First Break March 2025 - Reservoir Monitoring

SPECIAL TOPIC

Reservoir Monitoring

EAGE NEWS GET 2025 returns to Rotterdam

INDUSTRY NEWS Norway confirms APA winners

TECHNICAL ARTICLE Unsupervised clustering for geomorphological interpretation

Engage with industry experts and researchers in a collaborative environment, exploring cutting-edge advancements shaping the future of near-surface geoscience and engineering. Register now to secure your spot at this premier event in scenic city of Xi'an, China!

CHAIR EDITORIAL BOARD

Clément Kostov (cvkostov@icloud.com)

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, Viridien (Stephen.Hallinan@viridiengroup.com)

• Hamidreza Hamdi, University of Calgary (hhamdi@ucalgary.ca)

Fabio Marco Miotti, Baker Hughes (fabiomarco.miotti@bakerhughes.com)

Susanne Rentsch-Smith, Shearwater (srentsch@shearwatergeo.com)

• Martin Riviere, Retired Geophysicist (martinriviere@btinternet.com)

• Angelika-Maria Wulff, Consultant (gp.awulff@gmail.com)

EAGE EDITOR EMERITUS Andrew McBarnet (andrew@andrewmcbarnet.com)

PUBLICATIONS MANAGER

Hang Pham (publications@eage.org)

MEDIA PRODUCTION

Saskia Nota (firstbreakproduction@eage.org) 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 PO Box 501711

Dubai, United Arab Emirates

• +971 4 369 3897

• middle_east@eage.org www.eage.org

EAGE ASIA PACIFIC OFFICE

EAGE Asia Pacific Sdn. Bhd.

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 LATIN AMERICA OFFICE

EAGE Americas SAS Av. 19 #114-65 - Office 205 Bogotá, Colombia

• +57 310 8610709

• americas@eage.org

• www.eage.org

EAGE MEMBERS’ CHANGE OF ADDRESS Update via your MyEAGE account, or contact the EAGE Membership Dept at membership@eage.org

FIRST BREAK ON THE WEB

www.firstbreak.org

ISSN 0263-5046 (print) / ISSN 1365-2397 (online)

Editorial Contents

29 Unsupervised clustering as a tool for geomorphological interpretation: a case study from a Sarmatian delta in Romania’s Dacian Basin

Luca Fava

Special Topic: Reservoir Monitoring

41 Twenty five years of monitoring the Troll gas and oil field with time-lapse gravity and seafloor deformation surveys

Siri Vassvåg, Felix Halpaap, Charlotte Faust Andersen and Laust Jørgensen

47 Musings on the Golden Fault and the Front Range Zone of Flank Deformation

Tom Davis

53 Ametista Block – An unusual prospect in Santos Basin – Albian atoll upon exhumed mantle

Pedro V. Zalan, Milos Cvetkovic, Henri Houllevigue, Kyle Reuber and Andrew Hartwig

61 Estimating petrophysical parameters from acoustic impedance and P to S wave velocity ratio using a simple rock physics model

Krishna Agra Pranatikta and Ignatius Sonny Winardhi

71 Implementation of time-lapse gravity and subsidence monitoring for optimising the development of the Scarborough gas field

Mohamad Yousof Hourani, Abid Ghous, Rabin Sridaran, Martha Lien, Siri Vassvåg and Hugo Ruiz

78 Calendar

cover: Reservoir monitoring at the Troll Field in the Norwegian North Sea. This month we present the results of time lapse gravity and seafloor deformation surveys at the field photo courtesy of Equinor. Musings

European Association of Geoscientists & Engineers Board 2024-2025

Near Surface Geoscience Circle

Andreas Aspmo Pfaffhuber Chair

Florina Tuluca Vice-Chair

Esther Bloem Immediate Past Chair

Micki Allen Contact Officer EEGS/North America

Hongzhu Cai Liaison China

Deyan Draganov Technical Programme Officer

Eduardo Rodrigues 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

Oil & Gas Geoscience Circle

Yohaney Gomez Galarza Chair

Johannes Wendebourg Vice-Chair

Lucy Slater Immediate Past Chair

Wiebke Athmer Member

Alireza Malehmir Editor-in-Chief Geophysical Prospecting

Adeline Parent Member

Jonathan Redfern Editor-in-Chief Petroleum Geoscience

Xavier Troussaut EAGE Observer at SPE-OGRC

Robert Tugume Member

Timothy Tylor-Jones Committee Member

Anke Wendt Member

Martin Widmaier 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 € 85.00 a year including First Break, EarthDoc (EAGE’s geoscience database), Learning Geoscience (EAGE’s Education website) and online access to a scientific journal.

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 First Break B.V., Journal Subscriptions, Kosterijland 48, 3981 AJ Bunnik, The Netherlands. Tel: +31 (0)88 9955055, E-mail: subscriptions@eage.org, www.firstbreak.org.

First Break is published by First Break B.V., The Netherlands. However, responsibility for the opinions given and the statements made rests with the authors.

COPYRIGHT & PHOTOCOPYING © 2025 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.

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).

Sanjeev Rajput Vice-President

Laura Valentina Socco President

Martin Widmaier Technical Programme Officer

Andreas Aspmo Pfaffhuber Chair Near Surface Geoscience Circle

Maren Kleemeyer Education Officer

Yohaney Gomez Galarza Chair Oil & Gas Geoscience Circle

Carla Martín-Clavé Chair Sustainable Energy Circle

Diego Rovetta Membership and Cooperation Officer

Peter Rowbotham Publications Officer

Christian Henke Secretary-Treasurer

GET 2025 returns to Rotterdam with four parallel conferences

EAGE’s Global Energy Transition (GET) Conference and Exhibition is scheduled to return to Rotterdam on 27-31 October 2025 with every hope of building on last year’s success which saw a 600% increase in attendance from the 2023 event.

GET 2025 will feature four concurrent conferences, each focusing on different aspects of the energy transition. This structure not only allows for field-specific discussions but also fosters an inter-disciplinary exchange of ideas, helping to bridge the gaps between various scientific and technical disciplines. The goal is to collaboratively tackle complex challenges and drive impactful, sustainable energy solutions.

The call for abstracts is now open, with submissions welcomed until the deadline of 15 June 2025. We invite contributions to all four technical conferences. These are the topics:

Carbon Capture and Storage Conference

Top of the agenda will be the operational aspects and challenges of delivering CO2 storage projects, addressing both scale and uncertainty in CO2 studies. Detailed discussions will include high-level reservoir screening techniques through to comprehensive reservoir characterisation. Monitoring strategies and technologies essential for ensuring CO2 remain contained within the storage complexes will also be a discussion point. The role of wells in CO2 storage, the impact of geochemistry and reactive transport modelling, as well as the application of AI and machine learning in enhancing CCS efficiencies, will be thoroughly explored. Additionally, the conference will investigate alternative reservoirs and innovative solutions for atmospheric CO2 removal, presenting a broad spectrum of approaches and technologies critical to advancing the CCS field.

Hydrogen and Energy Storage Conference

The intention is to review innovative subsurface energy storage technologies, such as hydrogen and underground gas storage, ammonia storage in wells, tunnels, and lined rock caverns, along with compressed air and underground thermal energy storage methods. Discussions will extend to underground pumped hydro storage, with examples from various pilot projects and case studies. Key topics will include the

Ian Jones new FWI short course at 2025 Annual 06

Introducing our Looking Ahead feature 09

Asian petroleum geoscience event leads the field 14

Engaged attendees in a full room during a key session at the GET 2024 CCS Conference.

GET 2024 Hydrogen & Energy Storage Conference chair, Karin de Borst, warmly greets attendees during the opening.

exploration and technological enablers for subsurface energy storage, addressing geological, mineralogical, microbiological, and geochemical considerations. The conference will also explore the natural mechanisms of hydrogen generation, its migration and trapping, and the modelling of hydrogen systems and basins. The integration of the energy storage and hydrogen value chain, including demand modelling, techno-economic assessments, and the social and regulatory frameworks necessary for advancing these technologies in global markets will all be part of the conversation.

Geothermal Energy Conference

Three major themes essential for advancing geothermal technology and its applications will form the basis of the conference. The first theme, Resource Assessment and Sustainable Development, will cover basin-scale geothermal resource assessments, reservoir characterisation, and

modelling, along with the integration of geophysics and multiphysics to enhance system efficiency. This segment also addresses decarbonising heating and cooling solutions. The second theme, Reservoir Development and Production Optimisation, will explore cutting-edge drilling technologies and well completion techniques. Discussions will extend to transitioning traditional oil and gas infrastructures to geothermal systems and the exploration of unconventional geothermal system developments. The final theme, Cross-Sectoral Solutions for Long-Term Sustainability, will examine the environmental and societal impacts of geothermal projects, highlight the role of data and digitalisation in geothermal energy, and discuss the potential of cogeneration and cascade use. This comprehensive approach aims to drive innovation and sustainable practices within the geothermal sector.

Offshore Wind Energy Conference

Here the spotlight will be on comprehensive advancements and strategies in offshore wind technology. The conference will cover a breadth of topics, starting with Geophysical Data Acquisition, featuring innovations in geophysical methods for near-surface and seabed data acquisition, and strategies for fixed versus floating offshore windfarms. The discussion will also explore the optimisation of acquisition strategies and future trends,

including acoustic underwater noise analyses and very shallow water data acquisition. The Seismic Data Processing discussion will look at advances in processing techniques, inversion of near-surface seismic data, and re-processing of oil and gas seismic legacy data tailored for offshore wind applications. The Seabed Mapping & Characterisation sessions will explore improvements in backscatter data processing, seabed target detection, and benthic habitat mapping. Meanwhile the Ground Modelling & Data Integration sessions will highlight case studies integrating geophysical, geological, and geotechnical data for comprehensive site analysis, while the conference will also discuss novel geophysical methods for seabed characterisation and the use of machine learning in geomodelling, stressing the ongoing evolution and integration of technologies in offshore wind development.

Whether you are a scientist, engineer, policy maker, industry leader, or a student, this conference offers a unique platform to learn from the best, share your knowledge, and contribute to shaping a sustainable energy future. We encourage you to explore the range of topics and take an active role by submitting your abstracts.

Carbon Capture and Storage Conference

Geothermal Energy Conference Hydrogen and Energy Storage Conference Offshore Wind Energy Conference

A presentation from the GET 2024 Geothermal Energy Conference.

A packed session at GET 2024 Offshore Wind Energy Conference.

DUG Elastic MP-FWI Imaging solves for reflectivity, Vp, Vs, P-impedance, S-impedance and density. It delivers not only another step change in imaging quality, but also elastic rock properties for quantitative interpretation and prestack amplitude analysis — directly from field-data input. A complete replacement for traditional processing and imaging workflows — we talk the talk and walk the walk!

info@dug.com | dug.com/fwi

OF SHELL

Conventional Workflow

DUG Elastic MP-FWI Imaging

Join Ian Jones for FWI course at EAGE Annual

For over a decade, Full Waveform

Inversion (FWI) has been reshaping the landscape of geophysical exploration. Renowned geophysicist Ian Jones, who has been at the forefront of migration and model-building advancements, is now focusing on this transformative technology. His new course, State of the Art in Full Waveform Inversion (FWI), to be held at the EAGE Annual 2025, aims to demystify FWI’s complexities and provide a comprehensive understanding of its capabilities and limitations.

‘I’ve been teaching the EAGE course on migration and model building for 15 years now,’ Jones explains. ‘Initially, the development of methods for both migration and model update was gradual and incremental. However, the industrial deployment of FWI represented more of a step-change.’

Jones underscores FWI’s revolutionary potential to streamline workflows, bypassing the months of preprocessing traditionally required to remove noise, ensure wavelet stationarity, and suppress multiples. ‘When fully implemented, FWI offers a means to avoid all these precursory processing steps,

delivering a product that utilises more of the full recorded wavefield,’ he says.

The past decade has witnessed remarkable advancements in FWI, primarily fuelled by rapid increases in cost-effective computational power. While the foundational ideas of FWI date back to the early 1980s with pioneers like Lailly and Tarantola, it is only in recent years that these concepts have been fully realised in practice.

Jones notes, however, that the journey hasn’t been without challenges. Early implementations faced limitations with error-prone least-squares minimisation methods and simplistic approximations. These hurdles, coupled with inflated expectations, initially hindered widespread adoption.

Today, the technology has evolved to incorporate robust alternatives such as anisotropic and visco-acoustic FWI, with ongoing progress toward elastic FWI. Yet, significant challenges remain. ‘The state of the art still deals with simultaneous inversion for only one or two parameters,’ Jones explains. ‘We still have a way to go in mathematical formulation, numerical implementation, and cost-effective delivery of true elastic attributes.’

One of the most exciting recent developments in FWI is the integration of machine learning. ‘Machine learning is making rapid

inroads into the formulation of numerical optimisation problems, including FWI,’ says Jones. This emerging technology has the potential to further refine and accelerate the inversion process, opening new avenues for exploration in challenging geological environments.

Jones’ course will provide participants with a thorough understanding of the diverse implementations of FWI, highlighting the motivations behind each method, their limitations, and their potential benefits. It’s a unique opportunity to learn from one of the field’s leading experts and gain insights into the future of geophysical exploration.

Whether you’re a seasoned professional or a newcomer to the field, this course promises to equip you with the knowledge and tools to navigate the evolving landscape of FWI. Don’t miss the chance to deepen your understanding and stay ahead in this dynamic discipline.

Join the course at EAGE Annual 2025

The State of the Art in Full Waveform Inversion (FWI) course will be held during the EAGE Annual 2025. Mark your calendars and secure your spot to explore the cutting-edge developments in FWI with Ian Jones. For more details and registration, visit eageannual.org.

Special transport offerings for Toulouse

When you come to Toulouse for the 2025 Annual, you will find that we’ve made special arrangements to make sure your transport needs are met. In collaboration with the Toulouse Convention Bureau, we are offering a complimentary public transport card to all attendees, which can be collected during badge pick-up. The card provides full access to Toulouse’s public transportation network, encouraging participants to use this eco-friendly option, easing congestion, thereby meeting the

conference’s commitment to sustainability.

We are also complementing the city’s existing public transport service with a dedicated, non-stop shuttle bus to and from the city centre to the MEETT conference and exhibition venue, a journey which can take anything between 25 and 45 minutes depending on traffic. The buses will run every 10-15 minutes. Exact details will be announced nearer the time of the event.

Attendees are invited to check eageannual.org/plan-your-visit to receive the latest information as it becomes available.

Credit: Ingrid Moya.

Ian Jones.

Enjoy massive savings with the All Access Pass

Registration for the 86th EAGE Annual Conference and Exhibition (2-5 June) in Toulouse is now open, and we recommend members take advantage of the All Access Pass, the way to experience everything this great event offers under the over-arching theme of ‘Navigating change: Geosciences shaping a sustainable transition’.

The All Access Pass not only grants admission to the Conference and Exhibition but also opens the door to a variety of interactive side activities. This includes entry to side activities like workshops, field trips, short courses, and hackathons. Each of these components is tailored to complement the learning and connections made during the main sessions, providing a holistic view of geosciences’ role in sustainable development.

Workshops and Short Courses

Participants can engage in any of the 22 workshops and a variety of short courses. These sessions are led by experts and cover a range of topics from geophysical monitoring of CO2 storage and AI in seismic processing to the latest in reservoir engineering for hydrogen storage and

full waveform inversion techniques. They offer to provide the route to advance skills and knowledge in specific areas of the geoscience disciplines.

Field Trips

The longstanding tradition of field trips offers educational and enjoyable guided visits to sites of relevance to the interests of members, such as natural hydrogen sources in the Pyrenees, dynamic analogues in groundwater reservoirs near Montpellier, and space exploration facilities in Toulouse. Each trip is designed to enhance theoretical knowledge with a practical view of realworld applications.

Hackathon

Included at the conference is a two-day hackathon guided by the EAGE AI Com

mittee. This event challenges participants to form teams and develop AI-driven solutions for enhancing power efficiency or optimising workflows. It’s a perfect

To fully experience what the EAGE Annual has to offer, the All Access Pass is indispensable. Early bird registrants can save up to €370 if they secure their pass by 15 March 2025. This substantial discount makes the All Access Pass exceptional value, ensuring participants can enjoy the complete range of activities.

Check out more details and sign up for the event at eageannual.org.

Discussion during a workshop.

Hackathon from Oslo Annual.

Short Course on how to avoid committing visualisation crimes

Dr Steve Horne (Silixa UK) discusses his forthcoming short course on Data Visualization Principles for Scientists on 27 March at EAGE Digital 2025.

What sparked your fascination with scientific data visualisation dating back 20 years?

I can remember very clearly! I had just started a new posting and needed to write an academic paper. I thought that there must be a rigorous set of scientific principles that I could use in preparing the figures. As with many others, I quickly discovered Edward Tufte’s books and started to apply some of those principles, not only in my own work but when reviewing papers. I was really surprised to find that the plots in our literature weren’t consistent with the ideas I had been reading. And so, over the next couple of decades, I started collecting various articles and publications with the initial idea of writing some articles to promote better data visualisation hygiene. I soon realised that I couldn’t easily cover all the material in a couple of tutorial articles, as I had originally planned, and so I started to write a book that was written on my long daily train commutes or whilst waiting for my children to complete their sporting activities. The book is nearly complete and has been for the last five years. It will probably keep me busy when I retire!

Why is accurate data visualisation so critical?

We owe it to our readers and to respect their time by efficiently communicating our ideas. If we don’t do this then, at worst, our audience will be led to the wrong conclusion or, more commonly, left confused or uninformed. I reviewed a year’s worth of publications from a couple of the major geophysical publications and by far the biggest pitfall was the use of inappropriate colour schemes, specifically the overuse of the rainbow-like colour schemes. It is well known that these are a terrible choice for scientific purposes despite the many publications explaining their failings. It seems rainbow colour schemes are the zombies of the data visualisation world – they will not die!

Were there any surprises when you were putting the course together?

One fascinating discovery was that John W. Tukey was a pioneer in data visualisation. Not only did he invent the Fast Fourier Transform and coined the word ‘bit’ (short for binary digit), but he also promoted the concept of exploratory data analysis, and invented the box plot. Among Tukey’s many memorable quotes was: ‘Visualisation is often used for evil

— twisting insignificant data changes and making them look meaningful. Don’t do that crap if you want to be my friend.’ No surprise to learn to learn that the W in John W. Tukey stands for Wilder.

Where does visualisation go wrong?

I could give you many examples as they are so abundant. For example, I just thumbed through last month’s First Break and it was full of the usual viz crimes such as inappropriate colour maps (rainbow), 3D effects, and pie charts. As part of the course, I collected a representative sample of these published visuals. Participants will get the opportunity to identify how the plots could be improved and remake the plots based on what they have learnt.

What can we expect to learn from your course?

I want to improve the standard of data visualisation for our community; so that we can elegantly and efficiently communicate our ideas. Participants will get to use these principles in their daily lives, not only in communicating their own work, but also when they review papers, and even when watching social and media reports where visualisation crimes are far more prevalent.

Properly displaying figures is essential in scientific papers.

Does this graph have an appropriate colour scheme? Determine that with the course.

In this new section we will every month highlight some of the key upcoming conferences, workshops, etc. in the EAGE’s calendar of events. We will be covering separately our four flagship events – the EAGE Annual, Digitalization, Near Surface Geoscience (NSG), and Global Energy Transition (GET).

23rd EAGE European Symposium on Improved Oil Recovery (IOR+ 2025)

2-4 April 2025 – Edinburgh, UK

Improved Oil Recovery (IOR) remains the foundation of the event, the ‘+’ indicates the integration of broader topics to reflect how reservoir technologies and innovations developed for oil recovery can contribute to a sustainable energy transition. The programme will feature diverse technical sessions, providing updates and results from the latest research into IOR/EOR technologies together with assessments of recent pilots and field-scale deployment showcasing the successes, technical challenges and new technologies deployed. In addition, the conference will cover (re)emerging applications where IOR/EOR experience and knowledge are fundamental, and wrap up with a panel discussion titled ‘The Future of IOR+’.

Register today at www.ior2025.org

First EAGE/SBGf Workshop on Marine Seismic Acquisition

21-22 May 2025 – Rio de Janeiro, Brazil

This event will emphasise innovations and sustainable solutions shaping the future of marine exploration and monitoring. Topics will span seismic sources and sensors, pioneering geophysical methods for CCS, permanent reservoir monitoring systems and deepwater technologies, and latest technologies designed to enhance efficiency while minimising environmental impact. Key areas include advances in marine vibrators, broadband and simultaneous source techniques, and the integration of fibre optics and ocean-bottom nodes for data acquisition. AI-driven applications, drones, and satellite technology for wildlife monitoring, alongside sustainable approaches to seismic acquisition will also be discussed.

Join us in shaping the future in marine seismic acquisition!

EAGE/AAPG Workshop on Tectonostratigraphy 2-5 November 2025 – Riyadh, Saudi Arabia

Tectonic wonders of the Arabian Plate will be centre stage for those eager to expand their knowledge of regional geology. The proceedings are set to explore the intricate relationship between tectonics, lithostratigraphy, and the structural phases that have shaped this geologically significant region, addressing topics such as salt tectonics, sequence stratigraphy, structural evolution, and emerging challenges in petroleum systems and CO2 trapping. Contributions will be led by representatives of major companies such as Aramco, ADNOC, Shell, and more. A two-day field trip is also planned featuring the East Janadriyah and Umm-Shual anticlines, offering insights into the region’s folds, faults, and their impact on petroleum systems.

Abstract submissions deadline: 5 May 2025

Third EAGE Seabed Seismic Today Workshop 24-26 November 2025 – Manama, Bahrain

After the successful Milan event, the third workshop in Bahrain will reflect on the evolution of acquisition and processing over the past few years. This has led to a step-change in efficiency and quality, with seabed seismic being the sought-after solution for subsurface imaging for exploration and development. Presentations can be expected from industry experts on advances in acquisition technology on both the source and receiver side, acquisition methodologies, along with processing, FWI and imaging. Also look forward to discussions focusing on seabed seismic applications for site surveys, offshore wind, marine minerals, CCS/CCUS, transition zone, ultra-shallow water and co-located fields.

Abstract submissions deadline: 30 April 2025

Submissions sought for themed issues in EAGE journals

Two of EAGE’s journals – Geoenergy and Petroleum Geoscience – are planning special themed issues to which members are encouraged to contribute. This presents a valuable opportunity to share your research with a global audience and contribute to key discussions shaping the future of geoscience and energy. The collections’ details are summarised here.

Geoenergy : CCS in the AsiaPacific region

This collection explores carbon capture and storage (CCS) technologies, policy frameworks, and deployment strategies in the Asia-Pacific region, with a focus on regional implementation and cross-border industrial partnerships and collaboration.

Geoenergy

Geoenergy : The minerals-energy nexus in Africa

The idea of the theme is to explore the relationship between mineral resources and energy systems in Africa, highlighting sustainability challenges, economic opportunities, and the role of critical minerals in the energy transition.

Petroleum Geoscience: Geoscience driving the North Africa and Eastern Mediterranean energy hub

The intention is to investigate the geology, resource potential, and energy prospectivity of North Africa and the Eastern Mediterranean, emphasising the role of geoscience in exploration, development, and regional energy sustainability.

The submission deadline for all collections is 30 April 2025. We encourage you to take advantage of this opportunity to have your research featured in our renowned journals. For more information, visit the respective journal websites.

Geoscience relevance to energy transition explored at LC Norway event

Geoscience skills and techniques for the energy transition was the theme of a meeting last November held by Local Chapter Norway at the Oslo office of TGS in a welcome return to in-person presentations.

Carine Roalkvam, regional geophysical advisor and support manager from TGS, led proceedings in the talk on ‘Advanced 3D seismic crossover technologies between hydrocarbon exploration, CCS development and offshore wind’ highlighting innovative seismic acquisition techniques. Roalkvam showcased some fascinating examples of how these new techniques have significantly improved subsurface imaging, sparking engaging discussions and questions from the audience about seismic acquisition.

Benjamin Bellwald from NGI offered a deep dive into the challenges of understanding the shallow subsurface in glaciated margins. His presentation, ‘Resolution requirements for characterisation of sedimentary environments in glaciated margins: A geomorphological perspective,’ addressed the complexity of glacially influenced

geology. He illustrated how the shallow subsurface exhibits considerable lateral heterogeneity, with features like eskers and moraines leading to difficulties in geological correlation. He emphasised that these variations in glacial deposits require high-resolution data to accurately characterise the subsurface. This is crucial for environmental assessments and energy development projects such as wind projects. The audience showed a keen interest in this topic, as evidenced by the detailed questions and discussions that followed.

The event was well-attended, with over 30 participants enjoying an evening of insightful talks, pizza, and

refreshments. Its success was a testament to the community’s enthusiasm for reconnecting and exchanging expertise on the cutting edge of geoscience. Such meetings are essential for advancing the collective understanding of geoscience’s role in the energy transition, and with ongoing support from local companies and EAGE. The chapter will continue to facilitate these valuable opportunities for its members.

Connect with EAGE Local Chapter Oslo

EAGE Local Chapter Oslo resumed its in-person events with a session focused on utilising geoscience skills and techniques for the energy transition.

Active and passive surface wave surveying in hard rock sites

The EAGE Local Chapter Netherlands held its year-end event of 2024 at the Delft University of Technology campus. EAGE President Prof Valentina Socco, who recently joined TU Delft, was the guest speaker. Her presentation centered on active and passive surface wave surveying in hard rock sites, showcasing advancements in surface wave surveying techniques.

Dong Zhang (EAGE LC Netherlands) writes: The talk began with an introduction to surface waves, particularly Rayleigh waves, and their propagation in homogeneous and layered media. Prof Socco discussed methodological developments such as surface wave tomography for estimating shear (VS) and compressional (VP) wave velocities. She presented case studies demonstrating the application of these methods in geotechnical characterisation, mining sites, and geological modelling. Challenges in surveying hard rock sites were addressed through improved data acqui-

sition techniques, including interferometry and reciprocity principles. Prof Socco also covered inversion techniques like laterally constrained inversion and surface wave tomography, emphasising their role in building robust subsurface models. The W/D method’s application was detailed, showcasing its ability to estimate time-average velocities and its sensitivity to Poisson’s ratio. Case studies from the Middle East and the Siilinjärvi apatite site illustrated the efficiency and quality of these methods in various challenging environments. The talk concluded with a discussion on future directions and potential developments in surface wave surveying techniques.

After the talk, the audience engaged with numerous questions. One attendee was particularly interested in combining passive and active seismic surface wave inversion. The use of higher

modes was also discussed, along with the advantages of multi-component data compared to single-component data for surface waves. The role of full waveform inversion for surface waves was briefly mentioned and discussed. Many students and young professionals attended the event and enjoyed a networking dinner afterward. The efforts and contributions of our invited speakers are highly appreciated. Those interested in staying updated on chapter initiatives are encouraged to follow the Chapter’s LinkedIn page.

PIET GERRITSMA

Prof Socco speaking at Chapter meeting.

LC Netherlands photo op.

Promise of marine vibrators discussed at London LC meeting

For its penultimate technical lecture of 2024, EAGE Local Chapter London invited Dr Tiexing Wang (Shearwater) to present acquisition and processing results from the first broadband 3D marine vibrator data in the North Sea. Some of the community members gathered at Imperial College while others joined online, many of whom from overseas, making it a truly global gathering and allowing for students and professionals to meet experts in the field, exchange ideas and prompt discussion – all in a friendly and relaxed setting.

