This paper provides the learnings from a successful application of a smart completion in a complex heterogeneous carbonate reservoir. It details the study, planning, coordination, and implementation process of two pilot wells by a multidisciplinary team, and pilot production performance results, illustrating the success. First, to select an optimum completion design for the field, multi-segment well option and local grid refinement option were applied to the reservoir simulation model including calibration of faults/fractures. Second, based on the modified model, sensitivity analysis was conducted; 1) by selecting different types of completion including Open-hole, blank pipes (BP), compartmentalized slotted liners (SL), inflow control device (ICD) and hydraulic flow control valve (FCV); 2) by optimizing the number of compartments (packer and blank pipe placements for all cases), and ICD / FCV numbers and nozzle sizes. Using the data from the modeled cases, economic analysis was conducted, which indicated that the ICD in conjunction with sliding sleeves (SSD) was the best option. Two candidate wells were selected to cover the variation of reservoir characteristics: one well representing the heterogeneous part of the reservoir with high-density of faults, fractures and kurst, and another one representing the relatively homogenous part of the reservoir suffering from heel to toe effect. A multidisciplinary implementation team was set up to align all stakeholders on subsurface requirements, following up the completion design, coordinating material procurement and logistics for mobilizations, daily drilling operations follow-up, real-time logging data interpretations and completion design adjustment. Evaluation of the two pilots’ results based on predefined KPIs during the study, exceeded overall expectations.
{"title":"Agile Smart Completion Application for Effective Water Production Control in Heterogeneous Carbonate Reservoir in Abu Dhabi Offshore Reservoir","authors":"A. Abdelkerim, S. Bellah, A. Ziad, Kei Yamamoto","doi":"10.2118/208023-ms","DOIUrl":"https://doi.org/10.2118/208023-ms","url":null,"abstract":"\u0000 This paper provides the learnings from a successful application of a smart completion in a complex heterogeneous carbonate reservoir. It details the study, planning, coordination, and implementation process of two pilot wells by a multidisciplinary team, and pilot production performance results, illustrating the success.\u0000 First, to select an optimum completion design for the field, multi-segment well option and local grid refinement option were applied to the reservoir simulation model including calibration of faults/fractures. Second, based on the modified model, sensitivity analysis was conducted; 1) by selecting different types of completion including Open-hole, blank pipes (BP), compartmentalized slotted liners (SL), inflow control device (ICD) and hydraulic flow control valve (FCV); 2) by optimizing the number of compartments (packer and blank pipe placements for all cases), and ICD / FCV numbers and nozzle sizes. Using the data from the modeled cases, economic analysis was conducted, which indicated that the ICD in conjunction with sliding sleeves (SSD) was the best option.\u0000 Two candidate wells were selected to cover the variation of reservoir characteristics: one well representing the heterogeneous part of the reservoir with high-density of faults, fractures and kurst, and another one representing the relatively homogenous part of the reservoir suffering from heel to toe effect. A multidisciplinary implementation team was set up to align all stakeholders on subsurface requirements, following up the completion design, coordinating material procurement and logistics for mobilizations, daily drilling operations follow-up, real-time logging data interpretations and completion design adjustment.\u0000 Evaluation of the two pilots’ results based on predefined KPIs during the study, exceeded overall expectations.","PeriodicalId":10981,"journal":{"name":"Day 4 Thu, November 18, 2021","volume":"83 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74269350","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
P. Bimastianto, S. Khambete, Hamdan Mohamed Alsaadi, S. A. Al Ameri, Erwan Couzigou, A. Al-Marzouqi, F. A. Ameri, Said Aboulaban, Husam Khater, P. Herve
This project used predictive analytics and machine learning-based modeling to detect drilling anomalies, namely stuck pipe events. Analysis focused on historical drilling data and real-time operational data to address the limitations of physics-based modeling. This project was designed to enable drilling crews to minimize downtime and non-productive time through real-time anomaly management. The solution used data science techniques to overcome data consistency/quality issues and flag drilling anomalies leading to a stuck pipe event. Predictive machine learning models were deployed across seven wells in different fields. The models analyzed both historical and real-time data across various data channels to identify anomalies (difficulties that impact non-productive time). The modeling approach mimicked the behavior of drillers using surface parameters. Small deviations from normal behavior were identified based on combinations of surface parameters, and automated machine learning was used to accelerate and optimize the modeling process. The output was a risk score that flags deviations in rig surface parameters. During the development phase, multiple data science approaches were attempted to monitor the overall health of the drilling process. They analyzed both historical and real-time data from torque, hole depth and deviation, standpipe pressure, and various other data channels. The models detected drilling anomalies with a harmonic model accuracy of 80% and produced valid alerts on 96% of stuck pipe and tight hole events. The average forewarning was two hours. This allowed personnel ample time to make corrections before stuck pipe events could occur. This also enabled the drilling operator to save the company upwards of millions of dollars in drilling costs and downtime. This project introduced novel data aggregation and deep learning-based normal behavior modeling methods. It demonstrates the benefits of adopting predictive analytics and machine learning in drilling operations. The approach enabled operators to mitigate data issues and demonstrate real-time, high-frequency and high-accuracy predictions. As a result, the operator was able to significantly reduce non-productive time.
