Poy s»: 25 Years of Artificial uift _ Leaders’ Insights on the Past, Present, Future Experts’ Advice for Future .. Professionals ee. ¢ © Tell THE CONVERSATION ON THE PROGRESS OF FUTURE OFFSHORE PROJECTS See 5 SS See SR utertry a Cen sac amContents 4 PRESIDENT’S COLUMN « Maximizing Your Membership by Leveraging SPE’s Multimedia Content SPE President Terry Palisch is joined by Dana Oli, vice president of SPE's marketing and communication, to discuss how members can maximize their ‘membership by leveraging multimedia content 10 GUESTEDITORIAL « The Difference Between CO, EOR and CCS Injection Well Metallurgy The long successful history of various metallurgies in EOR wells has been cited as sufficient to allow te same completion for CS injection wells. The ack of actual data onthe long-term performance of these alloys in EOR welln combination withthe more-stringent requirements for Class VI wells suggests otherwise 75th Anniversary Commemoration 16 artificial Lift: 25 Years of Change Tracked in the Pages of JPT ‘The world of artical it nas witnessed a remarkable revolution over the past 25 years, with many ofthe events and technology trends recorded in the Journal of Petroleum Technology. 22 7 Industry Leaders Share Their Insights on the Past, Present, and Future of Artificial Lift Longtime leaders in artificial it discuss and share their insights on the rapidly evolving segment’ pas, present, and future 32_Experts Offer Advice for Future Artificial Lift Professionals: “Do Not Be Scared To Try Something Different” ‘There will aways be a need for good artificial lit engineers. So, what should the next generation of its professionals be trying to pursue? 34 As Hype Fades, LLMs Gaining Acceptance in Upstream as New Age Research and Coding Tool Ealy adopters of large language models praise the technology's promise to advance upstream research and software development while also offering cautionary notes. 42. Top Year for High-Impact Global Exploration in the Cards Majors and independents alike are targeting high-impact wells across the globe to boos future oll and gas production volumes. ey ‘nts Publication of he Seeety of Petroleum Engineers Volume 76 | Number 3 On the cover: More than 2.200 head of bison roam Oklahoma's Joseph H. Willams Tallgrass Pravie Preserve. Located in Osage County in the northeastern part of the state, he petroleum industry has explored ‘and produced from the area for over a century. Source: Jennifer Presiey. DEPARTMENTS 8 Comments 13. E&P Notes 50 Case study 99 SPE Technical Papers Available 100 SPE News 102 SPE EventsRecent Content from SPE Journal ‘To view these papers, go to OnePetro, [Link], and search by title or paper number ‘or navigate to the journal. Data Science and Engineering Analytics ‘SPE-218409-PA—A Target-Aware ‘Wall Path Control Method Based on ‘Transfer Reinforcement Learning ‘SPE-218416-PA—Sort Sensor Development for Real-Time interface “Tracking in Muitiple Product Pipelines. Based on Knowlodge and Data Drilling and Completion SSPE-218404-PA—Guantitative Evaluation of Liquid Permeabiity in (Cracked Oilwell Cement Sheaths 'SPE-719446-PASimuation of Fiter- (Cake Formations on Vertical and inclined ‘Wels Under Elevated Temperature ‘and Pressure (Open Access) Formation Evaluation ‘SPE-218411-PA—On the Influence of Fine Particle Migration and Deposition on Gas Hydrate Production: Insights trom the First “Tal Production in the South China Sea ‘SPE-218420-PA—-Improving Reservoir ‘Characterization and Prediction Via Machine Learning-Driven Integration ‘of Subseismic Geologic Concepts, ‘Geophysical Attributes, and Wells Production and Operations ‘SPE-218017-PA—A Systom to Detect Citwell Anomalies Using Deep Learing and Decision Diagram Dual Approach ‘SPE-218392-PA—A New Gradient- ‘Accelerated Two-Stage Multobjective ‘Optimization Method for CO; Alternating ‘Water njection in an Ol Reservoir Reservoir Engineering ‘SPE-218421-PA—Phase Behaviors ‘of Gas Condensate at Pore Scale: Direct Visualization via Microfluidics ‘and In-Situ CT Seanning ‘SPE-219448-PA—A One-Dimensional ‘Convolutional Neural Network for Fast Predictions of the OiL-CO, Minimum Miscbilty Pressure in ‘Unconventional Reservoirs ‘Sustainability and Energy Transition SPE-218412-PAA Comprehensive ‘Assessment ofthe Iniegration of ‘Solvent and Steam forthe Extraction ‘of Bitumen Through the Development ‘of Novel Process Models ‘SPE-219449-PA—In-Situ Hydrogen Production from Natural Gas, ‘Walls with Subsurface Carbon Retention (Open Access) Technology Focus ‘The complete SPE technical papers featured in this issue are available free to SPE members for 2 months at [Link]. 53 ARTIFICIAL LIFT Dennis Harris, SPE, Chevron 55 Hydrohelical ESP Gas Separator Increases Flow Range, Improves Reliability 58 Hybrid Energy Holds Promise for Artificial Lift for Offshore Brazilian Fields 61 Artificial Neural Networks Predict Discharge Pressures of ESPs 64 PRODUCTION MONITORING Mariela Araujo, SPE, Shell Global Solutions 66 Near-Real-Time Tracer Data Optimizes Well Cleanups in the Nova Field 69 Slickline-Deployed Fiber-Optic Cable Provides First Production Profile for High-Temperature Gas Well 72. Crosswell Strain and Production-Interference Testing Integrated in the Permian Basin 75 HIGH PRESSURE/HIGH TEMPERATURE Santanu Das, SPE, ONGC 77 Modeling Approach Estimates Temperatures in Growing High-Temperature Wells 80 Nigerian Case Study Explores Drilling Through Shales Below Depleted Sands 83 Enzyme Based Cleanup Fluid Effective for High-Temperature Filter-Cake Removal 86 GEOTHERMAL ENERGY Rita Esuru Okoroafor, SPE, Texas A&M University, 88 Study Explores Retrofitting Abandoned Petroleum Wells for Geothermal Production 91 Study Outlines Challenges of Repurposing Oil and Gas Wells for Geothermal Applications 95 Technical Framework Assesses Offshore Geothermal Opportunities in Existing Assets JOURNAL OF PETROLEUM TECHNOLOGY UP) (SSN 0149-2126) is published monthly by the Society of Petroleum Engineers, 222 Palisades Creek Drive, Richardson, TX 75080 USA. To contact Customer Sence, call 1.972.952 8393 or serviee@[Link] cr Update Your Profle athetps://[Link]/members/,I> 2024 President “ Terry Palsch CARBO Ceramics ae eer aa REGIONAL DIRECTORS ASIA PACIFIC TECHNICAL DIRECTORS HSE AND SUSTAINABILITY rent jsobe omer Presi, The ournel of Petroleum Tecoleay® Secretary of Petroleum Cercle] WN Ed lg ‘Alexey Borisenko, Schlumberger RP aya) eared ATE eae) RESERVOIR Rodolfo Comacho, UNAM Stephen Rssenfos Enea Teco Pat Bass Seymeza, Continfe CREATING OUR ENERGY FUTURE with Terry Palisch Maximizing Your Membership by Leveraging SPE’s Multimedia Content the podcast episode to hear the full conversation on your favorite podcast platform under “The SPE Podcast.” Editor's Note: This month we are introducing a summary of the full transcript. We encourage you to listen to | In this podcast episode, | am joined by Dana Otilio, vice president of marketing and communications at SPE, to discuss how members can maximize their membership by leveraging multimedia content. The episode covers various aspects, including SPE Live, webinars, online training, and the Distinguished Lecturer (DL) Program. The conversation delves into the history and evolution of SPE's, multimedia content and its impact on connecting members globally. l introduce the episode by highlighting the goal of helping members understand how to get the most out of their SPE membership. As | like to do, liken an SPE membership to a gym membership, emphasizing the importance of “Multimedia is now the future ... the esas Se eeu ecu Cmte 4 JPT | March 2024 exploring all available offerings to create your energy future. Our discussion delves into the birth of SPE Live during the pandemic, emphasizing its role in overcoming the challenges posed by the inability to hold in-person events. Its success in delivering timely, relevant technical knowledge led to it becoming a series on Linkedin providing conversational content on global topics. | also highlight its popularity and mention the addition of Tech Talks, sponsored content sessions that have garnered significant views. To date these have been very popular, streaming four to five SPE Lives a month and garnering 1,600 to 1,700 views live and on-demand per presentation. It's one of our many growing outlets for multimedia Evolution of Multimedia Content Our discussion then turns to the accessibility of multimedia content through the SPE Energy Stream website, launched in 2022. | describe howEnergyStream TOP 5 LIVE VIEWS BY TITLE NEW CONTENT FOR 2023 «Energy in2—@mnuto videos EWS [TLE views | Tue feature ble sized hvormation SPE LveDisinguted Lectrer ffoma recent SPE Lye, webinar, 5 acme SPE Lv Dtingshd Lecturer srs: fone eer 1100 |r taroterscas Sms 1, |e lna arnt es «Gp Live Resahe_cortain the eens 60, Storage: Two Favors in One bestof moments from SPE Live fpsodes dung he year wi he Move tom th Bene the Geothermal Vision For The Of And Gas SEG LTRO O1 Mowmate semen gg | Samer vant + Got to Know Serie miniseries Tear conversations with tears nung Wornen SPE Live Disingushed Lecturer Sua is ont Energy and SPE Regional eres Enreprencustip nthe Inotde n-dorand wows Average Been re nee S84 seen viows per episode: 76. itis a one-stop shop for all multimedia content, easily accessible from the [Link] homepage. The website offers webinars, online training, SPE Live, and podcast episodes showcasing the organization's commitment to providing a comprehensive multimedia experience. | like to think of SPE Energy Stream as the “OnePetro” for multimedia content. The conversation explores the integration of the DL program with SPE Live, allowing DLs to reach a broader audience through live broadcasts. We take a moment to review the DL program's significance in connecting with sections worldwide, and the addition of SPE Live as a feature expands the program's reach. We distinguish between SPE Live, which focuses on conversational content, and webinars, which are more technical, lasting 60 to 90 minutes. The availability of webinars on-demand and the opportunity to earn continuing education credits are emphasized. Online training courses, both SPE-developed and in partnership with IHRDC, are discussed as more extensive, allowing members to learn at their own pace. in 2019, SPE partnered with IHRDC to link their training content to SPE members. It's offered through the SPE website and provides a reduced rate for our members. | believe multimedia is now the future and the content helps us run the full gamut of our mission. It’s not just connecting our members to technology, but it also includes members to members. Multimedia also helps us connect to our external stakeholders as much as anything because we have a lot of external stakeholders that tune into SPE Live and some of the other offerings. The discussion continues about the development of PetroWiki as a collaborative alternative to the SPE Petroleum Engineering Handbook. Created in response to the popularity of Wikipedia, it allows members to update and add content. | point out that everything in PetroWiki is sourced, unlike ChatGPT where you have no idea where the source came from. The discussion also raises the question of potential integration with artificial intelligence (Al) tools like large-language models to enhance or complement Petrowiki’s functionality. Overall, there's an acknowledgment of the platform's significance and a call for members to contribute and improve its content actively. Many may not realize that Al has been employed by SPE for many years. It's the backbone of how we've taken our technical content or print content and similar media and made it iptspeorg 5Terry recently visited the University of Houston student chapter. Future, as well as visited the campus and participated in a student-led podcast. searchable in our Research Portal. When you go to OnePetro and use the Resource Portal, it's using Al to find Keywords and define the articles that you need. Creating Your Energy Future ‘We wrap up the episode by emphasizing the importance of utilizing the Energy Stream website and highlight various ways members can benefit from it. | encourage exploring webinars, SPE Lives, and staying updated on Linkedin. Members are urged to identify educational needs and use the webinar library for free technical training. Professionals are advised to attend local section events and consider contributing to PetroWiki, Seasoned professionals are encouraged to participate in SPE Live, host webinars, or apply 6 SPT | March 2024 en en eta to become a DL (don't forget the deadline for DL applications is 15 March), Finally, we underscore the overall message of how investing in personal and professional development through the various platforms will help create your energy future. Next Month Sneak Peek Next month, the conversation will focus on how SPE contributes to improving the industryss public image, including discussions about the Energy4me program and related initiatives. In summary, the podcast provides valuable insights into the diverse multimedia offerings of SPE, showcasing the organization's