In-situ combustion (ISC) involves compression and injection of air into heavy/extraheavy oil reservoirs for enhancing production and recovery. Initially, ISC was very popular due to its high theoretical thermal efficiency, though more failures than successes in the 1990s made this process unpopular. It is a fact that it is now widely considered archaic. However, Suplacu de Barcau (Romania) and Balol-Santhal (India) ISC projects have brought the process back into focus. Performance of the Balol-Santhal-Bechraji over the last 25 years provides clarity to answer the question “Failure to enhance oil production and recovery by ISC: Myth or fact?” The author appreciates the views, decisions, and efforts of all global scientists/engineers/operators associated with the ISC process in the laboratory/field. Opinions and views presented in this paper are solely based on the author’s experience, which may be in line or may differ. The discovery of heavy oil northwest of the Cambay Basin, India, in the 1970s led to the initiation of research and development in thermal processes. The depth, rock, and fluid characteristics, drive mechanism, and semi-arid area led to the testing of ISC over steamflood in Balol. Laboratory findings are key to understanding the reaction kinetics of oil and process manifestations. Upgrading of oil is the key manifestation of ISC in the laboratory, but it is not seen in the field due to blending in long-distance displacement methodology. The involvement of laboratory personnel in design and surveillance plays an important role in the success of the project. Over the last 25 years, the Balol-Santhal ISC projects demonstrate the rejuvenation of declining fields with sustained enhanced oil production and an increase in recovery. Lessons of the Bechraji field indicate that process does not succeed in all reservoir settings. It is particularly suited to relatively clean, mobile heavy oil reservoirs with structural relief. Long-distance displacement of oil (vertical injector-vertical/horizontal producer spaced apart) is effective in a mobile oil reservoir. With low mobility oils, a short-distance oil displacement process using a pair of vertical injector and horizontal producer (horizontal well placed below the air injector) can be the preferred way for exploitation. This methodology has also the potential to capture upgraded oil. The process attracts more value when it is designed as operator friendly and flexible, integrating with gravity. Appropriate ignition types, continuous surveillance, maintaining optimum air injection rates, and re-engineering are important for the success of ISC. Success depends on the fabric and architecture of the reservoir, the way it is designed and implemented, and by integration of knowledge gained in the journey from laboratory to field with the process. It can be concluded that the perception of the ineffectiveness of ISC to enhance oil production and recovery from mobile heavy/extraheavy oil reservoi
原位燃烧(ISC)技术是将空气压缩并注入稠油/超稠油储层，以提高产量和采收率。最初，ISC由于其理论热效率高而非常受欢迎，尽管在20世纪90年代失败多于成功使该过程不受欢迎。事实上，它现在被广泛认为是过时的。然而，Suplacu de Barcau(罗马尼亚)和Balol-Santhal(印度)ISC项目使这一过程重新受到关注。Balol-Santhal-Bechraji在过去25年的表现清楚地回答了“ISC未能提高石油产量和采收率:神话还是事实?”作者感谢所有与实验室/现场ISC过程相关的全球科学家/工程师/操作员的意见、决定和努力。本文提出的观点和观点完全基于作者的经验，可能是一致的，也可能是不同的。20世纪70年代，印度Cambay盆地西北部稠油的发现引发了热过程的研究和开发。根据Balol地区的深度、岩石和流体特征、驱动机理和半干旱地区，进行了蒸汽驱上ISC的测试。实验结果是理解油的反应动力学和过程表现的关键。油品的提质化是ISC在实验室中的关键表现，但由于长距离驱油方法的掺和，在实际应用中并未体现出来。实验室人员参与设计和监测对项目的成功起着重要作用。在过去的25年里，Balol-Santhal ISC项目通过持续提高石油产量和采收率，证明了衰落油田的复兴。Bechraji油田的经验表明，该方法并非适用于所有油藏。它特别适合于相对清洁、具有构造起伏的流动稠油油藏。在流动油藏中，远距离驱油(垂直注入器与垂直/水平采油器分开)是有效的。对于低流动性的原油，使用一对垂直注入器和水平采油器(位于空气注入器下方的水平井)进行短距离驱油是开发的首选方式。这种方法也有可能捕获升级后的石油。当它被设计成易于操作和灵活，并与重力相结合时，它会吸引更多的价值。适当的点火类型、持续的监测、保持最佳的空气喷射速率和重新设计是ISC成功的重要因素。项目的成功取决于储层的结构和建筑，设计和实施的方式，以及从实验室到现场的过程中所获得的知识的整合。综上所述，ISC对于提高流动稠油/超稠油油藏的产油量和采收率无效的看法不能被概括为事实。
{"title":"In-Situ Combustion: Myths and Facts","authors":"S. Sur","doi":"10.2118/210606-pa","DOIUrl":"https://doi.org/10.2118/210606-pa","url":null,"abstract":"\u0000 In-situ combustion (ISC) involves compression and injection of air into heavy/extraheavy oil reservoirs for enhancing production and recovery. Initially, ISC was very popular due to its high theoretical thermal efficiency, though more failures than successes in the 1990s made this process unpopular. It is a fact that it is now widely considered archaic. However, Suplacu de Barcau (Romania) and Balol-Santhal (India) ISC projects have brought the process back into focus. Performance of the Balol-Santhal-Bechraji over the last 25 years provides clarity to answer the question “Failure to enhance oil production and recovery by ISC: Myth or fact?” The author appreciates the views, decisions, and efforts of all global scientists/engineers/operators associated with the ISC process in the laboratory/field. Opinions and views presented in this paper are solely based on the author’s experience, which may be in line or may differ.\u0000 The discovery of heavy oil northwest of the Cambay Basin, India, in the 1970s led to the initiation of research and development in thermal processes. The depth, rock, and fluid characteristics, drive mechanism, and semi-arid area led to the testing of ISC over steamflood in Balol. Laboratory findings are key to understanding the reaction kinetics of oil and process manifestations. Upgrading of oil is the key manifestation of ISC in the laboratory, but it is not seen in the field due to blending in long-distance displacement methodology. The involvement of laboratory personnel in design and surveillance plays an important role in the success of the project. Over the last 25 years, the Balol-Santhal ISC projects demonstrate the rejuvenation of declining fields with sustained enhanced oil production and an increase in recovery. Lessons of the Bechraji field indicate that process does not succeed in all reservoir settings. It is particularly suited to relatively clean, mobile heavy oil reservoirs with structural relief. Long-distance displacement of oil (vertical injector-vertical/horizontal producer spaced apart) is effective in a mobile oil reservoir. With low mobility oils, a short-distance oil displacement process using a pair of vertical injector and horizontal producer (horizontal well placed below the air injector) can be the preferred way for exploitation. This methodology has also the potential to capture upgraded oil. The process attracts more value when it is designed as operator friendly and flexible, integrating with gravity. Appropriate ignition types, continuous surveillance, maintaining optimum air injection rates, and re-engineering are important for the success of ISC. Success depends on the fabric and architecture of the reservoir, the way it is designed and implemented, and by integration of knowledge gained in the journey from laboratory to field with the process. It can be concluded that the perception of the ineffectiveness of ISC to enhance oil production and recovery from mobile heavy/extraheavy oil reservoi","PeriodicalId":22066,"journal":{"name":"SPE Reservoir Evaluation & Engineering","volume":"40 1","pages":""},"PeriodicalIF":2.1,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79536180","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A. Turta, R. Sierra, Mohammad Safiqul Islam, A. Singhal
