137” type=”int2”/> <entry name=”tapertyp” byte=”139” type=”int2”/> <entry name=”aliasfil” byte=”141” type=”int2”/> <entry name=”aliaslop” byte=”143” type=”int2”/> <entry name=”notchfil” byte=”145” type=”int2”/> <entry name=”notchslp” byte=”147” type=”int2”/> <entry name=”lowcut” byte=”149” type=”int2”/> <entry name=”highcut” byte=”151” type=”int2”/> <entry name=”lowcslop” byte=”153” type=”int2”/> <entry name=”hicslop” byte=”155” type=”int2”/> <entry name=”year” byte=”157” type=”int2”/> <entry name=”day” byte=”159” type=”int2”/> <entry name=”hour” byte=”161” type=”int2”/> <entry name=”minute” byte=”163” type=”int2”/> <entry name=”second” byte=”165” type=”int2”/> <entry name=”timebase” byte=”167” type=”int2”/> <entry name=”trweight” byte=”169” type=”int2”/> <entry name=”rstaswp1” byte=”171” type=”int2”/> <entry name=”rstatrc1” byte=”173” type=”int2”/> <entry name=”rstatrcn” byte=”175” type=”int2”/> <entry name=”gapsize” byte=”177” type=”int2”/> <entry name=”overtrvl” byte=”179” type=”int2”/> <entry name=”cdp_x” byte=”181” type=”coor4”/> <entry name=”cdp_y” byte=”185” type=”coor4”/> <entry name=”iline” byte=”189” type=”int4”/>

<entry name=”xline” byte=”193” type=”int4”/> <entry name=”sp” byte=”197” type=”spnum4”/> <entry name=”sp_scal” byte=”201” type=”int2”/> <entry name=”samp_unit” byte=”203” type=”int2”/> <entry name=”trans_const” byte=”205” type=”scale6”/> <entry name=”trans_unit” byte=”211” type=”int2”/> <entry name=”dev_id” byte=”213” type=”int2”/> <entry name=”tm_scal” byte=”215” type=”int2”/> <entry name=”src_type” byte=”217” type=”int2”/> <entry name=”src_dir1” byte=”219” type=”int2”/> <entry name=”src_dir2” byte=”221” type=”int2”/> <entry name=”src_dir3” byte=”223” type=”int2”/> <entry name=”smeasure” byte=”225” type=”scale6”/> <entry name=”sm_unit” byte=”231” type=”int2”/> <extension name=”SEG00001”> <entry name=”linetrc” byte=”1” type=”linetrc8” if-non-zero=”1”/> <entry name=”reeltrc” byte=”9” type=”reeltrc8” if-non-zero=”1”/> <entry name=”ffid” byte=”17” type=”int8” if-non-zero=”1”/> <entry name=”cdp” byte=”25” type=”int8” if-non-zero=”1”/> <entry name=”relev” byte=”33” type=”ieee64” if-non-zero=”1”/> <entry name=”rdepth” byte=”41” type=”ieee64”/> <entry name=”selev” byte=”49” type=”ieee64” if-non-zero=”1”/> <entry name=”sdepth” byte=”57” type=”ieee64” if-non-zero=”1”/> <entry name=”rdatum” byte=”65” type=”ieee64” if-non-zero=”1”/> <entry name=”sdatum” byte=”73” type=”ieee64” if-non-zero=”1”/> <entry name=”wdepthso” byte=”81” type=”ieee64” if-non-zero=”1”/> <entry name=”wdepthrc” byte=”89” type=”ieee64” if-non-zero=”1”/> <entry name=”sht_x” byte=”97” type=”ieee64” if-non-zero=”1”/> <entry name=”sht_y” byte=”105” type=”ieee64” if-non-zero=”1”/> <entry name=”rec_x” byte=”113” type=”ieee64” if-non-zero=”1”/> <entry name=”rec_y” byte=”121” type=”ieee64” if-non-zero=”1”/> <entry name=”offset” byte=”129” type=”ieee64” if-non-zero=”1”/> <entry name=”nsamps” byte=”137” type=”uint4” if-non-zero=”1”/> <entry name=”nanosecs” byte=”141” type=”int4”/> <entry name=”dt” byte=”145” type=”ieee64” if-non-zero=”1”/> <entry name=”cable_num” byte=”153” type=”int4”/> <entry name=”last_trc” byte=”159” type=”int2”/> <entry name=”cdp_x” byte=”161” type=”ieee64” if-non-zero=”1”/> <entry name=”cdp_y” byte=”169” type=”ieee64” if-non-zero=”1”/> </extension> </SEG-Y-layout> At a minimum these are the key trace header entries that are required to read the dataset for three of the most common datatypes. Required traces to read a field dataset: <entry name=”ffid” byte=”9” type=”int4”/> <entry name=”chan” byte=”13” type=”int4”/> <entry name=”espnum” byte=”17” type=”int4”/> Required traces to read a pre-stack dataset: <entry name=”ffid” byte=”9” type=”int4”/> <entry name=”chan” byte=”13” type=”int4”/> <entry name=”espnum” byte=”17” type=”int4”/> <entry name=”cdp” byte=”21” type=”int4”/> <entry name=”cdptrc” byte=”25” type=”int4”/> <entry name=”trctype” byte=”29” type=”int2” <entry name=”co_scal” byte=”71” type=”int2”/> <entry name=”sht_x” byte=”73” type=”coor4”/> <entry name=”sht_y” byte=”77” type=”coor4”/> <entry name=”rec_x” byte=”81” type=”coor4”/> <entry name=”rec_y” byte=”85” type=”coor4”/> <entry name=”coorunit” byte=”89” type=”int2”/> Required traces to read a post-stack dataset: <entry name=”cdp” byte=”21” type=”int4”/> <entry name=”cdptrc” byte=”25” type=”int4”/> <entry name “offset” byte=”37” type=”int4”/> <entry name=”co_scal” byte=”71” type=”int2”/> <entry name=”sht_x” byte=”73” type=”coor4”/> <entry name=”sht_y” byte=”77” type=”coor4”/> <entry name=”rec_x” byte=”81” type=”coor4”/> <entry name=”rec_y” byte=”85” type=”coor4”/> <entry name=”coorunit” byte=”89” type=”int2”/> <entry name=”inline” byte=”189” type=”int4”/> <entry name=”xline” byte=”193” type=”int4”/>

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

mandate this format revision as it finds the current requests are just not working. There is also the issue of brick formats being put forward as standards such as Open VDS, Open VDS+, VDS, Openworks, ZGY and MDIO which are written and read via an API. There are some considerations here. The first is that these are application to application brick formats and lack the utility for seismic data exchange. The SEG standards are completely open binary published exchange formats which are used around the world. Anyone, whether they are a student, a government or a super-major can read the data to enormous advantage to the industry. Even if they do not have a SEG reader there are various free QC and viewing tools also available online to assist and enhance data use. Conclusion The SEG Technical Standards Committee has been manned by a voluntary committee representing operators and vendors who spend a large amount of time to checking on both specific

industry technology and IT changes to propose these format revisions. The national data repositories and the oil and gas companies have supported these updates, but those companies commonly referred to as the contractors seem reluctant to make any changes. The SEG Technical Standards Committee believe that this is the time to change that approach as we lose so much expertise due to changes in career and retirement. We need to capture knowledge and this is a very simple way to do so. With a very small amount of programming effort, we can all update software suites and launch updates to products that will handle this auto-read format version and improve the data exchange and ingestion internationally. The SEG’ Technical Standards Committee always welcomes new members and questions relating to the standards. If you have any questions then please make contact with the team and note that there is a meeting at the annual meeting, which will be at Image24 next August, open to anyone who is interested whether they are an SEG member or not.

