Accurate estimation of reservoir parameters such as fluid saturations and porosity is important for assessing petroleum volumes, economics and decisionmaking. Such parameters are derived from interpretation of petrophysical logs or time-consuming, expensive core analyses. Not all wells are cored in a field, and the number of fully cored wells is limited. In this study, a time-efficient and economical method to estimate porosity, water saturation and hydrocarbon saturation is employed. Two Least Squares Support Vector Machine (LSSVM) machine learning models, optimized with Particle Swarm Optimization (PSO), were developed to predict these reservoir parameters, respectively. The models were developed based on data from five wells in the Varg field, Central North Sea, Norway where the data were randomized and split into an unseen fraction (10%) and a fraction used to train the models (90%). In addition to the unseen fraction, a sixth well from the Varg field was used to assess the models. The samples are mainly sandstone with different contents of shale, while fluids water, oil and gas were present. The ‘seen’ data were randomized into calibration, validation and testing sets during the model development. The petrophysical logs in the study were Gamma-ray, Self-potential, Acoustic, Neutron porosity, bulk density, caliper, deep resistivity, and medium resistivity. The log based inputs were made more linear (via log operations) when relevant and normalized to be more comparable in the algorithms. Feature selection was conducted to identify the most relevant petrophysical logs and remove those that are considered less relevant. Three and four of the eight logs were sufficient, to reach optimum performance of porosity and saturation prediction, respectively. Porosity was predicted with R2 = 0.79 and 0.70 on the model development set and unseen set, for saturation it was 0.71 and 0.61, a similar performance as on the training and testing sets at the development stage. The R2 was close to zero on the new well, although the predicted values were physical and within the observed data scatter range as the model development set. Possible improvements were identified in dataset preparation and feature selection to get more robust models.
{"title":"Machine Learning Based Prediction of Porosity and Water Saturation from Varg Field Reservoir Well Logs","authors":"P. Andersen, Miranda Skjeldal, C. Augustsson","doi":"10.2118/209659-ms","DOIUrl":"https://doi.org/10.2118/209659-ms","url":null,"abstract":"\u0000 Accurate estimation of reservoir parameters such as fluid saturations and porosity is important for assessing petroleum volumes, economics and decisionmaking. Such parameters are derived from interpretation of petrophysical logs or time-consuming, expensive core analyses. Not all wells are cored in a field, and the number of fully cored wells is limited. In this study, a time-efficient and economical method to estimate porosity, water saturation and hydrocarbon saturation is employed. Two Least Squares Support Vector Machine (LSSVM) machine learning models, optimized with Particle Swarm Optimization (PSO), were developed to predict these reservoir parameters, respectively.\u0000 The models were developed based on data from five wells in the Varg field, Central North Sea, Norway where the data were randomized and split into an unseen fraction (10%) and a fraction used to train the models (90%). In addition to the unseen fraction, a sixth well from the Varg field was used to assess the models.\u0000 The samples are mainly sandstone with different contents of shale, while fluids water, oil and gas were present. The ‘seen’ data were randomized into calibration, validation and testing sets during the model development. The petrophysical logs in the study were Gamma-ray, Self-potential, Acoustic, Neutron porosity, bulk density, caliper, deep resistivity, and medium resistivity. The log based inputs were made more linear (via log operations) when relevant and normalized to be more comparable in the algorithms. Feature selection was conducted to identify the most relevant petrophysical logs and remove those that are considered less relevant. Three and four of the eight logs were sufficient, to reach optimum performance of porosity and saturation prediction, respectively. Porosity was predicted with R2 = 0.79 and 0.70 on the model development set and unseen set, for saturation it was 0.71 and 0.61, a similar performance as on the training and testing sets at the development stage. The R2 was close to zero on the new well, although the predicted values were physical and within the observed data scatter range as the model development set. Possible improvements were identified in dataset preparation and feature selection to get more robust models.","PeriodicalId":148855,"journal":{"name":"Day 4 Thu, June 09, 2022","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128030363","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}