An in-depth analysis was performed to determine the ignition delay via the enhanced spontaneous ignition (ESI) method on three well pairs (each pair constituted by one vertical injector and one horizontal producer) belonging to the Toe-To-Heel Air Injection (THAI) pilot in Athabasca. ESI consisted of preheating of the surroundings of injection wells by injecting a steam slug for 3 to 4 months just before starting air injection. At first, the ignition delay had been determined based on both the oil production and on the bottomhole temperatures (BHTs) recorded in the observation wells as well as at the toe of the horizontal producer. For the purposes of this paper, a more rigorous evaluation was carried out based on the variation in time of the apparent atomic hydrogen-carbon ratio (AAHCR) calculated from detailed gas analyses for a long period of time. AAHCR is a very strong synthetic parameter giving a direct indication of the peak temperature value before, during, and after the in-situ combustion (ISC) front is generated. Therefore, it provides complete information on the occurrence of high-temperature oxidation and low-temperature oxidation (LTO) reactions. Using the variation of the AAHCR, it was found that the ignition time was shorter than those determined by the previously mentioned methods. In the case of first well pair, ignition took 3 weeks as compared to the 1 month determined by the previous methods. The second well pair ignited in 1 month as compared to the previously calculated 2 months, and for the third well pair, ignition time was approximately 2 months in both cases. As an additional and complementary approach, estimation of the ignition time was also based on the variation of individual components of the produced gas. This allowed for the discovery of a new method for ignition time determination. This was possible in the THAI process, unlike conventional ISC processes, significant concentrations of hydrogen (H2) are produced, and the interpretation of its variation can give an indication of the ignition time. The new method is very simple to use, as the percentage of hydrogen in the produced gas starts to take off only after the full establishment of an ISC front, as hydrogen production is associated with high-temperature bond scission reactions in the ISC front. In general, the ignition delay is overestimated to some degree when using this method.
{"title":"Athabasca Toe-to-Heel Air Injection Pilot: Evaluation of the Spontaneous Ignition Based on Apparent Atomic Hydrogen-Carbon Ratio Variation","authors":"A. Turta, R. Sierra, Mohammad Safiqul Islam, A. Singhal","doi":"10.2118/212267-pa","DOIUrl":"https://doi.org/10.2118/212267-pa","url":null,"abstract":"\u0000 An in-depth analysis was performed to determine the ignition delay via the enhanced spontaneous ignition (ESI) method on three well pairs (each pair constituted by one vertical injector and one horizontal producer) belonging to the Toe-To-Heel Air Injection (THAI) pilot in Athabasca. ESI consisted of preheating of the surroundings of injection wells by injecting a steam slug for 3 to 4 months just before starting air injection. At first, the ignition delay had been determined based on both the oil production and on the bottomhole temperatures (BHTs) recorded in the observation wells as well as at the toe of the horizontal producer. For the purposes of this paper, a more rigorous evaluation was carried out based on the variation in time of the apparent atomic hydrogen-carbon ratio (AAHCR) calculated from detailed gas analyses for a long period of time. AAHCR is a very strong synthetic parameter giving a direct indication of the peak temperature value before, during, and after the in-situ combustion (ISC) front is generated. Therefore, it provides complete information on the occurrence of high-temperature oxidation and low-temperature oxidation (LTO) reactions. Using the variation of the AAHCR, it was found that the ignition time was shorter than those determined by the previously mentioned methods. In the case of first well pair, ignition took 3 weeks as compared to the 1 month determined by the previous methods. The second well pair ignited in 1 month as compared to the previously calculated 2 months, and for the third well pair, ignition time was approximately 2 months in both cases.\u0000 As an additional and complementary approach, estimation of the ignition time was also based on the variation of individual components of the produced gas. This allowed for the discovery of a new method for ignition time determination. This was possible in the THAI process, unlike conventional ISC processes, significant concentrations of hydrogen (H2) are produced, and the interpretation of its variation can give an indication of the ignition time. The new method is very simple to use, as the percentage of hydrogen in the produced gas starts to take off only after the full establishment of an ISC front, as hydrogen production is associated with high-temperature bond scission reactions in the ISC front. In general, the ignition delay is overestimated to some degree when using this method.","PeriodicalId":22066,"journal":{"name":"SPE Reservoir Evaluation & Engineering","volume":"5 1","pages":""},"PeriodicalIF":2.1,"publicationDate":"2022-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79121802","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
P. Aslanidis, S. Strand, T. Puntervold, K. Yeboah, I. Souayeh