Wang started his presentation with a brief history of the marine sources development and where on the timeline the marine vibrator (MV) is positioned. It will not come as a surprise that the concept of MV is as old as that of the air-gun. He reminded us that the many attempts to build a marine vibrator with more or less documented field results and pointed out that the development has certainly accelerated in recent years, due to air-gun permitting and environmental concerns in general.

To meet the energy transition goal and the demand for environmentally friendly sources, Shearwater has been developing more advanced MV technology and its associated data processing methods. Recently an alpha test was conducted in the

North Sea where among other field configurations a 3D source carpet swath of data was acquired with a single marine vibrator (sweep frequency 3-150Hz). In the same area a PRM data has been available and was used to benchmark the MV results. Wang described the specific processing workflow that was applied to the data generated by the continuous sweeping to generate an equivalent of an impulsive gather for a one-to-one comparison of the raw common receiver gathers. Subsequently shown processing results not only demonstrated quality comparable to traditional airgun arrays but also highlighted the MVs’ ability to generate signals down to 1.5Hz over large offsets. These findings indicate that MVs can act as a sustainable alternative to traditional sources for large-scale broadband acquisition. Moreover, MVs hold significant potential for enhancing 4D seismic monitoring, as they provide precise control over their emitted signal.

The future of this technology lies in advancing marine vibrators as a sustainable and precise alternative for

conventional seismic acquisition, leveraging greater operational efficiency and better control of emitted seismic frequencies, enabling enhanced 4D monitoring, improved ultra-low-frequency data quality, environmentally friendly energy exploration, and carbon storage monitoring.

PSDM Kirchoff image of UDD.

Shearwater marine vibrator technology.

On site gathering of LC London lecture participants.

WOMEN

IN GEOSCIENCE AND ENGINEERING

JOIN A NETWORK THAT MATCHES YOUR AMBITIONS

EAGE members at all stages of their career connect through a network of dedicated communities, unlocking new skills and new connections to empower their journey ahead.

MARTA CYZ

EAGE Women in Geoscience & Engineering Committee Member

I had the privilege of being part of the WGE team since 2019. Our focus has always been on empowering women and fostering a supportive community, and I did my best to contribute. During difficult times in my life, I was grateful to find strong support from the Committee. Being part of WGE not only helped me develop new skills but also, hopefully, positively impacted others and contributed to building a more inclusive environment.

EAGE’S LARGEST SPECIAL INTEREST COMMUNITY

YEARS OF MENTORING PROGRAMMES 9

INVEST IN YOUR CAREER DEVELOPMENT

STUDENTS CAN APPLY FOR THEIR FIRST YEAR OF MEMBERSHIP FOR FREE (RE)ACTIVATE YOUR MEMBERSHIP TODAY!

Asian petroleum geoscience event leads the field

‘Forging Collaboration and Innovation in Advancing Asia’ was the theme for the highly successful Asia Petroleum Geoscience Conference & Exhibition (APGCE) 2024 hosted by PETRONAS and organised by ICEP with EAGE as co-organiser. The event brought together more than 3500 participants from 21 countries, representing 108 organisations.

On the evidence of the 2024 proceedings, APGCE is rapidly becoming the premier event for Asia’s geoscience community to discuss the region’s opportunities and responsibilities. The theme underscored the critical need for cooperative efforts and creative approaches in the region’s petroleum geoscience sector to drive sustainable development and ensure energy security.

Ahmad Faisal Bakar, chairperson of APGCE 2024 in his opening address, made the point. ‘Together, we can share knowledge, create new technologies, and develop scalable solutions. The geoscientists, technology providers, and energy leaders here have the unique opportunity to shape the future. It is our responsibility

to push the boundary of innovative frontiers to achieve efficiency and deliver the discoveries that our society needs.’

The Opening Ceremony was attended by experts from various geoscience field with guest of honour Montri Rawanchaikul, CEO of PTT Exploration and Production (PTTEP), delivering an inspiring keynote address emphasising the hope for everyone to work together collaboratively using technology while sharing the knowledge.

Notable sessions included the executive plenary sessions which explored the perspectives of host authorities and operators, while the executive panel sessions focused on geophysics and geology respectively. The Technical Programme – 81 oral sessions and 70 poster sessions – offered participants a comprehensive look at emerging trends and developments in geoscience.

A staple feature of the conference was the Core Display, showcasing 13 Malaysian core booths that offered participants access to rocks and mineral samples from the region. This interactive exhibit provided delegates with a unique opportunity to engage in hands-on learning enabling them to examine core samples up close and apply theoretical knowledge to realworld materials.

The APGCE Exhibition featured 35 leading companies, including PETRONAS, Halliburton, SLB and TGS, presenting the latest geoscientific innovations and solutions. Over the course of two days, the exhibition welcomed more than 1700 visitors, creating a dynamic platform for knowledge sharing and collaboration.

APGCE 2024 hosted a series of pre-conference events that fostered innovation, talent development and collaboration, beginning with a field trip to Langkawi. A GeoHackathon also took place with 60 participants leveraging big data analytics and machine learning to solve real-world geoscience challenges.

Alongside the main conference, the Student Programme attracted 75 aspiring geoscientists in a mentorship and networking session with industry leaders, while the brand new NexGen Explorers Programme was attended by 47 young energy professionals to gain valuable networking opportunities and industry insights to advance their careers.

APGCE 2024 hopefully leaves a lasting foundation for knowledge-sharing, partnerships and innovation, influencing change in Asia’s geoscience landscape. Looking ahead, the next edition of APGCE expects to build on the momentum established in advancing the dialogue and collaboration needed to shape the future of Asia’s geoscience sector.

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!

Core display.

Visitors exploring exhibition booths at APGCE 2024.

Personal Record Interview

From teenage rebel to head of geoscience

Born in East Germany before unification, Susanne Rentsch remembers the Berlin Wall coming down. After high school she headed for Berlin, enrolling in geoscience at Freie Universität and working in a lawyer’s office to earn some money. Ultimately she joined SLB (WesternGeco) as a researcher focused for many years on the Isometrix seismic acquisition system. Today, she is head of geoscience at Shearwater GeoServices.

East Germany upbringing

I was born and raised in a tiny village in East Germany, near the Czech and Polish borders. My family had no academic background — my mum was an anaesthetic nurse, and my dad a tiler. East Germany heavily promoted youth programmes, particularly in sports. I spent my spare time training in cross-country and alpine skiing earning medals until being a rebellious teenager got in the way of structured training. I was too young to understand the restrictions that defined the political system but remember the joy of people when the Berlin wall came down. Most memorable to me was the sudden access to bananas and oranges without restriction.

Post-school education

A few weeks after finishing school, I packed a bag, moved to Berlin with only €100, and started building a life. I found a job, a small apartment, and enrolled at university. Torn between law and geoscience, I ultimately chose geophysics but kept law as an option by working at a law firm for quite a few years.

Joining WesternGeco

At an EAGE event in Paris, I met a Senior Researcher from SLB who later offered me an internship at its Cambridge Research Center. That internship turned out to be truly life-changing — it not only shaped my professional path but also my personal one, as I met my future husband who was also an intern there. After completing my PhD, I joined WesternGeco.

Isometrix

through the years

For nearly a decade, I worked on Isometrix. It was exciting, groundbreaking work that earned multiple awards and patents for our team. Later, I transferred to the engineering centre in Oslo to oversee and advance real-time processing. The Isometrix streamer remains an engineering marvel with capabilities we’ve not yet fully explored, all enabled by live access to over 600,000 traces in a conventional streamer spread.

Research trials and tribulations

Innovation comes with its share of disappointments. Some brilliant hardware designs from the engineering team never reached the market due to shifting priorities and the need to align with the energy transition. It’s tough to see excellent work not bear fruit, but the learnings are invaluable and will stay with me no matter what. On the success side, I’m proud of leading Shearwater’s fastest engineering project to date: bringing the Harmony low-frequency-rich broadband source to life and to market during Covid. I had limited experience with source technology at the time and the challenge pushed me far out of my comfort zone.

Transfer to Shearwater

The transition to Shearwater during SLB’s divestment of the Marine Acquisition business was difficult. Downsizing and uncertainty took a toll, but spotting Irene (Shearwater CEO) in the WesternGeco lobby reignited the spirit. Shearwater’s agile and empowering culture was refresh-

ing, and being part of shaping a growing organisation has been a highlight of my career.

Your current role

A year ago, I became head of geoscience at Shearwater. While the company already had strong geoscience branches, it wasn’t unified under a single leadership. My role is to bring them together, fostering close collaboration not only across all disciplines internally but also with our clients. With a one-team culture, we deliver expertise to business lines.

Energy transition thoughts  Affordable, sustainable, and reliable energy is crucial for geopolitical and environmental stability. Geoscience is indispensable in achieving this goal, playing a vital role in today’s energy supply and tomorrow’s innovations. The energy transition presents immense opportunities for growth and groundbreaking solutions, reinforcing the importance of our field. There is more to explore in securing energy now and into the future.

Life in Norway

Living in Norway is a privilege, especially for someone who loves the outdoors as much as I do. As a family we enjoy hiking, skiing, and simply spending time in nature throughout the seasons. Gardening is another hobby of mine, though it’s a challenge with Norway’s long winters and short growing season. Every strawberry, tomato, and green bean I grow brings immense satisfaction.

Susanne Rentsch

CROSSTALK

BY ANDREW M c BARNET

BUSINESS

Kalaallit Nunaat wants none of it

President Trump has an uncanny ability to expose and exploit the vulnerablity of others, a trait that is in full rein now that he is back in office. No need to rehearse here all the potential weaknesses in the US Constitution that he has already threatened in an unprecedented blitz of executive orders rolling back the power of federal government.

In this context President Trump’s musings about buying Greenland are entirely in character. Not for the first time he has touched on the increasingly fragile relationship between Denmark and Greenland (population 56,000) where calls for total independence are ever present. Currently Greenland (known as Kalaallit Nunaat to its indigenous Inuit people who make up 88% of the population) is, like the Faroe Islands, an autonomous territory of Denmark.

Trump’s timing is perfect because on 11 March Greenlanders go to the polls with independence again an issue. Prime Minster Múte B. Egede, leader of the pro-independence party Inuit Ataqatigiit, heads a coalition government and has made clear that Trump’s overtures are unwelcome stating ‘We do not want to be Danes. We do not want to be Americans’.

‘Greenlanders going to the polls’

Meantime Danish Prime Minister Mette Frederikse called Trump to tell him Greenland was not for sale.

Egede has suggested that a post-election referendum on full independence could be on the cards. This might seem reckless because a successful outcome could have major consequences. The Danish government’s position is that the wish of Greenlanders would be respected. The kicker is that Denmark’s block financial grant would be in jeopardy. This is a big deal. Half of the self-rule government’s revenue comes from Danish subsidy in an economy where over 40% of the workforce are employed by administration, and is basically dependent on its fish export industry, tourism and some mining. A 2019 poll showed that 67.8% of Greenlanders support independence from Denmark sometime in the next two decades, but in an apparent recognition of the economic repercussions, a 2017 poll showed that 78% would oppose independence if it implied a lower standard of living.

So we have it, Trump feeling out the weaknesses of an object of prey. He has even suggested that Denmark may not be the legitimate ruler of Greenland. This is a stretch but a good windup given the many sagas of nordic history over the centuries. It seems early Norse settlements including Eric the Red, who named the country, died out sometime in the early 14th Century during the so called Little Ice Age, leaving only local inuit inhabitants. Their ancestory is said to date back to the Thule people who arrived in Greenland between 1000 and 1300 A.D. A Lutheran missionary/ evangelist Hans Egede in the early 18th century helped the Dano-Norwegian government of the time claim Greenland as a colony. Things unravelled when the Denmark-Norway union was dissolved by the Treaty of Kiel in 1814 with Denmark assuming Greenland as part of its monarchy and keeping the land mass and its trade very much to itself. Norway continued to claim territorial rights until an international arbitration in 1931 went Denmark’s way.

Largely unnoticed in the Trump-generated furore is that the US first showed interest in Greenland around the time Secretary of State William Steward in 1868 negotiated the purchase of Alaska from Russia for $7.2 million. This expansionism was soon being mocked as ‘Steward’s folly’, so his enthusiam to also buy Greenland got the shoulder. However he presciently reported the bountiful natural resources especially coal, whale blubber and cryolite that could enable the US to ‘command the commerce of the world’. Nothing came of a further attempt to negotiate a deal with Denmark in 1910.

In 1946, a $100 million cash offer from the US was rejected, but the thinking behind it resonates today. During the Second World War, the Nazi occupation of neutral Denmark raised fears that Germany would invade Greenland. So, in 1941, the US agreed a ‘Defence of Greenland’ treaty with the Danish ambasador in New York. As a result the US established a military presence in Greenland which continues today on a modest basis under a treaty signed in 1951. At the peak of the Cold War, the US Thule Air Base (now

Pituffik Space base) housed 10,000 American troops. It now hosts a missile early warning system, space surveillance installation and is doubtless used for other military purposes.

Ironically climate change – one of Trump’s bête noires – is a major factor in bringing Greenland’s importance back into US strategic thinking. It may not be so long before global warming may conceivably free up shipping routes through the Arctic. It is not too far fetched to suggest that these options could emerge as plausible alternatives to transport through the Suez and Panama canals. That would put a more accessible Greenland in prime position to provide harbour for shipping, international trade and other roles.

For the US, additional maritime activity in the Arctic has more menacing implications. It fears Russia and to a lesser extent China can come uncomfortably closer to the US in a scenario where ports in Greenland would act as enablers. Russia’s increasing belligerence has prompted Finland and Sweden to join NATO which already includes the Arctic countries of Canada, Denmark, Finland, Iceland, Norway and the US.

in the era of climate change and would endanger traditional industries and the way of life. In 2021 the issuing of new oil and gas exploration was officially banned by the incoming Inuit Ataqatigiit administration.

A much more flexible approach is being adopted by the same government regarding the exploitation of Greenland’s massive mineral riches including unexploited critical minerals needed for the world’s energy transition goals. These are what President Trump is most likely talking about, not ruling out force or economic sanctions to get his way.

‘Natural resources of almost mythical proportions’

As geoscientists know better than anyone, all these high level considerations cannot disguise the more pressing reason for Trump eyeing up Greenland real estate. The world’s biggest island has a wealth of unrealised natural resources of almost mythical proportions, including offshore oil and gas, which will come more accessible if the big melt continues.

Most of the undiscovered oil, gas, and natural gas liquids are likely to be offshore associated with the Upper Jurassic Composite Total Petroleum System. A much quoted estimate from the US Geological Survey in 2007 guessed that the East Greenland Rift Basins Province contains approximately (mean) 31,400 million barrels oil equivalent of oil, natural gas, and natural gas liquids. During a high oil price period in the 1970s, some limited seismic persuaded oil companies to drill five wells, all dry, plugged and abandoned before turning to more hospitable climates (For more, see Proving up petroleum prospectivity, J.C Olsen, First Break, Vol 24, 4, 2006).

From 2007 TGS embarked on multi-year seismic acquisition programmes on behalf of licensed operators, when oil was supposed to be running out. In 2009 or thereabouts new marine geophysical contractor Polarcus announced the launching of a fleet of innovative design vessels including some purpose-built for seismic surveys in arctic conditions, an indicator of anticipated E&P industry interest. However, none of later substantial seismic shot around Greeland on behalf of TGS and others translated into successful exploration wells, e.g., Cairns Energy came up with nothing after a drilling campaign ending in 2011. Just about all the majors relinquished their licences discouraged by the drop in oil price in 2014, the rise of US shale production and of course the challenging arctic environment.

Even if the industry wanted to reengage, Greenland’s government has made clear that oil and gas activity is inappropriate

Coincidentally last month Greenland’s government published Mineral Resources Strategy for 2025-2029 outlining a vision for a sustainable and prosperous future driven by responsible development of its mineral wealth emphasising sustainability, attracting investment, capitalising on critical mineral resources and further geological mapping of areas of commercial interest. (As recently as 2019 under the Biden admistration, the US Department of State and the Greenland Ministry of Mineral Resources and Labour joined forces on a new aerial hyperspectral survey to boost mineral exploration investment in South Greenland.)

The takeaway is that Greenland is happy to collaborate with the likes of the US and EU countries but on its own terms. In essence this means not rushing things, i.e., taking account of the wishes and capabilities of its tiny population and the need for a gradual build-up of infrastructure investment. In fact Greenland has had plenty of experience with the ups and downs of mining investment, minerals have been a source of revenue since the 1700s, with cryolite, copper and graphite being the main early targets. From 1958 until 1980 uranium was also exported before Denmark opted against nuclear power. Numerous mining licences have been issued over the years with varying outcomes, see Greenland mineral exploration history Flemming G. Christiansen, Mineral Economics (online), October 2022.

Currently Greenland has two operating mines, one producing anorthosite, the other rubies and pink sapphires with other projects in the works including zinc and titanium mines. Rare earths are definitely on the horizon but all eyes are on what happens at the large-scale Kvanefjeld project in southern Greenland proposed by Energy Transition Minerals. The company claims over one billion tonnes of mineral resources have already been delineated making it the most significant western world producer of crtical rare earths. However the project has fallen foul of legislation passed in 2021 that precludes extraction of the uranium associated with the other minerals at the site.

So far Greenlanders have been able to lay down the law but how long it can hold out against international commercial and geopolitical pressures to open up its mineral resources, not to mention the unwanted attention of the US president, is another matter.

Views expressed in Crosstalk are solely those of the author, who can be contacted at andrew@andrewmcbarnet.com.

INDUSTRY NEWS

Norway hands out 53 licences in APA 2024

Twenty companies have been offered ownership interests in a total of 53 production licences on the Norwegian continental shelf (NCS) in the 2024 awards in pre-defined areas (APA).

The authorities reviewed applications from 21 companies in autumn 2024. Of these 53 production licences, 33 are in the North Sea, 19 in the Norwegian Sea and 1 in the Barents Sea, while 20 licences are additional acreage for existing production licences.

‘This year’s awards show that the companies on the NCS are still very confident in their ability to make more discoveries near existing oil and gas infrastructure,’ said Kalmar Ildstad, director of licence management in the Norwegian Offshore Directorate.

Equinor was the biggest winner in the APA round, scooping 27 production licences, 20 in the North Sea, six in the Norwegian Sea and one in the Barents Sea. Equinor is the operator of seven of the licences and a partner in 20.

‘We will continue to make robust investments, and our ambition is to drill around 250 exploration wells by 2035. In order to do this, we need regular access to acreage,’ said Jez Averty, Equinor’s senior-vice president for subsurface, the Norwegian Continental Shelf. ‘We have a significant portfolio of smaller discoveries near existing infrastructure. We’re working alongside the supplier industry to accelerate developments and reduce costs, which will ensure that several of these discoveries can come on stream even earlier. Despite most exploration wells being drilled near existing infrastructure, it is important that we also explore new areas and new ideas and concepts with the potential for more major discoveries. Our confidence in the Norwegian shelf remains strong.’

Aker BP has won ownership interest in 19 exploration licenses in APA 2024. For 16 of the licenses Aker BP is also granted operatorship. It has won licences in the North Sea and the Norwegian Sea and an operatorship on the former Frigg field, where the partnership plans to drill an exploration well in Q2 2025.

Per Øyvind Seljebotn, SVP exploration and reservoir development at Aker BP, said: ‘Although the NCS is maturing, we manage to continuously identify new opportunities. Leveraging new technology, digitalisation and investments in new data are crucial for creating good exploration opportunities for many years to come. Our strategy is to have a portfolio

of exploration licences that provides a good balance between exploration wells near existing fields and infrastructure, and wells that, if successful, can form the basis for standalone developments,’ said Seljebotn.

Norwegian oil and gas operator DNO has been awarded participation in 13 exploration licenses, of which four are operatorships. Of the 13 new licences, 10 are in the North Sea and three in the Norwegian Sea.

Vår Energi has been offered nine licences in the North Sea, six licences in the Norwegian Sea, and one in the Barents Sea. Most of the licences are close to existing infrastructure.

OKEA has been offered interests in eight new production licences on the NCS. The awards strengthen OKEA’s portfolio of near-field exploration opportunities around the Draugen, Gjøa, Brage and Ivar Aasen production hubs.

PL 1266 and PL 1252 are awarded with OKEA as the operator, located close to the Draugen field in the Norwegian Sea, and close to the Brage Field in the North Sea.

Sval Energi has been awarded seven new exploration licences, two as operator and five as partner.

The Norwegian Offshore Directorate is launching an APA 2024 dashboard showing acreage, work programmes and companies with associated ownership interests and roles with GIS interfaces (Esri platform).

TGS expands CO2 storage assessment to seven new basins

TGS has announced an expansion of its CO2 Storage Assessment initiative for 2025 to enhance understanding of CO2 sequestration potential across critical regions along with basin-scale stratigraphy, reservoir properties, formation penetration and the associated risks related to pressure and seals. TGS is expanding its CO2 Storage Assessments to include seven additional basins across the Gulf Coast and West Midwest. These new basins include the Central Gulf Coast–Haynesville, Uinta Basin, Piceance Basin, Greater Green River Basin, Wind River Basin, Powder River Basin and the Greater Williston Basin. This expansion complements the company’s existing models in the Midwest-Northeast regions (Illinois, Michigan, an Appalachia basins) and the Gulf Coast (East Gulf Coast and Gulf Coast regions), further solidifying TGS’ comprehensive data coverage.

Will Ashby, executive vice-president of new energy solutions at TGS, highlighted the importance of these assessments: ‘Our expanded CO2 Storage Assessments

provide detailed mapping of stratigraphic architecture and petrophysical properties, delivering actionable insights into potential storage sites. By leveraging quad combo log data, inferred curves, core samples, bottom hole temperatures, and wireline formation tests, we are equipping the industry with precise evaluations of storage capacities. These assessments also incorporate analyses of seal thicknesses to evaluate containment potential, ensuring a robust understanding of storage feasibility.’

Meanwhile, TGS has won a 30-day offshore wind site characterisation contract on the UK Continental Shelf. The vessel Ramform Vanguard will mobilise for the project in Q2 2025.

Kristian Johansen, CEO of TGS, said: ‘Our geophysical approach for mapping the shallow subsurface layers with an ultra-high resolution 3D streamer is significantly more efficient than conventional site survey solutions. Energy companies value the shorter lead time we can offer to access high-quality data.’

The vessel will start the project after completing a 60-day offshore wind site characterisation contract, also in the UK continental shelf for a repeat customer in the first quarter.

The vessel is equipped with ultra-high-resolution 3D (UHR-3D) streamers. The streamer technology samples the seismic wavefield at a high spatial and temporal rate providing high-resolution data of the shallow subsurface targets for wind farm development.

Johansen said: ‘This project further highlights the integral role UHR-3D acquisition has in providing our clients with better seismic data and helping them make data-driven decisions for their wind farm developments.’

Energy Drilling to acquire SeaBird Exploration

Energy Drilling has agreed in principle a share-for-share acquisition of seismic vessel provider SeaBird Exploration. ‘The combined company will be a diversified offshore oil and gas services provider with strong cash flows and significant capacity for near-term shareholder distributions,’ said SeaBird.

The transaction will be carried out by issuing approximately 651 million new SeaBird shares to Energy Drilling shareholders. The seismic and drilling businesses will continue to operate as Seabird Exploration and Energy Drilling.

Headquartered in Singapore, Energy Drilling controls approximately 38% of the world’s actively marketed tender rigs, strategically positioned to address Southeast Asia’s growing demand for natural gas to fuel its growth. With approximately 80% of available days contracted for

2025 and 2026 Energy Drilling has a firm revenue backlog of $490 million.

The combined company will have a pro-forma market capitalisation of $381 million and net debt of $44 million.

‘Merging with an operationally and financially robust market leader provides our shareholders with increased scale and reduced operational risk. Energy Drilling shares our focus on niche market leadership with strong profitability and a capital allocation strategy that prioritises distribution of all excess cash to shareholders. Based on already solid backlog visibility, our ambition is to create a leading dividend platform within offshore oil and gas services,’ said Ståle Rodahl, executive chairman of Seabird Exploration.

The agreed exchange ratio will result in Energy Drilling shareholders owning 89% of the combined company.

The transaction is supported by the board of directors of both companies and the shareholders of Energy Drilling. In addition, the five largest shareholders in SeaBird; MH Capital, Anderson Invest, Alden, Grunnfjellet and Storfjell have expressed support for the transaction. Together with shares held by the BoD, this constitutes approximately 39% of the shares outstanding in SeaBird.

The financial transaction remains subject to final documentation, approval of the transaction by a SeaBid shareholders meeting, confirmatory due diligence by both parties, relevant regulatory approvals and consents. The transaction is expected to close in Q2 this year.

Should the deal go ahead, the merged company will have ‘ample financial flexibility to pursue growth opportunities’.

Ramform Vanguard.

Up to 75 high impact wells will be drilled around the world this year, says Westwood

Some 65-75 high impact exploration wells are expected to be completed this year, compared to the 69 completed in 2024, according to the energy analyst Westwood.

As of the end of January, seven high impact wells have been completed, 22 wells are drilling, 20 firm wells have been identified and an additional 23 wells are classed as probable.

Twenty one frontier wells are expected in 2025, a small increase on the 19 wells in 2024. Eleven of these 21 are targeting frontier basins, whilst new plays will be tested in the proven Sabah, Rio Muni, Western Black Sea, Suriname-Guyana and Cauvery basins. Emerging play wells are expected to account for ~30% of the high impact inventory in 2025 down from 36% in 2024. High value, >100mmboe prospects in mature and maturing plays are forecast to make up 40% of the high impact programme, an increase on last year. The Arabian, Campos, Gulf of Mexico, Kutei, Norwegian Sea, Santa Cruz and Santos basins will all have multiple high impact maturing/mature play prospects drilled.

QatarEnergy is expected to be the most active explorer in 2025, participating in 13 high impact wells. All of these are operated by supermajors apart from one well in Brazil, operated by Petrobras. Chevron will rank second, participating in seven wells. Six companies are expected to drill six high impact wells in 2025, including all of the remaining supermajors.

In Africa 14 high impact wells are expected to be drilled. There are expected to be 7-10 wells in the newly opened Orange Basin. Key wells for the Orange Basin include Olympe-1X and Sagittarius-1X. Azule Energy is expected to drill the Kianda-1 well in the outboard area of the Congo Basin, Angola in 2H 2025, and there are potential high impact wells being drilled offshore in the Namibe, Rio Muni and Tano basins, as well as potential frontier onshore tests in the Cabora Bassa and Kavango basins.

High impact drilling in North America continues to decline, with only five high impact wells currently anticipated to be

drilled in 2025, down from 13 completed in 2024 and 20 in 2023. Four high impact sub-salt Miocene tests are expected to complete in the US GoM, including the Far South well which is currently drilling. In Alaska, the Armstrong-led JV is expected to drill the Sockeye-2 well in the 20242025 drilling season attempting to extend the Cretaceous topset play further east.

There could be 17 high impact wells in South America in 2025, making it the busiest region globally. There will be key wells at Andorinha in the Campos Basin, south of Marlim Sul, and the Bumerangue well in the Santos, which will attempt to

in 2025, with Elektra currently drilling, testing a significant extension of the Nile Delta Miocene play and Pegasus testing the emerging Cretaceous carbonate play. Matsola offshore Libya is testing the offshore extension of the Sirte Basin. In the Western Black Sea, OMV Petrom will drill the Vinekh well aiming to extend the Sakarya play from Turkey into Bulgaria. Elsewhere, high impact wells will be drilled in Kuwait, Kazakhstan, and the UAE.