Low salinity waterflood (LSWF) injection is an enhanced oil recovery (EOR) method proven effective through extensive experimental studies. Correct implementation of this method in reservoir-scale simulations requires reliable estimation of changes in relative permeability data associated with LSWF. For this purpose, a few models have been suggested based on geochemical interactions, such as the cation exchange capacity of clay, which are case dependent and cannot be applied to all systems. This study presents a novel semi-empirical model based on incremental oil recovery measured during low salinity injection. Therefore, it can be applied to all rock types, fluid systems, and wettability conditions regardless of the active mechanism. Some mechanisms proposed in the literature relate the additional oil recovery during low salinity injection to measurable parameters such as micro-dispersion. As a result, the kr curves can be constructed using this new methodology by measuring the micro-dispersion. This method has been validated against five sets of secondary and tertiary coreflood experiments published in the literature. First, the high salinity kr data is obtained by history matching using the CMOST module of CMG software. Then the proposed method and the measured value of additional oil recovery were used to estimate the kr data of low salinity injection. The results showed that the suggested method could predict the oil recovery and pressure drop in secondary and tertiary modes. The high-salinity relative permeability was shifted towards a more water-wet condition in tertiary mode. The kr curve of secondary LSWF showed a significant shift towards a more water-wet condition than tertiary mode, implying lower residual oil saturation. Since the additional oil recovery versus micro-dispersion curve was reported for this rock sample, one can simply predict the kr values of LSWF for other values of micro-dispersion. Due to the ongoing debate regarding the dominant mechanism during LSWF, there is no universal model for estimating the relative permeability of LSWF in all systems. The model presented in this paper provides a powerful tool for engineers to simulate the LSWF kr data in both tertiary and secondary flooding regardless of the active mechanism.
{"title":"A Universal Method for Predicting the Relative Permeability Data of Low Salinity Injection","authors":"Abdulla Aljaberi, S. Aghabozorgi, M. Sohrabi","doi":"10.2118/209661-ms","DOIUrl":"https://doi.org/10.2118/209661-ms","url":null,"abstract":"\u0000 Low salinity waterflood (LSWF) injection is an enhanced oil recovery (EOR) method proven effective through extensive experimental studies. Correct implementation of this method in reservoir-scale simulations requires reliable estimation of changes in relative permeability data associated with LSWF. For this purpose, a few models have been suggested based on geochemical interactions, such as the cation exchange capacity of clay, which are case dependent and cannot be applied to all systems.\u0000 This study presents a novel semi-empirical model based on incremental oil recovery measured during low salinity injection. Therefore, it can be applied to all rock types, fluid systems, and wettability conditions regardless of the active mechanism. Some mechanisms proposed in the literature relate the additional oil recovery during low salinity injection to measurable parameters such as micro-dispersion. As a result, the kr curves can be constructed using this new methodology by measuring the micro-dispersion.\u0000 This method has been validated against five sets of secondary and tertiary coreflood experiments published in the literature. First, the high salinity kr data is obtained by history matching using the CMOST module of CMG software. Then the proposed method and the measured value of additional oil recovery were used to estimate the kr data of low salinity injection. The results showed that the suggested method could predict the oil recovery and pressure drop in secondary and tertiary modes. The high-salinity relative permeability was shifted towards a more water-wet condition in tertiary mode. The kr curve of secondary LSWF showed a significant shift towards a more water-wet condition than tertiary mode, implying lower residual oil saturation. Since the additional oil recovery versus micro-dispersion curve was reported for this rock sample, one can simply predict the kr values of LSWF for other values of micro-dispersion.\u0000 Due to the ongoing debate regarding the dominant mechanism during LSWF, there is no universal model for estimating the relative permeability of LSWF in all systems. The model presented in this paper provides a powerful tool for engineers to simulate the LSWF kr data in both tertiary and secondary flooding regardless of the active mechanism.","PeriodicalId":148855,"journal":{"name":"Day 4 Thu, June 09, 2022","volume":"75 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115310063","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}