Smart water or low-salinity (LS) water injection are environmentally friendly methods for efficient hydrocarbon recovery. Wettability alteration toward more water-wet conditions and increased spontaneous imbibition (SI) of water are responsible for enhanced oil production. Wettability alteration and SI to expel oil from the low-permeability matrix are time-dependent processes and both injection rate and oil viscosity are important factors affecting the contribution of capillary and viscous forces to oil production. Low flooding rate must be applied in laboratory corefloods to allow for SI and improved sweep to take place. Residual oil saturation by waterflooding and SI has previously been determined in low-permeability limestone and in higher permeability sands under various flooding rates, wetting conditions, and initial oil saturations. In this study, the effect of flooding rate on oil displacement from low-permeability, water-wet Bandera Brown outcrop sandstone cores has been examined. Viscous forces have been varied by injection at two different rates in addition to SI experiments and using mineral oils with different oil viscosities. The results showed small differences in oil recovery by SI and viscous flooding at high and low rates, indicating that capillary forces contribute significantly to the oil mobilization and production process from this low-permeability, water-wet rock. By varying the oil viscosity, the results indicated that capillary forces were especially important for oil displacement at higher oil viscosity as the ultimate oil recovered by low-rate injection was higher than that from the high-rate injection. Capillary number calculations indicated that viscous forces should be dominant in the recovery tests; however, the experiments showed that capillary forces were important for efficient oil displacement from the low-permeability, water-wet cores used in this study. There was no direct link observed between generated pressure drops at high and low injection rates, including SI, and the ultimate oil recovery. Thus, to simulate oil production in the middle of the reservoir, it was concluded that low-rate waterflooding is needed in laboratory tests to allow SI into the matrix to displace oil by positive capillary forces. The combination of using oils that differ in viscosity at different injection rates could add some additional information to the literature on how to increase the efficiency of waterflooding with a low injection rate.
{"title":"Oil Recovery by Low-Rate Waterflooding in Low-Permeability Water-Wet Sandstone Cores","authors":"P. Aslanidis, S. Strand, T. Puntervold, K. Yeboah, I. Souayeh","doi":"10.2118/209688-pa","DOIUrl":"https://doi.org/10.2118/209688-pa","url":null,"abstract":"\u0000 Smart water or low-salinity (LS) water injection are environmentally friendly methods for efficient hydrocarbon recovery. Wettability alteration toward more water-wet conditions and increased spontaneous imbibition (SI) of water are responsible for enhanced oil production. Wettability alteration and SI to expel oil from the low-permeability matrix are time-dependent processes and both injection rate and oil viscosity are important factors affecting the contribution of capillary and viscous forces to oil production.\u0000 Low flooding rate must be applied in laboratory corefloods to allow for SI and improved sweep to take place. Residual oil saturation by waterflooding and SI has previously been determined in low-permeability limestone and in higher permeability sands under various flooding rates, wetting conditions, and initial oil saturations. In this study, the effect of flooding rate on oil displacement from low-permeability, water-wet Bandera Brown outcrop sandstone cores has been examined. Viscous forces have been varied by injection at two different rates in addition to SI experiments and using mineral oils with different oil viscosities.\u0000 The results showed small differences in oil recovery by SI and viscous flooding at high and low rates, indicating that capillary forces contribute significantly to the oil mobilization and production process from this low-permeability, water-wet rock. By varying the oil viscosity, the results indicated that capillary forces were especially important for oil displacement at higher oil viscosity as the ultimate oil recovered by low-rate injection was higher than that from the high-rate injection. Capillary number calculations indicated that viscous forces should be dominant in the recovery tests; however, the experiments showed that capillary forces were important for efficient oil displacement from the low-permeability, water-wet cores used in this study. There was no direct link observed between generated pressure drops at high and low injection rates, including SI, and the ultimate oil recovery. Thus, to simulate oil production in the middle of the reservoir, it was concluded that low-rate waterflooding is needed in laboratory tests to allow SI into the matrix to displace oil by positive capillary forces.\u0000 The combination of using oils that differ in viscosity at different injection rates could add some additional information to the literature on how to increase the efficiency of waterflooding with a low injection rate.","PeriodicalId":22066,"journal":{"name":"SPE Reservoir Evaluation & Engineering","volume":"32 1","pages":""},"PeriodicalIF":2.1,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89387657","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
While much has been learned in the laboratory over the past four decades about the in-situ combustion (ISC) process, especially through carefully conducted physical model experiments, and many advancements in numerical simulation capability have been achieved, successful field application of ISC remains a rarity. This paper discusses challenges that have been faced in moving from laboratory to field and some strategies that may be used for improving the success rate. There is a brief discussion of the advantages and disadvantages of ISC as a recovery method and comparisons with steam injection, which is the dominant thermal recovery method used in the field. A discussion of the challenges and progress made in numerical simulation is provided with the suggestion that such mathematical modeling can now be a useful tool in designing field projects and can increase the probability of success. The needs of industry to operate safe, simple, and economically and environmentally sustainable projects are discussed along with the currently negative perception of the ISC process in industry. The paper makes some suggestions regarding how to address these issues. The main thesis of this paper is that air injection into a reservoir introduces a large amount of nitrogen that is detrimental to the displacement of oil, and oil recovery yet offers few, if any, advantages. Reducing the amount of noncondensable gas (NCG) associated with the process can be done mainly in two ways—by using oxygen-enriched air injection and furthermore by injecting a mixture of steam and oxygen-enriched air. The paper does not make a comprehensive review of past field projects but does include a summary of promising areas for future application of the ISC combustion recovery process.