commitment to connecting, educating, and empowering its members in the dynamic energy industry.LET NOT Geothermal Energy xploring Saudi Arabia’s Untapped Resources ¥ - Resource assessment consultancy - Exploration and development programs execution Geothermal development projects DISCOVER TAQA GEOTHERMAL| COMMENTS Al Accelerates Modeling of New Materials for Carbon Capture Avtificial intelligence (Al) is increasingly being ‘employed to assist in the development of materials, including metal-organic frameworks (MOFS), to develop carbon capture technologies. MOFS are modular materials made up of three building blocks: inorganic nodes such as zinc or copper; organic nodes; and organic linkers made up of carbon, oxygen, and other elements. By changing the relative positions and configurations of the building blocks, the potential combinations for creation of unique MOFS, are countless. The idea is to create a porous carbon dioxide “trap” to capture carbon from the air. The structure created by the building blocks can be thought of simplistically as a scaffolding with joints (linkers) that functions to absorb carbon. Ina recent paper, researchers from the Us Department of Energy Argonne National Laboratory, The University of illinois Chicago (UIC), University of Illinois, Northwestern University, and TotalEnergies described their use of generative A\ to identify “good carbon absorbers” among, “billions and billions of possibilities.” Although early work on developing MOFs began in the 1990s, the power of Al modeling can’ generate new models with desired properties such as optimal selectivity and capacity without the laborious, reiterative experimental and computational efforts once required. In a press release, Argonne said, ".. The team was able to quickly assemble, building block by building block, over 120,000 new MOF candidates within Pam Boschee, JPT Managing Editor 3 JPT | March 2024 30 minutes” on a supercomputer at the Argonne Leadership Computing Facility. “The race for capturing carbon hinges on finding needles in a haystack, and trial and error is too slow. You have billions and billions of possibilities, and then you must narrow down to candidates that are good carbon absorbers,” said paper coauthor Santanu Chaudhuri, professor of civil, materials, and environmental engineering at UIC and director of manufacturing science and engineering at Argonne, in a news release. "With this project, we have taken the first significant step towards closing that gap by using generative Al.” Coauthor Eliu Huerta, an Argonne computational scientist who helped lead the study, said, “The traditional methods have typically involved experimental synthesis and computational modeling with molecular dynamics simulations. But trying to survey the vast MOF landscape in this way is just impractical "We wanted to add new flavors to the MOFs that we were designing. We needed new ingredients for the Al recipe,” Huerta said, The Al framework, called GHP-MOFassemble, screened the newly built 120,000 MOFs in 40 minutes to identify nearly 79,000 with valid bonds (oints). Those were screened to identify 19,000 with valid chemistry within 205 minutes. From this group, 364 MOFs were identified in 50 minutes with CO, capacity higher than a selected value. From assembly to selection of MOFS, GHP-MOFassemble completed the analysis within S hours and 7 minutes. Powering the Al Future Its ironic that processing of massive volumes of data on supercomputers for emissions-relatedresearch, such as Al simulations, could contribute to the problem to be solved because of the electricity required. Supercomputers use the computing power of multiple interconnected processing cores, which require an immense amount of energy—and depending on the source of the electricity generation, also contribute to emissions. In a November 2023 commentary, International Energy ‘Agency (IEA) analysts estimated that Al uses more energy than other forms of computing. Training a single model uses more electricity than 100 US homes consume in an entire year. In 2022, Google reported that machine learning accounted for about 15% of its total energy use over the prior 3 years. Ina January report, IEA estimated that Google could see a tenfold increase in their electricity demand if Al is fully implemented in its search engine. The average electricity demand of a typical search is 0.3 Wh. Opendl's ChatGPT requires 2.9 Wh per request. Considering 9 billion searches per day, nearly 10 TWh of additional electricity would be required in a year. Google and other tech companies are shifting their data operations around the world daily or hourly to tap into excess renewable energy production. Google uses its “carbon- intelligent’ platform to analyze day-ahead predictions of how much a given grid will be relying on carbon-intensive energy. Itthen shifts computing globally in favor of regions where more carbon-free electricity is available. As Al advances in modeling complexity with a greater number of parameters, there is a growing emphasis on sustainability and optimizing power consumption, including developing energy-efficient algorithms, hardware, and data center management strategies. JPT peat ING A Generative Artificial Intelligence Framework Based on a Molecular Diffusion Model for the Design of Metal- ‘Organic Frameworks for Carbon Capture by H. Park, Yan, and R. Zhu, et al. Communications Chemistry. Why Al and Energy Are the New Power Couple by V. Rozite, J. Miller, and S. Oh, International Energy Agency. Electricity 2024, international Energy Agency. We Now Do More Computing Where There's Cleaner Energy by R. Koningstein, Google Research. JPT EDITORIAL REVIEW BOARD Patrick Miler—Charprson, Pons Iran Abbasy, Faison Petroleum ‘Anson Abraham, (NRL Amir Alwazzan, razon Ot Marea hruj, Sel ‘yrsnaAsarpota,aibuton ElabethBarsot URC abor.0y Linda Battalora, Colao Schoo of Mines ‘himing Chen Crna, of etm Graham Cole, rsitart Santana Das, ONGC Sandeep Dhawan, Weleroem, Ibrahim ohmy, orstant Michaet Faller, Crenon Jeff Gagnon, Exon Reza Garmeh, lon Fetrosun David Gurney, XU Subodh Gupta, ereteh Energy Dennis Hans, ceron Matin Hayes, gun i Paul Henderson, naiex Hussein Hotet, KWUST Siar ain, Sah NOC Swathik Jayakumnar, Cor Laborers Coin Johnston, Woe ited Machine ‘Muhammad Khan, A0NOC Celestina Kis, champont Martin astinka, Constant Mahdi Mahmoud, Yxigen Swamy Margo, Haiburon Mark Mure, sac Kristian Mogens, ADNOC Sandeep Mukierje, Clon Soujaya Mukherjee, Wires A ‘Anand Nagoo, Nag and Asodates Gopi Nal, Dole ard Macken EadorahOcha-Aud Shel Danny Ochoa Stisbeze Rita Okaroafor, esas EM Urivesty Vibha Pandey, ConocFiles ‘aos Pedros, aia rere Ramonden, hunter Michael Romer, Ecc Babatunde Slaw, Shel Débora Salomon, Dow Vitor Soran, YF Chaco syed Zahoor lah 5.8 Larissa walker, 6c June Yan, Sa hes Karam Yateem, Sous amo [Link] 9GUEST EDITORIAL Bruce Craig is an ‘SPE member ‘and subject matter expert in metallurgy ‘and corrosion vith MetCor in Denver, Colorado. He holds BS, Ms, and PhD degrees in metallurgical engineering from the Colorada School of Mines. For ‘more than 40 years, Craig has been involved with the selection of the ‘cilities metallurgy for oll and gas. projects around the world, including Mobile Bay in the Gutf of Mexico (shallow water), deepwater Gulf of Mexico, the Rocky Mountains, Asia, and the Middle East. In addition, he has most recently been involved in metallurgy selection for CCS and CCCUS injection well ‘Adam Rowe is an SPE ‘member and principal metallurgical engineer with Stress Engineering Services in Houston. He holds BS and MS degrees in metallurgical and materials engineering from the Colorado School of Mines. Rowe has specialized expertise in material evaluations, corrosion assessments, welding metallurgy, additive manufacturing, and metallurgical feilure analyses, He has worked on a large variety of projects across several different industries, and recent project experience includes addressing manufacturing, fabrication, and corrosion concems in CCS, hydrogen storage, and offshore wind developments. 40. JPT | March 2024 The Difference Between CO, EOR and CCS Injection Well Metallurgy Bruce Craig, MetCorr, and Adam Rowe, Stress Engineering Services he recent surge in development of CO, carbon capture and storage (CCS) projects has brought focus on the requirements for corrosion resistance in the injection wells as required by US Environmental Protection Agency (EPA) permitting. In some cases, the long successful history of various metallurgies in enhanced oil recovery (EOR) wells has been cited as sufficient to allow the same completions for CCS injection wells. The lack of actual data on the long-term performance of these alloys in EOR wells in combination with the more stringent requirements for Class VI wells suggests otherwise. Introduction The recent significant influx of large amounts of government, incentives for a variety of green initiatives including CCS and CCUS has created a rush to drill and complete CO; injection wells. However, the necessary corrosion data to make informed choices for corrosion resistance in these wells is minimal at best. Some oil and gas professionals have argued that there is no difference between the more than 40 years of petroleum experience with CO2 EOR and planned CCS wells. This comparison is not a valid one and can be risky considering the need for very long-term containment of CO. required by regulators. This article presents a comparison between CO EOR and CCS for injection well metallurgy and explains why this comparison is invalid.Background and Comparisons There is a 50-year successful history of CO, EOR in petroleum production that has led to a consensus, largely without merit, that corrosion of injection- well equipment is manageable and does not result in long-term degradation of certain components such as casing. The reason there is little or no merit to applying this experience to CO2 CCS wells is the lack of data specific to the long-term corrosion resistance of casing that remains in the EOR wells. Rather, the industry generally simply plugs these wells with cement, and in the majority of cases without examining the casing before abandoning. Furthermore, industry rarely returns to inspect them after years of abandonment. Ina few instances, logging and retrieval of cement cores and casing have been performed and analyzed. For example, analysis of cement. and casing from the SACROC field that had been. exposed to CO» for over 30 years showed corrosion along the casing-cement interface (Carey et al. 2007). Laumb et al. (2016) reported substantial casing corrosion for a CO EOR well in the Weyburn field based on a series of logs over several years. Koplos et al. (2007) provided survey data from numerous CO EOR wells and found that 11.1% of the wells failed mechanical integrity tests and GUEST EDITORIAL concluded that injection-well mechanical integrity is a concern for CO; injection and storage due to the corrosivity of CO2 on downhole materials. There are also many factors that are significantly different between these two methods of CO) injection. For one, EOR wells are typically shallower than CCS wells, and therefore the temperatures and pressures are often lower than for CCS. Moreover, the formations are often uniquely different. in the case of EOR, the formations are often partially depleted reservoirs containing some amount of oil in place that can moderate the corrosivity. In the case of CCS, the formations are typically high-salinity, nonpotable aquifers. Another important difference between EOR and CCS wells is that historically the source C02 for EOR was from gas plants that had few impurities such as typically HyS or occasionally small amounts of mercaptans. On the other hand, CCS injectate streams can contain numerous impurities depending on the source of the CO, that can have a significant impact on corrosion of downhole equipment. These impurities are rarely seen in EOR, Another important difference between EOR and CCS wells is the permitting requirements. In most cases EOR wells are covered under EPA Ded Cay (ccd cours ‘CO; Composition CO>, H2S, CHs, No ‘CO>, H:S, Oo, Hp, SOx, | Impurities have a significant effect on NOx, etc. (Craig et al)_| corrosion, especially carbon steels. Reservoir Type Depleted oil formation Saline formation The chloride content of the water Is often ‘much higher in saline reservoirs. Depths, fe Typleally, 3.000% 7,000 | 5,000 to 10,000 (Meyer and NETL) ‘Temperature Range, °F | <250 (NETL) 125 to 260 ‘Temperature Is more Important than <120 optimum pressure and depth from a materials, (Elhoshoudy and Desouky) standpoint. Pressure Range, psi | >1,200 to 1,500 (NETL) 72,000 to 6,000 >3,000 optimum (El-hoshoudy and Desouky) EPA Permitiing ‘lass 1) Clase Vi care? rere eee ear kone iptspeorg 44GUEST EDITORIAL Class II wells while CCS and CCUS are required to conform with EPA Class VI requirements. The EPA requirements for Class VI wells are: “Ensure that the materials planned for all well components are compatible with the planned injectate and formation