EAGE’S ONLINE EDUCATION PLATFORM

LIFELONG LEARNING OPPORTUNITIES IN VARIOUS ONLINE FORMATS

FREE LEARNING

EAGE COURSES

PARTNER COURSES

E-Lectures, Distinguished Lecturer Webinars, Student E-Lectures

Extensive, Self-Paced, Interactive, EAGE Education Tours

Online courses provided by EAGE partners

Explore all the courses in our catalogue!

www.learninggeoscience.org

Learning Geoscience 171x115,5.indd 1 FIRST

BREAK

VOLUME

DECEMBER

2023

09/11/2023 16:34

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Data agility: Innovative approaches to subsurface data management Jose Chapela1* explores the challenges associated with data management and reviews solutions for overcoming them. Abstract Data management has become a critical component of oil and gas exploration. The industry has spent years collecting, storing, analysing and interpretating subsurface data. This data is invaluable for making informed decisions on where to explore for hydrocarbons, but properly managing subsurface data presents a multitude of challenges, each of which impacts on the quality, accessibility, and utility of this crucial resource. In this article, we explore the challenges associated with data management and review solutions for overcoming them. Introduction Dealing with data in the upstream oil and gas sector has always presented distinct and complex hurdles. This industry handles a vast array of data, each marked by its own unique formatting intricacies. Consider file types like SEG-D, SEG-Y, ACSII, UKOOA, Multibeam, LAS, GeoTiff, and more, encompassing seismic data, well logs, horizons, interpretations, and various other data categories. Furthermore, a significant portion of these data formats include spatial components that demand meticulous handling to ensure accurate geolocation. This article will primarily address techniques for effectively managing your seismic data, whether it’s in SEG-Y, SEG-D, or even older formats like SEG-A, SEG-B, SEG-C, or SEG-X. As you contemplate the next phase of data management, it’s crucial to acknowledge that the industry is currently shifting from

its historical focus solely on hydrocarbons as the primary energy source to integrating renewables like wind and solar energy. This transition will bring entirely new challenges to our existing data storage and retrieval systems. If today’s data management already poses formidable challenges, one can only imagine the obstacles that tomorrow’s data managers will encounter. When considering the future of data management, here are the key questions you should ponder: 1. How can you devise a storage strategy that is both manageable and cost-effective, considering the exponential surge in data volumes, primarily seismic, witnessed over the past decade? 2. How can you amass sufficient metadata pertaining to your data, making it not only usable for your data management team but also for your processing, interpretation, exploration, machine learning, and artificial intelligence teams? Does this metadata support eventually align with the Open Subsurface Data Universe (OSDU) standards? 3. How do you facilitate the efficient transfer of data, not only within your internal teams but also with external partners and collaborators? 4. How can you organise and cleanse your data, gathered over decades, to ensure your machine learning and artificial intelligence initiatives are poised for success? 5. How can you design the next generation of adaptable data management solutions to effectively address these challenges while remaining flexible enough to accommodate future data sets?

Figure 1 Size per shot on left axis and number of products on right axis.

TGS

Corresponding author, E-mail: Jose.Chapela@tgs.com

DOI: 10.3997/1365-2397.fb2023103

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Addressing storage — the elephant in the room Primarily, we must address the limitations of contemporary seismic data storage practices. In our industry seismic data is conventionally stored and shared in the SEG-Y format, a standard introduced by the Society of Exploration Geophysicists (SEG) in 1975. The inception of this standard occurred during a time when the predominant storage medium for seismic data was nine-track tapes. Consequently, the SEG-Y file format is tailored for sequential access, lacking optimisation for random input/output (I/O) operations. To illustrate, if you need to extract the last five inlines, this happens quickly but the process becomes significantly more resource-intensive, particularly when extracting or viewing crosslines. The complexity escalates when numerous nodes are attempting to work with a single file for processing or machine learning, causing I/O bottlenecks. Many organisations turn to parallel file systems to mitigate these challenges. Those parallel file systems, in turn, add expense and complexity to an already complex ecosystem. Secondly, the ever-increasing volume of data poses a formidable challenge. Seismic data management grapples with the massive data sets generated at an unprecedented speed. Modern seismic surveys employ advanced sensors and technologies that yield vast datasets. The resulting datasets also generate a greater number of products at an increased density compared to previous projects. See Figure 1, which illustrates the size increase per shot and the increase in the number of products delivered to TGS data management per project from 1975 to 2020. The high volume and rapid velocity of seismic data place considerable strain on storage systems and overpower traditional data management infrastructures. Handling such extensive datasets necessitates substantial computational resources, resulting in elevated expenses and resource intensity. For instance, within a decade, traditional streamer surveys have grown fivefold in size. Presently, a large 3D streamer survey demands more than 2 petabytes (PB) of storage for a single copy. For more data-intensive Ocean Bottom Node (OBN) surveys, this figure balloons to over 6 PB of storage for a single copy. Best practice suggests storing a ‘production’ copy and a second copy for disaster recovery for every piece of managed data. According to the aforementioned surveys, this amounts to 16 PB of stored data! To address these issues, we have implemented a file format created by TGS and subsequently released to the open-source community, known as Multidimensional Input/Output (MDIO). A comprehensive explanation of the benefits of MDIO can be found in the October issue of First Break in the article ‘Integrating Energy Datasets: The MDIO Format’ (Sansal et al., 2023) or at www.mdio.dev In the context of our data management requirements, this format has enabled us to achieve an average of 41% disk space savings compared to storing the data in the conventional SEG-Y format. Furthermore, as the data is now in a format that permits fast concurrent and random access, we can develop routines to efficiently and swiftly subset the data in a scalable automated fashion. Additionally, it enabled us to store the data in a format that remains readily accessible by our processing nodes and machine learning initiatives while at rest. 70

FIRST

BREAK

VOLUME

DECEMBER

2023

Utilising the hierarchical storage management (HSM) system embedded in cloud object storage, we can achieve further cost reductions by setting up policies to move data that is not actively being accessed by applications to lower-tier storage subsystems. The policies are configured so that the data is always immediately accessible and the HSM takes care of the file movement between tiers automatically in the background. With a significant decrease in storage requirements for the production copy of the data, we are free to explore the best possible path forward for storing disaster recovery copies of our data. Conversion to MDIO (ingestion) Now that you have established a data storage method, the next phase involves transitioning (ingesting) your datasets into this new format to harness various enhancements. When converting from SEG-Y to MDIO, this process offers an excellent opportunity to capture crucial metadata and address any discrepancies within your dataset. Drawing upon four decades of experience in geological and geophysical data management, TGS is currently developing an automated quality control (QC) application. This application will swiftly extract vital information from each seismic product. The primary aim of this QC tool is to provide the team with an overview of the product’s condition and to identify any issues that need attention before ingestion into MDIO. It will also extract essential information to facilitate the conversion to MDIO. For instance, it will attempt to determine the storage locations for CDP X and CDP Y values. Furthermore, it will extract key data from the EBCDIC header and geographically position the data on a map. This functionality enables data management personnel to efficiently review the data, extract pertinent information, and then proceed with the ingestion into MDIO or send the file for updates and corrections.

Figure 2.1 Trace Header locations extracted from EBCDIC.

Figure 2.2 Byte definition screen.

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

cleaned and key metadata was captured, we are assured that the data being delivered to our clients is in the best condition possible. We can also output JSON files with key attributes included to ease data loading for our clients. MDIO also allows for the storage of non-seismic data sets as well. It can effectively store any multi-dimensional data set, so it is also employed on wind data sets that are now being added to the data library.

Figure 3 Section of seismic in MDIO format.

It is commonly known that adherence to the SEG-Y standard is often lax, making it vital to store crucial metadata alongside the file to ensure accessibility for consuming applications. For instance, when TGS was ingesting SEG-Y data for a US land survey, certain trace headers did not conform to the SEG-Y rev 1.02 specifications for locations such as 181-184 for CDP X and 185-188 for CDP Y. As per best practice, trace headers that deviated from the standard locations were specified in the EBCDIC header. As depicted in Figure 2.1, it is evident that within the EBCDIC header, CDP X was stored in positions 201-204, and CDP Y was stored in positions 205-208. Within MDIO, we have the capability to configure these parameters, ensuring that any future applications accessing this dataset are aware of the key trace header locations through our byte definition screen. See Figure 2.2. In this case CDP X, CDP Y, Iline, Xline, and the scalar for position coordinates, were noted and added to the metadata of the file. If there is a need to define additional trace header locations, this can also be accomplished through the attributes section. To enhance data viewing and sub-setting efficiency, we have employed MDIO to create indexes for inlines, crosslines, and time slices. This allows for the quickest retrieval of our data in all three dimensions. Figure 3 illustrates the concept well. The blue-shaded polygon is our area of interest. The application will simultaneously access only the blocks of data necessary to create the data cut, returning a result set in a fraction of the time of traditional data cutting methodologies. In addition to defining byte structures and indices, a standardised nomenclature was introduced, facilitating automated data tracking, versioning, and retrieval. This naming convention can be cross-referenced with a relational database containing extensive project and product information and metadata. The synergy between the data stored within the file and the metadata housed in the relational database provides a comprehensive view, serving the needs of internal data management applications and allowing for the retrieval of the data necessary to comply with Open Subsurface Data Universe (OSDU) Data Platform standards. Once in the MDIO format, we can easily output the data back into SEG-Y for delivery to our external clients. Since it was