Fourteen high impact wells are expected in Asia Pacific, including the Hai Su Vang-1X well which was completed as a discovery in early 2025. Key frontier

extend the pre-salt play further south. In Colombia, Petrobras is expected to drill the Buena Suerte well targeting assumed Miocene reservoirs over a basement high close to the multi-tcf Sirius discovery. A key well in Suriname in 2024 is Korikori, which will test a shallow-water Upper Cretaceous play inboard of the prolific deep-water play. Two wells are also expected in the Demerara area offshore Suriname targeting the Macaw and Araku Deep prospects.

In the Mediterranean, Black Sea and Middle East, 14 high impact wells are expected. Key wells to watch are the dry Khendjer well and the Nefertari gas discovery in the Herodotus Basin offshore Egypt. Two wells are expected offshore Cyprus

wells are expected offshore South Korea at Daewanggorae, as well as at Mailu offshore Papua New Guinea. Drilling will continue in the Kutei Basin offshore Indonesia, and at the Megah high impact well offshore Malaysia. There will be at least 5-6 high impact wells in India from Oil India and ONGC across the east coast and Andaman Islands basins

In Europe, eight high impact wells are expected. Two high impact wells are expected in the Barents Sea, including the frontier ILX well Elgol. In the UK, the Dabinett well will test an emerging Permian carbonate play in the Southern North Sea, and in Poland the frontier Wolin East-1 well is expected to complete in the coming months.

High impact exploration activity and projection for 2025. Right – split by play maturity. Source: Westwood Wildcat.

Viridien wins Oman land imaging contract

Viridien has won a three-year contract from Petroleum Development Oman (PDO) to provide advanced land seismic imaging services at its dedicated processing centre (DPC) in Muscat, Oman.

Viridien geophysical experts at the Muscat centre, its largest DPC worldwide, will work to deploy advanced proprietary algorithms to bring step-changes

in image quality to PDO’s ever-growing library of seismic data, said Viridien.

Oman land data is characterised by complex near-surface conditions and strong multiples. High-resolution velocity model building, and elastic full-waveform inversion will be key to overcoming these challenges and to enhancing subsurface understanding, said Viridien. The company also will address chal-

lenges such as increased data density, developing land 4D monitoring and reinforcing synergies between seismic imaging and reservoir characterisation. To support these capabilities, Viridien HPC & Cloud Solutions specialists will deliver the in-house high-performance computing (HPC) capacity required to implement the most advanced workflows.

GTGi acquires 30,000 STRYDE land seismic nodes

GTGi International has acquired 30,000 STRYDE nodes to be deployed as part of a series of contract wins to deliver 2D and 3D land seismic projects across Europe.

GTGi is delivering seismic data for a range of energy exploration and decarbonisation initiatives across the EU, including carbon capture and storage (CCS), geothermal energy, and hydrogen resource exploration.

Mike Popham, CEO of STRYDE, said: ‘This agreement further underscores our ability to transform seismic data acquisition by delivering solutions that not only enhance trace density for superior imaging but also do so at no additional cost and with unprecedented speed, even

in challenging urban environments and in difficult terrain.’

David Dupuy, director of GTGi, said: ‘Apart from delivering superior data quality, STRYDE’s nodes offer unparalleled operational advantages. They require 4 to 5 times fewer vehicles and personnel for transportation, can be deployed using light vehicles as opposed to heavy trucks, eliminate the need for receiver line cutting, and have minimal environmental impact due to their small and discreet footprint. Combined with our expertise in operating multiple vibrators simultaneously within a given area, these benefits significantly reduce acquisition time while ensuring the highest data quality.’

AI will ultimately lead to net zero, says Shell

Artificial intelligence could lead to significant energy demand growth as the technology is implemented across industries, but its use could ultimately halve carbon intensity by 2050, according to a report from Shell.

As productivity improvements resulting from AI — including automation, especially in manufacturing — enable major economic growth, consumption of oil will continue to expand by three to five million barrels a day into the 2030s, before peaking and then declining slowly over a long period, Shell said. Natural gas demand could increase into the 2040s while the use of petrochemicals is

likely to continue into the 22nd century.

‘In all scenarios, coal stops growing as a source of electricity before 2030, with natural gas following by 2035,’ the report says. ‘After these dates there is no further net addition of generation from these energy sources.’

The report is the first global energy outlook in two years from the London-based oil and gas giant. It is also the first time it has included AI outcomes, which dominate the 57-page document assembled by Shell scientists, economists and researchers.

Shell sees the pace of decarbonisation gaining in the decades ahead —

with great assistance from AI becoming immersed across energy systems from transport to manufacturing. That will help the world avoid an increase in average global temperatures of 3C or 4C, the report says.

Shell’s scenarios forsee widespread electrification of the energy system, and also a world where carbon-removal technologies scale and become more economical, helping reduce global carbon intensity over time. ‘Any analysis which seeks to reach net-zero CO2 emissions by 2050 must fully embrace the use of carbon management and carbon removal technologies,’ says the report.

Mike Popham (STRYDE) and Claudio Trobbia (GTG International) (source: STRYDE)

Global energy merger and acquisitions to slow this year, says Rystad

Upstream merger and acquisition (M&A) activity is expected to slow significantly in 2025 after two years of record-high transactions driven by US shale mergers, said Rystad Energy.

The global deal pipeline value stands at approximately $150 billion as much of the sector’s consolidation has run its course, said Rystad. Furthermore, geopolitical tension in the Middle East, the conflict in Ukraine and the UK’s challenging fiscal environment are expected to create notable headwinds for market participants, the company added.

North America will continue to lead global M&A activity, driven by nearly $80 billion in upstream opportunities on the market. Elsewhere in the Americas, South American deal value rose from $3.6 billion in 2023 to $14.1 billion in 2024 (excluding Chevron’s acquisition of Hess), largely due to regional exploration and production (E&P) growth ambitions — and despite Petrobras halting its divestment program.

‘Last year marked a significant year of consolidation in the US shale sector, with approximately 17 consolidation-focused

deals, compared to just three acquisitions in late 2023. Activity was always expected to fall after such dramatic highs, but there is still plenty of business to be done. There is potential for further upside if US shale gas M&A activity increases,’ said Atul Raina, vice-president, oil and gas research, Rystad Energy.

Meanwhile, the Middle East is emerging as a significant centre for M&A activity. Bolstered by liquefied natural gas (LNG) expansion plans, the region recorded its second-highest year of M&A

activity since 2019, with deal value reaching nearly $9.65 billion in 2024, following a five-year peak of $13.3 billion in 2022.

The North Field expansion aims to elevate QatarEnergy’s LNG production to 142 million tonnes per annum (Mtpa) by the early 2030s. ADNOC is reportedly considering awarding an additional 5% stake in Ruwais LNG to an international partner.

M&A deal value in Europe decreased by around 10% year-on-year, to $14 billion in 2024. Around 75% of the regional total centered on the UK, where majors have been adopting an autonomous model strategy to expand their presence in the North Sea. The largest deal this year involved Shell and Equinor merging their UK North Sea upstream portfolios, excluding some of Equinor’s cross-border assets.

Despite $8 billion worth of upstream opportunities in the region, the outlook for future M&A activity in Europe remains uncertain due to tighter fiscal terms for oil and gas in the UK, which accounts for 73% of the potential deals, valued at about $5.9 billion.

TGS signs deal to carry out CO2 storage assessment offshore Australia

TGS has entered into an agreement with deepC Store to carry out subsurface well location planning, storage resource verification and evaluation for carbon storage in the G-14-AP permit area in the Browse Basin, offshore northwestern Australia.

The assessment will take place throughout 2025 on the permit, which is co-owned by deepC Store and Azuli (Australia).

The G-14-AP permit is one of the two offshore CO2 storage acreages awarded to deepC Store and Azuli by the Australian government in August 2024, alongside the G13-AP permit. deepC Store serves as the designated operator of the G-14-AP

permit and the associated CStore1 project, with an estimated CO2 storage potential of approximately 1 gigatonne (1 billion tonnes) in the permit area. The integration of TGS’ advanced subsurface data and deepC Store’s expertise in CCS project development will be essential in assessing the feasibility and long-term security of carbon storage in the Browse Basin.

deepC Store’s managing director Daein Cha said: ‘We are very pleased to work with TGS for progressing one of its key technical activities for G-14-AP. This work demonstrates our commitment to establishing CStore1 as the first offshore floating

CCS hub project in the Asia Pacific region, and to advance Australia’s strategic position in the CCS business.’

deepC Store and Azuli are CCS project developers with their core area of activity in the Browse basin and Bonaparte basin in offshore Western Australia. deepC Store has championed its ‘CStore1’ project as an innovation in the Asia-Pacific region, a commercial-scale floating CCS hub that integrates with CO2 emission facilities and covers all of the CCS value chain, being capture and liquefaction of CO2 onshore; transport by ships to the floater hub; and injection from the floater hub.

Rystad Energy chief executive, Jarand Rystad.

Photo: Marie von Krogh/DN.

UK spending watchdog casts doubt on financial case for CCUS, but DNV backs UK energy transition as good value

The UK government’s backing of carbon capture, storage and utilisation projects is unproven and high risk, according to one of the country’s spending watchdogs.

The UK Public Accounts Committee (PAC) has called on the British government to assess whether its full carbon capture, usage and storage (CCUS) programme will be affordable for taxpayers and consumers, given wider pressures on energy bills and the cost of living.

The PAC’s inquiry heard that CCUS may not capture as much carbon as expected, with international examples showing that government’s expectations for its performance are far from guaranteed.

The report notes recent scientific evidence that producing liquid natural gas, which will be used to run several CCUS projects, leaks more greenhouse gases into the atmosphere than previously thought. ‘This could undermine the rationale for pursuing certain schemes, and the PAC calls on government to consider the impact of up-to-date scientific understanding on CCUS,’ it said in a statement.

The report praised the government for learning from previous failed attempts to support CCUS, with the first two projects expected to begin operating in 2028. However, it notes that three-quarters of the almost-£22 billion envisaged to support the projects will come from levies on consumers who are already facing some of the highest energy bills in the world – yet the report finds the government has not yet looked at the likely financial impact of CCUS on households.

The PAC’s inquiry further found that neither of the contracts for the two new CCUS projects include any provision for the government to share the profits or for consumers to benefit from lower energy bills should things go well. This is despite any such profits, which could be significant if the programme is successful, being a result of early public support. The report calls on government to introduce mechanisms to make sure taxpayers and consumers benefit financially from the

success of all future CCUS projects that they have supported.

The UK government downgraded its ambitions for CCUS in 2024, with a target of storing 20 to 30 million tonnes per year of CO2 by 2030 now seen as no longer achievable. No revised targets have yet been announced. The PAC’s report notes that this creates a shortfall in government’s pathway to net zero, with the now-abandoned targets leaving it unclear how the government will meet its legally binding goals. The PAC’s report calls for new targets to be set out as a matter of urgency.

Sir Geoffrey Clifton-Brown MP, chair of the committee, said: ‘All early progress will be underwritten by taxpayers, who currently do not stand to benefit if these projects are successful. Any private sector funding for such a project would expect to see significant returns when it becomes a success. We were surprised that the Government had not even considered this aspect. Most concerningly, last year’s downgrading of ambitions for CCUS has left a glaring shortfall in the path to net zero.’

Meanwhile, DNV’s 2025 UK Energy Transition Outlook (ETO) report says that the UK’s energy transition will deliver a cleaner, more efficient, and less expensive energy system.

The outlook predicts reductions in greenhouse gas emissions of 58% by 2030, 68% by 2035 and 82% by 2050, against 1990 levels, but not reaching net zero.

By 2050, energy demand is expected to decrease by 25%, primarily because large-scale electrification will create a more efficient energy system compared to one based on fossil fuels.

Decarbonising the UK economy will reduce average household energy costs for consumers by nearly 40% by 2050 compared to 2021 levels, mainly due to a more efficient energy system using electric vehicles and heat pumps.

The short-term Clean Power 2030 target sets an ambition to decarbonise the electricity system by decade’s end, but DNV forecasts that unabated gas will still generate 12% of UK electricity in 2030. Full decarbonisation is expected by 2035. Solar, onshore wind, and offshore wind capacity will nearly double to 90 GW by 2030. However, this remains 45 GW short of government targets to double onshore wind, triple solar, and quadruple offshore wind.

The UK has committed to reducing greenhouse gas emissions by 81% by 2035, compared to 1990 levels. DNV’s projections suggest it will reach only 68%, requiring steeper reductions to meet its pledge.

Hari Vamadevan, executive vice-president and regional director, UK & Ireland, energy systems at DNV, said: ‘Despite economic and geopolitical challenges, the UK’s trajectory remains positive. A substantial green prize for our economy – cleaner and more affordable energy, is there for the taking if we can grasp it.’

Low carbon sources are expected to surpass fossil fuels in the supply mix, with the latter falling from 75% of primary energy today to 34% by 2050. However, oil and gas will remain dominant across the next decade, with significant amounts still required to balance energy demand and ensure security of supply.

Sir Geoffrey Clifton-Brown, chairman of the UK Public Accounts Committee.

Oil majors results round-up

BP has promised to ‘fundamentally’ reset its strategy after posting a 48% drop in fourth-quarter underlying replacement cost profit (RC profit) of $1.169 billion, compared with $2.99 billion in the same period of last year and with an analyst forecast of $1.2 billion, according to an LSEG poll.

The company’s net debt was nearly $23 billion in the fourth quarter, increasing 10% year-on-year. Capex was $3.7 billion in Q4, down from the $4.7 billion of fourth quarter 2024.

As part of its reset, the company is expected to ditch plans cut oil and gas output and will scale back its low carbon spending plans.

Shell has reported Q4 2024 adjusted earnings of $3.7 billion reflecting lower prices and margins, higher exploration well write-offs. Focus on disciplined capital allocation drove down 2024 cash capex to $21.1 billion. Free cash flow for the full year of 2024 was $39.5 billion.

ConocoPhilips has reported fourth-quarter 2024 earnings of $2.3 billion compared with fourth-quarter 2023 earnings of $3 billion. Excluding special items, fourth-quarter 2024 adjusted earnings were $2.4 billion compared with fourth-quarter 2023 adjusted earnings of $2.9 billion.

Full-year 2024 earnings were $9.2 billion compared with full-year 2023 earnings of $11 billion.

Equinor has reported adjusted operating profit of $7.9 billion and $2.29 billion after tax in the fourth quarter of 2024. Reported operating profit was $8.74 billion, and profit for the period was $2 billion.

Exxon Mobil has announced fourth-quarter 2024 earnings of $7.6 billion. Cash flow from operating activities was $12.2 billion and free cash flow was $8 billion. Capital and exploration expenditures, and cash capital expenditures were both $7.5 billion in the fourth quarter, bringing the full-year expenditures to $27.6 billion and $25.6 billion, respectively – both in line with full-year guidance. For full-year 2024, the company reported earnings of $33.7 billion and $55 billion cash flow from operations – its third best year in a decade.

Chevron has reported earnings of $3.2 billion for fourth quarter 2024, compared with $2.3 billion in fourth quarter 2023. Included in the quarter were severance charges of $715 million and impairment charges of $400 million. Foreign currency effects increased earnings by $722 million.

TotalEnergies has reported fourth quarter net income of $4.4 billion, up from the third quarter but down 21% year on year. Net income for the full year of 2024 was $18 billion.

TGS wins 4D streamer contracts in North Sea and Barents Sea

TGS has won four 4D streamer contract acquisition projects, three in the North Sea and one in the Barents Sea. The 4D campaign in the North Sea is scheduled to commence in early Q2 with backto-back scheduling and a total duration of approximately 130 days. The 4D contract in the Barents Sea is scheduled to commence in late May with a total duration of approximately 50 days.

In June the company will carry out a further two 4D streamer contract acquisition projects, one in the North Sea and one

in the Norwegian Sea. The 4D projects are scheduled to be acquired back-to-back and commence in June. The total duration of the two surveys is approximately 80 days. Kristian Johansen, CEO of TGS, said: ‘Our GeoStreamer technology, combined with the Ramform acquisition platform, ensures efficient delivery of high-quality data. We are experiencing higher demand for contract work on the Norwegian Continental Shelf this year, compared to last year, including our multi-client programs.’

Oman’s Ministry of Energy and Minerals (MEM) has launched a competitive bid round for three oil and gas exploration blocks – Block 36, Block 43A and Block 66.

Trinidad and Tobago has launched the Deep Water Competitive Bidding Round 2025, which includes 26 offshore blocks along the eastern and northern coasts. Deadline for submission of bids is 2 July 2025. Successful bids will be announced three months later. Winners will be awarded production sharing contracts (PSC). A fee of $25,000 will give bidders access to a data package on the 26 blocks.

Iraq and BP will sign a deal to evaluate and redevelop four Kirkuk oil and gas fields and adjacent areas. The agreement is expected to surpass the scale of a $27 billion deal Iraq signed with TotalEnergies in 2023.

BP is to cut about 4700 staff, more than 5% of its total workforce, as part of its plans to cut costs. The company, which has a global workforce of about 90,000 people, also confirmed that about 3000 contractor positions will also be axed this year.

EMGS has won a contract for a CSEM survey in India. The survey is expected to have a contract value of approximately $10 million. The vessel Atlantic Guardian will commence acquisition after completing a previously announced survey in India.

SeaBird Exploration has reported fourth quarter 2024 revenues of $10.2 million with adjusted EBITDA of $4.6 million. Net interest-bearing debt was $9.4 million. Vessel utilisation was 99% and the company reported a ‘strong market outlook’.

Five non-executive directors have been appointed to the start up board of Great British Energy, the publicly owned energy company that will own and invest in clean energy projects across the UK. Frances O’Grady, Frank Mitchell, Kate Gilmartin, Dr Nina Skorupska, and Valerie Todd have joined Great British Energy’s board as non-executive directors.

Oil and Gas round-up

Impact Oil and Gas has completed the Tamboti-1X exploration well and spudding of the Marula-1X exploration in Block 2913B (PEL 56), offshore Namibia. Tamboti-1X was drilled to a total depth of 6450 m on Block 2913B, approximately 12 km northeast of the Mangetti-1X well and 25 km north-northwest of the Venus-2A well. Oil was encountered within 85 m of net reservoir of lower-quality Upper Cretaceous sandstones, belonging to the Mangetti fan system. On 3 February the Marula-1X exploration well was spudded within the southern part of Block 2913B. This well will target Albian-aged sandstones, within the Marula fan complex and has the potential to unlock further exploration targets across the south, which is an area lying at the heart of the prolific Kudu sourcerock kitchen.

Vår Energi and its partners have discovered gas in the Elgol prospect in the Barents Sea. Exploration well 7122/9-2 is the first in production licence 1131, which was awarded in 2021. The next well is 7122/8-3 S Zagato. The discovery is estimated to between 0.4-3 million standard cubic m of oil equivalent, which corresponds to 2.6-19 million barrels of oil equivalent. The objective of the well was to prove petroleum in Upper Permian reservoir rocks in the Ørret Formation.

Well 7122/9-2 encountered a 1.5-m gas column in a sandstone reservoir in the Ørret Formation, where reservoir properties were moderate. The remainder of the reservoir totalled 28 m, generally drilled to a vertical depth of 1844 m below sea level, and was terminated in the Ørret Formation in the Upper Permian. Water depth is 415 m.

Reconnaissance Energy Africa’s Naingopo exploration well within the Damara Fold Belt on Petroleum Exploration Licence 073 (‘PEL 73’), onshore Namibia has encountered more than 50 m net reservoir in the Otavi Group. Indications of oil were observed from the Damara Fold Belt. Follow-on drilling in the Damara Fold Belt is accelerated to drill Prospect I ahead of Kambundu, which is expected to spud in July 2025.

Zephyr Energy has started drilling the extended lateral on the State 36-2 LNW-CC-R well. The extended lateral will be drilled horizontally from near the base of the existing wellbore and will target an additional 5500 feet of the Cane Creek reservoir. Results from the production test on the well will be available by the end of March 2025.

TGS reprocesses 17,000 km2 of data in India

TGS has launched its Krishna-Godavari (KG) Basin Mega 3D reprocessing project to maximise the value of legacy data for exploration and licensing in one of India’s most productive hydrocarbon regions.

Spanning 16,900 km2 of legacy 3D surveys, the KG Basin Mega 3D project will utilise Pre-Stack Time Migration (PSTM), Pre-Stack Depth Migration

(PSDM), and Full Waveform Inversion (FWI). These advanced methods will ensure contiguous, high-resolution imaging, enabling detailed exploration of Tertiary stratigraphic plays and deeper syn-rift prospects.

The project is expected to deliver final advanced seismic products by April 2026, with fast-track data available in Q2 2025.

The KG Basin Mega 3D area encompasses 10,900 km2 of open acreage over the upcoming OALP-10 bid round blocks. This will allow exploration and production companies to evaluate opportunities with improved clarity and confidence.

Meanwhile, TGS has launched advanced imaging centres for Petrobras in Rio de Janeiro, dedicated to ocean bottom node (OBN) and 4D imaging for the Campos and Santos basins. These facilities will utilise both on-premises and cloud-based hardware to process data with unparalleled precision. Geoscientists from both companies will collaborate at Petrobras’s Rio de Janeiro offices.

‘TGS’ 4D Full Waveform Inversion (FWI) technology will provide enhanced subsurface clarity for reservoir monitoring and characterisation,’ said TGS. ‘The imaging centres will also integrate hybrid 4D solutions combining streamer and OBN, as well as OBC and OBN data, for a comprehensive subsurface view.’

India KG Basin Mega reprocessing map.

STRYDE provides nodes for onshore CO2 storage in Denmark

STRYDE has secured a deal to supply 42,000 autonomous nodes to Smart Seismic Solutions (S3).

The nodes will be deployed on a 3D high-density seismic survey for the CarbonCuts carbon dioxide (CO2) Storage Project onshore in Denmark.

‘The CarbonCuts Project is a high-profile initiative aimed at identifying optimal onshore subsurface storage solutions for CO2 sequestration,’ said STRYDE in a statement. ‘The seismic survey will be crucial in advancing Denmark’s ambitious goal of achieving net-zero emissions by 2045, utilising advanced seismic technology to map geological formations with unprecedented accuracy, ultimately helping to ensure safe and effective onshore CO2 storage.’

S3, a geophysical service provider, is at the forefront of the initiative, spearheading the acquisition of a high-density seismic dataset using STRYDE’s nodal seismic acquisition system.

Patrick Robert, CEO at S3, said: ‘The Rødby structure, an elongated anticlinal dome with a four-way closure that formed during the Mesozoic Era over a Zechstein salt pillow, is a fascinating target for this high-density seismic survey.

The dense deployment of STRYDE nodes provides exceptional clarity of the subsurface, enabling stakeholders to identify geological structures suitable for carbon storage with greater precision. This significantly reduces risks and enhances the reliability of long-term CO2 sequestration solutions.’

Italian 3D seismic campaign gets green light

Po Valley Energy (PVE) and Prospex Energy have obtained regional approval to start a 3D seismic campaign covering the entire Selva Malvezzi Production Concession Area in the Po Valley in the north of Italy.

The partners are planning to drill four wells on the concession, building upon the current Podere Maiar-1 well, which produced 7.02 MMscm of gas (2.6 MMscm net to Prospex).

Operator PVE (63%) has completed all preliminary works in relation to the design, planning and permitting of the campaign. With regional approval it will be able to formalise permitting and agreements with landowners.

The campaign is expected to take no more than three weeks. Funding for this 3D seismic campaign is fully covered from production income.

Once the seismic data has been acquired, the dataset will be processed and interpreted in-house with the aim of optimising the subsurface drilling locations targeted by the planned four new wells on the concession. The environmental impact

assessment to obtain the permits to drill these four new wells was lodged with the Ministry of Environment and Energy Security on 24 December 2024.

Mark Routh, Prospex’s CEO, said: ‘The 3D seismic dataset will be acquired and processed early this year in order to optimise the subsurface drilling targets for four planned wells – two development wells into structures classified as contingent resources, North Selva and South Selva, and two exploration wells classified as prospective resources, East Selva and Riccardina. These wells, which are targeting a total of 88 Bcf gross, are anticipated to receive permits to drill in the second half of 2025.’

ENERGY TRANSITION BRIEFS

SLB and Star Energy Geothermal, a subsidiary of Indonesia’s largest renewable energy company Barito Renewables, have announced a collaboration agreement to accelerate advanced technologies for geothermal asset development. The two companies will aim to deploy technologies to shift the economics of conventional geothermal projects and improve recovery rates of geothermal assets.

The US Bureau of Ocean Energy Management (BOEM) has approved the Construction and Operations Plan for the SouthCoast Wind Project, which wil generate up to 2.4 GW of offshore wind energy for Massachusetts and Rhode Island.The approved project includes up to 141 wind turbines.

ADNOC Gas, in partnership with Baker Hughes, has installed British climate technology firm Levidian’s patented LOOP technology at the Habshan Gas Processing Plant in the UAE. In the first-ever deployment of the technology at an operational gas processing site, carbon will be captured from methane and transformed into graphene, a material set to shape the future of multiple industrial applications. The LOOP unit is capable of producing more than 1 tonne per annum (tpa) of graphene and 1 tpa of hydrogen. Future industrial-scale installations are expected to deliver 15 tpa.

The US Department of Energy has finalised with Equinor a $225 million grant for the South West Arkansas (SWA) lithium project. In May 2024 Equinor entered a strategic partnership with Standard Lithium,`acquiring a 45% share in two lithium companies in Southwest Arkansas and East Texas. The grant will support construction of a processing facility for the SWA project, which in Phase 1 is targeting an annual production of 22,500 tonnes of lithium carbonate for use in battery production.

Renewable energy developer Grenergy has secured $299 million in financing for the Víctor Jara phase of the Oasis de Atacama project, the world’s largest battery energy storage system, located in Chile’s Atacama Desert.

Po Valley, northern Italy.

EMGS Q4 revenue rises to $10 million

EMGS has reported fourth quarter 2024 revenues of $9.7 million, up from $1.1 million in Q4 2023, and including $9.4 million in multi-client prefunding revenue. Adjusted EBITDA (including capitalised multi-client expenses and vessel and office lease expenses) was $7.9 million, up from a loss of $1.7 million in the fourth quarter of 2023.

At the end of the fourth quarter the company had one vessel on charter, The Atlantic Guardian, which completed multi-client

surveys in the North Sea and the Norwegian Sea in the quarter and started transit to India for an upcoming proprietary survey. Utilisation for the fourth quarter was 31% compared with 0% for the fourth quarter 2023.

EMGS had one vessel in operation and recorded three vessel months in the quarter.

In the fourth quarter 2023, the company recorded three vessel months.

Oil and gas production set to rise in UK and Norway

UK and Norway oil and gas production are both set to rise in 2025 even though the UK’s sector will struggle this year, according to analysis from the Westwood Global Energy Group.

In the UK production is forecast to rise by 3% in 2025 to 1.14 million boepd, with full-year production from the Talbot field and contributions following the start-up of the Penguins, Affleck, Jackdaw, Murlach, Victory and Teal West fields. However, the increase is more a result of project delays than an uptick in investment, said Westwood. Post-2026, production is expected to enter a steep decline as the impact of under-investment takes effect. ‘Fields without submitted EIAs are unlikely to gain approval in 2025 due to the lengthy review process,’ said Westwood. ‘Currently, only Avalon and Buchan Horst have EIAs under consideration, but it is unclear whether either will gain sanction in 2025.’