fluids that ‘may be encountered, and that they can resist corrosion for the duration of the project.” Similar language is included in 40 CFR Part 146 Subpart H for Class VI wells. “All well materials must be compatible with ‘fluids with which the materials may be expected to come into contact..." First of all, carbon-steel casing is never considered corrosion resistant under any circumstances, and secondly, the design of many CCS wells can be 30 years or longer. Specifically, the EPA requires post-injection site care for 50 years after injection. There are no such requirements specific to casing, tubing, packers, etc. for Class II wells. Under no circumstances can carbon-steel tubulars be characterized as corrosion resistant; therefore, corrosion-resistant alloys (CRAs) must be considered for components exposed to the reservoir fluids in contact with CO>. The selection of the appropriate CRA for a specific well environment is a complicated process as detailed elsewhere (Craig, Rowe, and Doll et al. 2023). Also, for Class VI wells but not required for Class Il and EOR well “The Class VI Rule requires Class VI permit applicants to develop and implement a comprehensive testing and monitoring plan for their projects that includes injectate monitoring, corrosion monitoring of the well’s tubulars, etc.” Similar language is again reflected in the CFR. It is required that corrosion monitoring be performed with corrosion coupons or loops according to Class VI permitting, and that mass loss and thickness loss be measured. This requirement implicitly assumes coupons are made from carbon steel which is not the case for the use of CRAS in the injection zone. By and 42. JPT | March 2024 large, CRAs don't suffer from mass loss or thinning in these CO; injection streams. Rather if they are susceptible to corrosion, it will take the form of pitting or crevice corrosion which in many cases will not be observed on coupons. Conclusions Since there are currently no standards or guidelines on what would be considered acceptable corrosion damage to casing and tubing in CCS wells, and since Class VI wells are required to resist corrosion for the duration of the project, the only choice for these components when water is expected to be present is CRAS. JPT [Gesemmsems] Analysis and Performance of Oil Well Cement With 30 Years of CO2 Exposure From the SACROC Unit, West Texas, USA by . William Carey, et al. international Journal of Greenhouse Gas Control (2007). Wellbore Corrosion and Failure Assessment for C02 EOR and Storage: Two Case Studies in the Weyburn Field by. Laumb et. al. International Journal of Greenhouse Gas Control (2016). A Review of Injection Well Mechanical Integrity by]. Koplos, et. al. IEA Greenhouse Gas R&D Programme 3rd Well Bore Integrity Network Meeting, 12-13 March Santa Fe, New Mexico (2007). Guidelines for the Selection of Corrosion Resistant Alloys for CCS and CCUS Injection Wells by 8. Craig, A. Rowe, M. Warmack, et al. International journal of Greenhouse Gas Control (2023). ‘Summary of Carbon Dioxide Enhanced Oil Recovery Injection Well Technology by). Meyer. API (2007). Carbon Dioxide Enhanced Oil Recovery. NETL (2010). C02 Miscible Flooding for Enhanced Oil Recovery by A. Ethoshoudy and S. Desouly. intechOpen (2018),Supermajors Win Big in UK's 33rd Oil and Gas Round Seventeen companies have been offered a total of 24 licenses in the second round of the North Sea Transition Authority's (NSTA) 33rd oil and gas licensing round. Of the 17 companies, several supermajors including Equinor, BP, Shell, and TotalEnergies secured licenses for 74 blocks and part-blocks in the Central North Sea, Northern North Sea, and West of Shetland areas. Remaining blocks in the Southern North Sea and East Irish Sea will be offered when environmental evaluations are finalized by the UK Offshore Petroleum Regulator for Environment and Decommissioning. These awards follow the 27 licenses offered in the first allocation made in October 2023 which consisted of 931 blocks and part-blocks available in the same locations. The application window closed in January 2023 with 115 bids coming in from 76 companies. NSTA analysis reports the average time between licensing and first production is § years. Hercules Continues Namibia Exploration Offshore Namibia continues to be the site of more exploration activity as Portuguese oil and gas on Sean Drilling’s Hercules rig will ralocat Pees see Tere E&P NOTES company Galp announced the relocation of the Hercules rig, managed by Odfjell Drilling, to the Mopane-2x well location. The announcement came shortly after the company announced the discovery ofa significant column of light oil in reservoir- bearing sands of high quality at the Mopane-1x Well. Both wells are part of license PELS3 in the Upper Orange Basin with drilling depths up to 2550 m, Galp is the operator with 80% interest, along with the National Petroleum Company of Namibia (10%), and Custos (10%). New Production Licenses Awarded on Norwegian Continental Shelf ‘The Norwegian Ministry of Energy awarded 62 new production licenses on the Norwegian Continental Shelf to 24 companies in the Awards in Predefined Areas (APA) 2023 licensing round. Sixty-two licenses were offered including 29 in the North Sea, 25 in the Norwegian Sea, and & in the Barents Sea. Companies that received licenses include Aker BP ASA, ConocoPhillips Skandinavia AS, Harbour Energy Norge AS, OMV (Norge) AS, and Wintershall Dea Norge AS. Equinor received the most licenses with 39 new production licenses throughout the Norwegian Continental Shelf including 18 in the North Sea, 13 in the Norwegian Sea, and 8 in the Barents Sea, The company reported that 80% of the licenses are in areas near existing infrastructure and in known geology, and 20% are considered as options based on the company’s offshore geology knowledge. Equinor is the operator of 14 of the awarded licenses and a partner in 25, TotalEnergies Acquires Additional Interest in Namibia TotalEnergies will increase its presence in Namibia as the company announced the acquisition of an additional 10.5% participating interest in block 29138 and an additional 9.39% participating, interest in block 2912 from Impact Oil and Gas iptspeorg 13E&P NOTES NAMIBIA 2912 7,884 km? NAMIBIA 29138 8,215 km? VENUS-4x & rattax 2B Totalenergies acquired Blocks 2912 and 29138 from Impact Oil and Gas Namibia. Source: impact Oll and Ga Namibia. Both blocks are within the northern part of the Orange Basin, where TotalEnergies’ Venus discovery was made in February 2022. Impact will retain a 9.5% interest in both blocks, with TotalEnergies, operator, owning a total of 45.25% interest in block 29138 containing the Venus discovery, and a 42.5% interest in block 2912. Impact will receive a $99 million closing payment for the transaction. Shell Secures Valaris 123 for Exploration Shell announced the signing of a drilling contract with Valaris 123 for the Selene exploration well and the Pensacola appraisal well in the North Sea. Shell partnered with Deltic Energy on the P2436 and P2252 licenses with the contract anticipated to begin in June 2024. The pair is expected to be drilled as a two-well sequence. 44 JPT | March 2024 Shell previously made a gas discovery at the Pensacola gas prospect in the Southern North Sea in 2023. Shells the operator and holds a 65% interest with Deltic Energy (30%) and ONE-Dyas (5%). TotalEnergies Takes Control in Malaysia TotalEnergies signed an agreement with OMV to acquire its 50% interest in Malaysian independent gas producer and operator SapuraOMV Upstream. Sdn (SapuraOMV). The agreement is valued at $903 million, including the transfer of a $350 million loan granted by OMV to SapuraOMV. The transaction is expected to close at the end of Q2 2024, SapuraOMVs assets include a 40% and 30% operating interest in blocks Sk408 and SK310, respectively, which are both located offshore Sarawak in Malaysia, as well as explorationinterests in Mexico, Australia, and New Zealand Last year, the company's operated production was about 500 MMcf/D of natural gas and 7,000 B/D of condensates, Akpo West Starts Production TotalEnergies and partners announced the start of production from the Akpo West field on the PML2 license in deepwater OML 130 block, located NIGERIA South atlantic f OML 58, E&P NOTES 135 km offshore Nigeria. This production is the latest from FPSO Akpo which started up in 2009 and produced 124,000 BOE/D in 2023. Akpo West is expected to contribute an additional 14,000 8/D of condensates, to be followed by up to 4 million m3 of gas per day by 2028. TotalEnergies is the operator with a 24% interest, along with CNOOC (45%), Sapetro (15%), and Prime 130 (16%). JPT NIGERIA Port Harcourt Amenami Gulf of Guinea OML 99 OML 130 DD Asset operated by TotalEnergies CAMEROON Ikike OML 102 OML 100 0 50 km — iptspeorg 45Oe Ce koe cea Poneto asIPT 75 | CELEBRATING 75 YEARS | Artificial Lift 25 Years of Change Tracked in the Pages of JPT odding donkeys. Thirsty birds. Rocking horses. All are playful nicknames for the oil and gas icon known as a pumpjack. To the uninformed, the pumpjack is a thing-a-majig that has something to do with oil, probably “fracking” because that's what drilling rigs do, right? But as an industry-educated and well-informed reader of /PT, you know this is inaccurate. By whatever name you call it, you know that the pumpjack is the visible manifestation of an invisible physics equation, a mechanism buried deep underground that lifts reservoir fluids to the surface. You also know itis one type of artificial lift available in a stable of systems with equally curious and technical names like progressive cavity, plunger, jet, gas lift, and electrical submersible pump (ESP). While there is no global repository of artificial lift statistics, industry observers estimate that 90 to 95% of the world's producing wells use artificial lift, as noted in a 2013 JPT story on the challenges and opportunities for artificial litt Artificial lift, as applied in oil and ga production, has existed for over 100 years. ‘2014 JPTstory laid out a brief history of the practice, sharing how the walking beam principle JENNIFER PRESLEY, Senior Technology Editor dates back to at least 476 CE when used in Egypt, and evidence of the use of sucker rods to lift fluids has been excavated from the homes of wealthier families in the early days of the Roman Empire. Even as we leap forward to the modern- day application of artificial lft systems, many techniques and technologies resemble their original designs. How these systems are applied has changed, especially in the past decade-plus with the emergence of shale. As highlighted in a JPT story on the past 25 years of technology advances, the industry underwent a robust era of technology development in its rush to crack the shale development riddle. Since 1999, the industry also delivered a 36% increase in global supplies, as noted in this PT story celebrating this industry's biggest innovations. 1999 represents an interesting turning point in the industrys technological journey. Also, itis the last time PT editors took the opportunity to celebrate the publication as it turned 50 years old that year. What follows are a few highlights from JP's coverage of advances and changes made in artificial lift applications and practices over the past quarter century. iptspeorg 17CELEBRATING 75 YEARS JIPT75 The Gas Lift Method of Pumping Oil Wells "ws aiae “gue He hae a sew and afar sand in bed racic and any, therfore sien ntact or dbo, The Does should Pease the tabjet of eopnesing discussions ene ta "hepel general iuaderteniog "The ga its eimply oil ore of tho si i, which has boon ‘ne or ay sere nig mater fo well faa to Out Coat Si operat asl was weed enaiderblyinCalforia «ster of years ‘en Tho es i dior rm th ein th abl pe ed ne ‘Gas bused iste of a ehlety because i age to sve and never tral govtinn and soe natural ga acrompaies he, ‘The yor 5 gs, would te mated eared ost of ol wel a be died ith the stad orn an exploive titre ‘hich eval not be handed by compeesore. ‘The saving of gai e tne ofthe oun pa which patent overing the proces man rand {oPaip Janes 815, (Gar bar farter advantage of ot exaing 7 eoroding the ing snd tango aides epeially ia the presen lr Wont It Does Ws gar Min able to raise more ll fom wel than any kind of a ‘Walle pumping 300 or 40 bi, daily have buen eau to pro ‘ore thn HKD BOL per day hy the gar without njory te the ele ‘Memon oF Aeros In order to understand the method of operation ad th mito of haat proces, sme consieraton mut be pve to theories tha Tyeben inentigatl in eonseton wlth he al it ‘sening hat 1 fea ompresied ps eters low ie 1000 tno these ofthe isl be under me of preity Rein Ree Seen! Sere tee trek Adaptation and Transformation A circular beauty is realized in studying the history of artificial lift. For example, its place in the pantheon of published papers dates to 1925, with its first paper on a technique that is still highly relevant almost 100 years after its publication. “The gas lft is able to raise more oil from a well than any kind of