{"title":"Methods to Enhance Success of Field Application of In-Situ Combustion for Heavy Oil Recovery","authors":"T. Harding","doi":"10.2118/210600-pa","DOIUrl":"https://doi.org/10.2118/210600-pa","url":null,"abstract":"\u0000 While much has been learned in the laboratory over the past four decades about the in-situ combustion (ISC) process, especially through carefully conducted physical model experiments, and many advancements in numerical simulation capability have been achieved, successful field application of ISC remains a rarity. This paper discusses challenges that have been faced in moving from laboratory to field and some strategies that may be used for improving the success rate. There is a brief discussion of the advantages and disadvantages of ISC as a recovery method and comparisons with steam injection, which is the dominant thermal recovery method used in the field. A discussion of the challenges and progress made in numerical simulation is provided with the suggestion that such mathematical modeling can now be a useful tool in designing field projects and can increase the probability of success. The needs of industry to operate safe, simple, and economically and environmentally sustainable projects are discussed along with the currently negative perception of the ISC process in industry. The paper makes some suggestions regarding how to address these issues. The main thesis of this paper is that air injection into a reservoir introduces a large amount of nitrogen that is detrimental to the displacement of oil, and oil recovery yet offers few, if any, advantages. Reducing the amount of noncondensable gas (NCG) associated with the process can be done mainly in two ways—by using oxygen-enriched air injection and furthermore by injecting a mixture of steam and oxygen-enriched air. The paper does not make a comprehensive review of past field projects but does include a summary of promising areas for future application of the ISC combustion recovery process.","PeriodicalId":22066,"journal":{"name":"SPE Reservoir Evaluation & Engineering","volume":"68 1","pages":""},"PeriodicalIF":2.1,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90760599","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Water imbibition, and the associated oil displacement, is an important process in shale oil reservoirs after hydraulic fracturing and in water-based enhanced oil recovery (EOR). Current techniques for water imbibition measurement are mostly “black-box”-type methods. A more explicit understanding of the water imbibition/oil recovery dynamics and geological controls is in demand. In this paper, a multiscale imaging technique that covers centimeter to nanometer scale (i.e., core to pore scale), integrating neutron radiography, microcomputed tomography (micro-CT), and scanning electron microscope (SEM) is applied to investigate the water imbibition depth and rate and the cause of heterogeneity of imbibition in shale samples. The dynamic processes of water imbibition in the 1-in. (25.4-mm) core sample were explicitly demonstrated, and the imbibition along the matrix and imbibition through microfractures are distinguished through neutron radiography image analysis. The causes of observed imbibition heterogeneity were further investigated through micro-CT and SEM image analysis for 1.5-mm diameter miniplug samples from different laminas of the 1-in. core samples. Imbibition depth and rate were calculated on the basis of image analysis as well. Estimation of oil recovery through water imbibition in shale matrix was performed for an example shale field. This innovative and integrated multiscale imaging technique provides a “white/gray-box” method to understand water imbibition and water-oil displacement in shale. The wide span of the length scale (from centimeter to nanometer) of this technique enables a more comprehensive, accurate, and specific understanding of both the core-scale dynamics and pore-scale mechanisms of water imbibition, oil recovery, and matrix-fracture interaction.
{"title":"Water Imbibition and Oil Recovery in Shale: Dynamics and Mechanisms Using Integrated Centimeter-to-Nanometer-Scale Imaging","authors":"S. Peng, J. LaManna, P. Periwal, P. Shevchenko","doi":"10.2118/210567-pa","DOIUrl":"https://doi.org/10.2118/210567-pa","url":null,"abstract":"\u0000 Water imbibition, and the associated oil displacement, is an important process in shale oil reservoirs after hydraulic fracturing and in water-based enhanced oil recovery (EOR). Current techniques for water imbibition measurement are mostly “black-box”-type methods. A more explicit understanding of the water imbibition/oil recovery dynamics and geological controls is in demand. In this paper, a multiscale imaging technique that covers centimeter to nanometer scale (i.e., core to pore scale), integrating neutron radiography, microcomputed tomography (micro-CT), and scanning electron microscope (SEM) is applied to investigate the water imbibition depth and rate and the cause of heterogeneity of imbibition in shale samples. The dynamic processes of water imbibition in the 1-in. (25.4-mm) core sample were explicitly demonstrated, and the imbibition along the matrix and imbibition through microfractures are distinguished through neutron radiography image analysis. The causes of observed imbibition heterogeneity were further investigated through micro-CT and SEM image analysis for 1.5-mm diameter miniplug samples from different laminas of the 1-in. core samples. Imbibition depth and rate were calculated on the basis of image analysis as well. Estimation of oil recovery through water imbibition in shale matrix was performed for an example shale field. This innovative and integrated multiscale imaging technique provides a “white/gray-box” method to understand water imbibition and water-oil displacement in shale. The wide span of the length scale (from centimeter to nanometer) of this technique enables a more comprehensive, accurate, and specific understanding of both the core-scale dynamics and pore-scale mechanisms of water imbibition, oil recovery, and matrix-fracture interaction.","PeriodicalId":22066,"journal":{"name":"SPE Reservoir Evaluation & Engineering","volume":"2 1","pages":""},"PeriodicalIF":2.1,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90183033","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
R. Reddy, Aditya Ojha, R. Nachiappan, S. Mengal, M. A. Al Hosani, A. A. Al Bairaq, M. Baslaib