Data movement Ever since we decided to transition our complete data library to the cloud, we have harnessed the data movement technology developed by the leading public cloud providers: Microsoft Azure, Google GCP, and Amazon AWS. These providers have invested substantial resources to ensure efficient data transfers, whether it is between clouds or from the cloud to on-premises systems. Consequently, their data movement applications are finely tuned to deliver the fastest and most seamless data transfers. For instance, with a 10 Gb/s internet connection, transfer speeds of 75 terabytes per day (24 hours) were consistently achieved. This translates to an impressive utilisation rate of 69% of the link speed, while sharing the link with the rest of the organisation’s normal internet traffic. What is noteworthy is that these data-moving utilities eliminate the need for us to concern ourselves with complex tasks such as multi-threading the copies, bandwidth throttling, encryption, checksums, and more, as these aspects are already integrated into the utilities themselves. To provide context, TGS undertook the transfer of a massive single dataset, moving a remarkable 1.5 petabytes from our facility to a client’s Azure cloud. The transfer, theoretically achievable within 20 to 25 days at speeds ranging from 70 to 75 terabytes per day, experienced some delays in practice. This was primarily due to the client’s need to adapt and optimise their data intake procedures to handle the substantial daily influx of data. Once the client successfully adjusted their data intake routines to accommodate these large daily transfers, the process proceeded seamlessly, with approximately 70 terabytes of data delivered within each 24-hour period. In a traditional delivery model, 1.5 PB delivery would have taken approximately 45-60 days before we would have been ready to ship, depending on the availability of resources for data copy procedures. It also would have required approximately 188 IBM 3592JD tapes or 150 12 TB USBs hard drives. Once the tapes or USBs arrived at our client’s datacentre, they would have had to read that information from the media onto their systems before it was accessible to their internal teams. With the cloud transfer, the data was available to the client in a significantly reduced time frame. Data management as a service Standardising our data sets has opened up new possibilities. With data now described in a universally consumable manner and metadata accompanying each product, seismic data seamlessly integrates into the larger TGS data lake without any adverse impact on performance through API calls. For instance, having full certainty about position information enables us to query the data lake for all data falling within a project’s boundaries, yielding FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

results such as acquisition reports, processing reports, OB logs, navigation files, contracts, well data, and more. Furthermore, in our MDIO format, ordering data for internal processing or interpretation projects no longer necessitates involvement from data management personnel. Users can effortlessly request the entire project or a specific subset, whether a single product or all products. The ordering process is streamlined, user-friendly, and entirely automated. Once an order is submitted, the system promptly allocates the essential cloud resources for data extraction and subsequent quality control. Upon completion, the requester receives an email containing the location of the requested data and tools for downloading it to a location of their choosing. TGS has recently introduced Data Verse, a Data Management as a Service offering that enables customers to take advantage of the storage savings, performance, ease-of-use, AI/ML readiness, and scalability available from TGS’ data management systems. Conclusion Now is the opportune moment to contemplate your data management strategy. Does your strategy not only provide solutions for the four questions presented but also accommodate any unique concerns specific to your role as a data manager? While you explore the intricacies of your current data management solution, search for a system that possesses the flexibility to address present challenges and the agility to adapt to the ever-evolving demands of our industry in the future. Consider how your organisation plans to

handle multi-client data and the management of your proprietary datasets. Don’t shy away from dealing with the elephant in the room, the SEG-Y format, and the large datasets we store in SEG-Y. Acknowledgements We would like to thank Dr Eugenio Maria Toraldo Serra, Vidar Mikalsen, Kevin Nicholls, Scott Tompkins, Karen Malick, Lisa Sanford, Daniel Phelps, Joao Gusso, David Garner, Gary Stafford, and the rest of the TGS Data Management team for the countless hours spent not only in design discussions but also executing on the vision on top of their normal duties. I would also like to thank Sathiya Namasivayam, Charles Nguyen, Mohammad Nuruzzaman, Altay Sansal, Alejandro Valenciano, Inah Arthasarnprasit, and the rest of the TGS Data and Analytics team for its support, and unending patience during the design and creation of these solutions for the next generation data management ecosystem. Without this talented team of men and women, this would still be a vision and not a reality. Special thanks goes to Jan Schoolmeester, and Laura Arti for supporting the vision and providing much-needed guidance. References Sansal, A., Lasscock, B. and Valenciano, A. [2023]. ‘Integrating Energy Datasets: The MDIO format’, The Leading Edge, 41, 69-75. SEG Technical Standards Committee, [2002]. SEG Y rev 1 Data Exchange format.

EAGE/AAPG Workshop on New Discoveries in Mature Basins 3 0 - 31 JA N UA R Y 2 0 2 4 • K UA L A L U M P U R , M A L AY S I A

7 2 NDM24F V3H.indd I R S T B R1 EAK

VOLUME

DECEMBER

2023

10/11/2023 09:33

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Affordably making the invisible unmissable Neil Hodgson1*, Karyna Rodriguez1, Helen Debenham1 and Lauren Found1 demonstrate how reprocessing vintage data can bring new prospectivity to basins in Oman and east coast India. Reprocessing the geology At midnight, a man stands under a streetlight staring at the ground, looking for his car keys. ‘Did you drop them here?’ asks a stranger to which he replies ‘I don’t know, but it’s the only place with any light’. In exploration for oil and gas, seismic is your light source. Access to more vintage seismic data switches on more street lights which helps the search yet it is reprocessing these data that makes those lights brighter. Whilst just re-looking at what our predecessors looked at isn’t going to change the story, reprocessing these data will reveal a new order of information – both in the detail and depth that the imaging extends to. This is just what the explorer needs; new information that is an antidote to uncertainty. The most material advances in the exploration toolbox for the last 40 years have been the inexorable improvements in seismic processing techniques, giving better data, quicker and cheaper. Such advances in seismic technology not only throw more spotlights on the ground but its cost effectiveness vs new acquisition simply allows you to switch on daylight, understanding new geological stories painted on a broader canvas in exquisite detail. Reprocessing the available data, or, better accessing the reprocessed data that is available through a multi-client company, is rarely a waste of precious time, as Daniel Boorstin (1984) and maybe Stephen Hawking (2001) have both put it; ‘The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge. Global seismic reprocessing Ahead of new acquisition getting a new story has to come from the ‘affordable solution’ playbook and where old datasets already exist – these may well be able to supply the light at the right price if they are reprocessed. As techniques in processing evolve quickly, ‘vintage data’ that can be vastly improved includes most seismic processed more than seven years ago. Seismic processing improvements that have almost ubiquitous efficacy are of broadly two types; the first getting rid of noise that should not be there and the second putting the remaining signal in the right place. Techniques involving removing artefacts from the data include source and streamer de-ghosting, multiple removal such as Surface Related Multiple Elimination (SRME), and improved noise removal techniques (Radon etc). Putting data in the right place includes using new ways to migrate

data either via better migration algorithms or by deriving and utilising a better velocity field for the algorithms to work with. Such techniques like Full Waveform Inversion that were once frontier cutting edge research tools (ie expensive and slow) are now almost a standard procedure, run fast and cost effectively, yet they are also constantly evolving and improved imaging results. To that end Searcher constantly seeks to be the engine of change globally by collecting vintage datasets under agreement with host-governments and reprocessing these data with our partners to image geology never imaged before. We use this data to reveal new oil and gas stories, which we share with investors (oil and gas explorers) to allow them to identify new opportunities. Turning on the lights offshore Oman There has been very little exploration offshore Oman’s NE coast – sometimes called the Sea of Oman or the Sohar Basin, where only three offshore wells have been drilled (Hodgson et al., 2022). The quality of the available seismic dataset plays a big part in this, as even the two earliest wells in the offshore (drilled in the late 1960s early 1970s (BM-A1, BM-B1)) found clastic reservoirs, with oil and gas shows and as recent geochemistry has demonstrated, oil prone source rocks of Late Cretaceous age. As you can imagine from that time the seismic used to locate the wells is pretty shocking to modern explorers and one has to salute the courage of earlier generations in their embracing of uncertainty. A later well drilled just onshore Oman (Barka-1) found oil-stained Eocene Nummulites, proving the working hydrocarbon system, and many studies have located repeating natural oil slicks/seeps using satellite data in the basin. However, the vintage dataset offshore had poor resolution of the complex structuring in the basin, and almost no imaging of the pre-Tertiary. Almost 40 years separates the first exploration from the last well drilled in the basin (EP-A1, 2010). The well was drilled on a small 3D dataset in the far north of the basin, distal from sediment entry points. The 3D images constrained the exciting structuring offshore, yet this well found, perhaps unsurprisingly, no reservoir. In 2022 Searcher reprocessed 4200 km of 1999 vintage 2D data located closer to coarse quartz clastic sediment entry points, with our partners DUG, taking the raw data through a modern processing sequence that removed artefacts and noise, and then used an iterative Kirchhoff algorithm to obtain a new velocity profile and Pre-Stack Depth Migrate the data to accurately place

Searcher

Corresponding author, E-mail: n.hodgson@searcherseismic.com

DOI: 10.3997/1365-2397.fb2023104

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 1 Comparison of offshore Oman vintage (1999) processing (LHS) with 2023 PSDM reprocessing (RHS).