The average recovery factor of current producing oil reservoirs is about 35-50% worldwide. Enhanced Oil Recovery (EOR) methods such as Water Alternating Gas (WAG) target the oil left in place and improve the final recovery of the developed fields. In a WAG injection plan, some reservoir blocks experience simultaneous gas and water flow. Therefore, Simultaneous Water And Gas (SWAG) injection experiments are performed to understand and simulate the fluid flow behaviour in these blocks more accurately. The experimental data we analyzed in this manuscript were obtained by performing a SWAG experiment using real reservoir rock and fluid (mixed-wet carbonate rock extracted from the Abu-Dhabi field). In miscible and immiscible experiments, the injected gas was Methane and CO2, respectively. We tried to simulate the experiments using Stone's, Baker's, and Stone's exponent models to evaluate the performance of these models in simulating SWAG experiments. It was shown that SWAG displacement can be simulated using Stone's first model and changing two-phase kr data as a matching parameter. The results showed that we do not need to correct the three-phase relative permeability in the low oil saturation region for simulating SWAG experiments. The study presented in this paper is novel in two aspects: first, the SWAG experiments were conducted in reservoir carbonate samples using real reservoir fluids; and second, even though many researchers have simulated the WAG experiments, not many have discussed the simulation of SWAG experiments. The results presented in this paper is of utmost importance for decision making, designing, and simulating CO2-EOR plans in giant Abu-Dhabi carbonate reservoirs.
{"title":"Investigation and Simulation of SWAG injections Performed in Mixed-Wet Carbonate Rocks.","authors":"Latifa Obaid Alnuaimi, S. Aghabozorgi, M. Sohrabi","doi":"10.2118/209651-ms","DOIUrl":"https://doi.org/10.2118/209651-ms","url":null,"abstract":"\u0000 The average recovery factor of current producing oil reservoirs is about 35-50% worldwide. Enhanced Oil Recovery (EOR) methods such as Water Alternating Gas (WAG) target the oil left in place and improve the final recovery of the developed fields. In a WAG injection plan, some reservoir blocks experience simultaneous gas and water flow. Therefore, Simultaneous Water And Gas (SWAG) injection experiments are performed to understand and simulate the fluid flow behaviour in these blocks more accurately.\u0000 The experimental data we analyzed in this manuscript were obtained by performing a SWAG experiment using real reservoir rock and fluid (mixed-wet carbonate rock extracted from the Abu-Dhabi field). In miscible and immiscible experiments, the injected gas was Methane and CO2, respectively. We tried to simulate the experiments using Stone's, Baker's, and Stone's exponent models to evaluate the performance of these models in simulating SWAG experiments. It was shown that SWAG displacement can be simulated using Stone's first model and changing two-phase kr data as a matching parameter. The results showed that we do not need to correct the three-phase relative permeability in the low oil saturation region for simulating SWAG experiments.\u0000 The study presented in this paper is novel in two aspects: first, the SWAG experiments were conducted in reservoir carbonate samples using real reservoir fluids; and second, even though many researchers have simulated the WAG experiments, not many have discussed the simulation of SWAG experiments. The results presented in this paper is of utmost importance for decision making, designing, and simulating CO2-EOR plans in giant Abu-Dhabi carbonate reservoirs.","PeriodicalId":148855,"journal":{"name":"Day 4 Thu, June 09, 2022","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115518162","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}
Although polymer flooding technology has been widely applied. Yet the "entry profile inversion" phenomenon occurs inevitably in its later stage, which seriously affects the development effect. In recent years, the micro-nano oil-displacement system is a novel developed flooding system. The oil-displacement system consists of micro-nano particles and its carrier fluid. After coming into porous media, it shows the properties of "plugging large pore and leave the small one open" and the motion feature of "trapping, deformation, migration". In this paper, physicochemical properties, reservoir adaptability, oil displacement mechanism of micro-nano oil-displacement system in pore throat is explored by using macroscopic physical simulation and CT scanning technology. Furthermore, the typical field application case is analyzed. Results show that, micro-nano particles have good physicochemical performance and transport ability in porous media. According to the reservoir adaptability evaluation, the matching relationships between particle size and core permeability is obtained, to provide guidance for field application scheme. By using NMR andCT techniques, its micro percolation law in porous media and remaining oil distribution during displacement process is analyzed. During the experiment, micro-nano particles presents the motion feature of "migration, trapping, and deformation" in the core pore, which can realize deep fluid diversion and expand swept volume. From 3D macro experiment, the sweep volume can be further expanded by injecting MNS and adjusting well pattern structure after polymer flooding. The dual goals of expanding sweep volume and improving oil washing efficiency can be achieved by using binary composite system (MNS and petroleum sulfonate) and ternary composite system (MNS, alkali and petroleum sulfonate). Finally, the micro-nano oil-displacement system conformance control technology has been applied in different oilfields, which all obtained significant oil increment effect. By using the research methods of interdisciplinary innovative, the oil displacement mechanism and field application of micro-nano oil-displacement system is researched. The research results provide guidance for oil companies to enhance oil recovery significantly.