a pump. Wells pumping 300 or 400 bbl daily have been caused to produce more than 1,000 bbl per day by the gas lift without injury to the wells,” said R.P. McLaughlin in a short paper titled “The Gas Lift Method of Pumping Wells.” Itis the oldest paper on artificial lft listed in OnePetro. There's an old maxim about how those who do not study history are doomed to repeat it. This rings true, especially when paired with artificial lift expert—and self-professed ESP guy—Mike Berry's 4g JPT | March 2024 comment in a 2017 Technology Focus overview in which he said, “As much as it pains me to admit, it, by the book, gas lift should be the default lift method for shale wells.” ‘And therein lies the circularity of artificial lifts modern-day history. itis a space populated by innovators and tinkerers, by those that see failure as an opportunity to try again until the system works, ‘A 2022 JPT story on the sector's ability to adapt and then re-adapt to meet the changing demands of production touched on how adaptation and experimentation have created anew toolbox. “Ten years ago, you wouldn't have seen much gas lift used in the Permian Basin. Old school, traditional methods of lift, like electrical submersible pumps or rod pumps, were used there then. Gas lift, progressing cavity pumps (PCPs), or some of the other traditional methods weren't used because these types were not as familiar to operators. They went with what they knew,” said Shauna Noonan, 2020 SPE President and senior director of global supply chain initiatives and Fellow for Occidental Petroleum, in the article. Industry events like those sponsored by SPE and other technical groups have brought greater awareness to the industry of the different types of lift systems used across the globe. “By bringing all the lift types together, we can learn from one another and have more tools in our toolbox to choose from. We're starting to see that change, especially in the growth of new wells where we're picking the right lift method at the time, and not what they knew,” she said. “Its an interesting time. All the gas-lft people that used to be in Louisiana working in the Gulf of Mexico are now in Midland,” Noonan added. "Gas lift is one of the fastest-growing lift methods out there, displacing the other types.” But gas lift didr’t get its start in the Cajun State. There's a 1926 article available in OnePetro on its application in the Mid-Continent oil fields of Oklahoma,Use of Gas Lifts in the Mid-Continent Field Foes hin he inty resp Aout twa year ago we ton ans aeater inate rene fo propped by the wit that this ental ald flow prnapes worked ‘0 Ly hn wile hang age ae pertins. Th th eae of well that ba ented to flow where a lange qty of qu wa rset the ‘nao thse pine wi at in lg iar of proton by {ing eomprse log oth wel a theese a 9 wel owing tna two tale ths form of contraling te ae ssited wihte ‘eng eter im sama income of prado on edlin ff the gus rs por hart oi dim oth eases would esi {in aoeater tina over ro the ne The Rosana Petroloum Corp, appreciating the advantages to bo tained ty a slate applieaton of the pines of Ue ai it too ‘ella eetly emgage, fr Linitd tine, the services of the weiter to Sei et acre ofthe poets iaveled. The werk en nein tat Brumat, Garber and Tonkawa in Oklahoma, an at Oxdord in Kane allowing some suggnstions of the writer wll was sito equipped ty the Rocann sompany at Smackover by A'S. Fld Isoumarn Paoneenion my Appuaamoy on Ae Lier ‘Te role of the we of thin msthod wee, ata whol, factory ins fare ws pile to ary out these ea time, A total of 24 wel predioed 6500 DOL before Intl the a and 1847 BB afterward, Ch giving am inca of 1 tent The range a ieeran warm abet 10) pr eem. where oniitons were favorable for Uh te of the ar Hit indicating tat funy ces well would yal greater prdetin of oi te ld were Feaved rapidly axitener (he wast of thn ncace wat monde tn we sendy povided with he ule of gas Mt Ta bringing stout Ui dg ie pati onions it wae are to remove all bacepestues whetever pune. All pda where ai freuen arcuate at ll orsutl owe were hinted inthe How le, the ga trap lg moved elon to the wel at plas wok a A contributor to this circularity was the limited avenues for knowledge sharing, as noted by a few in this article's companion piece, “7 Industry Leaders Share Their Insights on the Past, Present, and Future of Artificial Lift." Before 2014, when SPE held its first Artificial Lift Conference and Exhibition (ALCE) in Houston, most technical presentations were delivered at non- SPE events with no technical paper accompanying, them, according to Greg Stephenson, chief production engineer for Occidental Petroleum. “if you weren't physically present in the room Where the presentation was delivered, you had virtually no way to access the information being shared,” he said, noting that the ALCE events in the Americas and Middle East, along with the ESP symposium, have provided a wealth of papers to learn and build on CELEBRATING 75 YEARS JPT75 “The requirement to write a manuscript increased the technical rigor of these works and made them more useful to industry professionals,” Stephenson said The 2024 edition of the ALCE-Americas will be held from 20-24 August in The Woodlands, Texas, with the theme "Modern Artificial Lift—Adapting to a Changing Industry.” ‘A quick check of the OnePetro database shows that in the 10 years from 1999 to 2009, 272 articles on artificial lift were published in JP7. That number ‘grew 22% to 333 articles published between 2010 and 2020. There have been 86 articles published between 2021 and February. Look to those innovators who accentuated the positives of gas lft and plunger lift to create the gas-assisted plunger lift (GAPL) system. The evolution of hybrid systems built from gas-lift principles can easily be traced through the pages of OnePetro and jPT. Injust one of many examples, a technical paper (SPE 48840) featured in a May 1999 Technology Focus shared how a gas-liftjet-pump (GLP) application delivered a significant production increase with the same gas-injection volume obtained in two wells during a test in a field located offshore Lake Maracaibo in Venezuela, The GLP performed successfully when installed in a well containing sand with a high gas/oil ratio above the main liquid-producing zone, providing the capability to produce without surface gas injection. A technical paper (SPE 64466) featured in a May 2001 Technology Focus shared how an ESP and automated gas-lift completion was developed to overcome production challenges in the Stag field located offshore Dampier, Australia, The flow from the well’s horizontal section had a high gas fraction, continuously slugged, and carried large sand volumes. And the reservoir pressure depleted more rapidly than expected. It has the hallmark challenges for shale production, begging whether something similar would work for shale. Maybe it has, but a paper has not been published in OnePetro or reported on iptspeorg 19CELEBRATING 75 YEARS JIPT75 inJPT. A check of OnePetro reveals that SPE 64466 has been cited in seven papers for conferences held in the Asia Pacific and Middle East regions. Artificial Lift Tech Snapshots Perhaps the only good thing to come out of a revolution is that people talk more. This was certainly the shale revolution’s impact on artif lift. Finding solutions to arrest the rapid decline in production was just one of many challenges the industry had to wrangle. Other issues, like gas and solids entering the wellbore and long, undulating lateral wellbores, also present pumping challenges. ‘Add to this various business and operational goals, like minimizing capex, opex, and downtime while maximizing the well’ life and production output, the challenges of artificially lifting shale reservoirs are complex. Artificial lft experts warned about these challenges in a PT story in 2015 regarding long- term production in horizontal wells. Many of these challenges remain in 2024. However, with more than 130,000 shale wells drilled and put into production in the major US basins, the industry is improving with the data and the knowledge of what will and will not work when applying artificial lift technologies. Fortunately, new technologies are making great inroads towards improving production. improved sucker rod coatings, better downhole separation techniques, advances in data and automation, and increased use of permanent magnet motors are all making an impact. Improved Sucker Rods. AJPT story in November 2022 asks a question of change; namely, ifa better sucker rod is built, will buyers be willing to change. Shale wells commonly produce corrosive fluids, paraffin, and asphaltenes. ‘One example of a sucker rod improvement discussed in the story is a protective anodic coating for continuous sucker rods that prevents corrosion and reduces drag when the rod rubs against 20 SPT | March 2024 the surrounding tubing, The paper (SPE 209751) presented at the 2022 ALCE-Americas, highlights the Championx product which relies on a powdered metal coating that short-circuits electrochemical reactions that cause rapid corrosion. Better Downhole Separation. Addressing the challenge of gas and solids separation for sucker rod or ESP pumping, a relatively new design named the WhaleShark takes advantage of multiphase liquid flow reversals to ensure the pump remains full of liquid with every stroke while also protecting the pump from erratic load conditions. ‘SPE 209755 was presented at the 2022 ALCE and featured in a PT story in November 2022. The Cilfy product sits above the bend of the horizontal wellbore, in the vertical section. It uses an upward- facing, open-mouth collector intake that captures liquid that naturally falls backward due to gravity. To encourage more liquid fallback, the area above the collector is the gas separation region that has no annulus and uses an oval-shaped eccentric (on the side), rather than centric, pump intake tube designed to slow fluid velocities While improving liquid fallback downward into the collector. Advances in Data. Another of the significant, advances in artificial lift is the increasing use of digital systems and automation to capture, compute, monitor, and react to well data. Regardless of the companys size, from supermajor to smaller privately held operators, these data are critical to operations, as noted in a January 2023 JPT article on how digitalization projects are delivering returns for operators. Init, a pair of Texas-based private operators shared how they apply data using Datagration’s, PetroVisor to improve their artificial lift operations. Hess shared how it developed the digital twin of its Stampede production platform in the US Gulf of Mexico as part of a pilot project to enable process equipment condition monitoring on key topsides equipment.ee ee eee eed rod lift system in Weatherford's North American es Snot PMM. The growing adoption of permanent magnet motors (PMM) has played out on the pages of, JPT, starting with a July 2009 article on their application in oil production. In it, the benefits and application of PMMs rather than asynchronous submersible electric motors used to drive ESPs at the time were discussed, PMMs in ESPs were just beginning to achieve wider application at the time. In the more than 10 years since the article, PMMs have entered various applications. A March 2023 JPT technical synopsis shared results of how PMMs were used to help reduce the carbon footprint in artificial lift systems used in Colombia. The operator installed and operated PMMs on ESPs, PCPs with an ESP PIM, PCP PMM, and sucker rod pump. The implementation of the PMM technology was successful, delivering power savings of more IPT 75 | CELEBRATING 75 YEARS than 15% and lifting cost optimization. The energy- savings effect obtained by each system translated into a forecast reduction of 24,500 tons of carbon dioxide equivalent and aligned with the operator's sustainability objectives. In February, Weatherford International confirmed it had successfully installed a PMIM- powered reciprocating rod-lft system in the Bakken Shale as part of a seven-well pilot test for independent producer Chord Energy. The oilfield technology provider uses a PMM on its PCP systems. It said this installation is the first application of a PMMM on a reciprocating rod-ift system in its North American operations. The past 25 years have demonstrated what Mauricio Prado, professor emeritus at The University of Tulsa, wrote in his May 2003 Artificial Lift Technology Overview that “this area, that a few years ago was seen as a mature one, is proving to stil be a fertile place for innovation, research, and development.” While it is impossible to know what the next 25 years of artificial lift advances will look like, it will certainly continue to be an engaging space to watch and report on. JPT Re eau aa AU UNC} Al Firm Ambyint’s New Bakken Deal with Equinor Moves the Industry Another Step Closer to the Edge by T. Jacobs. JPT. Artificial Lift Experience in Malaysia and Indonesia and the Way Forward by F. Mehran, Halliburton; K Idris, Petronas; and