{"title":"Application of Artificial Intelligence and Machine Learning to Detect Drilling Anomalies Leading to Stuck Pipe Incidents","authors":"P. Bimastianto, S. Khambete, Hamdan Mohamed Alsaadi, S. A. Al Ameri, Erwan Couzigou, A. Al-Marzouqi, F. A. Ameri, Said Aboulaban, Husam Khater, P. Herve","doi":"10.2118/207987-ms","DOIUrl":"https://doi.org/10.2118/207987-ms","url":null,"abstract":"\u0000 This project used predictive analytics and machine learning-based modeling to detect drilling anomalies, namely stuck pipe events. Analysis focused on historical drilling data and real-time operational data to address the limitations of physics-based modeling. This project was designed to enable drilling crews to minimize downtime and non-productive time through real-time anomaly management.\u0000 The solution used data science techniques to overcome data consistency/quality issues and flag drilling anomalies leading to a stuck pipe event. Predictive machine learning models were deployed across seven wells in different fields. The models analyzed both historical and real-time data across various data channels to identify anomalies (difficulties that impact non-productive time). The modeling approach mimicked the behavior of drillers using surface parameters. Small deviations from normal behavior were identified based on combinations of surface parameters, and automated machine learning was used to accelerate and optimize the modeling process. The output was a risk score that flags deviations in rig surface parameters.\u0000 During the development phase, multiple data science approaches were attempted to monitor the overall health of the drilling process. They analyzed both historical and real-time data from torque, hole depth and deviation, standpipe pressure, and various other data channels.\u0000 The models detected drilling anomalies with a harmonic model accuracy of 80% and produced valid alerts on 96% of stuck pipe and tight hole events. The average forewarning was two hours. This allowed personnel ample time to make corrections before stuck pipe events could occur. This also enabled the drilling operator to save the company upwards of millions of dollars in drilling costs and downtime.\u0000 This project introduced novel data aggregation and deep learning-based normal behavior modeling methods. It demonstrates the benefits of adopting predictive analytics and machine learning in drilling operations. The approach enabled operators to mitigate data issues and demonstrate real-time, high-frequency and high-accuracy predictions. As a result, the operator was able to significantly reduce non-productive time.","PeriodicalId":10981,"journal":{"name":"Day 4 Thu, November 18, 2021","volume":"21 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74951079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Zurailey Bin Baharum, M. Rourke, A. Muhadjir, W. Andono, Eva Sarah Binti Zakaria, Suzie Binti Hamzah, Noor Rohaellizza Binti Hademi, Nurul Aida Binti Hamdan
Well operators often face various technical challenges when intervening and repairing older, mature field wells. The most common problem associated with aging wells are tubing and casing integrity. Uncertain sources of downhole leaks and data ambiguity often lead to incorrect diagnostics that can hinder repair work or even contribute to additional or worsened integrity issues. Operators continuously challenge service companies and technology providers to drive innovation. One such challenge is in finding efficient and comprehensive integrity diagnostics for dual-string wells. A basic and general diagnostic method to verify well integrity in dual-string wells involves setting plugs in the long and short strings and pressure testing the tubings. These operations are generally time consuming, and the test data does not usually pinpoint the location of the leak, if any. Since 2016 a new diagnostic solution for this challenge has been implemented using a slickline-deployed passive acoustic logging technique. Carefully designed intervention planning, combined with efficient data acquisition, led to significant time saving and improved data quality. A more complete assessment of the integrity of both strings is now more frequent and often necessary, while challenging the conventional thinking of having to assess the lower string only while assuming the upper string is in good condition. However, investigating dual-string integrity with uncertainty on the source of leak, restrictions on facilities and limitations on surveillance time will often waste more time and money if not approached carefully. This paper discusses two case studies, including a dual-string oil producer in the South China Sea that had sustained pressure in production casing annulus. The well operator initially considered that the long string had an integrity issue, while the short string did not, based on their surface-based annulus pressure diagnostics. Consequently, the operator decided to diagnose only the long string. The passive acoustic memory tool. combined with a fast-response temperature and spinner used for the diagnosis, identified a possible short string leak while logging through the long string. This result clearly demonstrated that surface analyses can be misleading, and a comprehensive downhole diagnostic should be the recommended method to identify leaks, especially in dual-string completions. This well operator has completed more than 100 integrity diagnostic runs in the last five years. The passive acoustic diagnostic interventions have resulted in an average 50-percent time saving compared to legacy methods, and data analysis results have led to significant improvements in well productivity.
{"title":"Efficient and Comprehensive Integrity Diagnostics for Dual Completion String Wells, Using Spectral Noise Analyzer Tool","authors":"Zurailey Bin Baharum, M. Rourke, A. Muhadjir, W. Andono, Eva Sarah Binti Zakaria, Suzie Binti Hamzah, Noor Rohaellizza Binti Hademi, Nurul Aida Binti Hamdan","doi":"10.2118/207814-ms","DOIUrl":"https://doi.org/10.2118/207814-ms","url":null,"abstract":"\u0000 Well operators often face various technical challenges when intervening and repairing older, mature field wells. The most common problem associated with aging wells are tubing and casing integrity. Uncertain sources of downhole leaks and data ambiguity often lead to incorrect diagnostics that can hinder repair work or even contribute to additional or worsened integrity issues. Operators continuously challenge service companies and technology providers to drive innovation. One such challenge is in finding efficient and comprehensive integrity diagnostics for dual-string wells.\u0000 A basic and general diagnostic method to verify well integrity in dual-string wells involves setting plugs in the long and short strings and pressure testing the tubings. These operations are generally time consuming, and the test data does not usually pinpoint the location of the leak, if any. Since 2016 a new diagnostic solution for this challenge has been implemented using a slickline-deployed passive acoustic logging technique. Carefully designed intervention planning, combined with efficient data acquisition, led to significant time saving and improved data quality.\u0000 A more complete assessment of the integrity of both strings is now more frequent and often necessary, while challenging the conventional thinking of having to assess the lower string only while assuming the upper string is in good condition. However, investigating dual-string integrity with uncertainty on the source of leak, restrictions on facilities and limitations on surveillance time will often waste more time and money if not approached carefully.\u0000 This paper discusses two case studies, including a dual-string oil producer in the South China Sea that had sustained pressure in production casing annulus. The well operator initially considered that the long string had an integrity issue, while the short string did not, based on their surface-based annulus pressure diagnostics. Consequently, the operator decided to diagnose only the long string. The passive acoustic memory tool. combined with a fast-response temperature and spinner used for the diagnosis, identified a possible short string leak while logging through the long string. This result clearly demonstrated that surface analyses can be misleading, and a comprehensive downhole diagnostic should be the recommended method to identify leaks, especially in dual-string completions.\u0000 This well operator has completed more than 100 integrity diagnostic runs in the last five years. The passive acoustic diagnostic interventions have resulted in an average 50-percent time saving compared to legacy methods, and data analysis results have led to significant improvements in well productivity.","PeriodicalId":10981,"journal":{"name":"Day 4 Thu, November 18, 2021","volume":"3 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82115076","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Epoupa Mengou Joseph, G. Chiara, Alessi Andrea, Terenzi Andrea, Vecchione Michela, Binaschi Marco, Di Salvo Salvatore R, N. Anglani