Eight hubs are expected to close this year. As a result, companies are expected to spend $2 billion on abandonment activities in 2025, with total abex spend over the next decade (2025-2034) estimated at over $26 billion. ‘The next five years are crucial in maximising recovery from the high number of late-life hubs across the UK, but based on existing investment plans, Westwood forecasts the number of hubs to reduce to 46 by 2030, 20 by 2035, with only four remaining in 2040,’ it said.

E&A activity in 2025 has the potential to pick up from 2024 when just one appraisal and three exploration wells completed. Westwood has ‘visibility’ of 10 exploration wells that could operate this year with pre-drill resources of c. 545 mmboe. ‘However, with the fiscal and political uncertainty in the UK, investor appetite has been impacted and there is a high chance that many of these wells will not be drilled in 2025,’ said Westwood. All but one of the planned exploration wells are in the Central North Sea, with Dabinett the only well planned in the Southern North Sea. There is also the potential for three appraisal wells, appraising a midpoint volume of c. 230 mmboe.

The consultation on environmental impact assessments and the inclusion of Scope 3 emissions closed on 8 January with new guidance expected in the Spring. Thereafter a consultation is expected on licensing that will set out the UK government’s appetite for future exploration and potential new developments. The third consultation is on the fiscal regime post-2030, which could have a significant impact on project economics, particularly for near-term developments.

Norway, meanwhile, remained firmly in development mode after the huge wave of project sanctions in 2023. Although E&A activity remained buoyant, with 29 exploration well programmes completing, discovered volumes were relatively low, said Westwood.

Production is forecast to rise to 4.2 million boepd in 2025 following the start-up of seven new field developments and four brownfield projects, totalling 1.1 bnboe in reserves. Equinor’s delayed Johan Castberg field is the largest, which is now expected online in Q1 2025.

With operators focusing on delivering these projects, there has been a lag in the number of firm new development projects moving forward for FID. Only seven new field developments in three greenfield projects are expected to be sanctioned in 2025, with combined reserves of c. 225 mmboe.

High levels of E&A activity are expected to be maintained in 2025. There are 44 exploration wells currently on Westwood’s list of planned wells with associated pre-drill resources of c. 3.2 bnboe. Many of these have been deferred from 2024 due to rig schedules. Historically, the Northern North Sea has been the most drilled basin, but in 2025, the Norwegian Sea and NNS both have 15 planned wells. Predrill resources, however, in the Norwegian Sea are three times higher than in the NNS. All wells in the NNS are infrastructure-led exploration whereas three wells in the Norwegian Sea are considered to be high impact. There is a continued focus on Barents Sea exploration, with eight planned wells. In addition, there are four appraisal wells appraising a midpoint volume of at least c. 145 mmboe.

Unsupervised clustering as a tool for geomorphological interpretation: a case study from a Sarmatian delta in Romania’s Dacian Basin

Luca Fava1*

Abstract

Seismic attributes are a valuable tool in identifying sedimentary facies. Machine learning (ML) techniques can be helpful in combining information coming from different attributes. The unsupervised clustering algorithm called selforganising map (SOM) has already been successfully applied in paleo deep-water settings in many basins defining seismic facies. To evaluate the applicability of this ML method to a paleo shallow-water setting the SOM workflow is applied to 3D seismic data acquired onshore Romania over the south Carpathian foredeep. Many attributes are extracted along an interpreted horizon. Principal Component Analysis (PCA) is performed on the list of attributes in two ways. Firstly, the main principal components are defined to use them as input for the first run of SOM clustering (SOM 1). Secondly, the most important attributes are identified and used as input for the second run of SOM clustering (SOM 2). The geomorphological interpretation of the SOM results is compared with the interpretation of conventional attributes. The results show that the SOM is a powerful tool in defining seismic facies as it allows the definition of details not discernible using other tools. This exercise shows also that ML techniques can be easily implemented by exploration geophysicists using standard interpretation software’s and open-source Python libraries.

Introduction

To identify reservoir distribution is one of the tasks of exploration geoscientists. In siliciclastic-dominated depositional environments seismic attribute analysis can be of major help. In deep-water clastic and coastal-plain settings the use of seismic attributes to identify sandstone has been a widespread practice in the last 30 years. Although the first seismic images showing meandering channels date to the early 1980s (Brown et al., 1982), the discipline of seismic geomorphology (Posamentier et al., 2007) became widely studied only at the end of last century (Xu and Haq, 2022). The conventional approach to seismic interpretation of geomorphological features involves horizon picking and the extraction of attributes to identify channels and differentiate seismic facies. The seismic attributes conventionally used to highlight shape and margin of channel are edge detection (or geometrical) attributes such as dip, variance (coherence), and curvature (Bahorich and Farmer, 1995; Luo et al., 1996; Marfurt, 2006; Gersztenkorn and Marfurt, 2006). Other attributes can help in identifying seismic facies that can reliably indicate lithologies. Among them are attributes such as sweetness (Hart, 2008), instantaneous frequency (Taner et al., 1979; Marfurt, 2006), and relative acoustic impedance (RAI). Spectral decomposition (Partyka et al., 1999; Castagna and Sun; 2006) is another powerful approach in interpreting bed thickness and discontinuities. The methodology consists of applying a suite of constant-bandwidth filters to the seismic data and displaying those as a blend of colours (RGB)

1 OMV Petrom

* Corresponding author, E-mail: luca.fava@petrom.com DOI: 10.3997/1365-2397.fb2025019

representative of frequencies most relevant to the targeted object. Although the idea of integrating the information coming from different seismic attributes to increase the definition of geomorphological features is not new, the recent exponential growth in computational power has made this goal more achievable. In fact, machine learning (ML) allows multi-attribute analysis in a fraction of the time previously required, and it can help in finding patterns within the data that are not obvious to the human eye (Sacrey and Roden, 2014; Tellez, 2015; Zhao et al., 2016; La Marca-Molina et al., 2019; La Marca and Bedle, 2022; Lubo-Robles et al., 2022). An unsupervised clustering technique called self-organising maps (SOM) can be used to cluster seismic attributes data with the goal of extracting information not immediately visible to the seismic interpreter (Roden et al., 2015). Recently, many papers have shown examples of application of SOM to the definition of geomorphological features in deep-water setting (Matos et al., 2003; 2004; 2007; Matos and Marfurt, 2009; Stecker and Uden, 2002; Roden et al., 2015; Zhao et al., 2015, 2016, 2017, 2018; Roden and Sacrey, 2016; Lubo-Robles and Marfurt, 2017; La Marca-Molina et al., 2019; La Marca and Bedle, 2022; Chopra et al., 2022). The present paper illustrates an example of application of SOM technique to a shallow-water setting. The aim of this paper is to interpret geomorphological features in a channel complex entering the Dacian basin (south Carpathian foredeep) during Sarmatian time (Miocene epoch) using SOM. The paper presents two different approaches to SOM and compares the results of this

ML technique to ‘conventional’ approaches. An additional goal is to show that ML techniques can be easily implemented using common interpretation software (PetrelTM) and publicly available coding resources such as Python libraries.

2 North-South line of the PSDM cube stretched back to time. a) Seismic line showing the quality of the image. The approximate location of this line is shown in Figure 1. On the northern side of the image the fold and thrust belt can be seen. In the middle of the sequence few strong amplitude reflectors are displayed. Those reflectors are either truncated or onlap on the thrust sheet. They represent the fluvial delta entering the Dacian Basin. b) Same line with interpretation. The black horizons are following soft kick (i.e., reduction of impedance) reflectors. The green horizon is the horizon along which all the seismic attribute extractions have been performed.

Geologic setting

Figure 1 a) Location map of the study area (red rectangle). The red dashed line shows the approximate location of the regional cross-section across the Dacian Basin (b). This cross-section has been modified from Krézsek and Olariu, 2021 (their Figure 5b). The cross section is derived from seismic interpretation. The white star indicates the studied stratigraphic interval.

The study area is located 120 km north-west of Bucharest in the south-central part of Romania, at the foothills of the Carpathian Mountains (Figure 1a). From a geological point of view the area is in the western Dacian Basin, including the Late Miocene to recent foreland of the South Carpathian Mountains. The Carpathians fold and thrust belt formed because of the closure of the Neotethys Ocean, a process that started during Late Jurassic period and ended in the Middle Miocene period (Krézsek et al., 2013). This paper focuses on Late Miocene (Sarmatian) depositional systems infilling the Dacian Basin, south of the frontal thrust (Figure 1b). From a paleogeographic point of view the Dacian Basin is part of the Paratethys seaway (Rӧgl, 1998, 1999), meaning a series of water masses isolated from the Mediterranean Sea by the emerging Alps – Dinarides – Balkans –Pontides (Popov et al., 2006). The Dacian Basin became isolated from the adjacent Tethyan basins and formed a reduced salinity lake by the end of the Sarmatian period (11 Ma, Leever et al., 2006; Jipa and Olariu, 2009). During the Sarmatian period a series of deltas supplied sediments to the lake from the north, where the Carpathians Mountains provided sediment input due to uplift and erosion. This study analyses one of those lacustrine deltas developed during the Sarmatian period in the eastern part of the western Dacian Basin.

Methods and dataset

The methodology applied in the present paper is divided into five steps that do not require specific software. Any geophysicists can perform these steps using PetrelTM software (the most common seismic interpretation tool in the industry), publicly available computing platforms such as Jupyter Notebook, and open-source Python libraries.

Figure

Available seismic data

The available seismic data is a 950 km2 PSDM cube acquired by OMV Petrom in 2022 and processed by Viridien (formerly CGG). The bin size for the data is 15 m x 15 m, with a sample interval of 4 m. The target level is around 2000 m below mean sea level. Many wells have been drilled in the area, but none in a critical position to calibrate or assess the geomorphological interpretation. Therefore, no wells have been included in the database.

Seismic interpretation

The interpretation has been performed on the PSDM volume stretched back in time to ensure the integration of regional interpretation performed on vintage 2D lines. During the first phase of reconnaissance various intervals showing geomorphological elements have been identified. A few soft kicks, i.e., low impedance reflectors have been interpreted (Figure 2), and a surface has been gridded for each of them. One of these maps (green horizon in Figure 2) is the object of this study because of the interesting geomorphological features displayed at this level. The interpretation suggests the evolution of a prograding delta supplying sediment into the foredeep from the growing fold and thrust belt. This phase of sediment input ends with deposition of a mass-transport complex (MTC) generated by the instability of the basin margin due to either a late phase of movement of the thrust, or the sediment overloading of the delta front.

Attribute extractions

Seismic attributes are any quantity calculated from seismic data; They represent a subset of the seismic data that can be ‘extracted’ from the total information with attribute computations decomposing the data into constituent attributes (Barnes, 2001). In the last 20 years there has been a proliferation of seismic attributes, so much so that some authors warned that ‘…many seismic attributes are redundant or useless and confuse seismic interpretation more than they help,’ Barnes (2007). Thus, it is important to define the scope of the study before starting extracting attributes from the seismic data. This paper’s goal is to identify seismic facies by defining channels and lobes within a delta entering a lake. The seismic attributes in this study have been selected accordingly. To analyse these attributes visually, they are plotted on a map view in PetrelTM. Recently, a few papers have stressed the importance of colour bars in scientific data display (e.g., Niccoli, 2014; Crameri et al., 2020). Perceptually uniform colour maps are the best way to display scientific data; they highlight true data variations even when they are subtle, and they can easily be read by colour-blind people too.

Figure 3 Graphic explanation of the self-organizing map algorithm. a) An initial 2D lattice formed by X * Y neurons. The picture also shows the connections between every input point (N-dimensional data) and all the neurons, and the concept of neuron’s neighbours. Modified lattice with clusters on the lattice related and representative of the clusters in the higher-dimensional data space. (b).

For this reason, the present paper uses perceptually uniform colour bars in all the figures. The first selection of attributes is based on the interpreter’s experience and on literature. Following the definition and extraction of the seismic attributes using PetrelTM, a Pandas dataframe (McKinney, 2010) has been created in Python; each line of the dataframe corresponds to a node of the map, and it is defined by the X and Y coordinates of the node. For each node there is a corresponding value for each of the extracted attributes, and the dataframe has a column for each attribute. A multivariate analysis calculating the cross-correlation values between all inputs is conducted in python to avoid redundant information.

Principal Component Analysis (PCA)

The third step of the methodology is the principal component analysis (PCA). PCA is a statistical technique invented at the beginning of the 20th century (Pearson, 1901). The exponential growth in computational power in the last decades has made PCA an extremely common tool within ML. PCA is a linear dimensionality reduction technique that can be used as a pre-processing stage for the unsupervised clustering self-organising maps (Cortés Arroyo et al., 2021). The PCA module used in this paper is from the free and open-source ML library scikit-learn (Pedregosa et al., 2011).

Unsupervised clustering (SOM)

The fourth step is the computation of the unsupervised clustering using the self-organising map algorithm MiniSom (Vettigli, 2018). This algorithm implements the Self Organising Maps (SOM) relying only on the Numpy library and emphasising vectorisation in coding. The SOM was introduced by the Finnish professor Teuvo Kohonen in the 1980s and therefore is called

Figure 4 TWT map of the selected horizon. The TWT values span between 1510 msec and 1935 msec. The contour interval is 20 msec. The colour scale is a perceptually uniform colour map called viridis (Crameri et al., 2020).

Kohonen map (Kohonen, 1982). The SOM is a type of Artificial Neural Network able to convert complex, nonlinear statistical relationships between high-dimensional data items into simple geometric relationships on a two-dimensional display. In other words, it is a technique that makes high-dimensional data easier to understand plotting them on a map. It is a competitive learning neural network in which the neurons are nodes of a two dimensional lattice (manifold). Each node is connected to all input points and is the centroid of a cluster (Figure 3).

A weight is assigned to each node; adjacent neurons (neighbours) have similar weight values. Data is randomly chosen, and the map of weights is scanned to find which weight best represents that sample. The chosen weight and its neighbours are rewarded by becoming more like that input. In fact, the neurons compete to include the higher amount of data points in their cluster. The weights assigned to every neuron-point connection are modified and this deforms the 2D plane into a manifold that

better fits the data (Figure 3b). Finally, the N-dimensional data are projected onto this manifold displayed using a 2D colour bar. This technique can be used for seismic facies classification as demonstrated by the many studies referenced in the Introduction. The present paper uses MiniSOM for that very purpose, generating a seismic facies map from multiple seismic attributes.

Geomorphological interpretation

The fifth step is the geomorphological interpretation of all the different outcomes (conventional seismic attribute extractions, PCA and SOM). The comparison of all these different deliverables is used to improve the geomorphological and sedimentological understanding of the depositional environment analysed, and to evaluate the added value of the ML techniques.

Results and discussion

The result of the seismic interpretation is the two-way-traveltime (TWT) map shown in Figure 4.

The selected horizon is part of a package of reflectors truncated towards the north by a thrust representing the front of the southern Carpathians (Figure 1). This stratigraphic sequence represents a fluvial delta spilling over the thrust and supplying sediments to the late Sarmatian foredeep (Figure 2). The two-way time (TWT) map shows a structural high in the south-western corner with a series of normal faults. A structural nose is trending from south-west toward the centre of the map. The northern part of the map shows a gentle monocline dipping toward south and the frontal thrust fault of the southern Carpathians. The conventional way of highlighting channels and lobes implies the extractions of seismic attributes such as Root Mean Square amplitude, Sweetness and Variance. Root Mean Square amplitude (RMS hereafter) is the most used attribute to assess the depositional

Figure 6 Spectral decomposition display: low (27Hz), medium (33Hz), and high (40Hz) frequencies extracted along the selected horizon and RGB co-rendered. The constant-bandwidth filters have been defined at the target level and computed using the matching pursuit algorithm.

Figure 5 Extractions of standard seismic attributes along the selected horizon. (a) RMS amplitude. The colour scale is a perceptually uniform colour map called CET-R4 (colorcet.com; Kovesi; 2015). (b) Sweetness. The colour scale is a perceptually uniform colour map called CET-C1 (colorcet.com; Kovesi; 2015). (c) Variance (coherence). (d) Sweetness co-rendered with variance.

Attribute Name

3D Curvature

Dip Magnitude

Variance

Root Mean Square (RMS) amplitude

Instantaneous Frequency

Sweetness

Relative Acoustic Impedance (RAI)

PSDM time 40Hz

Peak Value

Peak Frequency

Attribute Category Usage

Geometrical attribute

Geometrical attribute

Geometrical attribute

Instantaneous attribute

Instantaneous attribute

Instantaneous attribute

Instantaneous attribute

Spectral attribute

Spectral attribute

Spectral attribute

Grey Level Co-occurrence Matrix (GLCM) Texture attribute

GLCM Entropy Texture attribute

Channel definition

Channel definition

Channel definition

lithology discrimination

lithology discrimination

lithology discrimination

lithology discrimination

Stratigraphy and thickness changes

Thickness of channels

Depositional, diagenetic and structural patterns

seismic texture

seismic texture

Table 1 List of 12 seismic attributes used as input for the ML workflow. The choice of the attributes is based on the author’s experience in seismic interpretation and definition of sedimentological and geomorphological features in different basins, and on the existing literature on the application of SOM to seismic facies definition.

environment, and to identify amplitude anomalies related to the presence of hydrocarbons. Sweetness attribute is routinely used to identify porosity and sandstones in clastic environments (Hart, 2008). The variance attribute (coherence, similarity in other software packages) is a geometric attribute used to detect edges and delineate faults and channels boundaries (Luo et al., 1996; Gersztenkorn and Marfurt, 2006). Figure 5 shows these attributes extracted at the selected horizon indicated in figure 2. A colour blending of sweetness and variance (Figure 5d) can be used to enhance the interpretability of the geomorphological features (Hart, 2008). Figure 5 shows a system of channels north-south oriented. Most channels are straight; only one (the brightest) displays a meandering pattern. The colour blending of sweetness and variance shows the most details with hints of fan deposits. The overall picture suggests the presence of a delta prograding from north to south (Figure 2). The delta deposition stops against a long meandering channel running east to west.

Another approach to improve the interpretation is the spectral decomposition (Partyka et al., 1999; Castagna and Sun, 2006). The technique consists of applying a suite of constant-bandwith filters

to the seismic data. The low, medium, and high frequencies calculated using the matching pursuit algorithm are extracted along the selected horizon, and then colour-blended (Figure 6). The spectral decomposition colour-blending confirms the observations drawn from the analysis of RMS, sweetness, and variance extraction.

Some of the features are more visible and better defined – in particular, the long meandering channel running east to west and a lobe in the north-western corner of the image. These conventional extractions have been produced to benchmark the results of the ML approach.

ML approach

The first step of the ML approach is the multivariate analysis performed using the cross-correlation of the selected attributes to drop out the redundant ones. The list of the attributes selected based on the interpreter’s experience and on literature is in Table 1.

The table shows the complete list with, for each attribute, the category, and the main use in seismic interpretation. The result of the cross-correlation is shown in Table 2.

2

Table 3 Cross-correlation matrix of the final selection of seismic attributes. The redundant attributes identified with the cross-correlation matrix shown in Table 2 have been dropped off.

* RAI = Relative Acoustic Impedance ** GLCM = Grey Level Co-occurrence Matrix

The redundant attributes are highlighted by the red colour that indicates values of cross-correlation above the 0.75 threshold. The final selection of attributes is shown in Table 3, and PCA is performed on these attributes. PCA can be used in two ways.

As a dimensionality reduction tool, and as a selection tool. In this project PCA has been used in both ways. In the first case the results are displayed on the so-called ‘scree plot’ (Figure 7), that shows the eigenvalues against the corresponding number of principal components. In other words, it shows how the proportion of variance of the data changes with the number of principal components.

How many principal components can be kept retaining the maximum variance but discarding the residual noise in the data is indicated by a distinct change in the slope of the line. Figure 7 shows the break in the slope at four principal components. Using four principal components we retain around 92% of the variance of the data (Figure 7), and the reduced dataset is used as input for a first run of the unsupervised clustering (SOM 1). To use the PCA as a screening tool for the seismic attributes we analyse the matrix (Figure 8) showing which input is contributing the most to the different principal components. This allows the selection

Figure 7 Scree plot showing the eigenvalues against the corresponding number of principal components. This plot is an output of the PCA, and it is used to define to ideal number of principal components to be used to reduce the dimensionality of the database.

of the most significant attributes, and, in turn the reduction of the dimensionality of the data. In this phase the selection of the attributes is up to the geophysicists who would use the result of the PCA together with their experience.

Figure 8 shows that the main contributors to the first four principal components are 3D curvature, instantaneous frequency, dip magnitude, variance, sweetness, and Grey Level Co-occurrence Matrix (GLCM) entropy (Table 4).

Considering that three of these attributes are geometrical attributes and that the variance extraction can be used in the display phase, only the following attributes have been selected as input for a second run of unsupervised clustering (SOM 2): dip magnitude, 3D curvature, sweetness, instantaneous frequency, GLCM entropy, and RAI. RAI contributes little to the first

Figure 8 Principal components matrix. This matrix shows how the principal components (vertical axis) are composed. In other words, it shows how much (the proportion) of any single principal component is formed by each single attribute (feature).

four principal components, but it is the main contributor to the fifth principal component, thus this attribute has been included to increase the amount of variance into this exercise. Figure 9 compares the result of the SOM 1 and SOM 2. To further improve the interpretability a blending of the SOM maps and the principal components maps with the variance attribute extraction have been performed (Figures 9c and 9d).

SOM 1 is the poorest result as it recognises only the two main north-south channels. SOM 2 instead, not only recognises all the geomorphological features displayed by the conventional attributes and by the spectral decomposition, but it adds details inside the channels and the main lobes.

For display purposes the PCA has been used in a third way, namely a colour blending of the first three principal components.

In this way 86% of the data variance is shown (see Figure 7). To evaluate the ML techniques used in this project the three principal components blending, the SOM1 and SOM2 are plotted side by side with the conventional spectral decomposition (Figure 10).

Of the four pictures the ones with more interpretable geomorphological features are the spectral decomposition (Figure 10a) and the SOM 2 colour blended with the variance extraction (Figure 10d). The colour blending of the first three principal components (Figure 10b) and the SOM 1 blended with the variance (Figure 10c) show some features, but they lack the details of the others two displays. Spectral decomposition is an extremely useful tool to define geomorphological features, but SOM 2 reaches comparable results and adds some details within the channels and the lobes that could be used to discriminates

Table 4 List of ten seismic attributes selected as input for the PCA ordered in terms of their cumulative contribution to the first four principal components. Highlighted in yellow are the six attributes used as input for the second SOM run. Only the most impactful attributes within the first four principal components have been selected with two exceptions. Variance has been excluded for two reasons: (a) the other two geometrical attributes have more impact, and (b) variance can be displayed co-rendered with the result of the SOM clustering. The second exception is the RAI that has minimal impact within the first four principal components but have a major impact on the fifth principal component.

Figure 9 Results of the ML exercise. (a) SOM 1: self-organizing map generated from the first four principal components. The result is expressed in 256 classes. The colour scale is a 2D perceptually uniform colour map expressed in 1D. (b) SOM 2: self-organizing map originated from a selection of seismic attributes based on the result of the PCA (see text for explanation). The result is expressed in 256 classes. The colour scale is a 2D perceptually uniform colour map expressed in 1D. (c) SOM 1 co-rendered with variance. (d) SOM 2 co-rendered with variance.

11 Comparison of the geomorphological interpretation of the (a) spectral decomposition, and (b) SOM 2. (c) Interpretation on the spectral decomposition with labelled depositional elements. (d) Interpretation on the SOM 2 with labelled depositional elements. See text for explanation.

sedimentological facies. Spectral decomposition has some advantages over the SOM approach, but on the other hand the SOM clustering provides more details within the geomorphological features and can be plotted on top of other attributes, something that is not possible with the spectral decomposition. In Figure 11 spectral decomposition and SOM 2 are shown with the performed geomorphological interpretation on top. The two pictures have been interpreted independently to highlight the differences and the inherent uncertainties of this kind of interpretation.

The long east-west meandering channel is visible and interpretable on both displays (Figures 11a and 11b), even if its margins are sometimes difficult to be exactly defined on the SOM 2 (i.e., blue arrow in Figure 11b). All the north-south channels are similarly interpretable on the two displays. Within these distributary channels (Figures 11c and 11d) the SOM 2 shows more details compared to the spectral decomposition colour-blending. One example is shown by the green arrow in Figures 11a and 11b. The outline of the feature indicated by the white arrows can be easily

Figure 10 Comparison of the ML results with spectral decomposition. (a) Spectral decomposition RGB rendering. (b) RGB rendering of the first three principal components. (c) SOM 1 corendered with variance. (d) SOM 2 corendered with variance.

Figure

picked in both displays (Figures 11a and 11b); this feature is interpreted as a barrier-beach (Figures 11c and 11d), meaning delta front sands reworked by wave action. The two depositional lobes in the east and in the west are equally interpretable in both displays, and yet some significant differences in the internal organisation and on the outline can be identified (see yellow arrows). The two bodies are interpreted as mouth-bars (Figures 11c and 11d). In the western mouth-bar the SOM 2 results allow to clearly discriminate between the mouth-bar (thicker part with coarser sands) and the mouth-bar fringe. This discrimination is not evident with the spectral decomposition (Figure 11c). The main differences in the interpretation performed on the spectral decomposition and the SOM 2 results are evident in the lower part of the meandering distributary channel running form north to south in the centre of the image (see red arrow in Figures 11a and 11b), and in the depositional area downslope of the many distributary channels in the central part of the area. This depositional lobe has been interpreted as a mouthbar (Figures 11b and 11d). The interpretation of this area is more complicated in the spectral decomposition where light-coloured patches interpreted as sands deposits seem to be detached from the distributary channels. The meandering channel is better defined by the SOM result (Figure 11b). The southernmost part of it, in the spectral decomposition, shows a light-coloured patch quite difficult to interpret within a delta facies model (Figure 11c). In addition, within the channel it is possible to recognise different facies on the SOM 2 result, with zone of deposition and zone of by-pass.

Conclusions

The unsupervised clustering technique of SOM is a powerful tool in deciphering sedimentary environments and geomorphological features in different geological settings. It reaches comparable results to spectral decomposition, and it adds details not visible with conventional attributes. A key factor is the attributes selection. This requires the seismic interpreter’s experience and knowledge of the different attributes with their strengths and limitations. The ML approach cannot be blindly applied; simply crunching the data is not good enough. The seismic interpreter involvement is paramount. Seismic interpreters and exploration geophysicists need to be accustomed to ML techniques. They are the ones who can better understand and interpret the results and they need to be the ones to feed-in meaningful input to the algorithms. Showing that these techniques can be implemented with little effort and no additional resources in terms of software is an important encouragement for those explorationists that feel a bit intimidated by ML jargon and buzzwords. Geophysicists are familiar with terms like regression, clustering, and neural network. These data analysis techniques have been an integral part of their work for decades. Thus, every geophysicist can and should drive the implementation of ML into the future of seismic interpretation.