{"title":"Research Progress and Field Trail of a New Micro-Nano Oil-Displacement System Flooding Technology","authors":"Zhe Sun, Xiujun Wang","doi":"10.2118/209656-ms","DOIUrl":"https://doi.org/10.2118/209656-ms","url":null,"abstract":"\u0000 Although polymer flooding technology has been widely applied. Yet the \"entry profile inversion\" phenomenon occurs inevitably in its later stage, which seriously affects the development effect. In recent years, the micro-nano oil-displacement system is a novel developed flooding system.\u0000 The oil-displacement system consists of micro-nano particles and its carrier fluid. After coming into porous media, it shows the properties of \"plugging large pore and leave the small one open\" and the motion feature of \"trapping, deformation, migration\". In this paper, physicochemical properties, reservoir adaptability, oil displacement mechanism of micro-nano oil-displacement system in pore throat is explored by using macroscopic physical simulation and CT scanning technology. Furthermore, the typical field application case is analyzed.\u0000 Results show that, micro-nano particles have good physicochemical performance and transport ability in porous media. According to the reservoir adaptability evaluation, the matching relationships between particle size and core permeability is obtained, to provide guidance for field application scheme. By using NMR andCT techniques, its micro percolation law in porous media and remaining oil distribution during displacement process is analyzed. During the experiment, micro-nano particles presents the motion feature of \"migration, trapping, and deformation\" in the core pore, which can realize deep fluid diversion and expand swept volume. From 3D macro experiment, the sweep volume can be further expanded by injecting MNS and adjusting well pattern structure after polymer flooding. The dual goals of expanding sweep volume and improving oil washing efficiency can be achieved by using binary composite system (MNS and petroleum sulfonate) and ternary composite system (MNS, alkali and petroleum sulfonate). Finally, the micro-nano oil-displacement system conformance control technology has been applied in different oilfields, which all obtained significant oil increment effect.\u0000 By using the research methods of interdisciplinary innovative, the oil displacement mechanism and field application of micro-nano oil-displacement system is researched. The research results provide guidance for oil companies to enhance oil recovery significantly.","PeriodicalId":148855,"journal":{"name":"Day 4 Thu, June 09, 2022","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126570908","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}
Depleted oil and gas fields may provide important locations for Carbon Capture and Storage (CCS). However, injection of carbon dioxide into pressure depleted oil and gas fields can be problematic due to the low reservoir pressure and the phase change behavior of carbon dioxide. The change of carbon dioxide from a liquid into a gas can trigger physical phenomena, such as significant cooling of the fluid as a result of the Joule-Thomson effect and the latent heat of vaporization, which can cause material embrittlement and loss of equipment functionality, and unstable or surging injection rates. Current mitigations restrict the quantity of carbon dioxide able to be injected by use of multiple injection tubing strings that can be costly or technically prohibitive. A more attractive alternative may be the use of downhole variable flow restricting devices which will autonomously respond to the changing well conditions, without the need for intervention or a workover in later well life. There is limited software currently available to model flow control to ensure carbon dioxide remains in liquid form in the completion. Through nodal analysis, the CCS simulator developed in this study can simulate the choking effect of downhole flow control devices placed at intervals in the completion that are sized and numbered to achieve the desired pressure distribution and CO2 injection rate. The modelling can then illustrate the required operating parameters of the downhole flow control solution with the results indicating the equivalent orifice sizes required for the flow control devices. The adjustable flow control devices can be removed or fully opened when the reservoir pressure increase and injection rate climbs and thus deemed to be no longer necessary. The use of downhole flow control devices can replace the need for a multiple string completion as the reservoir pressures and injection rates vary over the life of the well.