Energy storage is entering in the energy distribution supply chain due to the global goal of achieving carbon neutrality in human activities, especially those related to energy production. Renewable energies integrated with energy storage play an important role in this framework [1]. The purpose of the study is to evaluate through simulations the impact of new renewable energy technologies in a microgrid to minimize fossil fuels consumption. The case study considers a hybrid microgrid including: a gas microturbine, organic photovoltaic panels (OPV), a point absorber wave energy converter, a vanadium redox flow battery and a load. The microgrid is placed in an offshore hydrocarbon plant near the northern coast of Australia. Firstly, Australian meteorological data have been studied and three seasons identified (named ST1, ST2 and ST3). Then a correlation has been established between meteorological data and OPVs performances, analyzing data collected on OPVs panels installed. This relationship has been used to assess OPVs potential production at the site of interest. Similar correlation was made between the performances of a wave energy converter placed in the Adriatic Sea and the wave power matrix, to determine a suitable power data reference for the potential production of a wave energy converter to the Australian coast. Finally, the behavior of the microgrid was modeled. Different scenarios have been considered and the best one with optimal meteorological conditions enables lead to drastically decrease of the use of gas micro turbine resulting in lowest CO2 emissions. In fact, the consumption of natural gas has been summarized as follow: Season 1 (ST1): during this season the load is entirely fed by the renewable sources and by the battery, with consequent zeroing of the daily consumption of natural gas. Season 2(ST2): the battery is charged from 09:00am to 07:00pm with the exceeding power from the renewable sources. This configuration involves a daily natural gas consumption of 10.73 Sm3/d, which is equivalent to 987.16 Sm3/ ST2 (accounting for 92 days). Season 3(ST3): the battery is charged from 09:00am to 07:00pm with the exceeding power from the renewable sources. This configuration involves a daily natural gas consumption of 6.58 Sm3/d, which is equivalent to 1006.74 Sm3/ ST3 (accounting for 120 days). The avoided CO2 emissions are 2062 tons/year. This case study showed how the new renewable technologies, such as organic photovoltaics and wave energy converter, coupled with a long duration storage system, can be conveniently applied in sites with limited space for the decarbonization purpose of an offshore platform.
{"title":"A Case-Study for the Reduction of CO2 Emissions in an Offshore Platform by the Exploitation of Renewable Energy Sources Through Innovative Technologies Coupled with Energy Storage","authors":"Epoupa Mengou Joseph, G. Chiara, Alessi Andrea, Terenzi Andrea, Vecchione Michela, Binaschi Marco, Di Salvo Salvatore R, N. Anglani","doi":"10.2118/207864-ms","DOIUrl":"https://doi.org/10.2118/207864-ms","url":null,"abstract":"\u0000 Energy storage is entering in the energy distribution supply chain due to the global goal of achieving carbon neutrality in human activities, especially those related to energy production. Renewable energies integrated with energy storage play an important role in this framework [1].\u0000 The purpose of the study is to evaluate through simulations the impact of new renewable energy technologies in a microgrid to minimize fossil fuels consumption. The case study considers a hybrid microgrid including: a gas microturbine, organic photovoltaic panels (OPV), a point absorber wave energy converter, a vanadium redox flow battery and a load. The microgrid is placed in an offshore hydrocarbon plant near the northern coast of Australia.\u0000 Firstly, Australian meteorological data have been studied and three seasons identified (named ST1, ST2 and ST3). Then a correlation has been established between meteorological data and OPVs performances, analyzing data collected on OPVs panels installed. This relationship has been used to assess OPVs potential production at the site of interest. Similar correlation was made between the performances of a wave energy converter placed in the Adriatic Sea and the wave power matrix, to determine a suitable power data reference for the potential production of a wave energy converter to the Australian coast.\u0000 Finally, the behavior of the microgrid was modeled.\u0000 Different scenarios have been considered and the best one with optimal meteorological conditions enables lead to drastically decrease of the use of gas micro turbine resulting in lowest CO2 emissions. In fact, the consumption of natural gas has been summarized as follow:\u0000 Season 1 (ST1): during this season the load is entirely fed by the renewable sources and by the battery, with consequent zeroing of the daily consumption of natural gas.\u0000 Season 2(ST2): the battery is charged from 09:00am to 07:00pm with the exceeding power from the renewable sources. This configuration involves a daily natural gas consumption of 10.73 Sm3/d, which is equivalent to 987.16 Sm3/ ST2 (accounting for 92 days).\u0000 Season 3(ST3): the battery is charged from 09:00am to 07:00pm with the exceeding power from the renewable sources. This configuration involves a daily natural gas consumption of 6.58 Sm3/d, which is equivalent to 1006.74 Sm3/ ST3 (accounting for 120 days). The avoided CO2 emissions are 2062 tons/year.\u0000 This case study showed how the new renewable technologies, such as organic photovoltaics and wave energy converter, coupled with a long duration storage system, can be conveniently applied in sites with limited space for the decarbonization purpose of an offshore platform.","PeriodicalId":10981,"journal":{"name":"Day 4 Thu, November 18, 2021","volume":"2011 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82594611","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Long-xin Li, Yuan Zhou, Limin Li, J. Tinnin, Xian Peng, C. Cranfield, Yu Luo, R. Guises, Yuchao Zhao, Xia Wang, F. Gui, Christopher Burns, Huijuan Yu, Ahmad Reza Younessi Sinaki
Underground gas storage (UGS) will be key to addressing supply and demand dynamics as natural gas consumption grows during the coming decades in response to cleaner energy initiatives. The XGS facility began UGS operations in a depleted gas field located in SW China in 2013. Following this initial period of utilization, the site was reassessed to safely increase deliverability during winter months to meet future peak gas demand. The XGS field is located in a high tectonic stress region and has a structurally complex and highly faulted geological setting. The carbonate reservoir is heterogeneous and naturally fractured. Initial assessment steps involved determination of maximum storage capacity and estimation of required working gas and cushion gas volumes using fully integrated geological, geophysical, petrophysical frameworks. Geomechanical modeling was embedded into the analysis to determine the long-term impact inferred by cyclical variations of pressures on the reservoir performance and cap rock containment and evaluate both safe operating pressure limits and monitoring requirements. The coupling of complex reservoir and geomechanical parameters was required to create a dynamic model within the stress regime that could be history-matched to the early gas depletion phase and subsequent gas storage cycles. Such a holistic approach allows the operator to optimize the number of wells, their placement, trajectories and completion designs to ensure safe and efficient operations and develop strategies for increasing withdrawal rates to meet anticipated future demand. Additionally, tight integration of subsurface understanding with surface requirements, such as turbo-compressors, is critical to meet the UGS designed performance and deliverability objectives and ensure sufficient flexibility to optimize the facility usage. A further important task of the final phase of UGS facilities design involves enablement of sustainable operation through a Storage Optimization Plan. The results of the analyses serve as a basis for the design of this plan, in combination with fit-for-purpose surveillance systems of the reservoir and cap-rock seal recording pressure, rock deformation and seismicity in real time, along with regular wellbore inspection.