Figure 2a and 2b. 2a Zoom into the detail of the Gravity-Driven Fold and Thrust Belt as imaged on reprocessed PSDM data. 2b. Zoom in to the new Cretaceous or Jurassic basins imaged in deeper water again on reprocessed PSDM data.

signal from whence it came. Figure 1 shows a comparison of vintage data with reprocessed data. Reprocessed data has far less noise and multiple energy in it – this has been removed by de-ghosting and SRME techniques. The fidelity of the data is improved as the frequency spectrum of the data is now much wider, with lower frequency and much wider high-frequency signal present. This quality of the data has been called ‘broadband’ in the past, a term that is almost redundant as all reprocessed data is like this now. When migrated using the correct (geologically consistent) velocity profile steeply dipping data is imaged correctly, and reflectors below this irregular steeply dipping section are also enhanced. Although a crisper, cleaner image is useful, the key to this exercise is what extra geological information the data now brings to the interpreter. Of course, the individual thrust slices are now well imaged (Figure 2a), and were they to provide prospective targets (such as we see in the Orange Basin of Namibia), one could map these in a way impossible on the vintage version. 74

FIRST

BREAK

VOLUME

DECEMBER

2023

Bright events above, beyond and within the gravity-driven fold and thrust belt here can also be analysed on flattened gathers for Amplitude vs Offset responses where shallow gas and oil targets in clastic reservoirs have strong type III anomalies. However, the prize lies below the gravity-driven fold and thrust belt, where organic mudstones of Late Cretaceous to Early Tertiary age lie beneath the decollement. Actually, maturation of the organic material in these mudstones reducing viscosity and allowing horizontal shear may indeed be the cause of the decollement as the clastic wedge of the Sohar basin tilted into the Makran subduction zone in the east. Amplitudes vary a lot in this sub-decollement section. Some of these might represent clastics pouring into the basin soon after the emplacement of the Oman Ophiolite to the west where wadi s rapidly cut through the fractured ophiolite to erode basement granites. Such quartz-rich sediments are well represented in the Late Cretaceous and Early Tertiary sections onshore (ref Andy Racey 2023 Report available through Searcher).

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

What the reprocessed data also tells us that the vintage data could not, is that the decollement surface is not regular and that structures can now be mapped on the slope in modest water depth today (see Figure 1). Such prospects where source reservoir and structure are entwined are very prospective. Out in deeper water – an older basin, perhaps a Jurassic continental fragment from a lost Tethyan Sea is now imaged with potential for clastics and carbonates, and new source rock systems (Figure 2b). This area should have just been oceanic crust. However, what appears to be syn-rift is now imaged on reprocessed data suggesting an exotic terrain may be present. Because exploration of such geology within the Tethyan margin of the Persian Gulf to the north has been so rewarding, one is hard pressed to discount the presence source – reservoir pairs being imaged only on the reprocessed data from this basin. Such revelatory imaging is part of what is bringing this area into the spotlight again, where a forthcoming licensing round presents a new generation of explorers with the tools to better understand this area than ever before. An additional light source offshore east coast India An ongoing licensing round also underlies the 10,000 km of 2D data that Searcher and partner Shearwater reprocessed offshore SE India in 2022. This margin is partly magma poor, partly transform, and partly magma rich, but is only lightly explored outside the Krishna Godavari Basin (Hodgson et al, 2022). Although almost unexplored in deep water, a few brave wells have been drilled, including some of the deepest water wells ever drilled by man. The geology in deepwater reflects an interplay of a complex passive margin with several sediment input points along its margin, dominated by lateral drowning by sediments coming from the mighty Ganges in the north. Shallow clastic biogenic gas accumulations in the northeast have been discovered but the deeper geology remains largely untouched. Geologic plate reconstructions suggest that the deep basin on this margin could have a basal Early Cretaceous (Aptian) source rock, deposited

in a narrow, restricted basin between India and Antarctica. This situation is not unlike the early opening of the Atlantic, where a narrow Early Cretaceous basin, with an active spreading centre allowed anoxic Aptian source rock to be deposited, later to provide the source for the Venus, Graff and La Rona plays of Namibia. In frontier settings hunting and high-grading source rock is key to success. Whilst sediment entry points, and crucially from that sediment, provenance can be deduced on vintage data, and surfaces can be mapped on 2D cubed data (demigrated vintage 2D that is remigrated into 3D datasets), hunting source rock requires access to flattened gathers, and that requires full reprocessing to achieve. Searcher’s SE India regional lines reprocessed by Shearwater in India in 2023 come from two vintages. Firstly, an IndiaSpan dataset acquired in 2006, and secondly the East Coast 2D dataset acquired in 1995. These vintage datasets are spectacular when reprocessed (see Figure 3). The de-ghosting of the Ganges sourced Late Tertiary channel-levee systems allows the mapping in fantastic detail of the development of basin floor channel levee systems, the levee failure and beltway channel migrations (Figure 4a). These systems can be very cryptic on the vintage data but stand out gloriously in reprocessed data. The Searcher reprocessed data set is a complement to existing datasets to be used in combination with the vintage 2D and patches of vintage 3D and 2D cubed data in the basin, where the detail revealed by the reprocessed data, and the AVO analysis possible with properly flattened gathers can be combined with regional surfaces to evaluate and constrain the structuration and onlap of the sedimentary system in this immense deep-water basin. However, the power of fully reprocessed data is not limited to sedimentology or structure, but this data excels in AVO studies of lithology, particularly organic content, and in direct hydrocarbon detection. Positive Type 2/3 AVO anomalies within the fringing talus slopes of the 85o Ridge (Figure 4b) are tantalising yet – more so are the flattened gathers over the Aptian unit at the base of slope along this margin which display a positive Type IV AVO response which suggests the presence of an organic-rich mudstone. Yet

Figure 3 Offshore SE India vintage (1995) processing (LHS) and 2023 reprocessing through PSTM (RHS).

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 4a and 4b. 4a. Zoom into the detail of the Levee-constrained channel beltway on reprocessed PSTM data. Compare to Figure 3 vintage data. 4b. Zoom in to the prograding carbonate shelf edge on the 85o Ridge on reprocessed PSDM data.

this basin also has another secret; out beyond the shelf far offset stacks in the Early Tertiary (Paleocene) are dimmed significantly compared to near offsets. The response is not AVO Type IV, it is Type II and therefore could be a series of stacked sand reservoirs which have been charged with hydrocarbon from a source rock potentially below this package. The AVO Type II response could be due to a low acoustic impedance contrast between the sands and the shales overlying them such that the dimming in the far angle could be the response of hydrocarbon charge. Fluid substitution modelling is being undertaken to investigate this hitherto unreported feature, widespread over the basin and present over many of the open blocks in the current licensing round. The identification of the Early Tertiary plays and their relation to sediment entry systems from quartz-rich hinterlands to the shelf or the lateral Ganges-related sediments, is only possible within the new framework of reprocessed data adding a new dimension to exploration on this most prospective and as yet unexplored margin.

aggressive bidding. As L. Frank Baum said that ‘No thief, however skillful, can rob one of knowledge. Which is why knowledge is the best and safest treasure to acquire’. Reprocessed vintage seismic can allow explorers to reimage, reprocess and better the geology, finding brave new worlds and reducing uncertainty in their interpretations of them and increasing confidence in their analytical conclusions at a cost that matches the maturity of the investment. This principle can be applied world-wide (Hodgson and Rodriguez, 2022) to the reprocessing of vintage data which affordably brings such strong light to the search for new ideas and new insight. It can turn night into day, and make the hitherto invisible into the henceforth unmissable.

Reprocessing vintage data is given a new life, yielding new information In both Oman and SE India the combination of these reprocessed data into the evaluations of 2023/24 licensing round blocks is proving to be an essential aid to the explorers evaluating these blocks. These are the right data available at the right price to increase confidence in analysis and facilitate stronger and more

Hodgson, N., Rodriguez, K., Davies, J., Hoiles, P. and Al Albani., S.

References Boorstin, D. [1984]. January 29, The Washington Post, The 6 O’Clock Scholar: Librarian of Congress Daniel Boorstin And His Love Affair With Books by Carol Krucoff, Start Page K1, Quote Page K8, Column 2, Washington, D.C. (ProQuest). [2022]. Offshore Oman; Stunning hydrocarbon geology from a closing ocean. GeoExpro., 19(2), March 2022. Hodgson, N. and Rodriguez, K. [2022]. Fantastic Basins and Where to Find Them. First Break, 40(5), 67-72. Hodgson, N., Rodriguez, K. and Hoiles, P. [2022]. Revealing the potential of East Coast India. First Break, 40(6).