Gas cap pressure maintenance while developing the associated oil rim is a critical aspect for optimum recovery. Preventing gas cap pressure dropping below dewpoint by injecting lean gas is essential for concurrent gas cap-oil rim development. Reservoir heterogeneity aggravates lean gas override causing preferential movement of lean gas plume. Thus, it is important to track lean gas plume while recycling and understanding the breakthrough potential of lean gas. This paper demonstrates a new workflow to track lean gas plume by estimating phase saturations with a case study from one of the giant oil and gas fields, Onshore, Abu Dhabi. Pulsed neutron capture (PNC) tools are used for reservoir monitoring and surveillance. However, sigma log evaluation is insufficient to derive individual hydrocarbon phase saturations to monitor lean gas plume. Neutron response modeling (NRM) is devised to differentiate between lean and rich gas. NRM is a probabilistic solver with input of mineral and fluid phase parameters into tool response functions in petrophysical evaluation. To distinguish with discrete neutron fluid response between lean and rich gas, pressure/volume/temperature (PVT) data are utilized to derive hydrogen index, capture cross section, thermal decay length, and neutron macro parameters, such as neutron slowing down length and migration length. Neutron response is investigated for lean and rich gas with sensitivity of invasion effects on neutron log by calibrating to core porosity. The response for each phase under thermal neutron and capture modes with corresponding raw neutron log statistics is reviewed in both openhole and casedhole environments in known lean/rich gas intervals. Thirty-five wells spread across gas cap and oil leg with quality neutron log data are modeled and individual phase saturations are estimated. The target reservoir is under development with over three decades of lean gas injection to support oil production. NRM results and phase saturations are validated with recent formation sampling, which enhanced the confidence in the overall workflow. Later, the results are verified to be in excellent agreement with lean gas injection and production history of the target reservoir. The identified movement of lean gas highlights nonuniform geology and gravity segregation of injected lean gas into upper members of the target reservoir. The results also emphasized the need for better injection support to lower members of the target reservoir where gas cap development is ongoing. The solution presented is unique, particularly for lean gas injection projects by utilizing PVT for NRM based on neutron transport mechanism in pore fluids. Existing workflows require a special nuclear modeling platform with computationally expensive processing on data sets acquired using advanced logging technology. In spite of these prerequisites, existing workflows are not able to distinguish lean gas over rich gas. This paper effectively demonstrates NR
{"title":"Neutron Response Modeling to Track Lean Gas Plume in Recycled Gas Cap Reservoir in Concurrent Gas Cap-Oil Rim Development: A Step Forward","authors":"R. Reddy, Aditya Ojha, R. Nachiappan, S. Mengal, M. A. Al Hosani, A. A. Al Bairaq, M. Baslaib","doi":"10.2118/210566-pa","DOIUrl":"https://doi.org/10.2118/210566-pa","url":null,"abstract":"\u0000 Gas cap pressure maintenance while developing the associated oil rim is a critical aspect for optimum recovery. Preventing gas cap pressure dropping below dewpoint by injecting lean gas is essential for concurrent gas cap-oil rim development. Reservoir heterogeneity aggravates lean gas override causing preferential movement of lean gas plume. Thus, it is important to track lean gas plume while recycling and understanding the breakthrough potential of lean gas. This paper demonstrates a new workflow to track lean gas plume by estimating phase saturations with a case study from one of the giant oil and gas fields, Onshore, Abu Dhabi.\u0000 Pulsed neutron capture (PNC) tools are used for reservoir monitoring and surveillance. However, sigma log evaluation is insufficient to derive individual hydrocarbon phase saturations to monitor lean gas plume. Neutron response modeling (NRM) is devised to differentiate between lean and rich gas. NRM is a probabilistic solver with input of mineral and fluid phase parameters into tool response functions in petrophysical evaluation. To distinguish with discrete neutron fluid response between lean and rich gas, pressure/volume/temperature (PVT) data are utilized to derive hydrogen index, capture cross section, thermal decay length, and neutron macro parameters, such as neutron slowing down length and migration length. Neutron response is investigated for lean and rich gas with sensitivity of invasion effects on neutron log by calibrating to core porosity. The response for each phase under thermal neutron and capture modes with corresponding raw neutron log statistics is reviewed in both openhole and casedhole environments in known lean/rich gas intervals. Thirty-five wells spread across gas cap and oil leg with quality neutron log data are modeled and individual phase saturations are estimated.\u0000 The target reservoir is under development with over three decades of lean gas injection to support oil production. NRM results and phase saturations are validated with recent formation sampling, which enhanced the confidence in the overall workflow. Later, the results are verified to be in excellent agreement with lean gas injection and production history of the target reservoir. The identified movement of lean gas highlights nonuniform geology and gravity segregation of injected lean gas into upper members of the target reservoir. The results also emphasized the need for better injection support to lower members of the target reservoir where gas cap development is ongoing.\u0000 The solution presented is unique, particularly for lean gas injection projects by utilizing PVT for NRM based on neutron transport mechanism in pore fluids. Existing workflows require a special nuclear modeling platform with computationally expensive processing on data sets acquired using advanced logging technology. In spite of these prerequisites, existing workflows are not able to distinguish lean gas over rich gas. This paper effectively demonstrates NR","PeriodicalId":22066,"journal":{"name":"SPE Reservoir Evaluation & Engineering","volume":"270 1","pages":""},"PeriodicalIF":2.1,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83021754","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper investigates the computational behaviors of simple-to-use, relatively fast, and versatile machine learning (ML) methods to predict apparent viscosity, a key rheological property of nanoparticle-surfactant-stabilized CO2 foam in unconventional reservoir fracturing. The first novelty of our study is the investigation of the predictive performance of ML approaches as viable alternatives for predicting the apparent viscosity of NP-Surf-CO2 foams. The predictive and computational performance of five nonlinear ML algorithms were first compared. Support vector regression (SVR), K-nearest neighbors (KNN), classification and regression trees (CART), feed-forward multilayer perceptron neural network (MLPNN), and multivariate polynomial regression (MPR) algorithms were used to create models. Temperature, foam quality, pressure, salinity, shear rate, nanoparticle size, nanoparticle concentration, and surfactant concentration were identified as relevant input parameters using principal component analysis (PCA). A data set containing 329 experimental data records was used in the study. In building the models, 80% of the data set was used for training and 20% of the data set for testing. Another unique aspect of this research is the examination of diverse