{"title":"Underground Gas Storage Process Optimization Using Integrated Subsurface Characterization, Dynamic Modeling and Monitoring - A Case Study","authors":"Long-xin Li, Yuan Zhou, Limin Li, J. Tinnin, Xian Peng, C. Cranfield, Yu Luo, R. Guises, Yuchao Zhao, Xia Wang, F. Gui, Christopher Burns, Huijuan Yu, Ahmad Reza Younessi Sinaki","doi":"10.2118/207941-ms","DOIUrl":"https://doi.org/10.2118/207941-ms","url":null,"abstract":"\u0000 Underground gas storage (UGS) will be key to addressing supply and demand dynamics as natural gas consumption grows during the coming decades in response to cleaner energy initiatives. The XGS facility began UGS operations in a depleted gas field located in SW China in 2013. Following this initial period of utilization, the site was reassessed to safely increase deliverability during winter months to meet future peak gas demand.\u0000 The XGS field is located in a high tectonic stress region and has a structurally complex and highly faulted geological setting. The carbonate reservoir is heterogeneous and naturally fractured. Initial assessment steps involved determination of maximum storage capacity and estimation of required working gas and cushion gas volumes using fully integrated geological, geophysical, petrophysical frameworks. Geomechanical modeling was embedded into the analysis to determine the long-term impact inferred by cyclical variations of pressures on the reservoir performance and cap rock containment and evaluate both safe operating pressure limits and monitoring requirements.\u0000 The coupling of complex reservoir and geomechanical parameters was required to create a dynamic model within the stress regime that could be history-matched to the early gas depletion phase and subsequent gas storage cycles. Such a holistic approach allows the operator to optimize the number of wells, their placement, trajectories and completion designs to ensure safe and efficient operations and develop strategies for increasing withdrawal rates to meet anticipated future demand. Additionally, tight integration of subsurface understanding with surface requirements, such as turbo-compressors, is critical to meet the UGS designed performance and deliverability objectives and ensure sufficient flexibility to optimize the facility usage.\u0000 A further important task of the final phase of UGS facilities design involves enablement of sustainable operation through a Storage Optimization Plan. The results of the analyses serve as a basis for the design of this plan, in combination with fit-for-purpose surveillance systems of the reservoir and cap-rock seal recording pressure, rock deformation and seismicity in real time, along with regular wellbore inspection.","PeriodicalId":10981,"journal":{"name":"Day 4 Thu, November 18, 2021","volume":"19 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77360777","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Luai Alhamad, Basil M. Alfakher, Abdulla A. Alrustum, Sajjad Aldarweesh
Acidizing deep carbonate formations by Hydrochloric acid (HCl) is a complex task due to high reaction and corrosion rates. Mixing organic acids with HCl is a typical method to reduce the acid's reactivity and corrosivity. Lactic acid has not been investigated completely in the area of carbonate acidizing. Lactic acid has a dissociation constant similar to formic acid, which is approximately 10 times larger than acetic acid. Therefore, the objective of this work is to compare lactic/HCl blends with plain HCl and formic/HCl blends. Corrosion tests were conducted at high temperature on C-95 steel coupons to investigate associated corrosion damage. Coreflood tests were performed on Indiana limestone cores to mimic matrix acidizing treatment and to investigate amount of pore volumes required to breakthrough. All blends were prepared to be equivalent to 15 wt% (4.4 M) HCl for comparison. Lactic and formic acid concentrations were set to be (0.5 or 1 M), and HCl concentration was calculated as appropriate to reach a blend with strength of 4.4 M. In terms of corrosivity evaluation, blends of lactic and HCl acids showed a corrosion rate of up to 1.97 lb/ft2 at 300°F. The formic and HCl blend showed a corrosion rate of 1.68 lb/ft2 at the same temperature. The difference in corrosion rates between the two mixtures is due to molecular weight difference between lactic and formic acids. When both acids were prepared at 1 M, lactic acid blend required more HCl to be equivalent to 15 wt% HCl acid which was associated with an increase in corrosion rate. Coreflood results established acid efficiency curves for lactic/HCl acid blends. The curves highlighted the correlation between acid-core reactivity, injection rate, and dissolution pattern. Lactic/HCl blend was less reactive than formic/HCl mixture as the last required lower injection rate to obtain optimum pore volume to breakthrough at 300°F. Lactic/HCl blend was able to generate an optimum dissolution pattern as a dominant wormhole was shown on tested core plugs inlet face. This study expands the investigation of lactic acid utilization in carbonate acidizing. Major advantages rendered by using lactic acid with HCl include: (1) favorable dissolution pattern due to lactic acid being less reactive than HCl or formic acids, and (2) less corrosion rates comparing to HCl, that can reduce allocated costs for maintenance and replacements.