Dadik Kusuma, Petramina. JT. Permian Operators Collaborate to Improve Understanding and Performance of Unconventional Reservoirs by ¥. Pradhan. JPT. Artificial Lift Market To Reach $14 Billion in 2025 by]. Presley. JPT. Now Is The Time for Gas Lift To Live Up to Its Potential by S. Rassenfoss. JPT. ipt-speorg 24| IPT75 | CELEBRATING 75 YEARS 7 Industry Leaders Share Their Insights on the Past, Present, and Future of Artificial Lift aN aaa Aa unos gs nthe past 25 years, the artificial itt industry has seen incredible changes, making hydrocarbon production smarter and more efficient. in this Q&A article, JPT talks with seven industry leaders from this, field about its past, present, and future. This conversation showcases the technological progress such as materials, digital tools, and automation and the strategic leadership that has guided the industry to new heights. Join us as we dive into the world of artificial lift and discover the innovation and expertise shaping the future of energy. This roundtable Q&A includes the following Participants: + Greg Stephenson, chief production engineer, Occidental Petroleum + Michael Romer, principal artificial lift engineer, ExxonMobil + Laura Labrador, senior production engineer, Ecopetrol, and 2023-2024 chairman of the SPE Artificial Lift and Gas Well Deliquification Technical Section + Shauna Noonan, Oxy Fellow and senior director global supply chain initiatives, Occidental Petroleum 22. JPT | March 2024 + Jose Ernesto Jaua, global product champion, SLB + Kevin Leslie, vice president artificial lft solutions, Weatherford International + Dana Meadows, global portfolio director, artificial lift systems, Baker Hughes JPT: Can you recalll an experience, technical Paper, or mentor that significantly influenced your early career in artificial lift? How did it shape your approach? Stephenson: Numerous individuals impacted my early career, the most noteworthy being Herald Winkler, who was in the first class of SPE'S Legends of Artificial Lift awarded in 2014. | traveled to my first ATCE in New Orleans asa first-year petroleum engineering student. | distinctly remember walking the exhibit floor and seeing this little guy get mobbed by people asking him questions. | asked one of my fellow students, ‘Who is that guy? Tom Cruise” He told me, ‘No. That's Wink:| then learned that Wink was one of the pioneers of gas-lift technology and wrote the first definitive book on the subject. At that moment, | realized that attifcial lft might offer a viable career path for me. Eventually, I got to know him personally first as a student and then as an artificial lft professional. One of the most impactful conversations I had with him was one in which he told me, am not a ‘gas-lft expert. You cannot be an expert in gas lift— the field is too complex. 'm still learning things, and Ive been doing this for over 60 years!’ That encounter taught me to be humble in approaching, my craft and never assume | knew everything. Romer: My first role in artificial lift was 2 years after | started with ExxonMobil. There was a US production artificial lift team at the time. Some people | worked with closely were Pat Underwood, Mike Johnson, Rodney Bane, and Mark Johnson. Pat was our subject matter expert (SME) for ESPs and rod pumps, and Mike was the SME for gas-lift systems Rodney was the group manager and had 4 strong artificial lit background. He was responsible for turning the group into a global one. Mark was a global artifical lift engineering resource, Most of the artificial lift team came up through the field and had great hands-on experience. My background is in electrical engineering, not petroleum, so | was fortunate to learn from them. Labrador: Although it is not exactly an early influence ‘on my career, I can say that its the one that most marked my history in the industry. ‘Around a decade ago, when | was working for an operator company in Colombia, IPT 75 | CELEBRATING 75 YEARS 2 pivotal moment occurred when | presented our progressive cavity pump (PCP) results to the worldwide artificial lift system (ALS) team. In the audience was a remarkable woman, Shauna Noonan, the head of the worldwide ALS team for that company. Her knowledgeable comments and insightful questions left an indelible impression on me. From that day forward, | was inspired to keep working on ALS professionally and take her as a role model. Her guidance continues to shape my career, driving me to strive as a professional woman. Noonan: During one of my co-op internships, Kevin Grondin was my manager with Mobil Canada, and the fields he was responsible for were gas- lifted wells. My assigned project from him was to install the first plunger-lift system in the area. That was my introduction to artificial lift; his enthusiasm made me want to learn more about it My very first SPE event was the ESP Symposium, and it was women-led. Karen Draper and Carol Magney-Grande convinced me to join the committee That started with attending and coordinating artificial lift events and realizing that there were strong women leaders in the artificial lift space Jaua: | have been fortunate in my career to have had many great experiences in the artificial lift world. 1 ‘grew up professionally in the discipline, starting in the field and working up to managing entire geographies and products at a global scale IF must select one incredibly enriching period, it was in 2011 when | was assigned to workin Ecuador, managing the artifical lift business covering the Andean region. | had the incredible opportunity to iptspeorg 23IPT75 | CELEBRATING 75 YEARS be part of the board of directors of the SPE local chapter, meeting many top industry and government authorities. | developed many initiatives that positively impacted the industry and promoted technologies and local talent. | also took that platform to coach and mentor young professionals. Those moments in which I have been able to impact and contribute to a higher cause have always been the ones | have the best memories of. That experience was transformative for me, and since then, every time | change position or location, linvest effort and time to be engaged with the communities and mentor others whenever I can. Leslie: Thinking back to my early career, one person stood out as an influence in my artificial lift career: Larry Hoes, the vice president of oil field for Lufkin industries. Larry spent most of his career at Cameron but moved to Lufkin in the early 90s. During my career transition, Larry allowed me to showcase my abilities while working as the scheduler/planner of the Pumping Unit Facility in Lufkin, Texas, and then moved me to west Texas to manage sales and business development. Larry was a strong mentor who allowed employees to do what was needed as long as the results came through in the end. By providing this freedom, | developed my talent for sales, operations, and growth in the Permian Basin while learning multiple forms of artifical lift and the benefits of automation in the field. Meadows: Early in my artificial lift career, a mentor taught me to look for the value of just one more run day to an operator. That person taught me to listen and work to shape a tailored March 2024 24 SPT | solution around the customer's needs—not just sell them widgets. That made me view technology and development differently by asking the right questions, focusing on the business case from end to end, and framing it in a real-world application to provide a disruptive solution. JPT: How have you seen knowledge sharing and best practices evolve within artificial lift applications over the years? Stephenson: Artificial lift has experienced a metamorphosis while Ive been in the industry. During the early portion of my career, virtually every technical presentation on artificial lift was delivered at a non-SPE event. This meant that most knowledge sharing was via PowerPoint, and there was never an accompanying manuscript that might find its way into OnePetro or PT. If you weren't physically present in the room where the presentation was delivered, you had virtually no Way to access the information being shared. That all changed with the creation of the SPE Artificial Lift Conference and Exhibition (ALCE) events in the Americas and the Middle East. These events were the brainchild of Shauna Noonan during her first term on the SPE Board of Directors, ‘Those who work with artificial lift owe Shauna an enormous debt for this. When Shauna was the JPT technical editor for the topic she had relatively few papers to draw from. My job was significantly easier when | served in that role. The ALCE events and the ESP Symposium provided a wealth of papers from which to choose. The requirement to write a manuscript increased the technical rigor of these works and made them more useful to industry professionals. Romer: There are more avenues for sharing now, including SPE technical sections, SPE regional artificial lift conferences, the Artificial Lift Research and Development Council (ALRDC), and through joint industry projects like C-FER's ESP Reliability Information and Failure Tracking System, LSU'sGas Lift Valve Performance Clearinghouse, and the ‘ALRDC's Horizontal Well Downhole Dynamometer Data Acquisition Project. internally, knowledge sharing has increased with virtual events with video. Unfortunately, this has made in-person events less likely. Labrador: Knowledge sharing has positively increased over the years. Itis evident when you look back and see we have more events related to artificial lit systems and we recently got a specific SPE technical section for artificial lift (and gas well deliquification). The operators are working together to share the learning curve. Also, both operators and oilfield service (OFS) companies are documenting and publishing the lessons learned, best practices, and the knowledge acquired over time more frequently than before. Noonan: initially, most artificial lift events were not SPE-organized events, and the presentations/ papers were not publicly available. | made it my mission to change that when | joined the SPE Board as the Production & Operations Technical Director. | created new AL events, converted the SPE Gulf Coast Section ESP event to a symposium so papers could be formally published, advocated for the AL Tech Section, and created the Legends of artificial Lift Awards. Jaua: Knowledge sharing has grown tremendously in the past 20 years, along with the evolution of digital technologies. In the ‘not that far’ past, we had to travel to congresses to meet peers and be exposed to newer technological developments. Nowadays, we have all the information and much more at the tip of our fingers and in every format we can imagine. From microlearning content to fully immersive technical conferences, all delivered remotely. In parallel, more communities and interest groups have been formed because now there are more domains and subdomains within the artificial lift space and, in general, in industry. CELEBRATING 75 YEARS JPT75 Leslie: From personal experience, the oil field grows through knowledge sharing and best practices in all facets of the industry. in the artificial lift space, engineers and supply chain utilize their knowledge of products to secure better pricing and activity. in contrast, other OFS companies utilize knowledge sharing and best practices to produce wells at a higher volume with a better, mean time between failures. By sharing the learned knowledge, the OFS and E&P companies can produce at a higher production level with fewer problems. Without knowledge sharing and best practices, todays oil and gas market would be totally different. Meadows: Unlike in previous years, where the person who had the document in their filing cabinet was considered the expert, we now have multiple applications for data sharing on global internal operations that allow answers to pop up, lessons learned to be elevated quickly, and field instances to be shared and trended with actions. This allows us to help our customers with solutions gained from actual cases worldwide and to develop the best technologies to address their needs using real-time operational feedback. Eagerness to share, learn, and collaborate is a large part of our culture at Baker Hughes. Technology has allowed that to advance quickly in the artifical lift field and beyond. JT: |s there a specific paper that helped you or your team solve a challenging problem in artificial lift? Could you elaborate on the problem and the solution? Stephenson: The paper that has specifically influenced my work the most in recent years is an old one: SPE 179, Evaluation of Valve Port Size, Surface Chokes, and Fluid Fall-Back in Intermittent Gas Lift Installations by Kermit Brown and Frank Jessen. This paper was published in March 1952, and more than 60 years later, itis still the most iptspeorg 25CELEBRATING 75 YEARS JIPT75 extensive exploration of intermittent gas-lft performance available to the industry. Brown explored the relationship between slug height, gas passage, surface opening pressure, and