ONLINE LEARNING

EDUCATION PACKAGES 76

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Advanced imaging of hybrid acquisition data: Exploring new frontiers Sylvain Masclet1*, Fang Wang1, Guillaume Henin1, Loic Janot1, Olivier Hermant1, Hao Jiang1, Nicolas Salaun1, David Le Meur1 and Daniela Donno1 illustrate how the leveraging of sparse node data through an interferometry approach and the use of elastic FWI can enhance streamer seismic imaging. Introduction In recent years, two predominant acquisition trends have emerged: multi-source dense streamer acquisition, ideal for shallow targets requiring high-resolution imaging, and ocean bottom node (OBN) acquisition, initially designed for complex sub-salt imaging. Dense streamer acquisitions enhance subsurface image resolution, particularly in regions with challenging conditions such as very hard seafloor and fast shallow sediments. This improvement is achieved by towing streamers closer to wide-towed sources (Long et al., 2017) or deploying source-over-spread technology (Lie et al., 2018). OBN data, featuring rich low-frequency signals, full-azimuth coverage, and long-offset illumination, offer significant advantages for enhancing seismic velocity models and imaging. However, the operational cost of using dense nodes generally limits their deployment to known reservoirs with restricted shot coverage and maximum offset. Consequently, during the exploration phase, narrow-azimuth towed-streamer (NATS) acquisitions are often preferred. However, NATS data can lead to inadequate imaging of complex targets, increasing uncertainty and risk in prospect estimation. Increasing node spacing, even up to 1 km intervals (sparse nodes), has emerged as a potential solution to economically expand the imaged area while leveraging the benefits of node data for velocity model-building with full-waveform inversion (FWI) (Dellinger et al., 2017). The extended offsets allow for deeper diving-wave penetration, a crucial aspect of FWI updates. The full-azimuth coverage enhances inversion constraints, benefiting from diverse travelpath angles, while rich low-frequency content compensates for starting model inaccuracies (Michell et al., 2017; Shen et al., 2017). While node sparsity does not affect low-frequency FWI for model building, imaging with nodes at 1 km spacing may still face challenges in providing the necessary resolution for reservoir-level details. To mitigate this limitation, a hybrid acquisition approach combining sparse nodes and dense streamers has emerged as a cost-effective solution for large exploration areas. Such multi-input data acquisitions require innovative processing and imaging techniques to fully exploit their potential. From the challenging areas of the Nordkapp Basin in the Barents Sea to those of the Northern North Sea, this paper illustrates

how the leveraging of sparse node data through an interferometry approach and the use of elastic FWI can enhance streamer seismic imaging. As a result, building a high-resolution velocity model with an accurate low-frequency background reduces exploration risk and pushes the limits of imaging technologies one step further forward. Imaging the salt domes in the Nordkapp basin Geology and challenges

The Nordkapp Basin, located in the southern part of the Barents Sea, is a large, underexplored salt basin with a proven petroleum system containing mature Triassic source rocks with expected hydrocarbon traps located at a depth of approximately 3 km. Salt diapirism is prolific with diapirs penetrating all the way up to the seafloor (Figure 1). The imaging challenges of this exploration region are also related to the huge Tertiary uplift which brings high-velocity strata up to very shallow depths. This creates very strong multiple energy and also reduces the incidence angle in the very shallow section. Based on the geologic understanding of the Nordkapp Basin and the corresponding geophysical challenges, a hybrid blended acquisition was performed, combining 18-cable streamer data with sources located in front of and above the streamers and sparse OBN (Dhelie et al., 2021). With such a design, high-

Figure 1 3D seismic section across the Nordkapp Basin central area, showing the distinct salt diapirism.

CGG

Corresponding author, E-mail: sylvain.masclet@cgg.com

DOI: 10.3997/1365-2397.fb2023105

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 2 Raw node gathers before a) and after b) a comprehensive deblending workflow allowing recovery of diving waves above 3 Hz. Active c) and virtual d) seismic gather at low frequency (1-3 Hz); while the active node exhibits residual blended energy, refracted and surface waves are visible on the virtual gather.

resolution and high-fidelity imaging can be achieved by combining the dense streamer data with the wide-azimuth and long-offset (up to 40 km) receiver gathers provided by the OBN records. As a result of the use of a very sparse node carpet with 1200 m between nodes, a large-scale OBN acquisition covering an exploration area of 3700 km2 was economically viable. Ultra-low frequency data reconstruction by interferometry

To achieve a high-resolution subsurface sampling within a limited amount of time, a dense shooting carpet was required. With seven sources shooting almost simultaneously, and one shot-point every 2.5 seconds on average, severe blending noise appears on the recorded OBN data, affecting its signal-to-noise ratio. To use long-offset refraction energy for FWI, we needed to de-blend the data down to a 20-second interval (Figure 2a). Achieving this required the removal of more than 12 orders of blending noise, which we accomplished through a node-by-node inversion-based deblending approach, revealing weak diving wave energy along the 40 km offset (Figure 2b). However, due to the small dithers (±200 ms) used for blending noise randomisation, recovering low frequencies below 3 Hz through deblending was challenging, preventing FWI from starting below this frequency (Figure 2c). Conventional velocity modelling methods attempt to compensate for this by incorporating more accurate initial velocity models based on salt interpretation. But in such complex geology, even starting at 3 Hz, FWI may fall into local minima (cycle skipping) problems due to potential inaccuracies in the initial model. Fortunately, OBNs were continuously deployed on the water bottom for three months without redeployment. This continuous recording enabled interferometry to reconstruct ultra-low frequencies (Figure 2d). Interferometry uses continuous seismic data to create virtual

sources at receiver locations, a technique proven to be effective in Middle East land datasets (Le Meur et al., 2020) for velocity model building. By combining cross-correlations and temporal stacking, we generated virtual source gathers, capturing both active sources and environmental noise contributions. This approach allowed us to produce ultra-long offset sparse gathers with 1200 m spacing, effectively reconstructing surface and diving waves down to 0.5 Hz. Interferometry provides a solution for recovering low-frequency signals that conventional active data cannot capture. The virtual gathers also captured ultra-low frequency surface waves with a minimum frequency of 0.3 Hz. Unlike active data, where picking of surface-wave dispersion curves can be challenging (Figure 3a), virtual shots enabled phase velocity picking from 0.4 Hz to 1 Hz for the surface waves and also from 1 Hz for the guided waves (Figure 3b). As the surface-wave frequency content reflects its penetration depth (Strobbia et al., 2011), we combined these dispersion curve picks with first-break data to perform multi-wave inversion (Bardainne, 2018) for updates of both P-wave (Vp) and S-wave (Vs) velocity in the shallow section down to 300 m in depth. Figure 3c shows the result of 200 Hz FWI of the top of salt (Espin et al., 2022), while Figure 3d displays Vp/Vs information at a shallow level through the top salt region, revealing the diapir’s heterogeneous composition of salt and carbonate. Virtual data provides valuable information for geotechnical evaluation of subsurface layer consolidation and can serve as a starting point for creating an input Vs model for elastic FWI when well information is unavailable. Full-waveform inversion using virtual and active seismic data

Using 991 nodes with virtual shots at node positions, interferometry-based virtual gathers offered offsets up to 40 km, allowing Vp

Figure 3 Dispersion panels for surface and guided waves in the case of active (a) and virtual gathers (b). Vp model obtained from 200 Hz Acoustic FWI (c) and inverted Vp/ Vs at a shallow depth slice going through a salt diapir (d), which shows its heterogeneous composition.

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 4 a) Ultra-long offset OBN gather after deblending. Comparison of migrated seismic overlaid with velocity models of a) streamer and b) node FWI.

Figure 5 a) 3D view of 40 Hz FWI velocity model and FWI image over Nordkapp survey. b) Depth slice going through the target level, with the high-resolution model giving clear insight into the extent of the Carnian sands. Comparison of 40 Hz RTM stack image c) and 40 Hz FWI Image d) around the salt edges.

updates down to 6 km in depth. FWI using virtual data was run up to 3 Hz due to data sparsity at higher frequencies. Interferometry ensured FWI convergence starting at 1 Hz, even with complex salt structures. Active seismic nodes were then used for FWI above 3 Hz. After deblending, reflections and diving waves were effectively recovered at offsets up to 40 km and 20 s of record length (Figure 4a). Node data conveniently supplemented streamer data which are dense but limited in offset and diving-wave penetration depth. In the Nordkapp basin, where salt diapirs and high seismic velocities pose challenges, the FWI result obtained using only streamer data shows a limited depth update owing to the limited offset range (Figure 4b). Our new FWI approach, using both active and virtual node data, improved salt contour definition from low-frequency updates (Figure 4c) without the need for manual salt body interpretation. This method offers an alternative to ray-based tomography and reduces cycle skipping risks by starting at 1 Hz.