{"title":"Experimental Results to Design Lactic Acid for Carbonate Acidizing","authors":"Luai Alhamad, Basil M. Alfakher, Abdulla A. Alrustum, Sajjad Aldarweesh","doi":"10.2118/207273-ms","DOIUrl":"https://doi.org/10.2118/207273-ms","url":null,"abstract":"\u0000 Acidizing deep carbonate formations by Hydrochloric acid (HCl) is a complex task due to high reaction and corrosion rates. Mixing organic acids with HCl is a typical method to reduce the acid's reactivity and corrosivity. Lactic acid has not been investigated completely in the area of carbonate acidizing. Lactic acid has a dissociation constant similar to formic acid, which is approximately 10 times larger than acetic acid. Therefore, the objective of this work is to compare lactic/HCl blends with plain HCl and formic/HCl blends.\u0000 Corrosion tests were conducted at high temperature on C-95 steel coupons to investigate associated corrosion damage. Coreflood tests were performed on Indiana limestone cores to mimic matrix acidizing treatment and to investigate amount of pore volumes required to breakthrough. All blends were prepared to be equivalent to 15 wt% (4.4 M) HCl for comparison. Lactic and formic acid concentrations were set to be (0.5 or 1 M), and HCl concentration was calculated as appropriate to reach a blend with strength of 4.4 M.\u0000 In terms of corrosivity evaluation, blends of lactic and HCl acids showed a corrosion rate of up to 1.97 lb/ft2 at 300°F. The formic and HCl blend showed a corrosion rate of 1.68 lb/ft2 at the same temperature. The difference in corrosion rates between the two mixtures is due to molecular weight difference between lactic and formic acids. When both acids were prepared at 1 M, lactic acid blend required more HCl to be equivalent to 15 wt% HCl acid which was associated with an increase in corrosion rate. Coreflood results established acid efficiency curves for lactic/HCl acid blends. The curves highlighted the correlation between acid-core reactivity, injection rate, and dissolution pattern. Lactic/HCl blend was less reactive than formic/HCl mixture as the last required lower injection rate to obtain optimum pore volume to breakthrough at 300°F. Lactic/HCl blend was able to generate an optimum dissolution pattern as a dominant wormhole was shown on tested core plugs inlet face.\u0000 This study expands the investigation of lactic acid utilization in carbonate acidizing. Major advantages rendered by using lactic acid with HCl include: (1) favorable dissolution pattern due to lactic acid being less reactive than HCl or formic acids, and (2) less corrosion rates comparing to HCl, that can reduce allocated costs for maintenance and replacements.","PeriodicalId":10981,"journal":{"name":"Day 4 Thu, November 18, 2021","volume":"49 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88683447","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
K. Manikonda, A. Hasan, C. Obi, R. Islam, Ahmad K. Sleiti, M. Abdelrazeq, M. A. Rahman
This research aims to identify the best machine learning (ML) classification techniques for classifying the flow regimes in vertical gas-liquid two-phase flow. Two-phase flow regime identification is crucial for many operations in the oil and gas industry. Processes such as flow assurance, well control, and production rely heavily on accurate identification of flow regimes for their respective systems' smooth functioning. The primary motivation for the proposed ML classification algorithm selection processes was drilling and well control applications in Deepwater wells. The process started with vertical two-phase flow data collection from literature and two different flow loops. One, a 140 ft. tall vertical flow loop with a centralized inner metal pipe and a larger outer acrylic pipe. Second, an 18-ft long flow loop, also with a centralized, inner metal drill pipe. After extensive experimental and historical data collection, supervised and unsupervised ML classification models such as Multi-class Support vector machine (MCSVM), K-Nearest Neighbor Classifier (KNN), K-means clustering, and hierarchical clustering were fit on the datasets to separate the different flow regions. The next step was fine-tuning the models' parameters and kernels. The last step was to compare the different combinations of models and refining techniques for the best prediction accuracy and the least variance. Among the different models and combinations with refining techniques, the 5- fold cross-validated KNN algorithm, with 37 neighbors, gave the optimal solution with a 98% classification accuracy on the test data. The KNN model distinguished five major, distinct flow regions for the dataset and a few minor regions. These five regions were bubbly flow, slug flow, churn flow, annular flow, and intermittent flow. The KNN-generated flow regime maps matched well with those presented by Hasan and Kabir (2018). The MCSVM model produced visually similar flow maps to KNN but significantly underperformed them in prediction accuracy. The MCSVM training errors ranged between 50% - 60% at normal parameter values and costs but went up to 99% at abnormally high values. However, their prediction accuracy was below 50% even at these highly overfitted conditions. In unsupervised models, both clustering techniques pointed to an optimal cluster number between 10 and 15, consistent with the 14 we have in the dataset. Within the context of gas kicks and well control, a well-trained, reliable two-phase flow region classification algorithm offers many advantages. When trained with well-specific data, it can act as a black box for flow regime identification and subsequent well-control measure decisions for the well. Further advancements with more robust statistical training techniques can render these algorithms as a basis for well-control measures in drilling automation software. On a broader scale, these classification techniques have many applications in flow assurance, production, and