slug recovery in intermittent gas lft wells. This is particularly relevant today because the industry has thousands of wells in unconventional resource plays that produce via continuous flow gas lift and experience rapidly declining production rates in early life. As the wells mature, the artificial lift system becomes less efficient and requires some lift revision. One of the most likely is conversion to intermittent gas lift, often using a pilot valve. Unfortunately, the methods for designing and analyzing these wells are often antiquated and rely heavily on rules of thumb. The results can be unpredictable and/or suboptimal well performance. When | first read this paper, it inspired me to explore the issue further, ultimately leading me to Ali Hernandez’ textbook, Fundamentals of Gas Lift Engineering, published in 2016. Using the knowledge gained from Ali's textbook and the original paper by Kermit Brown, | built an in-house simulator for intermittent gas lift wells. This work also inspired the formation of the newly created Texas Tech Gas Lift Consortium in Lubbock, Texas. Romer: We were looking for ways to improve gas lift and came across SPE 21639 (1991), Foamed Gas Lift, by M.G. Bernadiner from the University of Michigan, The project presented an interesting way to improve gas lift using surfactants, with both laboratory and oilfield experiments showing that it could be an effective means for producing oll from depleted reservoirs. We thought it could be an interesting way to improve gas lit, and it led to research testing at FER in Edmonton, a pilot in the Gulf of Mexico, a presentation at ALRDC, and later pilot testing and application in the Delaware and Midland basins. No SPE papers were written on the ExxonMobil 26 JPT | March 2024 project, but the original paper demonstrated how somebody had done it once and that it should work. Leslie: One of the most influential technical papers impacting my team and colleagues to solve the goal for lifting higher production with rod lift continues to be SPE 176041 (2015), Electric Submersible Pumping Systems vs. Long-Stroke Pumping Units: A Case Study of Economic Comparison in a Low- Volume Well. The paper captures the topic of shifting to rod lift sooner by transitioning from an ESP to a long-stroke Rotaflex pumping unit, and the benefit of saving workover costs while increasing production well above traditional beam pumping systems. The shift to rod lift sooner had been successful due to the long, slow stroke of the vertical system, allowing for increased pump fillage and higher production rates. In this case, the long-stroke system proved to reduce power costs by 37%, reduce workover costs by 91% and combined reduced opex by 62%, all while producing a fluid rate of 1,180 STB/D in either system. This has been a fantastic paper to show the benefits of transitioning to rod lift sooner with huge success. Meadows: Permanent magnet motors (PPM) are a step change in the ESP industry. They generate high torque on startup, are highly efficient, and have a wider operating range than traditional motors. However, they can generate electricity anytime fluid flows through the pump and forces system rotation, even when powered off, energizing the power cable to the surface. A major limitation to global field adoption is the danger of this hazardous back-EMF associated with motor rotation and energization during installation, retrieval, and workover operations In response to this industrywide HSE problem, our team has developed the industry's first engineered PMM safety barrier that prevents the ESP pump from spinning the PMM, eliminating voltage generation at the surface. We were excited to present this HSE innovation at the 2023 SPEGulf Coast Section—Electric Submersible Pumps Symposium in SPE 214745, Permanent Magnet Motor Safety—The Big Short Question. JPT: How have advancements in technology impacted artificial lift? Are there any technologies that you find particularly promising for the future? Stephenson: Mostly, advancements in artifical lift technology have been incremental improvements. The most-used artificial lift methods have existed for decades or even longer. Technologies | think could make a difference include permanent magnet motors in ESP, sucker rod pump, and PCP installations, improved downhole gas separation and gas-handling devices, improved surveillance software, and the application of artificial intelligence/machine learning techniques. Romer: | see technology advancements have impacted artificial lift in several ways, including + Remote monitoring and control improvements, including accessibility and cost—variable frequency drives versus pumpoff controllers, for example Ability to optimize more with less effort, providing quicker transfer of knowledge Data storage, communication, and sensing improvements + Once unheard of to run bottomhole pressure ‘gauges onshore and now theyre almost standard with gas-lft completions Extra digital surface sensors once required justification, but now they're expected Surface-controlled (electric or hydraulic) gas lift has the potential to bring optimization to a lift method that's notoriously difficult to optimize with additional benefits over mechanical valves. The decreasing cost of downhole electrical devices and improving reliability are spurring adoption. + High-speed ESPs, particularly for offshore installation through-tubing, which may be able CELEBRATING 75 YEARS JPT75 to increase the life of offshore wells as the technology increases in maturity + Electric rod pumps are in early development, but a few vendors are working on them Labrador: Artificial lift technology advances have allowed us to install and run the systems in more complex trajectory wells and produce them reliably. Additionally, it has allowed us to be aligned with the commitment to care for the environment. Also, AL technology has impacted the artificial lift world positively. Who could imagine we would have more efficient motors installed at the surface and the bottom of the well? Or that we could challenge the boundaries of each AL according to the needs of the field and sufficiently produce them? We must keep challenging our boundaries and making bigger efforts in this important part of the oil and gas industry. The relationship between OFS and E&P companies plays an important role: the operator's needs and the service companies’ constant solutions to save on downtime and lost production Noonan: Instrumentation advancements allowed Us to understand the physics of how artificial lft systems work internally. Before, we didn't have the advancements in instrumentation to put sensors inside the various pumps or lift systems to really measure and understand what was happening inside. It was always theories and rules of thumb. But once we understood the fundamentals of each basic lift system, we knew how to design it properly—whether it was how we sized it or the materials used—to ensure it was the right design for the right environment. Also, permanent magnet and linear motors will make an impact (extinction of the rod string). Jaua: We see the dramatic increase in lifting systems’ average run life and their survivability operating under extreme environments in the past 20 years; itis evident that technological iptspeorg 27CELEBRATING 75 YEARS JIPT75 advancements have had a huge impact on the artifical lift industry. Artificial lft systems are installed faster, safer, last longer, and operate in wells that were impossible to operate not long ago. All thanks to improved designs, advanced materials, and evolved processes and practices. The lifting system with the most promising growth potential is PCP, particularly the rodless version powered by downhole permanent magnet motors, which is more suitable for deviated wells. Current elastomer compositions and manufacturing processes prevent the PCPs from surviving temperatures above 200°F and produce fluids with high aromatics content, frequently found in lighter oils. New technologies and manufacturing processes are under development that will allow PCPs to perform in hotter wells and produce lighter oils. This will result in a viable alternative to efficiently producing wells, replacing rod lifting systems with a higher frequency of failures in deviated wells. Leslie: The artificial lift market continues to see new technology in many shapes and forms. Whether itis advancement in a long-stroke system, a beam pumping system, advancement in ESP technology, or advancement in downhole technology, there continue to be new developments and products every year. The one technology that continues to advance rapidly with huge impacts on field operations is the digitalautomation space. The future of field operations will be centered around automation. The future will continue remote monitoring with autonomous control and failure prediction well ahead of where we are today. The autonomous field of the future will utilize sensors and monitors to detect failures before they happen, allowing operators to save on downtime and lost production. Digital integration and automation are the most impactful technologies for artificial the future. 2g JPT | March 2024 JPT: What are the most pressing challenges and opportunities facing the artificial lift sector today? Stephenson: The most pressing challenges | see include the application of artificial lft in unconventional shale oil wells and in subsea/ deepwater installations. Shale oil wells can be challenging to lift due to their high gas-to-liquid ratios, sand production, sharp production decline in early life, and horizontal well trajectory. All these issues can decrease equipment reliability and/or deferred production. Applying artificial lft systems in subsea installations can be challenging because of the high intervention costs associated with these wells. In a perfect world, we would like to install an artificial lift system in these applications that can survive for 10 years or more while boosting production. Romer: | see the most pressing challenges in the following areas: + Unconventional wells: How do you extend the economic life of wells when they aren't designed for late-life lift: crooked wells, high gas, small casing, low rates, and long horizontals Lift methods designed for rapid reservoir depletion as current methods are not Lift method changes: knowing when it’s the right time and what to do + Subsea: improving through-tubing options and applying artifical lift at the beginning of the well’ life because you may not be able to come back later to retrofit Labrador: The challenge will always be to be better than yesterday and to have better performance and efficiency in ALS equipment than the previous year. Reducing the carbon footprint through ALS is, a commitment we all have right now for the planet and future generations. The opportunity is to continue developing technology to make artificial lift systems more efficient and cleaner for theenvironment. Moreover, digital integration and automation technologies will help us to impact the future for artificial lift. Also, being competitive in quality, performance, and price is something that can't be left behind, Noonan: Advancements in manufacturing have not kept pace. Many facilities have old equipment and still rely on too much human interaction, which can lead to quality-assurance issues. The technical differentiation between vendors has shrunk, so the battle for business will rely on manufacturing, assembly, and service quality Jaua: All oil wells eventually require artificial lift systems to produce, meaning millions of wells operating globally, requiring a constant power source available 24/7. This makes lifting, systems one of the industry largest emitters of carbon dioxide. The greatest challenge will be the evolution of lift-system selection and sizing that puts efficiency levels as a top priority. It also represents an opportunity for those technology companies committed to developing new solutions to reduce their carbon footprint. Leslie: There are several key challenges facing the artificial lift sector today, including limited resources, the need for consolidation, and, most importantly, a consistent supply chain with minimal reductions in delivery. The oil and gas market has received a bad reputation as dirty and cyclical, driving young talent to avoid the industry. This reputation limits the amount of new talent in the industry, creating challenges for both E&P and OFS companies. We must find a way to recruit and retain talent better going forward. The artificial lift landscape has too many competitors. There is a major need for consolidation, ultimately benefiting the end user and OFS companies. Finally, most artificial lift market products are produced outside North America due to the high production cost and the demand CELEBRATING 75 YEARS JPT75 for lower-cost products. This creates an incredible challenge for OFS companies to work diligently to find alternative countries and manufacturing centers to supply products to all customers in artificial lift. These three areas