Moving to high-resolution full-waveform inversion and imaging

Recent advances in FWI (Zhang et al., 2018) enable us to fully leverage the full wavefield, including diving waves, reflections, and their multiples and ghosts. These components are especially valuable for high-frequency updates and enhance the velocity model details. By increasing the inversion frequency, we can reveal reservoir information that conventional seismic reflection methods cannot capture. To achieve the highest FWI resolution, we utilised all types of waves in the records, ensuring balanced illumination and avoiding migration artifacts and spatial sampling issues. Node gathers accurately captured the long-wavelength trend of the velocity model down to a 6 km depth. However, the lack of near offsets, related to the sparse node grid for the recorded primaries and multiples, prevented FWI updates with nodes above 10 Hz. For higher-frequency FWI, the streamer dataset, with its dense spatial sampling and FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

additional near-offset data, outperformed sparse OBN data, effectively addressing spatial aliasing issues at high frequencies for the shallower part of the velocity model. The high-frequency FWI utilised the entire records up to 40 Hz, making it possible to obtain model updates down to depths of 4 km, although this required substantial computational resources to cover the total volume of over 16,000 km3 (Figure 5a). The 40 Hz inversion already revealed clear details of the Carnian sand target (Figure 5b, white arrow) in the velocity model. For conventional imaging, we used the 10 Hz FWI update as the migration velocity model for running reverse-time migration (RTM) with streamer data, which delivered an accurate image of the salt bodies (Figure 5c). Our next step was to generate a FWI Image from the 40 Hz velocity model (Zhang et al., 2020). The FWI Image (Figure 5d) significantly enhanced imaging of the complex vertical salt wall’s steep dips compared to RTM (Figure 5c) and provided better delineation of salt bodies, thanks to the least-squares process of the FWI engine and utilisation of the entire wavefield. Unlocking new targets below the Base Cretaceous Unconformity (BCU) in the Northern North Sea: Fram and Oseberg surveys Geology and challenges

Exploration in the Northern North Sea has a long history, with ongoing discoveries and evolving exploration models. Recent finds in the Upper Jurassic sands (Dugong in 2020, Echino Sør in 2019), just below the Base Cretaceous Unconformity (BCU), have underscored the need for precise structural mapping of fault traps and, consequently, more accurate velocity models down to the sub-BCU level. Challenges arise from high-velocity injectites in the shallow section, a limestone/carbonate sequence with high velocities covering the targeted lower velocity mudstone units, and extensively faulted Jurassic blocks (Figure 6b). These challenges become more significant because they typically occur at depths beyond the reach of diving waves for FWI when using streamer-based data with a maximum 8 km offset. In response to the growing focus on near-field exploration and the search for less obvious deeper targets in the Northern Viking Graben (NVG) region, two successive hybrid blended acquisitions were recently completed in 2021 and 2022 over the NVG area. In addition to

an existing north-south 3D seismic survey spanning 44,000 km2 (blue polygon in Figure 6a), a blended hybrid acquisition was deployed. This approach combined an east-west oriented triple-source streamer survey with a 900 m x 900 m node grid. It provided a maximum offset of up to 24 km for the Fram survey (~50 km2, white polygon) and up to 60 km for the Oseberg survey (~2,000 km2, black polygon). Exploiting diving waves below the BCU

In 2018 the Fram and Oseberg areas were initially imaged using the north-south streamer data. A velocity model was built with visco-acoustic FWI (Q-FWI; Xiao et al., 2018) and tomography, and the image was obtained with attenuation-compensating Kirchhoff pre-stack depth migration (Q-KPSDM) as shown in Figure 7b for the Fram survey and Figure 8c for the Oseberg survey. With a maximum offset of 8 km, streamer-only FWI provided reliable updates down to ~2 km depths, partially resolving velocity contrasts caused by injectites. However, reaching sub-BCU levels, from 2.5 km in Fram to 5 km in Oseberg, is beyond the streamer data’s FWI capability. Tomography is needed for a deeper velocity update but is prone to inaccuracies due to multiples and the single-arrival assumption, leading to uncertain prospect interpretations. Sparse node data, with ultra-long offsets, is crucial for extending low-wavenumber FWI updates to sub-BCU levels (Lie et al., 2022). Diving-wave analysis using the 2018 legacy model shows that increasing the maximum offset from 8 km to 24 km in the Fram area deepens the diving wave penetration from 2 km to 4.5 km. A comparison between streamer-only and streamer and node diving wave FWI in the Oseberg area (Figures 8a and 8b) highlights the deep penetration of the longer offset, node-recorded, diving waves, capturing lateral velocity variations caused by complex sediment basins and faulted blocks. Starting with a smoothed 2018 legacy velocity model, we conducted acoustic FWI with an advanced time-lag cost function (A-FWI; Zhang et al., 2018) using only diving waves to obtain a reliable low-wavenumber update, essential for the higher-frequency model refinement. In the Fram survey, the 6 Hz FWI using node data managed to capture the velocity inversion below the BCU and provides a good background velocity down to 4.5 km depth (Figure 7a).

Figure 6 a) Location map showing legacy north-south and new east-west seismic data coverage, and sparse OBN layouts in the NVG area. b) 3D section across the Fram survey illustrating common geological features in the region and the associated geophysical challenges.

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 7 Fram survey: a) velocity profile at well location with sonic log (dark blue), 2018 legacy model (cyan), after 6 Hz A-FWI with streamer data (pink) and after 6 Hz A-FWI with node data (yellow). KPSDM stack images using b) the 2018 legacy model and c) the 6 Hz A-FWI model with node data.

Figure 8 Oseberg survey: 3D view of streamer-only FWI perturbation in a) and diving wave node and streamer FWI perturbation in b). Legacy model and its corresponding migration in c) is compared with the 60 km-offset node and streamer FWI and its corresponding migration in d).

In comparison with the 2018 legacy model, the 6 Hz node FWI follows the sonic log more closely and, as a result, provides an improved migrated image at the target level with an uplift in fault imaging (Figure 7c). In the Oseberg area, we addressed the presence of both shallow and deep gas anomalies using a Q-FWI and FWI-guided Q tomography to obtain a high-resolution Q model, accounting for amplitude attenuation and phase dispersion (Latter et al., 2023). This Q model was incorporated into data modelling for subsequent FWI velocity inversions, ensuring more accurate inversion convergence by reducing multi-parameter crosstalk. Even though node data were not initially used for Q-FWI, diving waves penetrating down to the deep Q anomalies enabled kinematic updates of gas reservoirs (as visible in the black circle in Figure 8b). Figure 8d illustrates the results of the 60 km-offset streamer and node diving wave FWI, effectively capturing the complexity of the velocity model across the entire survey area. This approach resolved lateral and vertical velocity variations caused by injectites, gas reservoirs in the Martin Linge and Oseberg fields, the chalk layer, heavily faulted blocks, and graben flanks, significantly improving the structural image (Figure 8d) compared to the legacy one (Figure 8c).

Elastic full-waveform inversion

North Sea geology exhibits a range of velocity anomalies and contrasts due to gas, injectites, chalk layers and the seabed. These contrasts can generate a range of elastic effects on the recorded seismic data, such as kinematic effects on pressure waves, and mode conversions between P- and S-waves. As a result, the diving-wave cone, which we usually rely on for the low-wavenumber FWI updates, may be contaminated by elastic effects (Malcolm and Trampert, 2011; Plessix and Krupovnickas, 2021). Together with other elastic effects in the wavefield, these issues can produce biased acoustic model updates for which elastic FWI is needed (Masmoudi et al., 2022; Wu et al., 2022; Zhang et al., 2023). Over the Fram and Oseberg surveys, long recorded offsets provide diving wave penetration below the sub-BCU level, which offers an opportunity to investigate the potential elastic effects around the velocity contrast at the BCU level in the Fram survey and for a shallow gas anomaly in the Oseberg survey. For the Fram survey, we used an edited 2018 legacy model in which the sharp velocity contrast at BCU was reinforced to run both acoustic and elastic modelling. The results in Figures 9a and 9b reveal that this initial velocity model provided a better fit FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Figure 9 Fram survey: comparison of a) acoustic and b) elastic modelling using the initial velocity model. 6 Hz acoustic FWI with node data and corresponding migrated gathers (a, right) and 6 Hz elastic FWI with node data and corresponding migrated gathers (b, right).

Figure 10 Oseberg survey: comparison of acoustic (a) and elastic (b) FWI results and their corresponding migrations.

between real data and synthetics generated by elastic modelling, particularly for longer offsets (over 12 km). This indicates the advantages of using elastic FWI when dealing with data acquired at such long offsets, where elastic effects are significant, and acoustic approximations fall short. Elastic FWI, although more challenging due to the need to account for Vs, was implemented by inverting for Vp using hydrophone data with a fixed Vp/Vs ratio from well logs, which is a pragmatic solution for this geological area. The acoustic and elastic inversion results (Figures 9a and 9b) demonstrate that elastic FWI with node data improves the velocity profile by enhancing the velocity contrast at the BCU level (Figure 9b, lower white arrow) and providing flatter migrated gathers compared to acoustic FWI (Figures 9a and 9b). Additionally, in the shallow section of Figure 9b (upper white arrow in the velocity profile), we observe a velocity increase recovered with elastic FWI, indicating the presence of injectite features. In the Oseberg survey, we ran acoustic and elastic node FWI honouring the attenuation model in an area where a shallow gas 82

FIRST

BREAK

VOLUME

DECEMBER

2023

anomaly is located right above the complex injectite level. While the acoustic FWI failed to fully resolve the velocity anomaly, elastic FWI captured accurately the velocity variation induced by the gas anomaly, reducing at the same time the halo effect visible with the acoustic update (velocity profiles in Figures 10a and 10b). As a result, the distortion on the seismic image just below the gas anomaly and even deeper in the section is attenuated in the elastic case compared to the acoustic one (orange versus green arrow in Figures 10a and 10b). High-resolution velocity model and FWI imaging