{"title":"Application of Machine Learning Classification Algorithms for Two-Phase Gas-Liquid Flow Regime Identification","authors":"K. Manikonda, A. Hasan, C. Obi, R. Islam, Ahmad K. Sleiti, M. Abdelrazeq, M. A. Rahman","doi":"10.2118/208214-ms","DOIUrl":"https://doi.org/10.2118/208214-ms","url":null,"abstract":"\u0000 This research aims to identify the best machine learning (ML) classification techniques for classifying the flow regimes in vertical gas-liquid two-phase flow. Two-phase flow regime identification is crucial for many operations in the oil and gas industry. Processes such as flow assurance, well control, and production rely heavily on accurate identification of flow regimes for their respective systems' smooth functioning. The primary motivation for the proposed ML classification algorithm selection processes was drilling and well control applications in Deepwater wells.\u0000 The process started with vertical two-phase flow data collection from literature and two different flow loops. One, a 140 ft. tall vertical flow loop with a centralized inner metal pipe and a larger outer acrylic pipe. Second, an 18-ft long flow loop, also with a centralized, inner metal drill pipe. After extensive experimental and historical data collection, supervised and unsupervised ML classification models such as Multi-class Support vector machine (MCSVM), K-Nearest Neighbor Classifier (KNN), K-means clustering, and hierarchical clustering were fit on the datasets to separate the different flow regions. The next step was fine-tuning the models' parameters and kernels. The last step was to compare the different combinations of models and refining techniques for the best prediction accuracy and the least variance.\u0000 Among the different models and combinations with refining techniques, the 5- fold cross-validated KNN algorithm, with 37 neighbors, gave the optimal solution with a 98% classification accuracy on the test data. The KNN model distinguished five major, distinct flow regions for the dataset and a few minor regions. These five regions were bubbly flow, slug flow, churn flow, annular flow, and intermittent flow. The KNN-generated flow regime maps matched well with those presented by Hasan and Kabir (2018).\u0000 The MCSVM model produced visually similar flow maps to KNN but significantly underperformed them in prediction accuracy. The MCSVM training errors ranged between 50% - 60% at normal parameter values and costs but went up to 99% at abnormally high values. However, their prediction accuracy was below 50% even at these highly overfitted conditions. In unsupervised models, both clustering techniques pointed to an optimal cluster number between 10 and 15, consistent with the 14 we have in the dataset.\u0000 Within the context of gas kicks and well control, a well-trained, reliable two-phase flow region classification algorithm offers many advantages. When trained with well-specific data, it can act as a black box for flow regime identification and subsequent well-control measure decisions for the well. Further advancements with more robust statistical training techniques can render these algorithms as a basis for well-control measures in drilling automation software. On a broader scale, these classification techniques have many applications in flow assurance, production, and","PeriodicalId":10981,"journal":{"name":"Day 4 Thu, November 18, 2021","volume":"218 2 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90754540","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
V. Suleymanov, Abdulhamid Almumtin, G. Glatz, J. Dvorkin
Generated by the propagation of sound waves, seismic reflections are essentially the reflections at the interface between various subsurface formations. Traditionally, these reflections are interpreted in a qualitative way by mapping subsurface geology without quantifying the rock properties inside the strata, namely the porosity, mineralogy, and pore fluid. This study aims to conduct the needed quantitative interpretation by the means of rock physics to establish the relation between rock elastic and petrophysical properties for reservoir characterization. We conduct rock physics diagnostics to find a theoretical rock physics model relevant to the data by examining the wireline data from a clastic depositional environment associated with a tight gas sandstone in the Continental US. First, we conduct the rock physics diagnostics by using theoretical fluid substitution to establish the relevant rock physics models. Once these models are determined, we theoretically vary the thickness of the intervals, the pore fluid, as well as the porosity and mineralogy to generate geologically plausible pseudo-scenarios. Finally, Zoeppritz (1919) equations are exploited to obtain the expected amplitude versus offset (AVO) and the gradient versus intercept curves of these scenarios. The relationship between elastic and petrophysical properties was established using forward seismic modeling. Several theoretical rock physics models, namely Raymer-Dvorkin, soft-sand, stiff-sand, and constant-cement models were applied to the wireline data under examination. The modeling assumes that only two minerals are present: quartz and clay. The appropriate rock physics model appears to be constant-cement model with a high coordination number. The result is a seismic reflection catalogue that can serve as a field guide for interpreting real seismic reflections, as well as to determine the seismic visibility of the variations in the reservoir geometry, the pore fluid, and the porosity. The obtained reservoir properties may be extrapolated to prospects away from the well control to consider certain what-if scenarios like plausible lithology or fluid variations. This enables building of a catalogue of synthetic seismic reflections of rock properties to be used by the interpreter as a field guide relating seismic data to volumetric reservoir properties.