References

Bahorich, M.S. and Farmer, S.L. [1995]. 3D Seismic discontinuity for faults and stratigraphic features: the coherency cube. The Leading Edge, 14(10), 1053–1058. https://doi.org/10.1190/1.1887523

Barnes, A.E. [2001]. Seismic Attributes in Your Facies. Recorder, 26(7), 33-38. https://csegrecorder.com/articles/view/seismic-attributes-in-your-facies

Barnes, A.E. [2007]. Redundant and useless seismic attributes. Geophysics, 72(3), 33-38. https://doi.org/10.1190/1.2716717

Brown, A.R., Graebner, R.J. and Dahm, C.G. [1982]. Use of horizontal seismic sections to identify subtle traps. AAPG Memoir, 32, 47–56. https://doi.org/10.1306/M32427C4

Castagna, J.P. and Sun, S. [2006]. Comparison of spectral decomposition methods. First Break, 24(3), 75–79. https://doi.org/10.3997/13652397.24.1093.26885

Chopra, S., Sharma, R.K., Marfurt, K.J., Bedle, H. and Verma, S. [2022]. Attempts at seismic characterization of a deepwater turbidite channel in Taranaki Basin, New Zealand. First Break, 40(10), 27-39. https:// doi.org/10.3997/1365-2397.fb2022080

Cortés Arroyo, O.L., Steuer, A. and Preugschat, B. [2021]. Principal component analysis as pre-processing stage for self-organized maps. Protokoll über das 29. Schmucker-Weidelt-Kolloquium für Elektromagnetische Tiefenforschung: virtuell, 29 September - 1 October 2021, Extended Abstract, 2-10. https://gfzpublic.gfz-potsdam.de/ pubman/item/item_5010114

Crameri, F., Shephard, G.E. and Heron, P.J. [2020]. The misuse of colour in science communication. Nature Communications, 11, 5444. https://doi.org/10.1038/s41467-020-19160-7

Gersztenkorn, A. and Marfurt, K.J. [2006]. Eigenstructure-based coherence computations as an aid to 3-D structural and stratigraphic mapping. Geophysics, 64(5), 1468-1479. https://doi.org/10.1190/1.1444651

Hart, B.S. [2008]. Channel detection in 3-D seismic data using sweetness. AAPG Bulletin, 92(6), 733-742. https://doi.org/10.1306/02050807127

Jipa, D.C. and Olariu, C. [2009]. Dacian Basin. In: Geo-Eco-Marina Special Publication, 3, National Institute of the Marine Geology and Geo-Ecology, Bucharest, 264. https://geoecomar.ro/website/docs/ cartea-jipa.pdf

Kohonen, T. [1982]. Self-Organized Formation of Topologically Correct Feature Maps. Biological Cybernetics, 43(1), 59–69. https://doi. org/10.1007/BF00337288

Kovesi, P. [2015]. Good Colour Maps: How to Design Them. ArXiv, abs/1509.03700, https://api.semanticscholar.org/CorpusID:947809

Kovesi, P. Perceptually Uniform Colour Maps, https://colorcet.com/

Krézsek, C., Lăpădat, A., Maţenco, L., Arnberger, K., Barbu, V. and Olaru, R. [2013]. Strain partitioning at orogenic contacts during rotation, strike–slip and oblique convergence. Paleogene–Early Miocene evolution of the contact between the South Carpathians and Moesia. Global and Planetary Change, 103, 63-81. https://doi.org/10.1016/j. gloplacha.2012.11.009

Krézsek, C. and Olariu, C. [2021]. Filling of sedimentary basins and the birth of large rivers. The lower Danube network in the Dacian Basin, Romania. Global and Planetary Change, 197, 103391. https://doi. org/10.1016/j.gloplacha.2020.103391

La Marca, K. and Bedle, H. [2022]. Deepwater seismic facies and architectural element interpretation aided with unsupervised machine learning techniques: Taranaki Basin, New Zealand. Marine and Petroleum Geology, 136, 105427. https://doi.org/10.1016/j.marpetgeo.2021.105427

La Marca-Molina, K., Silver, C., Bedle, H. and Slatt, R. [2019]. Seismic facies identification in a deepwater channel complex applying seismic attributes and unsupervised machine learning techniques: A Case Study in the Taranaki Basin, New Zealand. 89th Annual International Meeting, SEG, Expanded Abstracts, 2059–2063. https://doi. org/10.1190/segam2019-3216705.1

Leever, K.A., Matenco, L., Bertotti, G., Cloetingh, S. and Drijkoningen, G.G. [2006]. Late orogenic vertical movements in the Carpathian Bend Zone - Seismic constraints on the transition zone from orogen to foredeep. Basin Research, 18(4), 521–545. https://doi.org/10.1111/ j.1365-2117.2006.00306.x

Lubo-Robles, D. and Marfurt, K.J. [2017]. Delineation of thick incised canyons using spectral-decomposition analysis, curvature and Self-Organizing Maps in the Exmouth Plateau, Australia. 87th Annual Meeting, SEG, Expanded Abstract, 2420-2424. https://doi.org/10.1190/ segam2017-17790008.1

Lubo-Robles, D., Devegowda, D., Jayaram, V., Bedle, H., Marfurt, K.J. and Pranter, M.J. [2022]. Quantifying the sensitivity of seismic facies classification to seismic attribute selection: An explainable machine-learning study. Interpretation, 10(3), SE41-SE69. https://doi. org/10.1190/INT-2021-0173.1

Luo, Y., Higgs, W.G. and Kowalik, S. [1996]. Edge detection and stratigraphic analysis using 3-D seismic data. 66th Annual Meeting, SEG, Expanded Abstracts, 324-327. https://doi.org/10.1190/INT-20210173.1

Marfurt, K.J. [2006]. Robust estimates of reflector dip and azimuth. Geophysics, 71(4), 29-40. https://doi.org/10.1190/1.2213049

Matos, M.C. and Marfurt, K.J. [2009]. Seismic Color Self-Organizing Maps. 11th International Congress of the Brazilian Geophysical Society, Expanded Abstracts, 914-917. https://doi.org/10.1190/sbgf2009-198

Matos, M.C., Osorio, P.L.M. and Johann, P.R.S. [2003]. Using Wavelet Transform and Self Organizing Maps for Seismic Reservoir Characterization of a Deep-water Field, Campos Basin, Offshore Brazil. 8th International Congress of the Brazilian Geophysical Society, Expanded Abstracts, 1-3. https://sbgf.org.br/mysbgf/eventos/expanded_abstracts/8th_CISBGf/arq_144.pdf

Matos, M.C., Osorio, P.L.M. and Johann, P.R.S. [2004]. Unsupervised Seismic Facies Classification Using Matching Pursuit and Self Organizing Maps. 66th EAGE Annual Conference & Exhibition, Extended Abstracts. https://doi.org/10.3997/2214-4609-pdb.3.C016

Matos, M.C., Osório, P.L.M. and Johann, P.R.S. [2007]. Unsupervised seismic facies analysis using wavelet transform and self-organizing maps. Geophysics, 72(1), P9–P21. https://doi.org/10.1190/1.2392789

McKinney, W. [2010]. Data structures for statistical computing in Python. Proceedings of the 9th Python in Science Conference, 51–56. https:// doi.org/10.25080/Majora-92bf1922-00a

Niccoli, M. [2014]. How to evaluate and compare color maps. The Leading Edge, 33(8), 910–912. https://doi.org/10.1190/tle33080910.1

Partyka, G., Gridley, J. and Lopez, J.A. [1999]. Interpretational applications of spectral decomposition in reservoir characterization. The Leading Edge, 18(3), 353–360. https://doi.org/10.1190/1.1438295

Pearson, K. [1901]. On Lines and Planes of Closest Fit to Systems of Points in Space. Philosophical Magazine, 2(11), 559–572. https://doi. org/10.1080/14786440109462720

Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J., Passos, A., Cournapeau, D., Brucher, M., Perrot, M. and Duchesnay, E. [2011]. Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12, 2825–2830. https://www.jmlr.org/ papers/volume12/pedregosa11a/pedregosa11a.pdf

Popov, S.V., Shcherba, I.G., Ilyina, L.B., Nevesskaya, L.A., Paramonova, N.P., Khondkarian, S.O. and Magyar, I. [2006]. Late Miocene to Pliocene paleogeography of the Paratethys and its relation to the Medi-

terranean. Paleogeography, Paleoclimatology, Paleoecology, 238(1-4), 91–106. https://doi.org/10.1016/j.palaeo.2006.03.020

Posamentier, H.W., Davies, R.J., Cartwright, J.A. and Wood, L. [2007]. Seismic geomorphology- an overview, in R.J. Davies, H.W. Posamentier, L.J. Wood, and J.A. Cartwright, eds., Seismic Geomorphology: Applications to Hydrocarbon Exploration and Production. Geological Society of London, Special Publications, 277, 1-14. https://doi.org/10.1144/ GSL.SP.2007.277.01.16

Roden, R., Smith, T. and Sacrey, D. [2015]. Geologic pattern recognition from seismic attributes: Principal component analysis and self-organizing map. Interpretation, 3, SAE59–SAE83. https://doi.org/10.1190/ INT-2015-0037.1

Roden, R. and Sacrey, D. [2016]. Seismic interpretation with machine learning. GeoExpro, 13(6), 50–53. https://geoexpro.com/seismic-interpretation-with-machine-learning/ Rögl, F. [1998]. Palaeogeographic considerations for Mediterranean and Paratethys seaways (Oligocene to Miocene). Annalen Naturhistorisches Museum Wien, 99, 279–310. https://www.zobodat.at/pdf/ ANNA_99A_0279-0310.pdf

Rögl, F. [1999]. Mediterranean and Paratethys. Facts and Hypotheses of an Oligocene to Miocene Paleogeography (Short Overview). Geologica Carpathica, 40(4), 339–349. https://nhm-wien.ac.at/jart/prj3/nhm-resp/ data/uploads/mitarbeiter_dokumente/roegl/1999_Roegl_Palgeo_GeolCarp.pdf

Sacrey, D. and Roden, R. [2014]. Understanding Attributes and Their Use in the Application of Neural Analysis- Case Histories Both Conventional and Unconventional. Search and Discovery Article, #41473. https:// www.searchanddiscovery.com/pdfz/documents/2014/41473sacrey/ ndx_sacrey.pdf.html

Stecker, U. and Uden., R. [2002]. Data mining of 3D poststack seismic attribute volumes using Kohonen self-organizing maps. The Leading Edge, 21(10), 1032-1037. https://doi.org/10.1190/1.1518442

Taner, M.T., Koehler, F. and Sheriff, E.E. [1979]. Complex seismic trace analysis. Geophysics, 44, 1041–1063. https://doi.org/10.1190/1.2785048

Tellez, J.J. [2015]. Seismic Sequence Stratigraphy and Architectural Elements for Cenozoic Strata at the Rankin Platform Sub Basin, North Carnarvon Basin, Australia. M.S. thesis, University of Oklahoma. Vettigli, G. [2018]. MiniSom: minimalistic and NumPy-based implementation of the Self Organizing Map. https://github.com/JustGlowing/minisom/ Xu, G. and Haq, B.U. [2022]. Seismic facies analysis: Past, present, and future. Earth-Science Reviews, 224, 103876. https://doi.org/10.1016/j. earscirev.2021.103876

Zhao, T., Jayaram, V., Roy, A. and Marfurt, K.J. [2015]. A comparison of classification techniques for seismic facies recognition. Interpretation, 3, SAE29–SAE58. https://doi.org/10.1190/INT-2015-0044.1

Zhao, T., Zhang, J., Li, F. and Marfurt, K.J. [2016]. Characterizing a turbidite system in Canterbury Basin, New Zealand, using seismic attributes and distance-preserving self-organizing maps. Interpretation, 4(1), 79–89. https://doi.org/10.1190/INT-2015-0094.1

Zhao, T., Li, F. and Marfurt, K.J. [2017]. Constraining self-organizing map facies analysis with stratigraphy: An approach to increase the credibility in automatic seismic facies classification. Interpretation, 5(2), T163–T171. https://doi.org/10.1190/INT-2016-0132.1

Zhao, T., Li, F. and Marfurt, K.J. [2018]. Seismic attribute selection for unsupervised seismic facies analysis using user-guided data-adaptive weights. Geophysics, 83(2), O31-O44. https://doi.org/10.1190/ geo2017-0192.1

Special Topic

RESERVOIR MONITORING

Submit an article

Innovation in reservoir monitoring remains crucial as geophysicists continue to innovate to optimise existing fields and explore for new fields. Meanwhile, monitoring and characterisation of reservoirs for carbon capture, utilisation and storage is becoming at least as important as the CCUS industry continues to gather pace and governments hand out licences to operators.

This month we showcase new technologies including time-lapse gravity and subsidence monitoring and use of AI to estimate porosity and fluid saturation and gain insights into the latest potential plays that are benefiting from technological advances. Other vital technologies are the latest generation of ocean bottom nodes, seabed deformation measurements as well as P-Wave and S-Wave technology. The sector continues to evolve rapidly.

Laust Jørgensen et al present time-lapse gravity and subsidence monitoring as a key component of a broader geophysical monitoring program that integrates multiple methods, tailored to support the production strategy throughout the different phases of the Troll field’s lifetime.

Tom Davis tells the monitoring story of the Denver Basin, leading to the shale production.

Pedro V. Zalan et al present a new, distinctive prospect in the southern Santos Basin, constituted of a possible Albian atoll developed upon exhumed mantle.

Ignatius Sonny Winardhi et al offer a simple method to estimate porosity and fluid saturation directly from AI and VP /VS ratio typically obtained from the pre-stack seismic inversion.

Mohamad Yousof Hourani et al summarise the results of the modelling work conducted to assess the feasibility of time-lapse gravity and subsidence modelling at Scarborough gas field.

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.

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

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

Special Topic overview

January Land Seismic

February Digitalization / Machine Learning

March Reservoir Monitoring

April Underground Storage and Passive Seismic

May Global Exploration

June Navigating Change: Geosciences Shaping a Sustainable Transition

July Modelling / Interpretation

August Near Surface Geo & Mining

September Reservoir Engineering & Geoscience

October Energy Transition

November Marine Acquisition

December Data Management and Processing

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

Photo courtesy of Equinor

Twenty five years of monitoring the Troll gas and oil field with time-lapse gravity and seafloor deformation surveys

Siri Vassvåg1, Felix Halpaap1, Charlotte Faust Andersen2 and Laust Jørgensen2* present time-lapse gravity and subsidence monitoring as a key component of a broader geophysical monitoring program that integrates multiple methods, tailored to support the production strategy throughout the different phases of the Troll field’s lifetime.

Introduction

The time-lapse gravity and seafloor deformation surveying method underwent its initial trial at the Troll field more than 25 years ago. Since then, a total of nine surveys have been performed, with the latest one conducted in 2023, to provide insight into the flow dynamics within the reservoir.

Troll is one of the biggest petroleum fields in the Norwegian sector of the North Sea, located offshore the west coast of Norway. It contains both oil and gas, where gas accounts for 80% of the total quantity and constitutes approximately 40% of Norway’s total gas reserves.

1 Reach Subsea | 2 Equinor Energy AS * Corresponding author, E-mail: lajor@equinor.com DOI: 10.3997/1365-2397.fb2025020

The field is divided into three large, rotated fault blocks and consists of two distinct regions: Troll East and Troll West (see Figures 1 and 2). The eastern fault block, known as Troll East, contained an initial gas column up to 230 m thick, with a negligible oil leg. Troll West comprises the two westernmost fault blocks, which have a similarly substantial gas column, reaching an initial thickness of more than 200 m. Beneath the gas column in Troll West lies a thin oil leg, ranging from 11 to 26 m in thickness pre-production.

The reservoir in both segments consists of shallow marine sandstones from the Late Jurassic Sognefjord formation. The reservoir quality is excellent, characterised by high porosity and permeability that make it highly producible.

Troll development and geophysical monitoring

The development of the Troll field has been divided into three phases. Phase I, which began in 1996, focuses on gas production from Troll East and is expected to continue for several decades. Phase II, initiated in 1995, targets the oil column in the Troll West area. This phase is expected to conclude by the end of the decade, with the final well for oil extraction drilled in 2023. Phase III, which commenced in 2021, involves gas cap production at Troll West and, like Phase I, is anticipated to continue for many years. Due to the distinct reservoir dynamics and production strategies

the three

with the

to the east corresponding to Troll East and the other two to Troll West. The depth axes annotate metres below sea level. TVDMSL is true vertical depth, mean sea level.

Figure 1 Map showing the rim of the Troll field. The background colour represents the producible hydrocarbon type: red for gas at Troll East and red and green for oil and gas at Troll West. The black dots mark the position of the concrete pads on the seafloor, which are used for gravity and subsidence monitoring. The grey lines are all the drilled wells.

Figure 2 Cross section of the Troll reservoir. It shows

fault blocks,

one

in each phase, the geophysical reservoir monitoring strategy has been tailored accordingly. Figure 3 provides an overview of this comprehensive monitoring approach.

In Phase I, gas production from the Troll East reservoir is carried out through vertical wells drilled from the Troll A platform. This production has caused significant pressure depletion and a rise in the gas-liquid contact by over 50 m, leading to a shift in the reservoir’s mass balance. Time-lapse gravimetry is an effective monitoring method in this phase, as the resulting mass and pressure changes manifest as variations in gravity and subsidence over time. While 4D seismic surveys have also been conducted during Phase I, their primary role has been to complement gravity monitoring.

In Phase II, the thin oil column is produced through long horizontal multilateral wells drilled from templates distributed across the entire Troll West area. The primary goal of geophysical monitoring in this phase has been to optimise well placement and identify zones and layers with remaining producible oil. 4D seismic data has been instrumental in guiding the planning of infill wells by pinpointing undrained sand layers. In contrast, gravity data has had limited value in Phase II due to its inherent constraints in vertical and lateral resolution. Additionally, the lateral resolution, with stations spaced 1-3 km apart, is insufficient for identifying localised areas of undrained sands.

Troll Phase III marks an exciting new chapter in the development of the Troll field. In this phase, the gas cap at Troll West is produced from new seabed templates, with the first well clusters coming onstream in 2021. Drawing on lessons learnt from previous phases, time-lapse gravity monitoring has been chosen as the primary method for tracking contact movements and aquifer influx. While other alternatives, such as 4D seismic, are under consideration, time-lapse gravity monitoring will serve as the initial approach. The most recent survey conducted in 2023 is the first to assess the impact of Phase III production, and it is hence too early to finalise a long-term geophysical monitoring strategy.

Time-lapse gravity and seafloor deformation monitoring method

The method is based on periodic surveying, with normally one baseline and multiple repeat surveys conducted during the lifetime of a field. Each survey entails the acquisition of measurements of relative gravity and water pressure at predetermined locations on the seafloor. These measurement sites are demarcated by pre-deployed concrete pads (CPs), with the shape of a truncated cone and a top plate with a diameter of 80 cm. Figure 1 presents the spatial distribution of CPs deployed on the seafloor above the Troll field.

The surveys are conducted using a vessel equipped with a remotely operated vehicle (ROV). The ROV carries a sensor frame that typically contains three relative gravimeters and three pressure sensors. Each measurement cycle commences with the vessel navigating to the designated CP location, followed by the ROV’s descent to the seafloor. Upon reaching the CP, the sensor frame is carefully positioned atop it and left to perform a measurement for 15 minutes. Subsequently, the ROV is retrieved to the surface, and the vessel proceeds to the next CP location to repeat the process. Figure 4 shows a picture of a modern sensor frame being carried by an ROV.

During the survey, tide gauges are deployed across the field to facilitate the removal of tidal effects from each measurement recorded at the CPs. This ensures a consistent and accurate time-lapse analysis across all surveys. Additionally, at least two measurements are performed at each CP at different times during the survey. This multiple measurement approach serves two purposes: firstly, it enables the correction of potential drift in the gravimeters and pressure sensors; secondly, it allows for the evaluation of measurement uncertainty. For the latter, repeatability estimators (Ruiz et al., 2022) are obtained by analysing the variance between different gravity and pressure measurements taken at the same CPs, after being corrected for drift and tidal effects. The repeatability parameter obtained this way estimates

Figure 3 Timeline showing the geophysical reservoir monitoring (GRM) strategy on the Troll field, historical and planned, subdivided in the three specific development phases

Figure 4 The sensor frame (white cylinder) being carried by the ROV, while coming out of water after a dive (left), and at the parking position at the deck (right).

the uncertainty associated with individual single 15-minute measurements on a CP.

Relative gravity and water pressure measurements from subsequent surveys are processed to produce measurements of, respectively, time-lapse changes in the gravitational field and vertical deformation of the seafloor, that is, seafloor subsidence or uplift. These time-lapse variations are closely related to key properties of the reservoir: gravity informs about fluid saturations and densities, and vertical seafloor deformation informs about reservoir compaction or expansion. The technology has been used for decades to monitor gas fields in Norway (Sølbu et al., 2023, Ruiz et al., 2022), in some cases as the only time-lapse monitoring technology, and in others, complementing 4D seismic monitoring. The cost efficiency of the technology makes it also a key element for monitoring carbon sequestration projects (Borges et al., 2024).

Technology development: from pilot deployment to unmanned operations

The time-lapse gravity and seafloor deformation method underwent its pilot deployment in 1998 at the Troll field (Sasagawa et al, 2003). That inaugural baseline survey involved measurements on 42 CPs. Since then, eight repeat surveys have been conducted at Troll, reflecting continuous improvement in both measurement accuracy and operational efficiency.

Additional CPs have been installed over time to augment the coverage of the measurement grid. Notably, before the 2017 survey, the density of CPs across Troll West was increased to enhance lateral resolution, particularly to monitor gas production within Phase III.

Table 1 summarises the history of surveys performed at Troll, including the estimate of the single-measurement uncertainty as provided by the repeatability parameter defined above. The measurement accuracy improves steadily with time, with the slightly deteriorated repeatability in 2023 being caused by adverse weather conditions, which led to an increased frequency of operational interruptions and marginally deteriorated the quality of tidal corrections.

The accuracy in the measurement of changes over time of the gravitational field and depth at the CPs depends on the uncer-

Table 1 Time-lapse gravity and seafloor deformation surveys performed at Troll. The gWatch equipment, ready for operations with unmanned vessels, was introduced in 2021. Repeatability is computed by analysing the variance of repeated measurements to the same CPs during a survey, and estimates the uncertainty of a single measurement on a CP.

tainty of the measurement within each survey and the accuracy of the in-situ calibration process (Agersborg et al, 2017). The measurement uncertainty within a survey at a particular CP is in turn a function of the repeatability presented in Table 1 and the number of measurements performed at that CP. The accuracy of the measurements of gravity changes and subsidence is estimated to be respectively 0.79 µGal and 3.4 mm for the 2021-2023 time lapse at Troll.

The recent significant improvement in accuracy in 2021 is due to the introduction of a new generation of the technology, gWatch. The new generation features several upgrades that improve the accuracy of the measurements. For example, the use of the newer CG6 gravimeter and updated electronics and mechanical stabilisation measures reduces the drift rate to typically a few µGal per day, compared to the order of 1 mGal/ day of drift in the first CG5 deployments (Sasagawa et al., 2003). In addition, gWatch features patented enhanced thermal stabilisa-

Figure 5 Left: an unmanned vessel capable of performing time-lapse gravity and seafloor

Right: Illustration of the ROV that will be deployed by the unmanned vessel, featuring a front drawer (in yellow) for securely carrying the sensor frame (centre of the drawer) and the required tide gauges (sides of the drawer).

tion, which reduces temperature sensitivity from both gravimeters and pressure sensors.

gWatch also features several enhancements that improve operational efficiency. The integrated active thermal stabilisation eliminates the necessity for external cooling while in transit between CPs. The lower weight and reduced volume enable simple manipulation of the sensor frame with a standard ROV manipulator arm rather than from a dedicated tool. The improved sensor performance and temperature stabilisation allow for reducing the measurement time at each CP while still improving the quality of the measurements. These advancements render the technology suitable for deployment from unmanned surface vessels, as depicted in Figure 5. The forthcoming introduction of unmanned surveys in 2025 is poised to reduce even further fuel consumption and CO2 footprint of the operations.

Reservoir changes measurable by gravity and subsidence data

Changes over time in the gravitational field at the seabed over a hydrocarbon-producing or CCS storage site have three components. The first one, of primary relevance for reservoir monitoring, is the change of fluid mass distribution within the reservoir. The second component arises from seafloor subsidence or uplift, which results in the measurements at subsequent surveys being performed at different depths. This component is easy to subtract utilising the measured values of subsidence or uplift and a correction factor that accounts for the free air gradient and the height of the water column above the CP. The third component is the effect of rock deformation, especially within the overburden. This component is negligible for fields with moderate reservoir compaction or expansion and can be subtracted to the required accuracy by means of subsurface deformation modelling.

The reservoir fluid mass changes measured by gravity have in turn two main components. The first, with a positive sign is denser water replacing hydrocarbons in the pore spaces. The second one, with a negative sign, is hydrocarbon depletion. This would be caused by mass out-take by liquid production and oil being replaced by lighter gas when the aquifer is weaker.

When the pore pressure decreases due to hydrocarbon production, the reservoir rock compacts and the effect propagates through the overlying rock layers, leading to a subsidence of the seafloor. Therefore, seabed subsidence is directly related to reservoir depletion and can be used to monitor pressure depletion

Figure 6 Measurements of average yearly subsidence rates at Troll in the time lapses 2012-2017 (left), 2017-2021 (middle) and 2021-2023 (right). Each bubble corresponds to a CP, and the bubble area is proportional to the measured subsidence. A bubble provided on the bottom left of each plot indicates the scale.

and pressure communication across the field, and to calibrate rock-mechanical reservoir models.

Results and prospects on gravity and subsidence monitoring at Troll

Figure 6 displays the yearly average subsidence measurements obtained from the four most recent surveys conducted at Troll. The figure reveals a subsidence rate of approximately 1 cm per year. Even if the most valuable application of the subsidence measurements at Troll is correcting gravity data in the manner described above, the measurements are also used to provide information on pressure depletion and to increase the confidence in the geomechanical model.

Troll East exhibits the most considerable subsidence, which has been attributed to gas production and the consequent pressure depletion. Meanwhile, Troll West has experienced significantly increased subsidence in the most recent time-lapse, which coincides with the start of gas production in Phase III in 2021. Subsidence measurements provide valuable insights into the state of reservoir depletion across the field. Notably, the picture provided by subsidence measurements agrees with direct pressure measurements at wells and reservoir model simulations, as illustrated by Figure 7. Note the acceleration of the depletion observed in Phase III.

The subsidence predicted by the geomechanical reservoir model was compared to the subsidence measured by gravity stations and GPS at the Troll A platform. The strong alignment

Figure 7 Pressure depletion and production over time. Depletion is shown for Troll East (orange) and Troll West (blue), with the time period for the three development phases marked. Historical yearly production is shown for oil and gas in the bar chart.

between both datasets and the model has improved confidence in the model’s ability to accurately estimate subsurface stress, thereby enhancing its predictive power.

Gravity change measurements, once corrected from subsurface deformation effects as described above, have been utilised to monitor fluid changes and update the reservoir model, particularly with regards to the strength and direction of the aquifer support.

Figure 8 illustrates the yearly average gravity changes corrected for seafloor subsidence observed in the three most recent surveys.

Negative gravity changes over Troll West indicate that the effect of oil depletion dominates the mass change, while the positive gravity signals in the southern region of Troll East shows that water influx dominates over the effect of gas depletion in this area, with magnitudes indicating a strong aquifer with a south to north direction.

In response to the increased gas production in phase III (figure 7), certain rim areas are exhibiting positive gravity values due to accelerated water influx. This shift has been confirmed by production wells that are experiencing a rapid increase in water cut (marked with yellow circles in figure 8).

We have also used the gravity and subsidence data to update the reservoir model for Troll East and West. This is achieved through a sophisticated ensemble modelling technique that is central to the assisted history matching process. By leveraging an array of observational data, including gravity metrics, pressures, and production history, ensemble modelling serves as a robust framework to assimilate these diverse data sets.