ensemble learning techniques for improving computational performance. We developed meta-models of the generated models by implementing various ensemble learning algorithms (bagging, boosting, and stacking). This was done to explore and compare the computational and predictive performance enhancements of the base models (if any). To determine the relative significance of the input parameters on prediction accuracy, we used permutation feature importance (PFI). We also investigated how the SVR model made its predictions by utilizing the SHapely Additive exPlanations (SHAP) technique to quantify the influence of each input parameter on prediction. This work’s application of the SHAP approach in the interpretation of ML findings in predicting apparent viscosity is also novel. On the test data, the SVR model in this work had the best predictive performance of the single models, with an R2 of 0.979, root mean squared error (RMSE) of 0.885 cp, and mean absolute error (MAE) of 0.320 cp. Blending, a variant of the stacking ensemble technique, significantly improved this performance. With an R2 of 1.0, RMSE of 0.094 cp, and MAE of 0.087 cp, an SVR-based meta-model ensembled with blending outperformed all single and ensemble models in predicting apparent viscosity. However, in terms of computational time, the blended SVR-based meta-model did not outperform any of its constituent models. PCA and PFI ranked temperature as the most important factor in predicting the apparent viscosity of NP-Surf-CO2 foams. The ML approach used in this study provides a comprehensive understanding of the nonlinear relationship between the investigated factors and apparent viscosity. The workflow can be used to evaluate the ap
{"title":"Ensemble Machine Learning for Predicting Viscosity of Nanoparticle-Surfactant-Stabilized CO2 Foam","authors":"T. Olukoga, Micheal Totaro, Yin Feng","doi":"10.2118/210577-pa","DOIUrl":"https://doi.org/10.2118/210577-pa","url":null,"abstract":"\u0000 This paper investigates the computational behaviors of simple-to-use, relatively fast, and versatile machine learning (ML) methods to predict apparent viscosity, a key rheological property of nanoparticle-surfactant-stabilized CO2 foam in unconventional reservoir fracturing.\u0000 The first novelty of our study is the investigation of the predictive performance of ML approaches as viable alternatives for predicting the apparent viscosity of NP-Surf-CO2 foams. The predictive and computational performance of five nonlinear ML algorithms were first compared. Support vector regression (SVR), K-nearest neighbors (KNN), classification and regression trees (CART), feed-forward multilayer perceptron neural network (MLPNN), and multivariate polynomial regression (MPR) algorithms were used to create models. Temperature, foam quality, pressure, salinity, shear rate, nanoparticle size, nanoparticle concentration, and surfactant concentration were identified as relevant input parameters using principal component analysis (PCA). A data set containing 329 experimental data records was used in the study. In building the models, 80% of the data set was used for training and 20% of the data set for testing.\u0000 Another unique aspect of this research is the examination of diverse ensemble learning techniques for improving computational performance. We developed meta-models of the generated models by implementing various ensemble learning algorithms (bagging, boosting, and stacking). This was done to explore and compare the computational and predictive performance enhancements of the base models (if any).\u0000 To determine the relative significance of the input parameters on prediction accuracy, we used permutation feature importance (PFI). We also investigated how the SVR model made its predictions by utilizing the SHapely Additive exPlanations (SHAP) technique to quantify the influence of each input parameter on prediction. This work’s application of the SHAP approach in the interpretation of ML findings in predicting apparent viscosity is also novel.\u0000 On the test data, the SVR model in this work had the best predictive performance of the single models, with an R2 of 0.979, root mean squared error (RMSE) of 0.885 cp, and mean absolute error (MAE) of 0.320 cp. Blending, a variant of the stacking ensemble technique, significantly improved this performance. With an R2 of 1.0, RMSE of 0.094 cp, and MAE of 0.087 cp, an SVR-based meta-model ensembled with blending outperformed all single and ensemble models in predicting apparent viscosity. However, in terms of computational time, the blended SVR-based meta-model did not outperform any of its constituent models. PCA and PFI ranked temperature as the most important factor in predicting the apparent viscosity of NP-Surf-CO2 foams. The ML approach used in this study provides a comprehensive understanding of the nonlinear relationship between the investigated factors and apparent viscosity. The workflow can be used to evaluate the ap","PeriodicalId":22066,"journal":{"name":"SPE Reservoir Evaluation & Engineering","volume":"87 1","pages":""},"PeriodicalIF":2.1,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81937667","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Rock composition can be related to conventional well logs through theoretical equations and petrophysical endpoints. Multimineral analysis is a formation evaluation tool that uses inversions to quantify rock composition from well logs. However, because of data errors and the multivariate selection of petrophysical endpoints, solutions from the multimineral analysis are nonunique. Many plausible realizations exhibit comparable data misfits. Therefore, the uncertainties in rock composition and petrophysical endpoints must be quantified but cannot be fulfilled by deterministic solvers. Stochastic Bayesian methods have been applied to assess the uncertainties, but the high run time, tedious parameter tuning, and need for specific prior information hinder their practical use. We implement Markov chain Monte Carlo with ensemble samplers (MCMCES) to assess the uncertainties of rock composition or petrophysical endpoints in the Bayesian framework. The resultant posterior probability density functions (PDFs) quantify the uncertainties. Our method has fewer tuning parameters and is more efficient in convergence than the conventional random walk Markov chain Monte Carlo (MCMC) methods in high-dimensional problems. We present two independent applications of MCMCES in multimineral analysis. We first apply MCMCES to assess the uncertainties in volume fractions with a suite of well logs and petrophysical endpoints. However, defining the petrophysical endpoints can be challenging in complex geological settings because the values of standard endpoints may not be optimal. Next, we use MCMCES to estimate petrophysical endpoints’ posterior PDFs when the endpoints are uncertain. Our methods provide posterior volume-fraction or petrophysical-endpoint realizations for interpreters to evaluate multimineral solutions. We demonstrate our approach with synthetic and field examples. Reproducible results are supplemented with the paper.