{"title":"Seismic Reflections of Rock Properties in a Clastic Environment","authors":"V. Suleymanov, Abdulhamid Almumtin, G. Glatz, J. Dvorkin","doi":"10.2118/207808-ms","DOIUrl":"https://doi.org/10.2118/207808-ms","url":null,"abstract":"\u0000 Generated by the propagation of sound waves, seismic reflections are essentially the reflections at the interface between various subsurface formations. Traditionally, these reflections are interpreted in a qualitative way by mapping subsurface geology without quantifying the rock properties inside the strata, namely the porosity, mineralogy, and pore fluid. This study aims to conduct the needed quantitative interpretation by the means of rock physics to establish the relation between rock elastic and petrophysical properties for reservoir characterization.\u0000 We conduct rock physics diagnostics to find a theoretical rock physics model relevant to the data by examining the wireline data from a clastic depositional environment associated with a tight gas sandstone in the Continental US. First, we conduct the rock physics diagnostics by using theoretical fluid substitution to establish the relevant rock physics models. Once these models are determined, we theoretically vary the thickness of the intervals, the pore fluid, as well as the porosity and mineralogy to generate geologically plausible pseudo-scenarios. Finally, Zoeppritz (1919) equations are exploited to obtain the expected amplitude versus offset (AVO) and the gradient versus intercept curves of these scenarios.\u0000 The relationship between elastic and petrophysical properties was established using forward seismic modeling. Several theoretical rock physics models, namely Raymer-Dvorkin, soft-sand, stiff-sand, and constant-cement models were applied to the wireline data under examination. The modeling assumes that only two minerals are present: quartz and clay. The appropriate rock physics model appears to be constant-cement model with a high coordination number. The result is a seismic reflection catalogue that can serve as a field guide for interpreting real seismic reflections, as well as to determine the seismic visibility of the variations in the reservoir geometry, the pore fluid, and the porosity.\u0000 The obtained reservoir properties may be extrapolated to prospects away from the well control to consider certain what-if scenarios like plausible lithology or fluid variations. This enables building of a catalogue of synthetic seismic reflections of rock properties to be used by the interpreter as a field guide relating seismic data to volumetric reservoir properties.","PeriodicalId":10981,"journal":{"name":"Day 4 Thu, November 18, 2021","volume":"143 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91451469","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
I. Mitrea, R. Cataraiani, M. Banu, S. Shirzadi, W. Renkema, O. Hausberger, M. Morosini, G. Grubac
This Upper Cretaceous reservoir, a tight reservoir dominated by silt, marl, argillaceous limestone and conglomerates in Black Sea Histria block, is the dominant of three oil-producing reservoirs in Histria Block. The other two, Albian and Eocene, are depleted, and not the focus of field re-development. This paper addresses the challenges and opportunities that were faced during the re-development process in this reservoir such as depletion, low productivity areas, lithology, seismic resolution, and stimulation effectiveness. Historically, production from Upper Cretaceous wells could not justify the economic life of the asset. As new fracturing technology evolved in recent years, the re-development focused on replacing old, vertical/deviated one-stage stimulations low producing wells with horizontal, multi-stage hydraulic fractured wells. The project team integrated various disciplines and approaches by re-processing old seismic to improve resolution and signal, integrating sedimentology studies using cores, XRF, XRD and thin section analysis with petrophysical evaluation and quantitative geophysical analyses, which then will provide properties for geological and geomechanical models to optimize well planning and fracture placement. Seven wells drilled since end of 2017 to mid-2021 have demonstrated the value of integration and proper planning in development of a mature field with existing depletion. Optimizing the well and fracture placement with respect to depletion in existing wells resulted in accessing areas with original reservoir pressure, not effectively drained by old wells. Integrating the well production performance with tracer results from each fractured stage, and NMR/Acoustic images from logs enhanced the understanding of the impact of lithofacies on stimulation. This has allowed better assessment and prediction of well performance, ultimately improving well placement and stimulation design. The example from this paper highlights the value of the integrating seismic reprocessing, attribute analysis, production technology, sedimentology, cuttings analysis and quantitative rock physics in characterizing the heterogeneity of the reservoir, which ultimately contributed to "sweet spot" targeting in a depleted reservoir with existing producers and deeper understanding of the development potential in Upper Cretaceous. The 2017-2021 wells contribute to more than 30 percent of the total oil production in the asset and reverse the decline in oil production. In addition, these wells have two to four times higher initial rates because of larger effective drainage area than a single fracture well. Three areas of novelty are highlighted in this paper. The application of acoustic image/NMR logging to identify lithofacies and optimize fracturing strategy in horizontal laterals. The tracers analysis of hydraulic fracture performance and integration with seismic and petrophysical analysis to categorize the productivity with rock types. The opti
{"title":"Development of Tight Upper Cretaceous Reservoir in Offshore Black Sea Adds Life to a Mature Asset","authors":"I. Mitrea, R. Cataraiani, M. Banu, S. Shirzadi, W. Renkema, O. Hausberger, M. Morosini, G. Grubac","doi":"10.2118/207428-ms","DOIUrl":"https://doi.org/10.2118/207428-ms","url":null,"abstract":"\u0000 This Upper Cretaceous reservoir, a tight reservoir dominated by silt, marl, argillaceous limestone and conglomerates in Black Sea Histria block, is the dominant of three oil-producing reservoirs in Histria Block. The other two, Albian and Eocene, are depleted, and not the focus of field re-development. This paper addresses the challenges and opportunities that were faced during the re-development process in this reservoir such as depletion, low productivity areas, lithology, seismic resolution, and stimulation effectiveness.\u0000 Historically, production from Upper Cretaceous wells could not justify the economic life of the asset. As new fracturing technology evolved in recent years, the re-development focused on replacing old, vertical/deviated one-stage stimulations low producing wells with horizontal, multi-stage hydraulic fractured wells. The project team integrated various disciplines and approaches by re-processing old seismic to improve resolution and signal, integrating sedimentology studies using cores, XRF, XRD and thin section analysis with petrophysical evaluation and quantitative geophysical analyses, which then will provide properties for geological and geomechanical models to optimize well planning and fracture placement.\u0000 Seven wells drilled since end of 2017 to mid-2021 have demonstrated the value of integration and proper planning in development of a mature field with existing depletion.\u0000 Optimizing the well and fracture placement with respect to depletion in existing wells resulted in accessing areas with original reservoir pressure, not effectively drained by old wells. Integrating the well production performance with tracer results from each fractured stage, and NMR/Acoustic images from logs enhanced the understanding of the impact of lithofacies on stimulation. This has allowed better assessment and prediction of well performance, ultimately improving well placement and stimulation design. The example from this paper highlights the value of the integrating seismic reprocessing, attribute analysis, production technology, sedimentology, cuttings analysis and quantitative rock physics in characterizing the heterogeneity of the reservoir, which ultimately contributed to \"sweet spot\" targeting in a depleted reservoir with existing producers and deeper understanding of the development potential in Upper Cretaceous.