The Ensemble Smoother with localisation is used to simultaneously update a broad array of parameters, such as geological features, aquifer strength and direction, communication pathways, relative permeability, and other parameters. This ensures a comprehensive and up-to-date representation of the reservoir (Evensen, G. et al., 2009).

Figure 8 Measurement of average yearly gravitychange corrected for subsidence at Troll in the time lapses 2012-2017 (left), 2017-2021 (middle) and 2021-2023 (right). Each bubble corresponds to a CP, and the bubble area is proportional to the measured value. The colour of the bubble indicates whether the changes are positive (blue) or negative (red). A grey bubble provided on the bottom left of each plot indicates the scale. Stippled circles mark rim areas in Troll East (black) and Troll West (yellow).

For Phase III, optimising the placement of new wells is crucial to ensure maximum production efficiency. Gravity monitoring will provide critical insights into the movement of gas-liquid contacts, reservoir depletion and identifying areas with high aquifer influx. This information can be utilised to determine the optimal position and timing of new wells. Additionally, a unique opportunity exists to further improve the interpretation of the gravity data in this particular setting through calibration with pressure and fluid contact observations from the active oil producers in Troll West.

Conclusions

The use of gravity and subsidence monitoring has been essential for assessing and managing production at the Troll reservoir since 1998. As we enter Phase III in the history of the field, this technology continues to be an important part of the geophysical monitoring strategy.

Time-lapse gravity and seafloor deformation data have proven to be effective in providing insights into subsurface stress, depletion patterns, and aquifer responses. These insights enhance the reliability of reservoir models and improve our understanding of fluid movements within the reservoir.

With these data, we can improve our reservoir models through history matching and make better decisions regarding production optimisation. Ongoing development and application of gravity and subsidence monitoring techniques can further improve how we manage field resources, promoting sustainable and efficient operations.

Overall, gravity monitoring will continue to play a vital role throughout the life of the Troll field. It helps with effective well placement, supports production estimates, reduces operational risks, and optimises production strategies. Integrating this technology into our decision-making processes is important for maximising the reservoir’s potential while ensuring environmental responsibility.

Acknowledgements

We would like to thank the Troll Unit licence partners Petoro AS, A/S Norske Shell, TotalEnergies EP Norge AS, ConocoPhilips Skandinavia AS and Equinor Energy AS (operator) for permission to publish this article. The views expressed in this paper may not reflect the views of all Troll licence partners.

Figure 9 Flowchart of the modelling workflow using the gravity data to update the reservoir model.

Musings on the Golden Fault and the Front Range Zone of Flank Deformation

Tom Davis1* tells the monitoring story of the Denver Basin, leading to shale production.

Abstract

The Golden Fault on the west side of Golden, Colorado has been the subject of debate in the geological community for several years. The source of the basalt that caps North and South Table Mountain is also of interest to the scientific community as the source of the basalt has never been identified. There may very well be a connection between these two major geological features. In addition, the structure of the Front Range Zone of Flank Deformation itself is of interest as it contains mineral resources that have helped to shape the economic development of the region. This natural resource setting also shaped the Colorado School of Mines from its creation in 1874 to become one of the most famous engineering universities in the world.

Introduction

The Denver Basin is an asymmetrical-shaped foreland basin whose centre is located under the city of Denver, Colorado (Figure 1). Shale development drilling and production is underway in the Niobrara Formation with the onset of horizontal drilling going back to the early 1990s (Davis, 2024). The earliest production from the Niobrara took place in the Old Boulder Oilfield, 20 miles north of Golden, discovered in 1906. The field is still producing. The Johnson well drilled 10 miles south of Golden near Soda Lakes in the mid-1950s produced close to 10,000 barrels out of the Niobrara and Codell Formations. All this activity spread over a hundred years indicates that

the Niobrara is far from dead as a resource play and that considerable potential exists along the eastern flank of the Front Range.

Seismic programs shot by the Colorado School of Mines over a five-decade time span have been used to interpret the structure of the Golden Fault, the Zone of Flank Deformation and faulting in the Denver Basin. These faults have exhibited recurrent movement over geologic time from their origin during Precambrian time. They are largely categorised as wrench faults but they also created fault detachment or slide planes and listric normal faults depending on the mechanical properties of the rocks involved and fluid pressures in the subsurface. These faults are fundamental to the current development of the Niobrara and Codell reservoirs in the Denver Basin and seismic plays a significant role in exploring for and developing these resources.

The Golden Fault

The Golden Fault on the western side of Golden, Colorado marks the front of the Rocky Mountains (Figure 2). There is at least 12, 000 feet of throw on the fault bringing Precambrian age rocks of the Front Range over the sedimentary rocks of the Denver Basin. The City of Golden resides in a valley between the Front Range on the west and the basalt-capped table mountains to the east. Golden was the territorial capital in the late 1850s when gold was discovered. The capital was later moved to Denver and the State of Colorado was formed in 1876.

1 Colorado School of Mines

* Corresponding author, E-mail: tdavis@mines.edu

DOI: 10.3997/1365-2397.fb2025021

Figure 1 The Denver Basin and the Niobrara Petroleum System. Courtesy of Steve Sonnenberg.

The structure of the Eastern Flank of the Front Range was first depicted by Boos and Boos (1957) and later by Berg (1962) from outcrop and subsurface studies in the vicinity of Turkey Creek, south of Morrison Colorado (Figure 3). Morrison is famous for the dinosaurs discovered in the Morrison Formation in the 1880s. Controversy has surrounded the early interpretations of the Golden Fault (Gries, 1983). Foreland structure is of interest to the geological community because of the potential for discovery of oil and gas resources. The Golden Fault is a complex fault zone in large part due to the mechanical properties of the rocks involved. Precambrian igneous and metamorphic rocks behave in a brittle fashion whereas sedimentary rocks tend to behave in a ductile manner. Precambrian rocks break under stresses in a shear fashion whereas sedimentary rocks tend to fold. This observation led Stearns (1971) to use the term ‘drape fold’ to categorise Berg’s earlier interpretation of the Golden Fault. Gries categorises the Golden Fault as a thrust fault that brings substantial overhang of Precambrian Basement over the sediments of the Denver Basin.

A seismic line shot by Domoracki (1986) as part of his Master of Science Thesis in Geophysics at Colorado School of Mines and illustrated in Figure 2 shows numerous seismic events below the surface outcrop projection of the Golden Fault (Figure 4). Generally, the events dip eastward in the footwall of the Golden Fault and westward in the hanging wall. There appears to be a large structure in the footwall but much of this apparent structure is caused by velocity pull-up. The Precambrian has a relatively constant interval velocity of 16,000 feet/second or 4800 m/s versus an average velocity of 9000 feet/second or 2700 m/s for the sedimentary section. This velocity discrepancy creates a half second of velocity pull-up on the near basement event in the footwall section of the seismic line.

The determination of real structure in the footwall section versus apparent structure is complicated by the lateral and vertical velocity field in the subsurface. Seismic modelling is a necessary and integral part of unravelling this complexity. Figure 5 shows a seismic model created by Domoracki. Sediments in the footwall model dip eastward at ½ degree with no subsurface anticlinal

Figure 2 The location of the seismic line shot across the Golden Fault.

Figure 3 A geological cross-section across the Golden Fault south of Golden, Colorado (From Berg, 1962). The Golden Fault is not a singular fault but a complex fault zone.

structure. With a slight datum shift the seismic model and the seismic line are a remarkably good match. The one event that doesn’t match is the strongest reflector in the hanging wall section. For Gries and others this event would represent the Golden Fault with a dip of 30 degrees west. The Golden fault emanates out of the basement complex as a relatively steep fault that flattens into the sedimentary section and then becomes steeper as it shallows. In other words, it is not a planar feature. The seismic model indicates the position of the fault is identified as the apex of the diffractions from the major reflections generated from the strong impedance contrast zones in the subsurface. The simple interpretation of the linear event as the Golden Fault may not be accurate and may be deceiving, especially if one is prospecting for potential reservoirs below the Golden Fault.

Anomalous reflector

The strongest reflector in the hanging wall of the Golden Fault is a linear event that dips westward at approximately 30 degrees

(Figure 6). It has been interpreted by Gries to represent the Golden Fault. The fault plane reflection exhibited from the seismic model is much steeper than the anomalous event and has led to the speculation that the linear event is an out of plane reflector or a reflected refraction. The problem with both of those hypotheses is that there is no justification for either of these hypotheses. The Golden Gate seismic line runs perpendicular to the mountain front and the fault itself cuts out any steeply dipping or overturned sediments that would cause a reflected refraction.

The most reasonable interpretation is that the event is real and occurs within the Precambrian basement rock as a shear zone which was also the conduit for extrusion of the basalts in Golden. The strongest case for this event being real is its projection from west to east coinciding with the basalt flows on top of North Table Mountain. The Table Mountains at Golden are considered herein to be formed by the extrusion of basalt from the shear zone in the Precambrian Basement of the Front Range. The reasons are several, including the fact that no other potential source has been

Figure 4 Golden Gate Canyon seismic line that traverses from on top of North Table Mountain on the east up Golden Gate Canyon to the west.

Figure 5 A seismic model of the Golden Gate Canyon seismic line.

identified, to the reflectivity and orientation of the anomalous seismic event. There is no visible evidence of the Golden Fault at the surface or the source of the linear event. The reason is the sedimentary cover at the base of the mountain front. There has been no drilling into the Precambrian Basement to confirm or deny the existence of a basalt-filled shear zone.

The following is a collage of photos taken from the eastern side of the Golden Gate Seismic Line to the western side (Figure 7-10).

The Table Mountains

The basalt flows of Paleocene age that cap the Table Mountains on the east side of the City of Golden are approximately 200 feet thick and are made up of four latite flows. They occur immediately above the Cretaceous/Tertiary Boundary at 66 million years ago, which is marked by a layer of iridium associated with the Chicxulub impact crater in the Yucatan Region, offshore Mexico. That impact led to the ultimate demise of the dinosaurs. The orogenic event that led to the Formation of the Rocky Mountains

is called the Laramide and extended from approximately 70 million years to 40 million years ago. The Laramide was preceded by the Ancestral Rockies uplift that occurred toward the end of Pennsylvanian time (320 my). That orogenic event was related to collision between the African and South American continents with the North American continent and its influence extended from the Gulf Coast all the way up into Colorado. That orogenic event led to the formation of the Red Rocks just south of Golden which is the site of one of the most famous natural amphitheatres in the world.

The instantaneous amplitude display shown in Figure 6 demonstrates the linear, westward dipping event that projects from the ‘basement’ of the Front Range directly to the basalt flows present on the Table Mountains. The polarity of this seismic event indicates it is generated by a high velocity and high density medium. The velocity of the Precambrian basement rock is 16,000 ft/sec (4800 m/s) and the density is 2.65 gm /cm3

Figure 6 Instantaneous amplitude display of the Golden Gate seismic line. Notice the linear event in the upper third of the seismic line.

Figure 7 The eastern side of the Golden Gate Canyon seismic line with North and South Table Mountains in the background.

Figure 8 Acquisition of the Golden Gate Canyon seismic line. The Eastern side of the line began on the basalt-capped North Table Mountain in the background traversing the Cretaceous sediments in the Golden Valley then onto the flank of the Precambrian-cored Front Range.

The basalt within the Precambrian shear zone has a velocity of 19,000 ft/sec (5700 m/s) and a density of 2.9 gm/cm3. The velocity and density contrast gives rise to a high, positive reflection coefficient in the basement rock resulting in a high amplitude event. The largest contributor to the high reflection coefficient from this basalt-filled shear zone is the density contrast. In sedimentary rocks the velocity contrast is the greatest generator of reflection coefficients whereas in basement rocks the density contrast is the dominant generator of reflections in the subsurface.

Prior to acquiring the Golden Gate Canyon seismic line three lines were acquired on the Table Mountains (Young, 1977). Speculation was that the source vent for the basalt flows was under North Table Mountain. Two lines were acquired on North Table Mountain that revealed no evidence for a source vent for the flows. One line was acquired on South Table Mountain. Continuous reflection events were correlated from both surveys revealing no significant disruptions of sediments in the subsurface under either of the Table Mountains. Young’s work indicated there was no source of the basalt under the Table Mountains.

Gravity surveys were conducted along the eastern side of the Front Range to see if the incorporation of gravity would contribute to our knowledge of the deformational flank of the Front Range. After gravity corrections were made the fault was best modelled as a wedge of basement rock dipping at 60 degrees to the southwest. That work was completed by Dr Tom LaFehr and incorporated into Domoracki’s thesis.

Front Range Zone of Flank Deformation

The zone of flank deformation is characterised by a complex fault system with the largest basement overhang in the Golden area. North of Golden the Golden Fault bifurcates into a series of faults. The easternmost fault is referred to as a basin margin fault and runs north-south for at least 10 miles. The fault runs along the western side of Rocky Flats which was the site of a facility that made plutonium triggers for nuclear weapons. Numerous seismic surveys were conducted by the Department of Geophysics, Colorado School of Mines to map the faults and investigate the potential for movement on this fault due to proximity to the Rocky Flats Plant. No faulting was observed directly under the

Plant and there was no evidence of shallow movement on the western fault beyond the end of Laramide time (40 million years). There is evidence of low levels of seismicity in the Front Range indicating the basement is still under stress. On the eastern side of Denver, the Denver Earthquakes felt by residents in the mid-1960s, were attributed to fluid injection near a basement fault that was critically stressed. Earthquakes with magnitudes of 4.0 and above on the Richter scale were recorded at one point in 1965. The source of the earthquakes was the injection well drilled to basement at the Rocky Mountain Arsenal by the Army Core of Engineers (Evans, 1966). This was the first observation of man-made earthquakes. The fault was mapped as having a northwest orientation based on seismicity studies and regional reflection profiling. Near Boulder, Colorado and north of the Rocky Flats wrench faults were mapped from the regional seismic profiling carried out by Mines. Northeast-striking faults trend toward the Wattenberg Field in the deep Denver Basin as shown in Figure 11. Wattenberg is the seventh-largest oil field in the nation. Numerous wrench faults have been mapped in this zone of deformation and are critical to petroleum exploration in the Niobrara (Davis, 2021). Fractures associated with these faults are critical to economic development of the Niobrara. The intersection of the Golden Fault and the northeast trending wrench fault south of Boulder led to a ‘pop-up’ structure that was drilled by Teton Energy in the early 1980s resulting in the discovery of the Superior Gas Field. Other fields have been found where fault intersections or bends occur in the subsurface giving rise to pop-up structures and natural fracturing in the subsurface. One of these is the Old Boulder Oil Field.

Summary

Five decades of seismic studies by the Department of Geophysics of the Colorado School of Mines led to the identification of faults in the Golden area and along the Front Range Zone of Flank Deformation. From the early reconnaissance work of Davis (1974) to present time the understanding of structure has evolved and has led to extensive resource development. From the early detection of oil seeps in the Golden area to the present-day development of the Niobrara and Codell Formations in the Denver Basin through horizontal drilling and multistage hydraulic fracturing geologic structure has played a significant

Figure 10 The recording truck on the western end of the seismic line. Towards the end of the seismic line an oil seep was found at Crawford Gulch.

Figure 9 Traversing up Golden Gate Canyon with the seismic line. The Basement rocks consist of metamorphic rock mainly, but there is a pegmatite dyke shown in pink that cross-cuts the metamorphic rocks.

role in petroleum resource development. Faults and fractures have controlled the emplacement and subsequent development of mineral resources in the Front Range Zone of Flank Deformation.

References

Berg, R.R. [1962]. Subsurface interpretation of the Golden Fault at Soda Lakes, Jefferson County, Colorado. American Association of Petroleum Geologists Bulletin, 46, 704-707.

Boos, C.M. and Boos, M.F. [1957]. Tectonics of eastern flank and foothills of the Front Range, Colorado. American Association of Geologists Bulletin, 41, 2603-2676.

Davis, T. L. [1974]. Seismic investigation of Late Cretaceous faulting along the east flank of the Front Range, Colorado. PhD thesis, Colorado School of Mines, T-1681, 65p.

Davis, T.L. [2024]. Bringing change to shale reservoir development-parts 1 and 2. American Association of Petroleum Geologists Explorer, January and February issues.

Figure 11 Wrench faults in the Denver Basin. The 3D (Merge) and 4D (Turkey Shoot) multicomponent surveys were shot to optimise horizontal well drilling and hydraulic fracturing in the Niobrara and Codell Formations during 2013-2016.

Davis, T.L. [2021]. Faults in the Denver Basin. First Break, 39(4), 57-62.

Domoracki, W. J. [1986]. Integrated geophysical survey of the Golden Thrust north of Golden, Colorado. MS thesis, Colorado School of Mines, T-3052, 134 p.

Evans, D. M. [1966]. The Denver area earthquakes and the Rocky Mountain Arsenal disposal well. The Mountain Geologist, 3, 23-36.

Gries, R. [1983]. Oil and gas prospecting beneath Precambrian foreland thrust plates in Rocky Mountains. American Association of Petroleum Geologists Bulletin, 67(1), 1-28.

Stearns, D. W. [1971]. Mechanisms of drape folding in the Wyoming Province. Wyoming Geological Association, 23rd Annual Field Conference Guidebook, 149-158.

Young, T. K. [1977]. A seismic investigation of North and South Table Mountains near Golden, Jefferson County, Colorado. MS Thesis, Colorado School of Mines, T-1947, 54p.

Ametista Block – An unusual prospect in Santos Basin – Albian atoll upon exhumed mantle

Pedro V. Zalan1*, Milos Cvetkovic2, Henri Houllevigue2, Kyle Reuber 2 and Andrew Hartwig2 present a new, distinctive prospect in the southern Santos Basin, constituted a possible Albian atoll developed upon exhumed mantle.

Abstract

The authors present a new, distinctive prospect in the southern Santos Basin, constituted of a possible Albian atoll developed upon exhumed mantle. The prospect was identified due to the acquisition of new high-quality 3D with the latest processing techniques. The prospect shows great petroleum potential and has excellent analogs in the present Pacific atolls and on an island in the Red Sea. A large buildup composed of stratified seismic facies exhibiting basin-edge offlaps/clinoforms, and capped by a massive structureless construction, is developed upon a protrusion of exhumed mantle. This four-way closed structure is surrounded by Aptian salt bodies that onlap its flanks and by the lowermost strata of the Drift Sequence of the Santos Basin, known to be the ACT source rocks. Considering the nearby geology, this buildup is interpreted as an Albian carbonate platform capped by a reef (rudist?). An adjacent on-trend structure is constituted by volcanos developed upon the exhumed mantle. The overall structure resembles the wellknown atolls of the Pacific Ocean.

Introduction

The Santos Basin in southern Brazil became famous for the fabulous reserves of oil and gas in the Aptian microbialite reservoirs of the so-called ‘Pre-Salt Play’. Super-giant fields such as Tupi and Búzios, and giant fields such as Sapinhoá and Mero, among others, produce close to 2 million boepd. These fields and numbers blur the modest existence of different, but still effectively producing, oil and gas fields in the so-called ‘Post-Salt Plays’. Numerous turbidite-hosted and carbonate-hosted oil and gas fields are also noteworthy in terms of reserves and production. The objective of this work is to present one quite unusual prospect in the Post-Salt Play.

The Ametista Block in southern Santos Basin is one of a kind; half of its area is inside the Pre-Salt Polygon, regulated by Production Sharing directives, and the other half is situated outside the polygon, regulated by Concession regime (Figure 1). Until recently, all the interpretation and basin evolution assessment have been done on gravity and magnetics data, aided by 2D seismic data, some of it acquired and processed 10-15 years ago.

1 ZAG Consulting | 2 TGS

* Corresponding author, E-mail: Pedro.Zalan@tgs.com DOI: 10.3997/1365-2397.fb2025022

Figure 1 Location of Ametista Block on the map of the crustal provinces of the Santos and Campos Basins (modified from Zalan, 2024). Notice its special location upon the V-shaped tip of exhumed mantle. Albian string of pearls refers to several producing Albian age carbonate-hosted oil and gas fields.

Recently acquired 3D data shot specifically to unravel the potential of Ametista Block (Figure 2) displayed a rather unusual, but highly prospective, petroleum system. Geologically, it is the only exploratory block in Brazil situated entirely over exhumed mantle (Figure 1, Zalan 2024). This unique economic basement is the tip of a V-shaped crustal feature termed ‘Propagator’ (Figure 1), considered to be a failed crustal zipper-like rupture, from south to

Figure 2 Outline of new 3D survey shot over Ametista Block, termed by TGS SAN 3D Ph4. See regional context of block in Figure 1.

Figure 3 NW-SE regional 2D seismic section (depth) (Pseudo-Relief/TecVa attribute) depicting the uniqueness of the Ametista Block among surrounding terrains. Central High is the expression of exhumed mantle, it is devoid of salt and displays a stratified buildup (Ametista prospect) developed upon it.

Figure 4 3D KPSDM section inside Ametista Block displaying the Central High developed upon a protrusion of exhumed mantle. It is constituted by two structural closures (A and B). Closure A presents a large buildup constituted by stratified seismic facies (platform) capped by a massive structureless construction (reef?). Closure A constitutes Ametista Prospect. Closure B is devoid of stratified facies, and it is interpreted as volcanoes. The Central High is surrounded by Aptian salt bodies that onlap its flanks and by the lowermost strata of the Drift Sequence of the Santos Basin, known to be the ACT source rocks.

north, formed during the breakup stage (syn-rift) of Western Gondwana. A strong V-shaped positive Bouguer gravity anomaly and direct visualisation in seismic sections set the foundation for this interpretation (Figure 3 and 4). The exhumed mantle forms an outstanding structural high along the centre of the block (Central High). Over this high, no salt is visible, although it is plentiful around it. On top of it, along its crest, lies the other unusual

aspect of this block: a possible Albian atoll constituting a possible prospect with all the characteristics that can be found in Pacific atolls of French Polynesia.

The Ametista Block has an area of 1587 km² and is slated to be offered to the industry in the 3rd Round of Production Sharing in 2025. The block was originally outlined by ANP because of the existence of a prominent high, that in the poor quality 2D seismic data appeared to be capped by either salt diapirs or volcanic constructions. In 2021, TGS acquired a 3400 km² 3D multi-client (MC) marine survey that completely covered the Ametista Block (Figure 2). The modern acquisition and processing build confidence in exploration geologists’ interpretation and correlation to other post-salt and pre-salt fields in proximal parts of Santos Basin – for an updated perspective of this unique opportunity.

In the southern Santos Basin, close to the Ametista Block, there are several discoveries and producing oil fields distributed as a string of pearls (Tubarão-Estrela do Mar-Coral-Caravela-Cavalo Marinho) (few tens of million boer each) (Figure 1) hosted in oncolite grainstones of Albian age. The producing petroleum system is the Cretaceous Marine Anoxic Shales-Albian carbonates (!). The reservoirs are Albian carbonates that are bountiful producers in Campos and Santos Basins; the source rocks are marine anoxic shales of Albian-Cenomanian-Turonian ages (ACT); and the traps consist of classic rollover structures developed upon mobile salt.

The same successful petroleum system is expected to be working in the Ametista Block. Inside the block, a prospect also termed Ametista, was mapped and an unusual hypothesis for its composition and formation is here presented.

Acquisition, processing, and model building

The new 3D Narrow Azimuth (NAZ) hydrophone data was acquired with 10 streamer setups, each 10,500 m long, 100 m apart and 13 m below sea surface. A dual source, 25 m flip-flop configuration with sources at 8 m below sea level was used. Outboard parts of Santos Basin often experience challenges such as rough weather, strong currents, and rapid barnacle growth on the cables throughout the year. The configuration used allowed for the best combination of data quality and most efficient data acquisition, which is both operationally and environmentally friendly.

The pre-processing workflow is modern broadband with rigorous QCs at every step to ensure the highest quality, AVO-compliant data. We use inversion-based de-blending, 3D de-ghosting and a robust set of de-multiple, de-noise and 4D regularising algorithms. Since several 3D surveys have been acquired in this area over the years, we also tried to match the data as close as possibly in phase and amplitude between different vintages.

Model building and imaging also reflect advanced workflows tailored for Brazil salt basins, where we are using a top-down approach with a combination of Tilted Transversely Isotropic Kirchhoff Pre-Stack Depth Migration (TTI KPSDM) and Reverse Time Migration (RTM) algorithms as well as reflection tomography and Dynamic Matching Full Waveform Inversion (DM FWI) (Cvetkovic et al. 2023).

We began with an initial 3D model created from a grid of underlying 2D data. By integrating all available regional data, the starting Vertical Transversely Isotropic (VTI) model was tied and calibrated with major regional horizons. The Vp model is updated with reflection-based tomography using non-parametric moveout picker and inversion for a global tomography solution with dip structural smoothing. We apply three to four passes of tomography to achieve overall flat gathers that allow for good structural imaging and salt interpretation.

Salt model building is performed in typical interpretation workflow where top of salt is interpreted, a constant salt interval velocity is ‘flooded’, followed by interpretation of salt flanks and regional base of salt. Overhangs and pre-salt section are updated by Common Offset RTM gather tomography and then later refined by DM FW’. DM FWI uses both refraction and reflection and with 10 km of offsets we have available here, we are getting good reliable updates up to the base of the Central High lithology.

Figure 5 shows models before and after DM FWI and respective KPSDM stacks. Here we see that the left part of the Central High (Closure A), with an opaque appearance, is updated with slower velocities, varying from 3600m/s up to 3900m/s. On the other side of the Central High (Closure B), a combination of speedup and slowdowns can be seen, and this is supported by the number of QCs we run in both the data and image domain. KPSDM gathers are flatter and the stacking response also improves. Other data domain QCs such as forward

Figure 5 3D KPSDM stack overlay with velocity model, before and after DM FWI (and b), respective KPSDM gathers (c and d). Arrows point to imaging improvements mentioned in text above.

Figure 6 Improvements in 3D imaging, leading to more certain interpretation. 2D vs 3D arbitrary line PSDM stacks overlay with velocity model over Ametista Central High. Arrows point to clear imaging improvement across the section.

Figure 7 3D KPSDM longitudinal section crossing Closures A and B in Ametista Central High. Noninterpreted, Interpreted, Velocity Overlay. Closure A displays a stratified seismic facies (sub-parallel layered reflectors) (platform); capped by a massive, structureless facies (reef). Arrows point to V-shaped collapse features, typical of karstic topography of carbonate terrains. Closure B presents the typical shape and facies of a cone volcano. Low velocity anomalies within the platform and reef facies may indicate porous lithologies, corroborating the presence of reservoirs.

model shots and correlation coefficients across different bands support the DM FWI models.

Figure 6 shows improvements in imaging compared to 2D data that most exploration geologists have been using historically. When comparing 2D and 3D data, we often see that incremental improvement in 3D pre-processing, model building and imaging leads to significant uplift in final data quality and interpretability. Besides much better signal to noise ratio, clearer and continuous seismic events are imaged across entire sections, and we can start interpreting different geologic facies and structural details that on 2D data cannot be distinguished. Deep crustal faulting and rift structures are better imaged as well. This new 3D data is now imaged at the same high-quality standard as most of the 3D datasets covering inboard parts of the basin.