{"title":"Practical Bayesian Inversions for Rock Composition and Petrophysical Endpoints in Multimineral Analysis","authors":"Liwei Cheng, G. Jin, R. Michelena, A. Tura","doi":"10.2118/210576-pa","DOIUrl":"https://doi.org/10.2118/210576-pa","url":null,"abstract":"\u0000 Rock composition can be related to conventional well logs through theoretical equations and petrophysical endpoints. Multimineral analysis is a formation evaluation tool that uses inversions to quantify rock composition from well logs. However, because of data errors and the multivariate selection of petrophysical endpoints, solutions from the multimineral analysis are nonunique. Many plausible realizations exhibit comparable data misfits. Therefore, the uncertainties in rock composition and petrophysical endpoints must be quantified but cannot be fulfilled by deterministic solvers. Stochastic Bayesian methods have been applied to assess the uncertainties, but the high run time, tedious parameter tuning, and need for specific prior information hinder their practical use. We implement Markov chain Monte Carlo with ensemble samplers (MCMCES) to assess the uncertainties of rock composition or petrophysical endpoints in the Bayesian framework. The resultant posterior probability density functions (PDFs) quantify the uncertainties. Our method has fewer tuning parameters and is more efficient in convergence than the conventional random walk Markov chain Monte Carlo (MCMC) methods in high-dimensional problems. We present two independent applications of MCMCES in multimineral analysis. We first apply MCMCES to assess the uncertainties in volume fractions with a suite of well logs and petrophysical endpoints. However, defining the petrophysical endpoints can be challenging in complex geological settings because the values of standard endpoints may not be optimal. Next, we use MCMCES to estimate petrophysical endpoints’ posterior PDFs when the endpoints are uncertain. Our methods provide posterior volume-fraction or petrophysical-endpoint realizations for interpreters to evaluate multimineral solutions. We demonstrate our approach with synthetic and field examples. Reproducible results are supplemented with the paper.","PeriodicalId":22066,"journal":{"name":"SPE Reservoir Evaluation & Engineering","volume":"52 1","pages":""},"PeriodicalIF":2.1,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73791757","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Physics-based and empirical rate-time models inherently assume constant bottomhole flowing pressure (BHP), an assumption that may not hold for many unconventional wells. Hence, applying these models without accounting for BHP variations might lead to inaccurate (a) flow regime identification, (b) estimation of the parameters of these models, and (c) estimated ultimate recovery (EUR) and drainage volumes. This study evaluates and compares the predictions of rate-time relations including and ignoring corrections for time-varying BHP for both synthetic and shale gas wells. We generate a real gas synthetic case with errors in the time-varying BHP. First, we convert pressures into pseudopressures. Second, we deconvolve the pseudopressure history by applying the regularized exponential basis function inverse scheme to obtain an equivalent rate—the unit-pseudopressure-drop rate at standard conditions—at constant BHP. Third, we history match the production using the scaled single-phase compressible fluid physics-based model for three different approaches: (a) using rate-time-pressure data with rate-pseudopressure deconvolution, (b) using rate-time-pressure data using just rate-pressure deconvolution, and (c) using only rate-time data. Finally, we compare the results in terms of their history matches and estimated reservoir parameters. We conclude by illustrating the application of this procedure to shale gas wells. For the synthetic case, the fit of the single-phase compressible fluid rate-time model using rate-pseudopressure deconvolution can accurately estimate the original gas in place, characteristic time, gas permeability, and fracture half-length. In contrast, considerable errors are noted when either using rate-pressure deconvolution or failing to account for variable BHP. Regarding the shale gas examples, the rate-pseudopressure deconvolution scheme accurately identifies the flow regimes present in the well, which can be difficult to detect by only analyzing rate-time data. For this reason, the fits of the scaled single-phase compressible fluid model using only rate-time result in unreasonably large estimates of the reservoir parameters and EUR. In contrast, the application of rate-pseudopressure deconvolution constrains the fits of the single-phase compressible fluid model yielding more realistic estimates of the time of end of transient flow, and EUR. This paper illustrates the application of a workflow that accounts for variable BHP by estimating an equivalent constant unit-pseudopressure-drop gas rate (at standard conditions). We illustrate the workflow for a particular decline-curve model, but the workflow is general and can be applied to any rate-time model. The approach history matches and forecasts the production of unconventional gas reservoirs using rate-time models more accurately than assuming constant BHP.