\u0000 The 2017-2021 wells contribute to more than 30 percent of the total oil production in the asset and reverse the decline in oil production. In addition, these wells have two to four times higher initial rates because of larger effective drainage area than a single fracture well.\u0000 Three areas of novelty are highlighted in this paper. The application of acoustic image/NMR logging to identify lithofacies and optimize fracturing strategy in horizontal laterals. The tracers analysis of hydraulic fracture performance and integration with seismic and petrophysical analysis to categorize the productivity with rock types. The opti","PeriodicalId":10981,"journal":{"name":"Day 4 Thu, November 18, 2021","volume":"28 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84899144","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Antonio Andrade Marin, Issa Al Balushi, Adnan Al Ghadani, Hassana Al Abri, Abdullah Khalfan Said Al Zaabi, K. Dhuhli, I. Al Hadhrami, Saif Hamed Al Hinai, Fahad Masoud Al Aufi, Aziz Ali Al Bimani, Rahul Gala, Eduardo Marín, Nitish Kumar, Apurv Raj
Failure Prediction in Oil and Gas Artificial Lift Systems is materializing through the implementation of advanced analytics driven by physics-based models. During the Phase I of this project, two early failure prediction machine learning models were trained offline with historical data and evaluated through a blind test. The next challenge, Phase II, is to operationalize these models on Real-Time and re-assess their accuracy, precision and early prediction (in days) while having the assets focusing on either extending the runtime through optimization, chemical injection, etc. or proactive pump replacement (PPR) for high producers wells with triggered early prediction alarms. The paper details Phase II of live prediction for two assets consisting of 740 wells to enable data-driven insights in engineers’ daily workflow. In Phase I, a collaboration between SMEs and Data Scientists was established to build two failure prediction models for Electrical Submersible Pumps (ESP) using historical data that could identify failure prone wells along with the component at risk with high precision. Phase II entails the development of a Real-Time scoring pipeline to avail daily insights from this model for live wells. To achieve this, PDO leveraged its Digital Infrastructure for extraction of high-resolution measured data for 750 wells daily. A Well Management System (WMS) automatically sustains physics-based ESP models to calculate engineering variables from nodal analysis. Measured and engineered data are sampled, and referencing learnt patterns, the machine learning algorithm (MLA) estimates the probability of failure based on a daily rolling data window. An Exception Based Surveillance (EBS) system tracks well failure probability and highlights affected wells based on business logic. A visualization is developed to facilitate EBS interpretation. All the above steps are automated and synchronized among data historian, WMS and EBS System to operate on a daily schedule. From the Asset, at each highlighted exception, a focus team of well owners and SME initiate a review to correlate the failure probability with ESP signatures to validate the alarm. Aided by physics-based well models, action is directed either towards a) optimization, b) troubleshooting or c) proactive pump replacement in case of inevitable failure conditions. This workflow enables IT infrastructure and Asset readiness to benefit from various modeling initiatives in subsequent phases. Live Implementation of Exceptions from Predictive Analytics is an effective complement to well owners for prioritization of well reviews. Based on alarm validity, risk of failure and underperformance – optimizations, PPRs or workover scheduling are performed with reliability. This methodology would enable a Phase III of scaling up in Real-Time with growing assets wherethe system would be periodically retrained on True Negatives and maintained automatically with minimum manual intervention. It is experienced that
{"title":"Real Time Implementation of ESP Predictive Analytics - Towards Value Realization from Data Science","authors":"Antonio Andrade Marin, Issa Al Balushi, Adnan Al Ghadani, Hassana Al Abri, Abdullah Khalfan Said Al Zaabi, K. Dhuhli, I. Al Hadhrami, Saif Hamed Al Hinai, Fahad Masoud Al Aufi, Aziz Ali Al Bimani, Rahul Gala, Eduardo Marín, Nitish Kumar, Apurv Raj","doi":"10.2118/207550-ms","DOIUrl":"https://doi.org/10.2118/207550-ms","url":null,"abstract":"Failure Prediction in Oil and Gas Artificial Lift Systems is materializing through the implementation of advanced analytics driven by physics-based models. During the Phase I of this project, two early failure prediction machine learning models were trained offline with historical data and evaluated through a blind test. The next challenge, Phase II, is to operationalize these models on Real-Time and re-assess their accuracy, precision and early prediction (in days) while having the assets focusing on either extending the runtime through optimization, chemical injection, etc. or proactive pump replacement (PPR) for high producers wells with triggered early prediction alarms. The paper details Phase II of live prediction for two assets consisting of 740 wells to enable data-driven insights in engineers’ daily workflow.\u0000 In Phase I, a collaboration between SMEs and Data Scientists was established to build two failure prediction models for Electrical Submersible Pumps (ESP) using historical data that could identify failure prone wells along with the component at risk with high precision. Phase II entails the development of a Real-Time scoring pipeline to avail daily insights from this model for live wells. To achieve this, PDO leveraged its Digital Infrastructure for extraction of high-resolution measured data for 750 wells daily. A Well Management System (WMS) automatically sustains physics-based ESP models to calculate engineering variables from nodal analysis. Measured and engineered data are sampled, and referencing learnt patterns, the machine learning algorithm (MLA) estimates the probability of failure based on a daily rolling data window. An Exception Based Surveillance (EBS) system tracks well failure probability and highlights affected wells based on business logic. A visualization is developed to facilitate EBS interpretation. All the above steps are automated and synchronized among data historian, WMS and EBS System to operate on a daily schedule. From the Asset, at each highlighted exception, a focus team of well owners and SME initiate a review to correlate the failure probability with ESP signatures to validate the alarm. Aided by physics-based well models, action is directed either towards a) optimization, b) troubleshooting or c) proactive pump replacement in case of inevitable failure conditions. This workflow enables IT infrastructure and Asset readiness to benefit from various modeling initiatives in subsequent phases.\u0000 Live Implementation of Exceptions from Predictive Analytics is an effective complement to well owners for prioritization of well reviews. Based on alarm validity, risk of failure and underperformance – optimizations, PPRs or workover scheduling are performed with reliability. This methodology would enable a Phase III of scaling up in Real-Time with growing assets wherethe system would be periodically retrained on True Negatives and maintained automatically with minimum manual intervention.\u0000 It is experienced that ","PeriodicalId":10981,"journal":{"name":"Day 4 Thu, November 18, 2021","volume":"12 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2021-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72766667","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：