New interpretation

With new and improved imaging, we can strongly infer that the basement of the prominent high was a protrusion of exhumed

Figure 8 3D KPSDM transversal section crossing Closure A in Ametista Central High. Non-interpreted, Interpreted, Velocity Overlay. Closure A displays a stratified seismic facies (sub-parallel layered reflectors) (platform); capped by a massive, structureless facies (reef) of several hundred meters thick. Striking low velocity anomaly within the reef facies may indicate porous lithologies, corroborating the presence of reservoirs.

mantle (Figure 3, 4, 7 and 8). Second, the sub-cropping mantle arch was completely devoid of salt cover. Salt was undoubtedly never deposited on top of it, although its flanks and surroundings are rich in salt domes and diapirs (confirmed during the model-building process). Not only is salt absent, but also any graben typical of the Pre-Salt section (Figure 4, 7 and 8). The exhumation of the mantle seems to have occurred during the rift phase and continued into the thermal subsidence phase during the salt deposition. This backbone high (Central High) is divided into two structural closures (A and B, Figures 4-7). In Closure B there are sharp conical constructions that are characteristic of volcanoes (Figure 4-7). In Closure A (Figure 4-8), the lithological cover of the exhumed mantle is startling.

A thick package (a few hundred metres) of strong, stratified, sub-parallel layered reflectors cover the exhumed mantle completely, in a large four-way closure. They strongly resemble carbonate beds, displaying clinoforms on the edges of the closure and several V-shaped collapse features (Figure 7). Massive, irregular structure-

less bodies occur at localised areas at the base and, especially, at the top of these stratified facies (Figure 4, 7 and 8). The deduced Albian age of these strata is favourable to the carbonate hypothesis, because of the well-known Albian carbonate platforms that are common throughout the eastern Brazilian offshore basins. Thermal subsidence that ensued after the cessation of rifting allowed the incursion of sea waters and provided increasing accommodation space in these basins. The encroaching seas are reflected under the form of a typical transgressive cycle, deposited during Early Albian to Coniacian. Albian carbonates developed during the first marine invasions into the Brazilian marginal basins. They started as shallow water oncolite grainstones, changing gradually into peloidal packstones/boundstones and carbonate mudstones, as the seas deepened.

Given the interpreted carbonate depositional environment and its Albian age, this upper massive structureless construction could tentatively be interpreted as a reef. A well-known reef-builder organism that thrived during the Albian were rudists. Rudists were bivalve mollusks, benthic marine organisms, constructors of very extensive and large reefs, that lived in tropical shallow seas during the Late Jurassic and Cretaceous (Neo-Tethys and Paleo-Pacific Oceans realms). They climaxed during the Albian and are famous for the numerous and voluminous reefs hosting large reserves of oil and gas in the Middle East and the Gulf of Mexico regions. Some reefs ran for hundreds of kilometres along the edge of carbonate platforms. They formed buildups that were hundreds of metres tall (Johnson, 2002). Their porosities and permeabilities are phenomenal.

In the past, these organisms were distributed along warm low paleo-latitudes, somewhat concentrated in boreal domains. Their presently known geographic distribution is somewhat confined to 400 North and 300 South latitudes. The southernmost reported occurrence of rudists is in Madagascar (300 South, Sha et al., 2020). The Ametista Prospect is situated at 270 South; thus, although unknown in the South Atlantic, the possibility of their occurrence in the Ametista Prospect is viable and cannot be eliminated. According to Johnson (2002) rudists larvae were dispersible and were carried along with plankton in surface currents, allowing them a cosmopolitan distribution.

The entire structural high is completely encased in thousands of metres of argillaceous Upper Cretaceous seismic facies providing an efficient seal to this unusual prospect. At the base of this Upper Cretaceous package, continuous, parallel and

moderately strong reflectors denounce the presence of the ACT source rocks completely surrounding the entire structural high. As a result, the Ametista prospect presents all the items of the successful Cretaceous Marine Anoxic Shales-Albian carbonates (!) petroleum system.

Modern analogs

Figure 9 displays the 3D view of the Central High structure (with Closures A and B) compared to a modern atoll in the Pacific Ocean (with elevations A and B). The coexistence of volcanoes in Closure B of the Central High and carbonate packages in Closure A reminds us of the famous atolls of the Pacific Ocean. In particular, the Raiatea Atoll in Frech Polynesia is very similar in shape, size and composition to the Ametista atoll. While in Closure B and elevation B volcanism was active and carbonate deposition scant, Closure A and elevation A received shallow water organisms that created a wide carbonate platform and, later, a possible rudist reef. Just for the sake of correctness, elevations A in Figure 9 are also composed of volcanoes, different from what is suggested by us to Closure A (elevations of exhumed mantle without volcanism).

The prominent protrusion of exhumed mantle in the middle of a nascent sea functioning as a shoal to receive carbonate deposition also has a modern analog, in the island of St. John`s (or Zabargad) in the Red Sea (Figure 10). The island is composed of exhumed mantle rocks (Bonatti et al. 1981 and 1983, Abu El Rus 2007), surfacing in the middle of the young Red Sea, in a purely extensional environment. These mantle rocks are rimmed by carbonate platforms and reefs. This is exactly the environment envisaged for the Ametista Prospect (Closure A) in the Early Albian transgressive sea in the young Santos Basin. The protrusion of exhumed mantle must have functioned just as a guyot in the present Pacific Ocean; as a shallow water paradise for carbonate deposition.

The petroleum potential of the Ametista Prospect (Closure A) is very high given its size, the notable four-way closure, the presence of carbonates as reservoirs, the presence of a thick package of the ACT organic-rich shales surrounding and onlapping the structure and the very thick seal of Late Cretaceous argillaceous strata enveloping the entire four-way closure.

An alternative hypothesis

We also have an alternative hypothesis to the Albian atoll interpretation, still based on another carbonate lithology. These could be pre-salt carbonates, microbialites, travertines and

Figure 9 Left — 3D visualisation of Ametista Central High and of Closures A and B. Right — The analogy with the present-day Atoll Raiatea (Leeward Islands, French Polynesia) with elevations A and B (image from Google Earth Pro) is remarkable. The large carbonate platform surrounding elevation A in the atoll would correspond to our layered seismic facies in Closure A (Ametista Prospect).

coquinas. Since there are no visible sedimentary grabens on top of the exhumed mantle and salt was never deposited on it, a syn-tectonic microbial platform could have developed during the continued exhumation of mantle during the late rifting phase/starting of thermal subsidence phase of basin development. The massive structureless facies could then be interpreted as travertines or coquina pile ups, like those found at the giant Mero and Búzios Pre-Salt fields. Different from a

Figure 10 Google Earth Pro images of St. John`s (Zabargard) Island in the Red Sea. The island is composed of exhumed mantle rocks (EM, outcrops indications based on Abu Le Rus, 2007). Carbonate platforms surround the mantle rocks and are rimmed by reef constructions/buildups, some of them displaying large proportions (arrows). This modern analog to Closure A demonstrates the plausibility of our hypothesis.

classic Pre-Salt prospect, the hydrocarbon charge would be from the ACT source rocks instead of from Barremian/Early Aptian lacustrine organic-rich shales.

Conclusions

The newly acquired and processed 3D NAZ dataset over Ametista block provides a step change in imaging of this part of Santos Basin. Whether the Ametista Prospect is a Post-Salt Albian atoll

capped by a rudist reef (preferred interpretation) or a Late Aptian Pre-Salt microbialite platform, this promising prospect and associated petroleum system deserves a more detailed subsurface investigation and a drilling campaign to unlock the southern Santos Basin frontier.

Acknowledgements

We would like to thank TGS Management for giving us permission to publish this material. We would also like to thank TGS Imaging and Operations teams for all the work they did to make this a successful project. This article is an extension of an oral presentation at Fourth EAGE Conference on Pre-Salt Reservoir, September 2024, in Rio de Janeiro.

References

Abu El-Rus, M.M.A. [2007]. Plagioclase lherzolites from Zabargad Island, Red Sea, and their bearing on the evolution of lithospheric mantle beneath an embryonic ocean. Proceedings of the Fifth International Conference on the Geology of Africa, V-181–V-213.

Bonatti, E., Clocchiatti, R., Colantoni, P., Gelmini, R., Marinell, G., Ottonello, G., Santacroce, R., Taviani, M., Abdel-Meguid, A.A., Assaf, H.S. and El Tahir, M.A. [1983]. Zabargad (St. John’s) Island:

an uplifted fragment of sub-sea Red Sea lithosphere. Journal of the Geological Society, 140(4), 677-690. https://doi.org/10.1144/ gsjgs.140.40677.

Bonatti, E., Hamlyn, P. and Ottonello, G. [1981]. Upper mantle beneath a young oceanic rift: Peridotites from the island of Zabargad (Red Sea). Geology, 9(10), 474-479. https://doi.org/10.1130/0091-7613.

Cvetkovic, M., Yong, S.L., Johnson, T., Armentrout, D., Brookes, D., Pralica, N., Ge, L, Soelistijo, B. and Baldock, S. [2023]. The value of uplift processing in frontier exploration basins – Offshore Brazil case studies. 84th EAGE Annual Conference and Exhibition, Extended Abstracts.

Deckelman, J.A., Chi, S., Hartwig, A.K., Reuber, K.R. and Fruehn, J. [2021]. The role of megaregional seismic data in Santos Basin pre-salt exploration and development. AAPG Memoir, 124, 487–516.

Johnson, C.C. [2002]. The rise and fall of Rudist reefs. American Scientist, 90(2), 148-153. DOI 10.1511/2002.2.148.

Sha, J., Cestari, R. and Fabbi, F. [2020]. Paleobiogeographic distribution of rudist bivalves (Hippuritida) in the Oxfordian-early Aptian (Late Jurassic-Early Cretaceous). Cretaceous Research, 108, 104289. https://doi.org/10.1016/j.cretres.2019.104289

Zalán, P.V. [2024]. Remaining petroleum potential of Brazil – State of the art 2023. Derbyana, 45. htpps://doi.org/10.14295/derb.v45.826

ADVERTISEMENT

Innovative Technology for Reservoir Optimization

FIFTH EAGE WELL INJECTIVITY & PRODUCTIVITY IN CARBONATES (WIPIC) WORKSHOP

13-16 APRIL 2025 • DOHA, QATAR

Elevate your experience by hearing from our distinguished keynote speakers, renowned experts at the forefront of their industries! Register now for this exclusive workshop and gain invaluable knowledge, inspiration, and networking opportunities!

Early bird deadline: 10 March 2025

Estimating petrophysical parameters from acoustic impedance and P to S wave velocity ratio using a simple rock physics model

Krishna Agra Pranatikta1 and Ignatius Sonny Winardhi1* offer a simple method to estimate porosity and fluid saturation directly from AI and VP/VS ratio typically obtained from the prestack seismic inversion.

Abstract

In reservoir characterisation, rock physics analysis has been frequently used to describe petrophysical properties of rocks based on their elastic behaviour. Recently, a new rock physics model that facilitates direct estimation of porosity and fluid saturation from acoustic impedance (AI) and P to S wave velocity ratio (VP/ VS) was introduced. The method has the flexibility to adapt the rock physics model to the data by simply adjusting two parameters without the need to consider the detailed elastic moduli. In this study, a slight modification is proposed by considering the influence of fluids on VP/VS ratio to better comply with the Gassmann equation. A workflow for obtaining optimal model parameters for the observed data by incorporating the curved pseudo elastic impedance (CPEI) approach is also demonstrated. The modified method successfully produces a model that fits the data. The proposed workflow effectively finds optimal model parameters, resulting in better estimation of petrophysical properties. However, a detailed examination of the results shows that variations in shale volume should be incorporated to obtain better petrophysical estimation results. This method offers a simple yet reliable way to estimate porosity and fluid saturation directly from AI and VP/VS ratio typically obtained from the pre-stack seismic inversion.

Introduction

In reservoir characterisation, petrophysical properties are typically described using rock elastic properties from wireline logs and seismic data. Two main reservoir petrophysical properties are porosity and hydrocarbon saturation. Rock physics analysis is commonly used to describe the target petrophysical properties based on the behaviour of the elastic properties. This analysis is performed using the rock physics template (RPT), introduced by Avseth and Odegaard (2004), as a guide to understanding the properties of reservoir rocks. However, the construction of the rock physics template requires detailed information on mineralogy, pore-filling fluids and insitu reservoir conditions (Johansen et al. 2013).

Recently, Fawad and Mondol (2022) simplified RPT generation by using only rock elastic properties, such as acoustic impedance (AI) and P to S wave velocity ratio (VP/VS ratio), to predict

1 Bandung Institute of Technology

* Corresponding author, E-mail: swinardhi@itb.ac.id

DOI: 10.3997/1365-2397.fb2025023

hydrocarbon saturation without using detailed information about the rock elastic moduli as given in Equation (1). In addition, they also developed a flexible RPT by adding a G factor that controls the vertical shift of the RPT and N that controls the inclination of the RPT with respect to the data.

where, are the P-wave velocities of the mineral matrix, brine, and apparent P-wave velocity of target fluid, are the density of mineral grains, brine, and apparent density of target fluid, AI is acoustic impedance, ϕ is porosity, and water saturation (SW) can be calculated using SW = 1–SFL

Adopted from Lee et al. (1996), the P to S wave velocity ratio (VP/VS) in this model is assumed to be inversely proportional to porosity only as given in Equation (2), where G is the mineralogy/ shaliness, N is stress/ cementation coefficient, and α is the VS/ VP ratio of the rock/mineral matrix. To better comply with the Gassmann equation, a slight modification is proposed by considering the influence of fluids on VP/VS ratio. A workflow for obtaining optimal model parameters, particularly the G and N factors, for the observed data is also suggested to apply the model efficiently.

Methods

Data and workflow

This study focuses on oil-bearing clastic reservoirs of the Early Oligocene Gabus Formation. The Gabus Formation is one of the potential reservoirs that consist of fluvial-deltaic deposits formed during the early Oligocene post-rift phase, with the depositional environment dominated by lacustrine and fluvial systems (Surjono et al. 2023). As part of a Paleogene system rich in good quality reservoir rocks, the Gabus Formation is a major target for hydrocarbon exploration in the West Natuna Basin. This study uses two main types of data, synthetic well

The modification is made by combining the Lee et al. (1996) velocity relation and Gassmann’s rule (1951). Gassmann (1951) assumes that a rock with the same matrix and porosity has a shear modulus value that is not affected by pore fluids.

Table 1 Matrix and fluid component elastic parameter information.

data and field well data. The synthetic well data is generated by mimicking the field well data pattern by combining Wyllie’s time average, bulk density and modified Lee velocity ratio equations. The matrix and fluid variables used to generate the synthetic data are listed in Table 1. The G and N factors, 1.10 and 3.00, are used as initial model parameters and as reference model parameters obtained from the proposed workflow. Field well data are restricted to depths associated with the Early Oligocene Gabus Formation clastic reservoir. This study is therefore divided into three stages, namely (i) modification of the Lee’s velocity ratio using the Gassmann equation, (ii) determination of the critical porosity value using the Nur et al. (1998) model, and (iii) implementation of the modified Fawad and Mondol rock physics model workflow.

Modifying Lee’s velocity ratio using Gassmann’s rule

Lee et al. (1996) stated a simple relationship between P-wave and S-wave which depends on porosity and can be adjusted by a scaling factor, as follows.

(2)

where, the scaling factor, G factor governs the vertical shift of the RPT, N factor governs the inclination of the RPT curve, and α is the VS/VP ratio of the rock/mineral matrix that defines the matrix-mineral pole in the AI vs VP/VS plot. This equation provides convenience and flexibility in predicting velocity that fits the data. Despite its convenience and flexibility, this equation does not comply with the rock physics rules of Gassmann (1951) as the VP/VS ratio is only sensitive to porosity.

Gassmann’s (1951) rule states that P-wave and S-wave are sensitive to porosity and fluid saturation (cf. Equations 3 and 4) so that Lee et al. (1996) velocity ratio requires a modification of the equation to include the porosity and fluid saturation components.

The resulting equation after reformulating Eqs. (5) and (6) is as follows.

The variable ρsat is the density log data (RHOB). The variable VS dry is expressed using the empirical equation of Phani and Niyogi (1986), which is stated as follows.

The empirical equation of Phani and Niyogi (1986) determines the dry rock S-wave as a function of porosity, S-wave matrix and N as a fitting parameter (Zhao and Li, 2021). Furthermore, the relationship between the equation of Phani and Niyogi (1986) and the definition of α from the equation of Lee et al. (1996) is synchronised as follows

The variable ρdry is expressed as ρdry = (1 – ϕ) ρma. By defining the variables VS dry and ρdry, the modified Lee equation is obtained as follows.

The modified form of Lee’s equation gives the relationship of the variable ϕ to predict hydrocarbon saturation, as follows.

Figure 1 The study is conducted in three main steps, as depicted in the flowchart.

Determine the critical porosity using Nur’s Critical Porosity model

Critical porosity is a rock physics concept that separates two distinct rock domains, namely consolidated frames and suspensions (Nur et al. 1998). To supplement the critical porosity information in this RPT, the Nur Critical Porosity model is used to determine the critical porosity value of the reservoir rock. This model is formed using Equation (12) to generate a distribution of critical porosity values, which is represented as a diagonal line in the cross plot of dry rock bulk modulus (Kdry) versus porosity (Ambarsari et al. 2018 and Ambarsari et al. 2020).

where, Kdry, K m are dry rock bulk modulus, matrix bulk modulus, dan ϕ, ϕc are porosity dan critical porosity. Kdry is determined using reformulated of Gassmann equation (Zhu and McMechan, 1990), as follows.

Create a workflow to implement modified Fawad and Mondol’s RPT

Fawad and Mondol (2022) do not provide a workflow for implementing their proposed approach. This study proposes a workflow that uses the integration of different methods to derive the G and N factors that result in model responses that match the data (see Figure 2). Given that the predicted hydrocarbon saturation is the desired product, the appropriate G and N factors are determined by forward modelling, which involves finding the variation of G and N factors that produces a model response that matches the true data (hydrocarbon saturation data from petrophysical analysis) by trial-and-error process, where the model response of G and N factors that are close to the true data is indicated by the least misfit value.

Regularisation of hydrocarbon saturation data

where, K sat, K m data, KFL data are saturated bulk modulus, matrix bulk modulus, fluid bulk modulus, dan ϕ data is effective porosity (PHIE). The limited information on matrix composition and reservoir fluid is addressed by assuming that the reservoir rock matrix is composed of sandstone and shale, while the pore fluid is composed of formation water and oil. The bulk modulus of each component is calculated from . The Voigt – Reuss – Hill (VRH) approach is used to calculate the bulk modulus of the combined matrix, while the Woods approach is used to calculate the bulk modulus of the combined fluid.

Finding the appropriate model response requires reference data, namely hydrocarbon saturation data from petrophysical analysis results. However, the scattered distribution pattern of hydrocarbon saturation data is considered inappropriate and complicates the process of finding the desired model response, and a more regular reference data is required.

Curved Pseudo Elastic Impedance (CPEI) is an approach that uses a combination of acoustic impedance (AI) and VP/VS ratio based on rock physics models to predict water saturation (Winardhi et al. 2023 and Pranatikta et al. 2024). This attribute rotation approach has been shown to mimic and produce data patterns close to water saturation, with the resulting data patterns being more regular compared to field data (Winardhi et al. 2023). This shows that the CPEI approach can be used to generate regularised data as a reference for finding the desired model response. CPEI attributes can be generated using Equation (16) as follows.

(12)

(13)

(14)

(15)

Figure 2 The proposed workflow for determining the optimal model parameters, G and N, in the modified Fawad and Mondol RPT consists of five main steps, as illustrated in the flowchart.

Figure 3 A schematic representation of the process for determining optimal m , n, and χ parameters in scheme of 3D CPEI.

Table 2 The parameters n, m, and χ within the CPEI attribute that exhibit the strongest correlation with water saturation.

(16)

where AI is acoustic impedance (km/s * gr/cc), m, n are fitting parameter, dan χ is rotation angle (o). The parameters m, n and χ are determined by obtaining the best correlation between the CPEI pattern variation and the associated petrophysical data, i.e. CPEI to water saturation data (see Figure 3). The best m, n and χ parameters are listed in Table 2, where the best CPEI attribute parameters successfully mimic the water saturation data pattern as indicated by the best correlation value and produce a pattern close to the water saturation data (see Figure 4).

Variation of model from the modified Fawad and Mondol

To find the model response that best matches the reference data, the forward modelling technique is carried out in a trial-and-error process by changing the value of the model parameters until a match is obtained between the theoretical data (model response) and the field data (Grandis, 2009). The model response in terms of predicted hydrocarbon saturation can be generated using the modified equation as follows.

where, matrix (VP ma, VS ma, ρma) and fluid (VP w, ρw, VP FL, ρFL) are listed in the Table 1. The G and N factors have a range of 0.5 – 1.5 (G) and 0.00 - 4.00 (N) to produce a varied model response. The variation of the G and N factors is then tested for compatibility with the reference data until the G and N factors that produce the model response closest to the reference data are obtained.

Histogram matching between model responses to reference data

In reality, there are significant differences in the histogram distribution characteristics of the reference data versus the model response, especially related to the different value ranges between the model response and the reference, which makes it difficult to determine the G and N factors that produce an appropriate model response to the reference data. This can be addressed by performing histogram matching or matching techniques of the model response histogram against the reference data histogram. This technique is applied by matching the CDF (Cumulative Distribution Function) value between the model response and the reference data.

Figure 4 As shown in the top panel, the CPEI attribute effectively regularises the data, leading to a more consistent fluid saturation distribution compared to the Keris - 1 well data in the AI vs V P/VS domain. The bottom panel further validates the CPEI attribute by demonstrating a strong correlation between its parameters (m, n, and χ) and the distribution of water saturation data.

The CDF (Cumulative Distribution Function) is a function that represents the cumulative probability or shows the accumulated probability from the lowest value to the highest value (x-value) of the value under consideration. Generally, the CDF has the lowest value of 0 and the highest value of 1. The concept of cumulative probability provides an index of the position of a number in the model response data distribution against the reference data distribution. This information facilitates the process of matching the value of the model response data to the reference data, so that it can be known how much value change is required for the model response data to match the reference data in order to obtain the matching statistical properties between the two. The histogram matching technique alters the value range of the model response that is close to the reference data to facilitate the process of determining the model factors that produce the best model response.

Finding and determining the best G and N factors

Finding and determining the best G and N factors is done by determining the misfit value between the model response varia-

Figure 5 The difference in data histogram distribution characteristics between the reference data (left) and the model response (right) makes it difficult to find the best model parameters, hence the need for histogram matching techniques.

Figure 6 The application of histogram matching succeeded in modifying the histogram distribution of the model response close to the histogram distribution of the reference data. The CDF technique is analogous to giving the relative position of the predicted data value element to the position of the reference data value element in the range 0-1.

tion and the reference data. Misfit is a measure of the response fit between the model response and the reference data, which is determined by satisfying the following equation.

Where the squares in the equation above are expressed without distinguishing between positive ( ) or negative ( ) differences (Grandis, 2009).

The best model response is associated with the smallest misfit value or the minimum error rate between the model response and the reference data. The process of finding the smallest misfit value is done by trial and error to produce a grid scheme as shown in Figure 7, where the smallest misfit value is associated with the model factors, G and N, that produce the best model response. The search for the best model parameters is carried out not only by determining the best model factor based on the smallest misfit value, but also by using the fit of the resulting RPT model to the field data tested.

Estimation of porosity and fluid saturation based on the best model parameters

Once the best G and N factors are obtained, they are combined with acoustic impedance (AI) and VP/VS ratio to estimate porosity and hydrocarbon saturation. Density has a linear relationship to porosity and fluid saturation so porosity can be determined using the following equation.

or

Results

Comparative analysis of Fawad and Mondol RPT models before and after modification

The original Fawad and Mondol RPT model uses the Lee et al. (1996) equation to relate the VP/VS ratio to AI as the basis for RPT generation. As explained earlier, the Lee et al. (1996) velocity ratio is expressed as a function of porosity only, in contrast to the velocity ratio in the Gassmann (1951) rule, which is expressed as a function of porosity and fluid saturation. The modified form of the Lee et al. (1996) equation to the Gassmann (1951) rule is shown in Equation 10, which results in a modified Fawad and

Figure 7 The proposed workflow produces G and N factors with the smallest misfit values for (a) synthetic data (G: 1.10, N: 3.00, and misfit: 0.030), (b) Keris – 1 well data (G: 0.90, N: 2.70, and misfit: 0.060), (c) Keris – 2 well data (G: 0.90, N: 2.70, and misfit: 0.049).

Mondol RPT model that differs from the original Fawad and Mondol RPT models.

The insensitivity of the original Fawad and Mondol RPT models to fluid saturation is shown by the same VP/VS ratio value for any variation in fluid saturation and sensitivity only to porosity (Figure 8), whereas the modified Fawad and Mondol RPT models have sensitivity to porosity and fluid saturation, resulting in VP/VS ratio that vary with changes in porosity and fluid saturation (Figure 9). When compared to the field data, it appears that the original Fawad and Mondol RPT models do not provide a curvature pattern that is close to the field data pattern, whereas the modified Fawad and Mondol RPT models provide a curvature pattern that is closer to the field data. Based on this, it can be understood that the modified Fawad and Mondol RPT models have successfully filled in the missing rock physics rules, which naturally leads to better prediction results compared to the original Fawad and Mondol RPT models.

Determination of critical porosity using the Nur Model

To complete the RPT, critical porosity plays an important role in describing the endpoint of an RPT. In addition to eliminating the suspension zone and focusing only on the real rock, the endpoint information can provide an indication of the quality of the reservoir under investigation. The results of critical porosity determination using the Nur model on field data are shown in Table 3 and Figure 10. The different critical porosity values are a

result of the different depths of the reservoirs, with the reservoir at Keris – 2 being deeper than Keris – 1. The deeper the reservoir, the greater the rock mass overlapping the reservoir, so the effect of overburden stress is larger. The effect of the overburden stress triggers a compaction process that changes the intrinsic structure of the rock, where the grains of the rock are brought closer together and the pore space in the rock is reduced. The reduced pore space indicates the volume of fluid that can be stored in the rock, so deep reservoirs tend to have poorer reservoir quality than shallow reservoirs.

Figure 8 The original Fawad and Mondol RPT model is insensitive to fluid saturation, exhibiting a constant V P/VS ratio regardless of saturation changes. This contrasts with observed data, which displays a different curvature pattern than the RPT model.

Figure 9 In contrast to the original formulation, the modified Fawad-Mondol RPT model exhibits sensitivity to fluid saturation, reflected in the variation of the V P/ VS ratio as a function of saturation. The resulting RPT curvature patterns show improved agreement with observed data, suggesting a better representation of the underlying rock physics.

Test results of the proposed workflow

The proposed workflow is tested using synthetic well data and field well data. The success parameter of testing the proposed workflow using synthetic well data is determined based on the difference in comparison between the proposed workflow model parameters and the original model parameters. The application of the proposed workflow to synthetic well data resulted in the best model parameters, namely G of 1.10 and N of 3.00 with a misfit value of 0.030. This result shows that the proposed workflow produces the best model parameters that are close to the initial

model parameters. These best model parameters produce perfect correlation values between predicted attributes and petrophysical parameters as shown in Table 4.

Application of the proposed workflow to the field well data resulted in the best model parameters for both wells, namely G of 0.90 and N of 2.70 with a misfit of 0.060 for Keris – 1 and G of

Table 4 Parameter compilation of the best model for each data and fit test against reference data.