are just some of the critical challenges facing the ALS market today. Meadows: In terms of challenges facing the artificial lift sector today, the main concerns we continue to hear from our customers are centered around increasing the run life and reliability of ESP systems, reducing costs and downtime of artificial lift systems in general, and maximizing and optimizing production. Our customers all want to unlock the lowest-carbon, shortest-cycle barrels possible. To do this, they need a partner to help them better manage power consumption and costs and boost pump speeds and production. We all know new drilling is slowing down while the world's energy demands grow. New energy sources can only pick up some of that demand, Which means welll need to get better at getting more out of the reservoirs we've already drilled, and ESPs will be a critical part of that. JPT: Where will artificial lift applications be heading in the next decade? Are there any emerging trends or technologies that will shape the future? Stephenson: In the next decade, | expect to see a greater emphasis on + Lifting low-rate, mature shale oil wells. + Reducing emissions. + Minimizing lift costs. + Improving overall system reliability. + Applying artificial lift and subsea boosting to extend the life of deepwater installations. Romer: | expect to see in the next decade + More PMIM applications on, for example, surface pumpjacks and PCPs. + Closed-loop optimization of surface-controlled gastlift systems. ipt-speorg 29CELEBRATING 75 YEARS JIPT75 + Through-tubing high-speed ESPs. + Data-driven, fieldwide artificial lft system optimization. + Optimization and artificial lift selection based on emissions/power usage. The industry is already seeing this in some parts of the world, for example, in thermal lift applications in Alberta, Labrador: The technologies that allow us to produce complex wells efficiently and cleanly for the ecosystem will be what will govern the next decade. The world is changing, and we must change with it. Noonan: Material advancements will make all the difference. Rod strings will be obsolete. Jaua: | see artificial lft applications accelerating the adoption and implementation of artificial intelligence and machine learning solutions. For instance, until today, designing, selecting, and optimizing lifting systems required a technical staff to make and control decisions. itis already happening today—but not fully adopted—that machines are gathering and processing vast sources of historical data and, in seconds, without human interaction, selecting the best lifting system, sizing, and features required for every well, Allowing a fully autonomous operation and increased failure predictability levels make the entire operation more efficient. Digital solutions are stil in the infant stages in the artificial lift sector, but without any doubt, they are destined to lead a more efficient and sustainable O&G industry. 30 JPT | March 2024 Leslie: The future for artificial lift will be centered around digital technology and operator energy consumption. The evolution of PM will be significant for operating all artificial lift forms more safely and energy efficient. PMMs have already been installed on basically every form of lift today with much success, but adoption has been slow as operators are still in the testing phase. In addition, the digital space will bring new technology with capacitors and storage of excess energy to maximize efficiency and productivity in the field. These technologies are just arriving to the market and producing significant energy savings, and the future will only get better as we move forward with this new technology. Meadows: Automation simplifies decision- making. Digital will continue to shape the future by optimizing and increasing production. Variable speed drives (VSDs) will also play an important role in reducing power consumption. Were testing a new VSD slated for release later this year to reduce power consumption by up to 20% over our most-efficient drive. A virtual VSD interface will also be integrated, enabling operators and service companies to adjust drive settings from anywhere in the world remotely. The switch from induction motors to PMMs is also where the future is headed. We want to ensure we're positioned to ensure this happens as quickly and safely as possible, so we are excited to champion the industry first engineered PMM safety module. This new module prevents the ESP pump from spinning the PMM, eliminating voltage generation at the surface for safer, simpler operations. JPTSO SPE/ICOTA Well Intervention Conference and Exhibition Attend the Premier Global Well Intervention Industry Event 19-20 March 2024. travoirctaa conentionconter,71 With Over 1,100 Attendees, 70+ Exhibitors, and More Than 50 Technical Sessions, This Is an Event You Won't Want to Miss! 1S=Ye) hel OV] 17 March & SVZI USD100| IPT 75 | CELEBRATING 75 YEARS Experts Give Career Advice for Future Artificial Lift Professionals: “Do Not be Scared to Try Something Different” nowing where to start can be daunting for those new to the oil and gas industry. With all the changes the artificial lift sector has experienced in the past 25 years, the constant is that a new challenge always needs a solution. The spirit of innovation and invention will serve the next generation of artificial lift professionals well, Here are a few nuggets of advice from experts to help newcomers navigate this dynamic sector while also making a difference in the energy world. Greg Stephenson, chief production engineer, Occidental Petroleum My advice is to go for it! While ours is a notoriously cyclical industry, artificial lift professionals tend to enjoy greater job security than many of their counterparts. Also, we get to experiment with new technologies and can often 32 JPT | March 2024 JENNIFER PRESLEY, Senior Technology Editor see immediate results from our efforts. | advise any young professional to do what you love and love what you do. If you follow that credo, you can't go wrong. | feel very fortunate to work in a field that | love. Michael Romer, principal artificial lift engineer, ExxonMobil Network, go to artificial lift events to learn from SPE and others. Artificial lift is a small world. You'll keep seeing the same people in other roles and places. Get to the field, into the shop, and figure out the challenges firsthand, as not all problems can be solved digitally or wont tell the entire story. Current lft methods have been around for more than 100 years and will continue. The more things change, the more they stay the same; there's a long history of learning from industry. Someone may have already solved the problem, and the answermay already exist. Or someone could have tried something you're considering, and it didn’t work. artificial lift is a great place to be, with lots of variety. It's important but not nearly as ‘right now as drilling: you can quickly see results, and the costs/risks are generally lower for trying new things. Figure out what you have passion for and own it because others will notice, Laura Labrador, senior production engineer, Ecopetrol Do not take everything as absolute truth. Ask questions before believing. Investigate and consult various opinions from people with experience. We do not need to know everything, but we do need friends who know what we do not know. Finally, teach and transfer your knowledge: this is how you will be able to continue growing. Shauna Noonan, Oxy Fellow and senior director of global supply chain initiatives, Occidental Petroleum Regardless of the cycles and when drilling and completions have halted due to volatile oil prices, the existing wells will still need to be on-production, and most will be on some form of artificial lift. There is always a constant demand for good artificial lft engineers. Jose Ernesto Jaua, global product champion, SLB First, get technically proficient fast. Second, develop strong collaboration and communication skills. Third, try to understand all areas of IPT 75 | CELEBRATING 75 YEARS the oil and gas industry; artificial lft is not isolated, and understanding how it is connected to everything else will make you a more-rounded professional. Last, remember that we are here to create value for our customers, our communities, and the planet in everything we do. | know that any young professional that follows this advice will be set for success! Kevin Leslie, vice president artificial lift solutions, Weatherford International ‘The production arm is typically the last leg of the well cycle when looking at drilling to completion to production, Artificial lift falls into this realm because ALS is ‘Always the Last Solution’ for a well. For all new young professionals entering, the artificial lft segment, | would challenge you to help find the next big technology to drive better and faster results in the field. Do not be scared to get your hands dirty; go to the field and learn while trying something different from everyone else. There are several creative ideas in the production space; sometimes, being the first to try might result in something greater. Do not be scared to try something different. Dana Meadows, global portfolio director, artificial lift systems, Baker Hughes My advice to future generations of artificial lift professionals would be to team up with experienced field personnel and ask questions. Question everything—there may be a better way. Try to get as much exposure in front of customers as possible to understand their pain points and solve their issues. Ask for a mentor to help you and listen to key employees to build upon their advice and experiences. JPT iptspeorg 33FEATURE | Langu: As Hype Fades, LLMs Gaining Acceptance in Upstream as New Age Research and Coding Tool SAN NGO) See Muss Ce ae ‘ chart size o) Pleo ia Pallfter using ChatGPT-4 for more than a year, Clinton Lott has made familiarity with the large language model (LLM), a requirement for new software programming hires. The director of Houston-based Spotter Tech gave this reasoning: ‘Taking 2 hours to do something that ChatGPT can do in 15 seconds is unacceptable today.” Its a line you're likely to hear from the most ardent champions of LLMs which have simultaneously reignited the world's interest in and ire for artificial intelligence (Al). And while Spotter Tech, a small hydraulic fracturing data specialist, may be an exception with its hiring practices, itis far from the only upstream company embracing the new technology. Both BP and Shell have recently given thousands of employees access to Microsoft's new cloud service called Copilot, which at its core is powered by GPT-4, OpenAt's latest and most capable LLM. Others including ExxonMobil have placed use of LLMs under lockdown while still signaling a long-term interest in their embrace. ‘As chair of SPE’s Data Science & Engineering Analytics Technical Section, Pushpesh Sharma has spent the past several months speaking with peers in the industry about how and where LLMs are being used. The product manager for field automation firm Aspen Technology told JPT that its clear from those conversations that a good number of energy firms are interested in tailoring the technology to their needs, albeit many are doing so quietly as it remains early days. “The big focus right now is on what they call fine tuning,” he said. "You take an existing base model that one of the big tech companies have created and make it more of a fit for your business. Its kind of lke taking a graduate student and then letting them work on their PhD with very specific subject matter.” While Opend\'s ChatGPT has become the popular symbol of LLMs, there are a growing FEATURE | Language Models number of alternatives in the mix. In February, Google introduced its latest model called Gemini 1.5 which seeks to rival GPT-<. There are also open source LLMs that oil and gas companies can use as templates for internal projects. Or they may turn to other big tech firms such as NVIDIA and Amazon Web Services (AWS) which are in the business of helping industrials build and train thelr own private, custom LLMs. SPE joined the trend after inking a Memorandum of Understanding (MoU) in October with Al firm i2k Connect and Aramco Americas. The collaboration is expected to result in a new domain-specific LLM that industry engineers and researchers can use to answer questions with more technical depth than can be done with the current slate of commercial offerings. One enabling aspect of this current phase of experimentation with LLM technology is retrieval- augmented generation (RAG). The method is considered essential for models to provide more- accurate and actionable responses by extending their scope beyond initial training datasets to include more authoritative and even real-time data. Spotter Tech The ansvsis or Treating Wel 3 Stage 8 has been successfully itiaed and ‘el take ew moments forthe data tobe prepare. Analysis ead. ‘The derivative ofthe cour coreepondng to OFFSET WELL 1H Catng 1 ‘pressure ha been caledated and saved to anew CSV tle. You can