As mentioned earlier in the Nordkapp case study, one of the main requirements before considering the full wavefield in the FWI is to have an accurate and reliable long-to-mid wavelength update to which the resolution will be added. Usually, the depth of this reliable model is a function of the maximum penetration depth of the diving waves. In the Fram case, the 8 Hz diving-wave elastic FWI with node data provides an accurate and reliable update down to

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

the Brent level at a depth of 4 km. In the Oseberg case, the sparse node FWI provides a reliable update right down to the basement structure at a depth of 12 km. To move to the high-frequency update, a joint node and streamer FWI was run, combining all the benefits brought by each dataset: the dense spatial receiver sampling from the streamer datasets and the long-offset and full-azimuth information from the node data. The joint node and streamer FWI was run up to 30 Hz over the Fram survey and up to 40 Hz for the Oseberg survey, providing high-resolution velocity models in both cases. Results from the Fram survey are shown in Figure 11, where we see that the 30 Hz FWI model aligns well with the sonic log down to the Brent level and effectively detects small injectite velocity contrasts (upper green arrow in Figure 11d). It

also identifies the thin limestone layer atop the target sandstone units at the BCU level (middle green arrow in Figure 11d). This high-resolution velocity model serves as a valuable attribute for geological interpretation. Below the fast limestone layer at the BCU level, the model distinguishes alternating faster and slower velocity layers within the low-velocity mudstone unit, representing shale and sandstone layers (lower green arrow in Figure 11d). Figures 11b and 11e demonstrate the seismic image improvement from the 2018 legacy model to the 30 Hz joint streamer and node FWI model. The migrated seismic image is cleaner with reduced distortion, extending from the Late Cretaceous level to the Brent (Figures 11b and 11e). Fault imaging in the tilted blocks below the BCU is also enhanced, resulting in sharper attributes as shown in Figures 11c and 11f.

Figure 11 Fram survey: 2018 legacy model (section with velocity profile overlaid on the sonic log) in a) its migrated image in b) and the similarity attribute below BCU in c). Equivalent plots for the 30 Hz joint node and streamer FWI velocity model are shown in (d), (e) and (f), respectively.

Figure 12 Oseberg survey: a) comparison of Q-KSDPM stack image filtered at 40 Hz with b) the 40 Hz FWI image. The 40 Hz FWI velocity model in c) highlights the geological complexity of the Oseberg area.

FIRST

BREAK

VOLUME

DECEMBER

2023

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Reflectors within the faulted blocks exhibit greater continuity up to the fault planes, which are crucial for accurate prospect delineation and volumetric calculations. Over the Oseberg survey, 40 Hz FWI was performed with the idea of directly outputting the reflectivity image to assess the benefits of FWI Imaging in exploration frontier areas, where conventional imaging suffers from limited S/N. Below the injectites level, FWI Imaging corrects for loss of illumination and for small-scale variations that induce distortion on the migrated images (purple arrow in Figure 12b). Thanks to the least-squares iteration process and full-wavefield utilisation, the FWI Image increases the S/N ratio between the Shetland and Tryggvason levels where primary energy is weak (orange arrow in Figure 12b). Wavefront noise along the Shetland, and the BCU horizons, is attenuated on the FWI Image (pink arrow in Figure 12b). Below the BCU level, some faults are better imaged with the FWI Image and the event continuity of the sediments is improved up to the fault plane (green arrow in Figure 12b). In addition to the reflectivity, the high-resolution velocity model can be used as an interpretation attribute by highlighting potentially interesting features (white arrow in Figure 12c).

unlock new targets in unexplored areas of the deeper part of the sedimentary basin. This innovative acquisition design enables the application of advanced subsurface imaging technologies such as interferometry and elastic FWI down to reservoir depths. It holds promise for near-field exploration and broader basin exploration. Depending on prospect depths, transitioning to a denser node grid, from 1 km to 500 m spacing, could be considered to harness node data for higher frequencies and refined FWI Images. Acknowledgements We thank CGG Earth Data, AkerBP and its partners DNO Norge AS and Petoro AS in PL1083, for permission to publish this work. We also thank our colleagues Olivier Leblanc and Mathieu Reinier for fruitful discussions on new imaging technologies and Jaswinder Mann-Kalil and Gustav Aagenes Ersdal for their geological insights. References Bardainne, T. [2018]. Joint inversion of refracted P-waves, surface waves and reflectivity. 80th EAGE Annual Conference & Exhibition, Extended Abstracts, We K 02. Dellinger, J., Brenders, A.J., Sandschaper, J.R., Regone, C., Etgen, J.,

Conclusion This paper showcases how advanced imaging techniques can be implemented to fully exploit the potential of a hybrid acquistion to adress the challenge of thin reservoirs in two geologically complex regions. In the Barents Sea exploration survey covering 3700 km2, FWI with node data combining active and virtual seismic helped to improve salt body delineation, overcoming streamer shortcomings. However, the node carpet sparsity ultimately prevented its usage for direct imaging or very high-frequency FWI updates. Dense streamer data remains the dataset of choice in those cases. Both data types had to work in tandem to achieve optimal velocity model building and imaging. A fully data-driven approach combining ultra-low frequencies, ultra-wide offsets and dense near offsets led to a high-resolution FWI velocity model and imaging with improved salt flank definition. Then, in the Fram and Oseberg examples, node data provide the ultra-long offsets required to obtain a reliable low-wavenumber velocity update down to the Brent level and to capture the sharp impedance contrast and the velocity inversion below the BCU where the reservoirs lie. In addition, with long recorded offsets, the expected elastic effects induced by the large impedance contrast at this level can be observed and inverted using elastic FWI, which results in a reliable long- to mid-wavelength update down to about 5 km and 12 km depths, respectively, for the Fram and Oseberg surveys. This update subsequently unlocked the use of the full wavefield in FWI at the sub-BCU level. Full wavefields from streamer and node data were then jointly inverted up to high frequencies, producing an accurate high-resolution velocity model down to sub-BCU level. Seismic imaging and, especially, fault imaging are significantly improved with more continuous events up to the fault plane. In addition, FWI Imaging shows the benefits of high-resolution FWI for providing new insights for the interpretation and may 84

FIRST

BREAK

VOLUME

DECEMBER

2023

Ahmed, I. and Lee, K.J. [2017]. The Garden Banks model experience. The Leading Edge, 36(2), 151-158. Dhelie, P.E., Danielsen, V. and Lie, J.E. [2021]. Combining nodes and streamers to tackle the imaging challenges of salt basins in the Barents Sea. First International Meeting for Applied Geoscience & Energy, SEG Technical Program Expanded Abstracts, 41-45. Henin, G. , Espin, I., Salaun, N., Masclet, S., Dhelie, P.E., Danielsen, V. and Lie, J.E. [2022]. Large-Scale 3D High-Resolution Near-Surface Imaging over Nordkapp. 83rd EAGE Annual Conference & Exhibition, Extended Abstracts, 1-5. Latter, T., Nielsen, K.M., Beech, A. and Cantu Bendeck, D. [2023]. Deriving a high-resolution regional scale Q model over the Northern Viking Graben. 84th EAGE Annual Conference & Exhibition, Extended Abstracts, 1-5. Le Meur, D., Solyga, D., Prescott, A., Courbin, J. and Donno, D. [2020]. Retrieving ultra-low frequency surface waves from land blended continuous recording data. SEG Technical Program Expanded Abstracts, 1855-1859. Lie, J.E., Danielsen, V., Dhelie, P.E., Sablon, R., Siliqi, R., Grubb, C., Vinje, V., Nilsen, C. and Soubaras, R. [2018]. A Novel Source-OverCable Solution to Address The Barents Sea Imaging Challenges. Marine Acquisition Workshop, EAGE, cp-560-00016. Lie, J.E., Vetle, V., Dhelie, P.E., Jiang, H., Danielsen, V. and Salaun, N. [2022]. Nordkapp Topseis/Node Acquisition—Lessons from a Modelling Study. First Break, 40(2), 41-49. Long, A. [2017]. Source and streamer towing strategies for improved efficiency, spatial sampling and near offset coverage. First Break, 35(11), 71-74. Malcolm, A.E. and Trampert., J. [2011]. Tomographic errors from wave front healing: More than just a fast bias. Geophysical Journal International, 185(1), 385-402. Masmoudi, N., Stone, W., Ratcliffe, A., Refaat, R. and Leblanc, O. [2022]. Elastic Full-Waveform Inversion for Improved Salt Model

SPECIAL TOPIC: DATA MANAGEMENT AND PROCESSING

Building in the Central North Sea. 83th EAGE Annual Confer-

Wu, Z., Wei, Z., Zhang, Z., Mei, J., Huang, R. and Wang, P. [2022]

ence & Exhibition, Workshop Programme, Extended Abstracts,

Elastic FWI for large impedance contrasts. Second International

1-3.

Meeting for Applied Geoscience & Energy, SEG/AAPG, Expanded Abstracts, 3686-3690.