0.90 and N of 2.70 with a misfit of 0.049 for Keris – 2. These best model parameters resulted in a good fit of the modified Fawad and Mondol RPT models to the data (Figure 11) and a perfect correlation between the predicted attributes and petrophysical parameters as listed in Table 4 and Figure 12. In addition, the predetermined critical porosity is included as an endpoint in the modified Fawad and Mondol RPT models so that the modified Fawad and Mondol RPT models have complete information in line with the conventional RPT models.

Conclusion

A simple rock physics model, based on the approach of Fawad and Mondol (2022), which facilitates direct estimation of porosity and fluid saturation from acoustic impedance (AI) and the ratio of P and S wave velocities, has been modified and implemented. This method has the flexibility to fit the rock physics model to the data by adjusting only two parameters without the need to consider the elastic moduli in detail. To obtain optimal model parameters against the observed data, a curved pseudo-elastic impedance (CPEI) approach has been incorporated into the workflow to better regularise the relatively noisy field data. The method has been successfully tested on both synthetic and field data and resulted in a model that fits the data well. The proposed workflow effectively finds the optimal model parameters, result-

Figure 10 The critical porosity determination using Nur’s model for each reservoir is 0.33 for Keris – 1 (left) and 0.3 for Keris – 2 (right). The difference in critical porosity is thought to be due to the relationship between reservoir depth and the effect of overburden stress, which gives Keris – 1 a greater critical porosity than Keris – 2.

Figure 11 The best model parameters, G and N, produced modified Fawad and Mondol RPT models for synthetic well data (A) and field well data, Keris – 1 (B) and Keris – 2 (C), as shown in the figures above. The resulting RPT model provides a curvilinear pattern that fits the data, indicating that the modified RPT model and proposed workflow have good implications for RPT modelling. In addition, the previously determined critical porosity is used as an endpoint in the RPT model in panels B and C.

Table 3 The results of determining the critical porosity of each field well data using Nur’s model.

ing in better estimates for petrophysical properties. A detailed examination of the results shows that shale volume variation should be incorporated to obtain better petrophysical estimation results. This requires further study. This method offers a simple yet reliable way to estimate porosity and fluid saturation directly from AI and VP/VS ratios typically obtained from pre-stack seismic inversion.

References

Ambarsari, D. and Winardhi, S. [2018]. Rock Physics Modelling for Estimating of Reservoir Tight Sand Quality in Bintuni Basin, West Papua, Indonesia. Jurnal Geofisika, 16(3), 14-18

Ambarsari, D., Winardhi, S., Prakoso, S. and Sukmono, S. [2020].

Reservoir quality determination by using critical porosity and volume of clay: a case study in the Talang Akar Formation in the NW Java Basin, Indonesia. First Break, 36(5), 63-70.

Avseth, P. and Odegaard, E. [2004]. Well log and seismic data analysis using rock physics templates. First Break, 23(10), 37-43.

Fawad, M. and Mondol, N.H. [2022]. Monitoring geological storage of CO2 using a new rock physics model. Scientific Reports, 12, Article 297.

Figure 12 The results of the RPT model modification and proposed workflow provided the best model parameters, G and N, which resulted in the predicted attributes, porosity and hydrocarbon saturation, correlating well with the associated petrophysical parameters.

Gassmann, F. [1951]. Über die elastizität poröser medien (Kathy, Christine, John, dan Jim, Translated in English). Vierteljahrss-chrift der Naturforschenden Gesellschaft in Zurich, 96, 1-21.

Grandis, H. [2009]. Pengantar Pemodelan Inversi Geofisika. Himpunan Ahli Geofisika Indonesia (HAGI), 1-186.

Johansen, T.A., Jensen, E.H., Mavko, G. and Dvorkin, J. [2013]. Inverse rock physics modelling for reservoir quality prediction. Geophysics, 78(2), M1-M18.

Lee, M.W., Hutchinson, D.R., Collett, T.S. and Dillon, W.P. [1996]. Seismic velocities for hydrate – bearing sediments using weighted equation. Journal of Geophysical Research, 101(B9), 347-358.

Nur, A., Mavko, G., Dvorkin, J. and Galmudi, D. [1998]. Critical porosity: A key to relating physical properties to porosity in rocks. The Leading Edge, 357-362.

Phani, K.K. and Niyogi, S.K. [1986]. Porosity dependence of ultrasonic velocity and elastic modulus in sintered uranium dioxide – a discussion. Journal of Materials Science Letter, 5, 427-430.

Pranatikta, K.A., Winardhi, S. and Prastama, R.A. [2024]. Direct estimation of petrophysical properties from elastic parameter: A Case

Study on the Batu Raja Formation. IOP Conference Series: Earth and Environmental Science, 1437, 1-6.

Surjono, S.S., Afandi, M., Arifianto, I., Mitasari, A.A. and Mahendra, F.H.M. [2023]. Sedimentology and reservoir characteristics of synrift to syn-inversion succession in Anoa half-graben, West Natuna Basin, Indonesia. Marine and Petroleum Geology, 152, Article 106258.

Winardhi, S., Dinanto, E., Sigalingging, A.S., Adityo, R., Utami, A.P., Ritonga, A.B., Satriawan, W. and Budiyono. [2023]. Curved Pseudo

Elastic Impedance based on Elastic Attribute Rotation Scheme for Reservoir Petrophysical Property Prediction. IOP Conference Series: Earth and Environmental Science, 1288, 1-6.

Zhao, R. and Li, C. [2021]. Porosity-dependent velocities of longitudinal and transverse waves in dry porous materials. Applied Acoustics, 176, 1-15.

Zhu, X. and McMechan, G.A. [1990]. Direct Estimation of the Bulk Modulus of the Frame in a Fluid-Saturated Elastic Medium by Biot Theory. SEG Technical Program, Expended Abstracts, 787-790

Implementation of time-lapse gravity and subsidence monitoring for optimising the development of the Scarborough gas field

Mohamad Yousof Hourani1*, Abid Ghous1, Rabin Sridaran1, Martha Lien2, Siri Vassvåg2 and Hugo Ruiz 2 summarise the results of the modelling work conducted to assess the feasibility of time-lapse gravity and subsidence modelling at Scarborough gas field.

Introduction

The Scarborough gas field is located in the Carnarvon Basin, approximately 375 km off the Pilbara coast of Western Australia. It lies primarily within the WA-61-L and WA-62-L production licences in water depths between 900 and 1000 m (Figure 1). The field, discovered in 1979 by the Scarborough-1 well, is a low-relief anticlinal structure at a depth of approximately 950 m below the mudline, with a lateral extent of about 40 × 20 km. The reservoir, located within the Lower Barrow Group formation, contains high-quality dry gas and consists of poorly consolidated high-quality reservoir sands exhibiting approximately 30% porosity and multi-Darcy permeability. The gas is sweet, with no hydrogen sulphide and only trace carbon dioxide. The reservoir consists of a three-tiered basin floor fan deep-water turbidite sequence (upper, midle and lower fans) with a gross gas column thickness of 110 m and an estimated 11.5 trillion cubic feet (2P) of natural gas reserves.

The fields’s development plan includes 13 wells across two phases, with Phase 1 consisting of eight high-rate subsea gas wells. The wells will be connected to a new semi-submersible floating production unit (FPU) via three carbon steel gathering flowlines. The normally unattended, moored FPU will provide gas separation, dehydration and export compression to shore via a 430 km trunkline.

The phased development will enable a dynamic assessment of the primary subsurface uncertainties such as the lateral extent of reservoir sands and the degree of vertical and lateral communication between these sands. Insights gained from Phase 1 will inform the optimisation of subsequent development phases, ensuring efficient hydrocarbon recovery. To maximise these insights, a comprehensive reservoir surveillance program is planned, incorporating both near-well and field-wide monitoring of pressure and saturation.

Time-lapse (4D) seismic surveys are typically acquired during field life to monitor changes in water saturation and pressure depletion caused by gas production. However, the explicit assessment of water influx using 4D seismic alone is challenging in Scarborough due to the non-linear sensitivity to fluid saturation

1 Woodside Energy | 2 Reach Subsea

* Corresponding author, E-mail: yousof.hourani@woodside.com DOI: 10.3997/1365-2397.fb2025024

(Landrø, 2001; Müller and Gurevich, 2004) coupled with the Scarborough’s reservoir properties and characterisitics. Similarly, whilst a robust pressure depletion signal is expected in Scarborough from 4D seismic, the quality of the signal can be impacted by rock deformation, and accurately quantifying the magnitude of the deformation remains challenging without additional calibration data (Hatchell and Bourne, 2005).

To address these limitations, an additional monitoring technology will be implemented at Scarborough: time-lapse gravity and subsidence surveys. This technology has been successfully applied in Norway for over 25 years but has yet to be utilised in the Asia-Pacific region. It involves periodic surveys, typically conducted every two years, to measure changes in the gravitational field and seafloor depth at a grid of precisely defined locations on the seabed. These locations are marked by concrete pads (CPs)

Figure 1 Location and rim of the Scarborough field.

deployed on the seafloor. The technology achieves measurement accuracies of approximately 1 μGal for gravity changes and 2 mm for seafloor subsidence (Ruiz et al., 2022).

Monitoring gravitational field variations over time can map hydrocarbon depletion and provide precise estimates of water volumes flowing in from surrounding aquifers, as these flows significantly impact the subsurface mass distribution. Seafloor subsidence is a direct result of reservoir compaction and can be linked to this important dynamic property of the reservoir through geomechanical modelling. The lateral distribution of subsidence offers crucial insights into pressure depletion across the field.

This paper summarises the results of the modelling work conducted to assess the feasibility of this technology at Scarborough. Notably, this will be the first application of the technology to a relatively shallow reservoir at a depth of only 950 m below mudline. Modelling indicates that this depth will allow for unprecedented lateral resolution from both time-lapse gravity and subsidence data.

The time-lapse and seafloor deformation monitoring method

Hydrocarbon production alters the distribution of fluid masses within reservoirs, leading to changes in the gravitational field at the surface. As gas or oil is depleted, the mass decreases, while influx from surrounding aquifers can cause localised mass increases. Similarly, pressure drops within the reservoir reduce pore volume, resulting in reservoir compaction and surface subsidence.

Conversely, monitoring changes in the gravitational field and seafloor subsidence over time provides valuable insights into mass and pressure variations in the reservoir. This is the principle behind time-lapse gravity and subsidence surveys, which have been commercially applied across major gas fields in Norway (Alsos et al., 2024; Ruiz et al., 2022; Solbu et al., 2023; Siri Vassvåg et al, 2025) and are being proposed as key monitoring tools for CCS projects (Borges et al., 2024).

The method requires smaller vessels than those used for 4D seismic acquisitions, potentially resulting in significantly lower costs and reduced environmental impact. Additionally, it relies on passive measurements, meaning it does not involve active sources that may encroach upon marine wildlife environments. Beginning in 2025, remotely operated vessels will be available for this application (Siri Vassvåg et al, 2025), virtually eliminating health, safety, and environmental (HSE) risks. The remotely operated vessels are much smaller than conventional ones, with the potential to reduce CO2 emissions by 85-90%.

In these surveys, both gravity and depth are measured at predefined seabed locations. Seawater pressure serves as the starting point for precise depth measurements (Siri Vassvåg et al, 2025). The measurement locations are marked by concrete pads (CPs), deployed prior to the first survey to ensure consistent measurement positioning across repeated surveys. During each survey, an instrument frame equipped with three relative gravimeters and three pressure sensors is placed sequentially on top of the CPs. The effect of tides on measurements is corrected using tide gauges deployed during the surveys.

To ensure time-lapse accuracy, some CPs, known as zero-level CPs, are positioned away from the reservoir’s edge. These are

crucial for measuring relative changes in gravity and pressure, and thus subsidence, with high precision. By assuming no timelapse signals occur at the zero-level CPs, the method can filter out non-reservoir-related effects, such as fluctuations in sea level or minor differences in sensor calibrations across surveys. Figure 2 shows one of the CPs deployed at the seafloor at Scarborough.

Forward modelling of time-lapse gravity and subsidence signals

To assess whether time-lapse gravity and subsidence measurements can support key reservoir management decisions in a timely way, four dynamic reservoir model realisations were evaluated, labelled 1 to 4. Realisation 1 serves as the base case, while the other three introduce variations in fault transmissibility and aquifer influx to different areas of the reservoir. All four models align with the current understanding of the reservoir, but each would necessitate different field development strategies, including the optimal placement of infill wells after Phase 1. Therefore, if time-lapse gravity and subsidence can distinguish between these scenarios in a timely manner, the measurements will be valuable for optimising infill well planning and improving recovery.

Forward-modelling of gravity signals from a dynamic reservoir model is made straightforward by using Newton’s law of universal gravitation. The effect of a mass change ∆mi in a reservoir cell on the vertical component of the gravity change ∆gj at the location of the j-th CP can be computed to the required precision by using the point mass approximation. The time-lapse gravity response at CP j is given by:

where G is the gravitational constant, and (xij, yij,z ij) represents the vector from CP j to the centre of reservoir cell i. The mass change in each grid cell is derived from variations in fluid saturations, densities, and pore volume.

The relationship between mass changes in the reservoir and gravity changes at the CPs is unambiguous, depending solely on their relative positions and the gravitational constant G. In contrast, the relationship between reservoir deformation and subsidence at the CPs is influenced by the complex distribution of mechanical properties in the overburden. However, for many

Figure 2 One of the 221 CPs deployed to serve as a base for time-lapse gravity and subsidence measurements at Scarborough.

Figure 3 4D gravity signal for models 1, 2, 3 and 4 (top left, top right, bottom left and bottom right respectively) for the first two years of production. Well positions are shown as yellow squares.

Figure 4 Seabed subsidence (in cm) for models 1, 2, 3 and 4 (top left, top right, bottom left and bottom right respectively) for the first two years of production. Well positions are shown as yellow squares.

field applications, simple semi-analytical modelling tools have proven effective to describe this relationship (e.g., Alnes et al., 2010; Van Thienen-Visser et al., 2015). These tools assume that vertical deformation at any seafloor point is the sum of contributions from the deformation of individual reservoir cells (nuclei of strain).

In the formulation by Van Opstal (1974), that assumes a homogeneous isotropic medium with Poisson’s ratio v and a stiff basement at a depth k, the vertical deformation Sj at CP can be computed as:

where T are analytical transfer functions and

For this study, pore volume changes ∆Vi are obtained from the dynamic reservoir models, which assume a pore compressibility of 34.8 • 10-5 Bar-1 for the upper and mid alluvial fan, and 58.0 • 10-5 Bar-1 for the lower fan across all four model realisations. Since the reservoir model assumes linear elastic compaction, multiplying these values by a scaling factor would proportionally rescale the predicted seafloor subsidence by the same factor.

Modelling results

Figure 3 presents the forward-modelled gravity change signals for the four reservoir model scenarios. The results assume an initial baseline survey conducted before gas production began, followed by a repeat survey two years after production commenced.

Figure 5 Contrast between the signals expected from models 1 and 2 (top), 1 and 3 (centre), and 3 and 4 (bottom), for time-lapse gravity (left) and subsidence (right) during the first two years of production. Well positions are shown as yellow squares.

Areas with negative gravity changes represent segments where hydrocarbon depletion has resulted in a net loss of mass. Conversely, positive gravity changes reflect zones where mass increase, due to aquifer water replacing gas, dominates. The distinct patterns across the four model realisations highlight the differences in reservoir behaviour and the high lateral resolution power of the distribution of gravity changes.

Figure 4 shows the forward-modelled seafloor deformation for the same four model scenarios, assuming identical timing for the baseline and repeat surveys as in the gravity modelling. Note that although the maximum expected subsidence is consistent across models, the lateral distribution of subsidence varies significantly, particularly in the central areas of the field.

Figure 5 illustrates the differences in time-lapse gravity and subsidence signals obtained from the different model realisations after two years of production.

To assess how confidently these measurements can differentiate between the models, the signal contrasts shown in Figure 4 must be compared with the expected uncertainties in the time-lapse measurements. Historical results from similar surveys provide insight into the factors that influence these uncertainties. Based on this record, the projected uncertainties for Scarborough are 2.5 mm for subsidence, 1 µGal for raw gravity change, and 1.3 µGal for subsidence-corrected gravity change. These values account for total noise, including imperfections in tidal corrections, sensor drift, and calibration.

Importantly, the differences between the forward-modelled time-lapse gravity and subsidence signals for the various realisations are significantly larger than the projected noise levels. This suggests that the two datasets can be confidently used to rule out certain models early in the project, providing critical information for decision-making and optimising field development strategies.

CP layout at Scarborough

The gravity and depth changes discussed above will be measured at a discrete grid of locations defined by the CPs deployed on the seafloor. A critical parameter for the CP layout design is the spacing between CPs, which must balance the amount of information gathered with the duration and cost of the survey. The modelling results show distinctive features in the signals with short lateral wavelengths, a consequence of the reservoir’s shallow depth. These time-lapse gravity and deformation measurements will provide accurate insights into the reservoir’s pressure depletion and flow patterns.

Fine-tuning the CP layout around regions with the most significant contrast between models risks overlooking unexpected features that could be revealed with a more uniform sampling of the signals. Therefore, a uniform CP distribution was chosen across most of the field, with slightly coarser spacing in the northeastern section, where lower permeability sands are expected, and no developments are planned. If reservoir surveillance reveals otherwise, additional CPs can be added in the northern area later.

Several CPs located at a distance from the field rim, known as zero-level CPs, play a crucial role in ensuring the accuracy of time-lapse changes of gravity and subsidence. These CPs are located in areas not expected to show significant timelapse signals related to reservoir depletion. Signals measured

on them enable us to capture and eventually filter out effects affecting time-lapse accuracy, such as varying sea levels or minor differences in sensor calibration across surveys (Agersborg et al., 2017). To properly calibrate subsidence and gravity measurements in situ, zero-level stations need to cover a range of depths and gravitational field values at the seafloor above the reservoir. A relatively uniform distribution around the field also reduces the likelihood of correlated noise, such as unexpected aquifer depletion.

For the initial phase, zero-level CPs at Scarborough have been installed within the WA-61-L and WA-62-L licences. Modelling indicates that these CPs will remain unaffected by production-related signals for the first five years of production. After this period, three CPs could potentially experience subsidence, and the deployment of additional stations may be necessary depending on early survey results. Forward modelling suggests that several of the northernmost field CPs could function as zero-level stations during the first two to three years of production, as their expected gravity and subsidence signals are negligible. This would enhance the robustness of the in-situ calibration process.

A total of 224 CPs were deployed on the seafloor, with an average spacing of 1.88 km between stations. Figure 6 illustrates the CP layout, showing that all stations are located within the Woodside titles (WA-61-L and WA-62-L). A disturbed, rugose area on the seafloor was avoided due to uncertain geotechnical properties and anticipated limited informational value. However, opportunities for infill in these areas remain available, pending future reservoir surveillance needs.

The CPs were deployed in Q1 2024, with the baseline survey scheduled for the first half of 2025. This will allow sufficient time for the CPs to settle into the sediments. If residual settlement varies between CPs, it could be misinterpreted as vertical seafloor

Figure 6 CP layout at Scarborough, overlayed on the gas rim, the outlines of Woodside titles, and the bathymetry.

deformation. Geotechnical modelling performed prior to the deployment concluded that the variability in soil parameters across the field will result in differences in settlement rates across the field of less than 1 mm/year. Therefore, this effect is not expected to significantly impact measurement noise levels.

Figure 7 shows the results of gravity changes and subsidence obtained from model 1 at the locations of the CPs.

Sensitivity to key reservoir properties

Time-lapse gravity and subsidence measurements are commonly used to quantitatively constrain reservoir properties through history-matching workflows (Alsos et al., 2024). Discrepancies between the measured values at CP locations and the forward-modelled predictions are incorporated into the objective function that is minimised to optimise the reservoir model. The modelling results presented here indicate that after two years of production, the range of allowed reservoir parameters will be significantly narrowed, as shown by the fact that at most one of the four considered realisations will remain unrefuted. We illustrate the ability of these datasets to quantitatively constrain model parameters by utilising two sensitivity calculations, focusing on vertical movement of the gas-water contact and pore compressibility.

Under the point mass approximation, a mass change of 200 kilotons occurring 950 m below the seabed produces a timelapse gravity signal of approximately 1.5 µGal at the seabed. In the case of the Scarborough field’s upper fan reservoir, assuming parameters such as 20% porosity, a water saturation change (∆S w) of 0.5, and fluid densities of ρw = 1010 kg/m³ for water and ρw = 160 kg/m³ for gas, this same 200-kiloton mass change corresponds to a 0.5-m rise in the gas-water contact over a 2 x 2 km² reservoir segment.

Given that subsidence-corrected gravity measurements have an expected accuracy of σ = 1.3 µGal, this allows for detecting vertical movements of the gas-water contact with a sensitivity of around 50 cm. This level of sensitivity has been validated through comparisons of time-lapse gravity measurements with direct well observations, confirming the method’s precision in tracking fluid migration within the reservoir (Siri Vassvåg et al, 2025).

Monitoring seafloor subsidence during the early stages of production will be critical for constraining the reservoir’s pore

Figure 7 Forward-modelled signals of changes of gravity (left) and subsidence (right) after two years of production for model 1. Magnitudes are given by the area of the circles, and the scale is provided as a circle on the top left of the figures. On the left plot, a circle is plotted surrounding zero-level CPs to facilitate their visibility.

compressibility, which in turn can be used to optimise the initial phases of production. As noted earlier, subsidence is proportional to reservoir compaction (Van Opstal, 1974), which, in the elastic regime, depends on the product of pore compressibility and pressure depletion in the reservoir.

The maximum observed subsidence provides, therefore, key information on the pore compressibility of the reservoir. Equation 2 shows that the forward-modelled subsidence using Van Opstal’s model depends on the Poisson’s ratio. However, this dependency is relatively mild. We performed a sensitivity analysis by assuming a subsidence measurement accuracy of 2.5 mm and a Poisson’s ratio uncertainty of ±0.1 which covers the expected variability in the overburden, assuming that pore compressibility varies as a consistent scale factor throughout the reservoir. The resulting compressibility sensitivity is ±4 • 10-5 Bar-1 for a repeat survey after one year of production. After two years of production, the sensitivity provided by subsidence measurements improves to ±2• 10-5. In other words, the uncertainty in this parameter can be reduced by a very large factor as early as one year after production start.

Conclusions

The decision to implement time-lapse gravity and subsidence monitoring at the Scarborough gas field represents a significant advancement in reservoir management. This technology, which has been effectively used in Norway, will be applied for the first time in the Asia-Pacific region. The shallow reservoir depth at Scarborough will allow for unprecedented lateral resolution and accuracy.

The detailed modelling indicates that time-lapse gravity and subsidence measurements will provide critical insights into the reservoir’s pressure depletion and fluid migration patterns, as well as required input data for calibration for 4D seismic acquisitions. The ability to discern between the different reservoir scenarios considered in a timely manner demonstrates that these data will enhance decision-making regarding well placements and future field development strategies, potentially optimising gas recovery.

In the longer term, the use of this technology may also facilitate a reduction in frequency of higher cost and environmental presence of 4D seismic surveys. Time-lapse gravity and

subsidence surveys require smaller vessels and fewer operational resources, thereby reducing the potential ecological footprint, including CO2 emissions due to the imminent deployment of remotely operated vessels.

References

Agersborg, R., Hille, L.T., Lien, M., Lindgård, J. E., Ruiz, H., Vatshelle, M. [2017]. Density changes and reservoir compaction from in-situ calibrated 4D gravity and subsidence measured at the seafloor. SPE Annual Technical Conference and Exhibition, Extended abstracts, PSE-187224-MS.

Alnes, H., Stenvold, T. and Eiken, O. [2010]. Experiences on seafloor gravimetric and subsidence monitoring above producing reservoirs. 72nd EAGE Annual Conference and Exhibition, Extended abstracts, L010.

Alsos, T., Osdal, B., Haverl, M., Skeie, S., Sørensen, H. and Holdahl, R. [2024]. Gravity and 4D seismic monitoring data in history matching of the Snøhvit reservoir model. 85th EAGE Annual Conference & Exhibition, Extended Abstracts

Borges, F., Basford, H., Calder, T., Lien, M., Vassvåg, S.C. and Ward, C. [2024]. Monitoring CO2 Storage in the Morecambe Depleted Gas Reservoirs through Seafloor Deformation and Time-Lapse Gravimetry Measurements. First Break, 42(3), 71-75.

Eiken, O., Stenvold, T., Zumberge, M., Alnes, H. and Sasagawa, G. [2008]. Gravimetric monitoring of gas production from the Troll field. Geophysics, 73(6), WA149-WA154.

Hatchell, P. and Bourne, S. [2005]. Rocks under strain: Strain-induced time-lapse time shifts are observed for depleting reservoirs. The Leading Edge, 24(12), 1222-1225.

Landrø, M. [2001]. Discrimination between pressure and fluid saturation changes from time-lapse seismic data. Geophysics, 66(3), 836-844.

Müller, T. M. and Gurevich, B. [2004]. One-dimensional random patchy saturation model for velocity and attenuation in porous rocks. Geophysics, 69(5), 1166-1172.

Ruiz, H., Lien, M., Vatshelle, M., Alnes, H., Haverl, M. and Sørensen, H. [2022]. Monitoring the Snøhvit gas field using seabed gravimetry and subsidence. First Break, 40(3), 93-96.

Solbu, Ø. H., Nyvoll, A., Alnes, H., Vassvåg, S. C., Lien, M. and Ruiz, H. [2023]. Time-Lapse Gravity and Subsidence Applied in History Matching of a Gas-Condensate Field. First Break, 41(9), 69-74.

Van Opstal, G. [1974]. The effect of base-rock rigidity on subsidence due to reservoir compaction. Advances in rock mechanics. Proceedings of the Third Congress of the International Society for Rock Mechanics, 2B:1102-1111.

Van Thienen-Visser, K., Pruiksma, J.P. and Breunese, J.N. [2015]. Compaction and subsidence of the Groningen gas field in the Netherlands. Proceedings of the International Association of Hydrological Sciences, 372, 367-373.

Vassvåg, S., Halpaap, F., Faust Andersen, C. and Jørgensen, L. [2025]. Twenty five years of monitoring the Troll gas and oil field with timelapse gravity and seafloor deformation surveys. First Break, 43(3), 41-46. See page 41 of this issue.

CALENDAR OF EVENTS

12-14

19-22 May InterPore 2025 www.events.interpore.org/event/56

7-11 Sep Near Surface Geoscience Conference and Exhibition (NSG) 2025 - 31st European Meeting of Environmental and Engineering Geophysics - 6th Conference on Geophysics for Mineral Exploration and Mining - 4th Conference on Geophysics for Infrastructure Planning, Monitoring and BIM - 1st Conference on Geohazards Assessment and Risk Mitigation - 1st Conference on UXO and Object Detection www.eagensg.org

7-11 Sep 32 nd International Meeting on Organic Geochemistry (IMOG) www.imogconference.org

9-11 Sep Second EAGE Conference and Exhibition on Guyana-Suriname Basin www.eage.org

14-18 Sep Seventh International Conference on Fault and Top Seals www.eage.org

15-18 Sep AOW Energy 2025 www.aowenergy.com

16-18 Sep The Middle East Oil, Gas and Geosciences Show (MEOS GEO) www.meos-geo.com

17-18 Sep First EAGE Workshop on Energy Transition in Latin America’s Southern Cone www.eage.org

22-24 Sep Sixth EAGE Borehole Geology Workshop www.eage.org

23-24 Sep EAGE Workshop on Geoscience in Development Enhancement www.eage.org

CALENDAR OF EVENTS