download ‘be eating He tom he folowing ine Offset [Link] ‘Tocaloust he detvetve of te “OFFSET WELL 1H1 Casing 1 | res (p column, |used the diference quent method which is apreprit for ‘umercl iferrtiaton. This invohed the flowing sop: 1. rapped any ay Se eater aed ‘with real-time data using different methodologies. area iptspeorg 35FEATURE | Language Models Question ‘Has anyone had success determining fracture dimensions? @ For each question order the answers via the drag and drop arrow on the lef side of each answer based on the answer's quay, with te top of th ist answer withthe highest qualty. ChatGPT Yes according fo the context, racture dimensions ean be determined by using modals to make decisions and us ’by analyzing tracture-treatment data. The design must always compare actual resuts with the predictions trom ‘mode! calculations. By “calfrating” the 2D modol with told results, te 2 models can be used to mako design Changes and improve the success of simulation treatments. Sth et a. discussed other issues and presonted BERT Pipeline| several caso histories in which fracture-reaiment data were analyzed fo determine the propped tracture dimonsions. Asm ot at cscussod ater issues and prosentd soveral case histories in which racturetratment data me fracture dimensions. wore analyzed fo determine the propped fracture dimensions. Smith statod that fracture height growth during ‘and after pumping operations, fluid loss In layered formations, and slurry viscosity all affect the propped = Rad more here ‘Source citation Sharma spent months poring over papers While working on his PhD and said he’s “jealous” of current and future students that will have access to new age Al research tools as they are inevitability refined through features like RAG to provide references. But not everyone in the business has stood on the sidelines waiting for industry-specific or enterprise-level solutions. it should come as no surprise given mainstream LLMs are free to use while individual access to the most powerful versions can be purchased for $20 to $30 per month. Preston Freisner, a senior petroleum engineer with VSO Petroleum Consultants in Dallas, is among those taking advantage of the low barrier to entry. With programming counting for only a small fraction of his coursework at Texas A&M University, the 2018 graduate uses ChatGPT to keep pace with a small team of engineers with significantly more computer science experience. “if | am getting an error in the code that Ive just written, instead of going through it all myself to find what's wrong, | can just throw it into ChatGPT 36 JPT | March 2024 and it takes what was a 3-hour problem-solving, process down to about 30 minutes,” he said. Using the LLM, Freisner and his teammates have developed and updated several in-house analytics tools for the firm that focuses on buyside reserve evaluations—a corner of the US industry in which timelines to deliver are often counted in days rather than weeks or months. The homegrown programs help with everything from connecting to SQL databases and auto- populating seller's data into spreadsheets to building interactive maps of US shale fields and detailed gun-barrel views of individual well pads. In combining its collective expertise with the capabilities of LLM technology, VSO has avoided the need to rely on the market for such software solutions. Further, the small firm with limited resources has also foregone the need to hire engineering techs or computer scientists that would otherwise be needed to write all the code. “Its a tool, and if we can learn how to use it right, it will be very hangy in speeding up our processes,” shared Freisner, adding, “I think of it kind of like a little best friend that sits on yourdesk—use it when you need to, ignore it when you dont.” All these examples suggest that the hype around LLMs may be giving way to practical Use cases, particularly in the areas of software development and research. But the path to industrywide adoption—or @ measurable industrywide impact—is never so. straightforward, There are bumps in the road and they include valid concerns over accuracy, data security, and Al complacency. In Need of Reliable Sources Those who see a need for new industry-specific LLMs argue that commercial versions provide petrotechnical answers that while good for a general audience fall short of being able to help run an oil and gas project. The limitations of these existing models are seen as the result of them having been trained largely on publicly accessible web content. And be it 2 major flaw or an unavoidable feature of their inner workings, there's also the well-known fact that LLM-generated answers are sometimes entirely incorrect. Thus it is believed an energy- focused LLM could meet the need for highly technical, reliable answers. This might make it seem like one obvious thing to do then is to dump the SPE’s paper repository into one of the existing LLMs and watch as it trains itself to become an upstream know-it-all machine. However, Joshua Eckroth, the chief architect at i2k Connect, explained that's not how it works: “if we did that, we'd be training the model how to write a research paper which is of no real value. What we need is something that answers questions, and to do that, we would have to generate question and answer pairs from the knowledge within OnePetro and that is a big process.” Another potential step to any such process would likely involve asking industry experts to check a model's work as it is developed. Eckroth FEATURE | Language Models said such reinforcement training would not only enhance the model's accuracy but would build industry confidence in the tool, serving as a sort of human-issued “stamp of approval.” The experts needed to make this possible are not expected to be hard to find after an early study into the capabilities of LLMs drew in about 150 oll and gas experts to volunteer as the Al's quiz masters. Eckroth helped lead that study and coauthored a recently published paper. Presented at SPE's Annual Technical Conference and Exhibition (ATCE), SPE 214888 describes a blind study run a year ago by staff from i2k and SPE to test a prototype LLM against the models developed by Google and OpenAl. The work counts as one of the first of only a handful of papers found in OnePetro that directly address the use of LLMs in the industry. The test it outlines offers important insight into how people interact with these advanced machines. Experts from various backgrounds took part by asking a selection of hundreds of petrotechnical questions to the LLMs and then ranking the answers from best to worst. When the votes were tallied, the undisputed winner was GPT-3.5. (GPT-4was not included since it was released after the bling study launched.) The runner up was a Petrowiki-enhanced version of GPT-3.5 designed to copy and paste relevant snippets of text from the SPE website with oil and gas information. Yet, the volunteers overwhelmingly preferred the answers that had no citations or sources to back them up. The result was a little surprising but Eckroth found rational explanation. First, the volunteers were not asked to vote as if they were using the answers in a hypothetical work scenario—something he thinks might have altered the outcome a bit. But in terms of the depth of knowledge, he added that the standalone GPT model outranked the others because though itis a centralized source of information, PetraWiki is surpassed by the iptspeorg 37FEATURE | Language Models FAQ BERT Pipeline a = 100, 1000. ° 8 3 8 6 (725), (4801) te rk Pett rit cig GP chart co 0 a 7 5 0 8 & 1000 ° oo ° Cee ee) FAQ and a BERT model based on Google LLM technology from 2018.1 Is given for the best answers, while 7 is assigned popcorn tent ns endear Se et oe peter) amount of upstream information that’s been spread across the common web. All this said, the early study clearly shows that good conversation and writing skills should not be discounted when assessing the value of LLMs. “ChatGPT is personalized, as init is responding to you and | think this is one of the real inventions behind the GPT revolution—whatever you ask it or tell it to do for you, it will do specifically that,” explained Eckroth. However, conversation skills and a wide but ‘general knowledge of the business won't cut it, said Eckroth. He believes for this new realm of Alto become trusted to make business or engineering decisions in the upstream industry, LLMs will need to prove they are relying on reliable sources. 3g JPT | March 2024 “if you are going to be using this technology for your job, you need a source and a way to verity things,” said Eckroth, adding that, “I dont think anyone should be using an LLM without having that kind of ability.” Researchers at the Ready ‘The budding potential of Al-based research aids took center stage at the recent PIVOT2023 geothermal conference run by the nonprofit Project innerSpace. An innovation competition run in collaboration with SPE during the conference asked people to showcase novel ways that LLMs could be used to improve some aspect of the geothermal value chain. Finalists included a prototype LLM designed to replace emails at the rigsite with real-timemessaging. Another proposed using an LLM for predictive drilling analytics. The winning entry was a program called ChatGeoPT which was designed to answer geothermal-specific questions by pulling in direct sources from a database with hundreds of geothermal papers. A proof of concept, it showed how engineers could use a chatbot to learn about various operational issues such as corrosion prevention or how to best service heat exchangers. Daniel Merino-Garcia, a chemical engineer and contest judge, said one of his favorite submissions was another chatbot that also drew on papers but in this case to help serve an even smaller niche: the planning and designing of wells used for ground- sourced heat pumps. “That program was built, in about a week, so imagine the capabilities that this could deliver if it was taken to a higher level,” he said, The research and development lead with Project innerSpace who previously researched technology at Repsol said the contest offered him a glimpse of a tipping point, on the other side of, which LLMs can be used for an increasing number of engineering tasks. While that implies human talent will be replaced to some degree by advanced models, he also argues we will see a premium placed on the “supervised collaboration’ between engineers in both the geothermal and oil and gas sectors. “it may be few years from now, but if you don't learn how to use and collaborate with this technology, you are going to be out of the game,” said Merino-Garcia. “What | tell people to do is read and learn as much as you can because the shift will be massive.” There is another point of view to consider as well, one that places even more emphasis on the human's role in unlocking the full potential of LLM technology. Recalling an Al panel discussion held during ATCE in October, Sharma paraphrased a comment that resonated with him: "We have to make sure to remember that LLMs are a tool—they are not FEATURE | Language Models the answer. Just like a hammer is useless unless a human is involved, we should see this as a tool for humans to use to get to an answer.” Get Smart, but Not Lazy Ruud Weljermars, a professor at King Fahd University of Petroleum & Minerals in Saudi Arabia and an expert in hydraulic fracturing and geomechanics, is an admitted skeptic of the contributions delivered so far to the industry by machine learning and Al. After years of applying computer science to oil and gas problems and reviewing papers about doing the same, he's been left with “only disappointment.” Weljermars presents 2 more upbeat yet measured view on LLMs in a recent study he coauthored. The research explored how GPT models specifically will change petroleum engineering and geoscience education. As others have pointed out, the study found GPTs currently fall short of being reliable tools for deep research. But the authors—a group of graduate students and their professors—also revealed the immediate promise they offer for those who write and edit technical works. This doesn't mean using LLMs to spin up paper content from whole cloth. It means using the technology for brainstorming, summarizing key points, and improving the readability of what are usually dense and verbose texts. ‘SPE has formed an initial policy that allows the use of LLMs such as ChatGPT to help write papers, but it also warns that authors should rigorously fact check and disclose the content that is generated by Al. SPE has said it plans to review this policy to ensure it remains “relevant and effective in the face of changing technologies and emerging concerns.” Weljermars said the introduction of GPTs have had a particular impact for students who lack paper-writing experience and/or are not yet fluent in English—the common language of most industry paper circuits. Beyond this, it has helped him and other professors simply reclaim many of the hours iptspeorg 39