{"title":"Rate-Pseudopressure Deconvolution Enhances Rate-Time Models Production History-Matches and Forecasts of Shale Gas Wells","authors":"L. R. Ruiz Maraggi, L. Lake, M. Walsh","doi":"10.2118/208967-pa","DOIUrl":"https://doi.org/10.2118/208967-pa","url":null,"abstract":"\u0000 Physics-based and empirical rate-time models inherently assume constant bottomhole flowing pressure (BHP), an assumption that may not hold for many unconventional wells. Hence, applying these models without accounting for BHP variations might lead to inaccurate (a) flow regime identification, (b) estimation of the parameters of these models, and (c) estimated ultimate recovery (EUR) and drainage volumes. This study evaluates and compares the predictions of rate-time relations including and ignoring corrections for time-varying BHP for both synthetic and shale gas wells.\u0000 We generate a real gas synthetic case with errors in the time-varying BHP. First, we convert pressures into pseudopressures. Second, we deconvolve the pseudopressure history by applying the regularized exponential basis function inverse scheme to obtain an equivalent rate—the unit-pseudopressure-drop rate at standard conditions—at constant BHP. Third, we history match the production using the scaled single-phase compressible fluid physics-based model for three different approaches: (a) using rate-time-pressure data with rate-pseudopressure deconvolution, (b) using rate-time-pressure data using just rate-pressure deconvolution, and (c) using only rate-time data. Finally, we compare the results in terms of their history matches and estimated reservoir parameters. We conclude by illustrating the application of this procedure to shale gas wells.\u0000 For the synthetic case, the fit of the single-phase compressible fluid rate-time model using rate-pseudopressure deconvolution can accurately estimate the original gas in place, characteristic time, gas permeability, and fracture half-length. In contrast, considerable errors are noted when either using rate-pressure deconvolution or failing to account for variable BHP. Regarding the shale gas examples, the rate-pseudopressure deconvolution scheme accurately identifies the flow regimes present in the well, which can be difficult to detect by only analyzing rate-time data. For this reason, the fits of the scaled single-phase compressible fluid model using only rate-time result in unreasonably large estimates of the reservoir parameters and EUR. In contrast, the application of rate-pseudopressure deconvolution constrains the fits of the single-phase compressible fluid model yielding more realistic estimates of the time of end of transient flow, and EUR.\u0000 This paper illustrates the application of a workflow that accounts for variable BHP by estimating an equivalent constant unit-pseudopressure-drop gas rate (at standard conditions). We illustrate the workflow for a particular decline-curve model, but the workflow is general and can be applied to any rate-time model. The approach history matches and forecasts the production of unconventional gas reservoirs using rate-time models more accurately than assuming constant BHP.","PeriodicalId":22066,"journal":{"name":"SPE Reservoir Evaluation & Engineering","volume":"15 1","pages":""},"PeriodicalIF":2.1,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73281752","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Oilfield development performance prediction is a significant and complex problem in oilfield development. Reasonable prediction of oilfield development performance can guide the adjustment of the development plan. Moreover, the reservoir will change slowly during reservoir development because of flowing water however, previous networks that forecast production dynamics ignored it, which leads to inaccurate predictions. Routine well-wise injection and production measurements contain important subsurface structure and properties. So, for the dynamic prediction of oil/water two-phase waterflooded reservoirs, we built a deep learning framework named adaptive correction interwell connectivity model based on graph convolutional networks (GCN) and gated recurrent unit (GRU). It includes two parts: The first part is the adaptive correction model based on GCN, which uses dynamic production data to automatically correct the initial interwell connectivity computed by permeability, porosity, interwell distance, and so on. The second part is the adaptive learning model based on GRU, which predicts the production performance of oil wells according to the time characteristics of production performance data. This framework considers the influence that changes in reservoir conditions have on production over time to solve the problem of inaccurate production dynamic prediction. It can also predict interwell connectivity. For oilfields with too many wells, using the embedding idea classifies similar wells into one category, saving time for training and avoiding overfitting problems. Applying the model to five different reservoirs to predict interwell connectivity, well oil production rate, and well water cut compare the results with artificial neural networks (ANN), GRU, and long short-term memory (LSTM) models and compare the interwell connectivity with numerical simulation software ,tNavigator® (Rock Flow Dynamics Llc), too. When the model is applied in Block B of Bohai A reservoir, the mean absolute percentage error of “Adaptive Graph convolutional network and GRU” (AG-GRU) is 2.1150% while the LSTM is 9.8872%. The error reduces by 78.6%. The injected water has a direction from the water injection well to the production well; this paper only considers the interwell connectivity without considering the direction. Further research is needed to consider the water injection direction and form a weighted directed graph.
{"title":"A Deep Learning Framework Using Graph Convolutional Networks for Adaptive Correction of Interwell Connectivity and Gated Recurrent Unit for Performance Prediction","authors":"Leding Du, Yuetian Liu, Liang Xue, Guohui You","doi":"10.2118/210575-pa","DOIUrl":"https://doi.org/10.2118/210575-pa","url":null,"abstract":"\u0000 Oilfield development performance prediction is a significant and complex problem in oilfield development. Reasonable prediction of oilfield development performance can guide the adjustment of the development plan. Moreover, the reservoir will change slowly during reservoir development because of flowing water however, previous networks that forecast production dynamics ignored it, which leads to inaccurate predictions. Routine well-wise injection and production measurements contain important subsurface structure and properties. So, for the dynamic prediction of oil/water two-phase waterflooded reservoirs, we built a deep learning framework named adaptive correction interwell connectivity model based on graph convolutional networks (GCN) and gated recurrent unit (GRU). It includes two parts: The first part is the adaptive correction model based on GCN, which uses dynamic production data to automatically correct the initial interwell connectivity computed by permeability, porosity, interwell distance, and so on. The second part is the adaptive learning model based on GRU, which predicts the production performance of oil wells according to the time characteristics of production performance data. This framework considers the influence that changes in reservoir conditions have on production over time to solve the problem of inaccurate production dynamic prediction. It can also predict interwell connectivity. For oilfields with too many wells, using the embedding idea classifies similar wells into one category, saving time for training and avoiding overfitting problems. Applying the model to five different reservoirs to predict interwell connectivity, well oil production rate, and well water cut compare the results with artificial neural networks (ANN), GRU, and long short-term memory (LSTM) models and compare the interwell connectivity with numerical simulation software ,tNavigator® (Rock Flow Dynamics Llc), too. When the model is applied in Block B of Bohai A reservoir, the mean absolute percentage error of “Adaptive Graph convolutional network and GRU” (AG-GRU) is 2.1150% while the LSTM is 9.8872%. The error reduces by 78.6%. The injected water has a direction from the water injection well to the production well; this paper only considers the interwell connectivity without considering the direction. Further research is needed to consider the water injection direction and form a weighted directed graph.","PeriodicalId":22066,"journal":{"name":"SPE Reservoir Evaluation & Engineering","volume":"48 1","pages":""},"PeriodicalIF":2.1,"publicationDate":"2022-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91179734","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","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：