{"title":"Using a CCS Simulator to Maintain Liquid CO2 in the Completion","authors":"Anna Helene Petitt, M. Konopczynski","doi":"10.2118/209705-ms","DOIUrl":"https://doi.org/10.2118/209705-ms","url":null,"abstract":"\u0000 Depleted oil and gas fields may provide important locations for Carbon Capture and Storage (CCS). However, injection of carbon dioxide into pressure depleted oil and gas fields can be problematic due to the low reservoir pressure and the phase change behavior of carbon dioxide. The change of carbon dioxide from a liquid into a gas can trigger physical phenomena, such as significant cooling of the fluid as a result of the Joule-Thomson effect and the latent heat of vaporization, which can cause material embrittlement and loss of equipment functionality, and unstable or surging injection rates. Current mitigations restrict the quantity of carbon dioxide able to be injected by use of multiple injection tubing strings that can be costly or technically prohibitive. A more attractive alternative may be the use of downhole variable flow restricting devices which will autonomously respond to the changing well conditions, without the need for intervention or a workover in later well life.\u0000 There is limited software currently available to model flow control to ensure carbon dioxide remains in liquid form in the completion. Through nodal analysis, the CCS simulator developed in this study can simulate the choking effect of downhole flow control devices placed at intervals in the completion that are sized and numbered to achieve the desired pressure distribution and CO2 injection rate. The modelling can then illustrate the required operating parameters of the downhole flow control solution with the results indicating the equivalent orifice sizes required for the flow control devices. The adjustable flow control devices can be removed or fully opened when the reservoir pressure increase and injection rate climbs and thus deemed to be no longer necessary. The use of downhole flow control devices can replace the need for a multiple string completion as the reservoir pressures and injection rates vary over the life of the well.","PeriodicalId":148855,"journal":{"name":"Day 4 Thu, June 09, 2022","volume":"80 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127188368","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}
Production from the Schoonebeek heavy oil steam flood in northeast Netherlands was historically curtailed because of limits on H2S and water production. The interdependence of various permit and facility constraints makes production optimisation for Schoonebeek extremely challenging. So much so, that the conventional IPSM approach does not apply. To understand the field's production potential and to reach it, the team developed a novel Production System Optimisation (PSO) workflow using techniques from machine learning and operations research. In this paper we explain the details of this PSO workflow, the mathematics behind it, and share our results and learnings. The algorithm runs in 5 minutes and is used in daily optimisation. The application of this new workflow in combination with the successful deployment of a novel H2S scavenger, resulted in a Schoonebeek production uplift of 50%.
{"title":"Artificial Intelligence for Production Optimization in Schoonebeek Thermal EOR Field","authors":"Mezlul Arfie, N. Ghodke, Kasper Groenbroek","doi":"10.2118/209670-ms","DOIUrl":"https://doi.org/10.2118/209670-ms","url":null,"abstract":"\u0000 Production from the Schoonebeek heavy oil steam flood in northeast Netherlands was historically curtailed because of limits on H2S and water production. The interdependence of various permit and facility constraints makes production optimisation for Schoonebeek extremely challenging. So much so, that the conventional IPSM approach does not apply. To understand the field's production potential and to reach it, the team developed a novel Production System Optimisation (PSO) workflow using techniques from machine learning and operations research. In this paper we explain the details of this PSO workflow, the mathematics behind it, and share our results and learnings. The algorithm runs in 5 minutes and is used in daily optimisation. The application of this new workflow in combination with the successful deployment of a novel H2S scavenger, resulted in a Schoonebeek production uplift of 50%.","PeriodicalId":148855,"journal":{"name":"Day 4 Thu, June 09, 2022","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129401915","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：