Michell, S., Shen, X., Brenders, A., Dellinger, J., Ahmed, I. and Fu, K. [2017]. Automatic velocity model building with complex salt: Can

Xiao, B., Ratcliffe, A., Latter, T., Xie, Y. and Wang, M. [2018]. Inverting

computers finally do an interpreter’s job? 87th Annual International

near-surface absorption bodies with full-waveform inversion: a case

Meeting, SEG, Expanded Abstracts, 5250-5254.

study from the North Viking Graben in the Northern North Sea. 80th EAGE Conference & Exhibition, Extended Abstracts, Tu A12 03.

Plessix, R.-E., and Krupovnickas, T. [2021]. Low-frequency, long-offset elastic waveform inversion in the context of velocity model build-

Zhang, Z., Mei, J., Lin, F., Huang, R. and Wang, P. [2018]. Correcting

ing. The Leading Edge, 40(5), 342-347.

for salt misinterpretation with full-waveform inversion. 88th Annual International Meeting, SEG, Expanded Abstracts, 1143-1147.

Shen, X., Ahmed, I., Brenders, A., Dellinger, J., Etgen, J., and Michell, S. [2017]. Salt model building at Atlantis with Full Waveform

Zhang, Z., Wu, Z., Wei, Z., Mei, J., Huang, R. and Wang, P. [2020]. FWI

Inversion. 87th Annual International Meeting, SEG, Expanded

Imaging: Full-wavefield imaging through full-waveform inversion.

Abstracts, 1507-1511.

90th Annual International Meeting, SEG, Expanded Abstracts,

Strobbia C., Laake, A., Vermmer, P. and Glushcenko, A. [2011]. Surface

656-660.

waves: use them then lose them. Surface wave analysis, inversion,

Zhang, Z., Wu, Z., Wei, Z., Mei, J., Huang, R. and Wang, P. [2023].

and attenuation in land reflection seismic surveying. Near-surface

Enhancing salt model resolution and subsalt imaging with elastic

Geophysics, 9(6), 503-513.

FWI. The Leading Edge, 42(3), 207-215.

Publish with EAGE To serve the interests of our members and the wider multidisciplinary geoscience and engineering community, EAGE publishes a range of books and scientific journals in-house. Our extensive, professional marketing network is used to ensure publications get the attention they deserve.

EAGE is continually seeking submissions for both book publishing and articles for our journals. A dedicated and qualified publishing team is available to support your publication at EAGE.

CONTACT OUR PUBLICATIONS DEPARTMENT AT PUBLICATIONS@EAGE.ORG!

20879-Filler Journal Media Publish with EAGE .indd 1

FIRST

BREAK

VOLUME

09/03/2021 14:41

DECEMBER

2023

CALENDAR

CALENDAR OF EVENTS 25-27 MARCH 2024

4th EAGE Digitalization Conference & Exhibition Paris, France www.eage.org

December 2023 4-6 Dec

4th EAGE Eastern Mediterranean Workshop www.eage.org

Athens

Greece

4-7 Dec

17th Workshop of the International Lithosphere Program ILP-Task Force VI on Sedimentary Basins ilp-abudhabi.com

Abu Dhabi

UAE

Kuala Lumpur

Malaysia

January 2024 EAGE/AAPG Workshop on New Discoveries in Mature Basins www.eage.org

30-31 Jan

February 2024 12-14 Feb

IPTC 2024 www.iptcnet.org

Dhahran

Saudi Arabia

19-21 Feb

EGYPES 2024 www.egypes.com

Cairo

Egypt

26-28 Feb

First EAGE Data Processing Workshop www.eage.org

Cairo

Egypt

29 Feb 1 Mar

GeoTHERM Expo & Congress 2024 www.geotherm-offenburg.de/en

Offenburg

Germany

4-6 Mar

EAGE Sub-Saharan Africa Energy Forum www.eage.org

Windhoek

Namibia

25-27 Mar

4th EAGE Digitalization Conference & Exhibition www.eagedigital.org

Paris

France

EAGE GeoTech 2024 - 3 rd EAGE Geoscience Technologies and Applications Conference

The Hague

The Netherlands

Kuwait City

Kuwait

March 2024

April 2024 8-10 Apr

(including 4 th EAGE Workshop on Distributed Fibre Optic Sensing and 4th EAGE Workshop on Practical Reservoir Monitoring)

www.eagegeotech.org

23-25 Apr

First EAGE Workshop on Advances in Carbonate Reservoirs from Prospects to Development www.eage.org

EAGE Events

FIRST

BREAK

Non-EAGE Events

VOLUME

DECEMBER

2023

CALENDAR

May 2024 13-15 May

Fourth EAGE/AAPG Hydrocarbon Seals Workshop www.eage.org

Al Khobar

Saudi Arabia

13-15 May

6th Asia Pacific Meeting on Near Surface Geoscience and Engineering www.eage.org

Tsukuba

Japan

13-16 May

InterPore 2024 events.interpore.org/event/46/

Qingdao

China

85 th EAGE Annual Conference and Exhibition www.eageannual.org

Oslo

Norway

June 2024 10-13 Jun

August 2024 12-13 Aug

3 rd EAGE Conference on Carbon Capture and Storage www.eage.org

Perth

Australia

14-15 Aug

4th EAGE Workshop on Fiber Optic Sensing for Energy Applications www.eage.org

Perth

Australia

18-23 Aug

Goldschmidt 2024 conf.goldschmidt.info/goldschmidt/2024/meetingapp.cgi

Chicago

United States

September 2024 2-4 Sep

Fourth EAGE Marine Acquisition Workshop www.eage.org

Oslo

Norway

2-5 Sep

ECMOR 2024 European Conference on the Mathematics of Geological Reservoirs www.ecmor.org

Oslo

Norway

8-12 Sep

Near Surface Geoscience Conference and Exhibition 2024 www.eagensg.org

Helsinki

Finland

12-13 Sep

First EAGE Workshop on The Role of AI in FWI www.eage.org

Cartagena

Colombia

17-19 Sep

Fourth EAGE Conference on Pre-Salt Reservoir www.eage.org

Rio de Janeiro

Brazil

23-24 Sep

Asia Petroleum Geoscience Conference and Exhibition (APGCE) icep.com.my/apgce

Kuala Lumpur

Malaysia

24-26 Sep

Eighth EAGE High Performance Computing Workshop www.eage.org

Jeddah

Saudi Arabia

October 2024 3-4 Oct

Third EAGE Workshop on EOR www.eage.org

Buenos Aires

Argentina

6-8 Oct

EAGE Workshop on Naturally Fractured Rocks (NFR) www.eage.org

Muscat

Oman

14-16 Oct

Third EAGE Conference on Seismic Inversion www.eage.org

15-16 Oct

EAGE Conference on Digital Twins and Artificial Intelligence www.eage.org

Kuala Lumpur

Malaysia

21-24 Oct

GEO 4.0: Digitalization in Geoscience Symposium www.eage.org

Al Khobar

Saudi Arabia

24-25 Oct

Third EAGE Workshop on Advanced Seismic Solutions in the Gulf of Mexico www.eage.org

Mexico City

Mexico

29-31 Oct

Fourth SPE/EAGE Geosteering and Well Placement Workshop www.eage.org

Al Khobar

Saudi Arabia

November 2024 4-6 Nov

First EAGE Workshop on Tectonostratigraphy of the Arabian Plate www.eage.org

Al Khobar

Saudi Arabia

5-7 Nov

First EAGE Conference on Energy Opportunities in the Caribbean www.eage.org

Port of Spain

Trinidad & Tobago

EAGE Events

Non-EAGE Events

FIRST

BREAK

VOLUME

DECEMBER

2023

2-5 SEP T EMBER 202 4 I O SLO, NOR WAY

TOPICS Physical Modelling Computational Methods Uncer tainty Quantification and Optimization Engineering of Open Simulation Software

SUBMIT YOUR A B S TR AC T BY 1 FEBRUA RY

W W W.ECMOR.ORG

#ECMOR2024

SHOWCASE YOUR

RESEARCH &

2024

O S L O | N O R WAY

WWW.EAGEANNUAL.ORG

TOPICS INCLUDED

EXPERTISE! GEOPHYSICS

GEOLOGY

RESERVOIR ENGINEERING

INTEGRATED SUBSURFACE

ENERGY TRANSITION

MINING AND INFRASTRUCTURE

DATA AND COMPUTER SCIENCE

SUBMIT YOUR ABSTRACTS BEFORE 15 JANUARY 2024! HOST

PARTNERS

RENEW YOUR EAGE

MEMBERSHIP FOR 2024 JOIN OUR COMMUNITY AND ENJOY EAGE’S BENEFITS!

FREE SUBSCRIPTIONS

DISCOUNTS

to First Break and to one of EAGE’s scientific journals

on global EAGE events, courses and journal subscriptions

FREE ACCESS to the latest content on EarthDoc, EAGE’s online repository of proceedings, and Learning Geoscience, EAGE’s online education platform

LEARN

W W W. E A G E . O R G