A. Shahkarami, Kimberly L. Ayers, Guochang Wang, Alivia Ayers
The Marcellus Shale has more than a decade of development history. However, there are many questions that still remain unanswered. What is the best inter-well spacing? What are the optimum stage length, proppant loading, and cluster spacing? What are the ideal combinations of these completion parameters? And how can we maximize the rate return on our investment? This study proposes innovative tools that allow researchers to answer these questions. We build these set of tools by utilizing the pattern recognition abilities of machine learning algorithms and public data from the Southwestern Pennsylvania region of the Marcellus Shale. By means of artificial intelligence and data mining techniques, we studied a database that includes public data from more than 2,000 wells producing from the aforementioned study area. The database contained completion, drilling, and production history information from various operators active in Allegheny, Greene, Fayette, Washington, and Westmoreland counties located in the Southwestern Pennsylvania. Extensive preprocessing and data cleansing steps were involved to prepare the database. Various machine learning techniques (Linear Regression (LR), Support Vector Machines (SVMs), Artificial Neural Networks (ANNs), and Gaussian Processes (GP)) were applied to understand the non-linear patterns in the data. The objective was to develop predictive models that were trained and validated based on the current database. The predictive models were validated using information originating from numerous wells in the area. Once validated, the model could be used in reservoir management decision-making workflows to answer questions such as what are the best drilling scenarios, the optimum hydraulic fracturing design, the initial production rate, and the estimated ultimate recovery (EUR). The workflow is purely based on field data and free of any cognitive human bias. As soon as more data is available, the model could be updated. The core data in this workflow is sourced from public domains, and therefore, intensive preprocessing efforts were necessary.
{"title":"Application of Machine Learning Algorithms for Optimizing Future Production in Marcellus Shale, Case Study of Southwestern Pennsylvania","authors":"A. Shahkarami, Kimberly L. Ayers, Guochang Wang, Alivia Ayers","doi":"10.2118/191827-18ERM-MS","DOIUrl":"https://doi.org/10.2118/191827-18ERM-MS","url":null,"abstract":"\u0000 The Marcellus Shale has more than a decade of development history. However, there are many questions that still remain unanswered. What is the best inter-well spacing? What are the optimum stage length, proppant loading, and cluster spacing? What are the ideal combinations of these completion parameters? And how can we maximize the rate return on our investment? This study proposes innovative tools that allow researchers to answer these questions. We build these set of tools by utilizing the pattern recognition abilities of machine learning algorithms and public data from the Southwestern Pennsylvania region of the Marcellus Shale.\u0000 By means of artificial intelligence and data mining techniques, we studied a database that includes public data from more than 2,000 wells producing from the aforementioned study area. The database contained completion, drilling, and production history information from various operators active in Allegheny, Greene, Fayette, Washington, and Westmoreland counties located in the Southwestern Pennsylvania. Extensive preprocessing and data cleansing steps were involved to prepare the database. Various machine learning techniques (Linear Regression (LR), Support Vector Machines (SVMs), Artificial Neural Networks (ANNs), and Gaussian Processes (GP)) were applied to understand the non-linear patterns in the data. The objective was to develop predictive models that were trained and validated based on the current database. The predictive models were validated using information originating from numerous wells in the area. Once validated, the model could be used in reservoir management decision-making workflows to answer questions such as what are the best drilling scenarios, the optimum hydraulic fracturing design, the initial production rate, and the estimated ultimate recovery (EUR). The workflow is purely based on field data and free of any cognitive human bias. As soon as more data is available, the model could be updated. The core data in this workflow is sourced from public domains, and therefore, intensive preprocessing efforts were necessary.","PeriodicalId":298489,"journal":{"name":"Day 4 Wed, October 10, 2018","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127490357","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}
Kriti Singh, S. Miska, E. Ozbayoglu, Batur Alp Aydin
Narrow annulus is frequently encountered in drilling operations as in Casing while Drilling, Liner Drilling etc. Hydraulics of narrow annulus is a relatively new topic of research in drilling. Current analytical solutions have limited applicability for complex flow regimes affected by casing motion, pipe rotation, eccentricity and cuttings. Therefore, the objective of this research is to develop data-driven statistical learning models which can be very effective in making pressure loss predictions for given operating conditions. The data for proposed supervised learning was obtained from large scale experiments conducted on a narrow annulus wellbore configuration on LPAT (Low Pressure Ambient Temperature) flow loop at TUDRP, Tulsa University Research Projects Group. Exploratory visualizations were used to determine the relationship between operational parameters and pressure drop. Resampling methods, such as cross-validation and bootstrapping, were used to split the dataset into training and test data. Shrinkage and Decomposition technique was applied to make multivariate regression more robust. Comparison was made between different algorithms to determine the best model in terms of Least Mean-Squared-Error (MSE) on test data prediction and interpretability. Multivariate exploratory plots were used for data inference. Relationships between different factors and annular pressure drop were mostly linear. As expected, pressure drop increased with increase in flow-rate, inclination angle, ROP and for non-Newtonian polymeric fluids. Principal Component Analysis (PCA) was performed to reduce the dimensionality of the data set. Approximately 98% of variance in data was explained by 5 principal components and the resulting model produced a MSE less than 1% of median pressure drop. Even though PCA regression model performed well on test data, final model was more difficult to interpret because it does not perform feature selection or even produce coefficient estimates. Therefore, Partial Least Squares (PLS) regression was used which gives better model interpretability as it is supervised by feature-outcome relationship. Shrinkage methods-Lasso and Ridge Regression were also used. These methods add an additional penalty term to Least Square Regression to get a bias-variance tradeoff. Cross-validation was used to select the penalty term that gave the lowest MSE. Both methods produced competitive MSE but performed better than PCA and PLS regression. In conclusion, Lasso-Regression performed the best with lowest error and good interpretability.
{"title":"Using Supervised Machine Learning Algorithms to Predict Pressure Drop in Narrow Annulus","authors":"Kriti Singh, S. Miska, E. Ozbayoglu, Batur Alp Aydin","doi":"10.2118/191794-18ERM-MS","DOIUrl":"https://doi.org/10.2118/191794-18ERM-MS","url":null,"abstract":"\u0000 Narrow annulus is frequently encountered in drilling operations as in Casing while Drilling, Liner Drilling etc. Hydraulics of narrow annulus is a relatively new topic of research in drilling. Current analytical solutions have limited applicability for complex flow regimes affected by casing motion, pipe rotation, eccentricity and cuttings. Therefore, the objective of this research is to develop data-driven statistical learning models which can be very effective in making pressure loss predictions for given operating conditions. The data for proposed supervised learning was obtained from large scale experiments conducted on a narrow annulus wellbore configuration on LPAT (Low Pressure Ambient Temperature) flow loop at TUDRP, Tulsa University Research Projects Group. Exploratory visualizations were used to determine the relationship between operational parameters and pressure drop. Resampling methods, such as cross-validation and bootstrapping, were used to split the dataset into training and test data. Shrinkage and Decomposition technique was applied to make multivariate regression more robust. Comparison was made between different algorithms to determine the best model in terms of Least Mean-Squared-Error (MSE) on test data prediction and interpretability. Multivariate exploratory plots were used for data inference. Relationships between different factors and annular pressure drop were mostly linear. As expected, pressure drop increased with increase in flow-rate, inclination angle, ROP and for non-Newtonian polymeric fluids. Principal Component Analysis (PCA) was performed to reduce the dimensionality of the data set. Approximately 98% of variance in data was explained by 5 principal components and the resulting model produced a MSE less than 1% of median pressure drop. Even though PCA regression model performed well on test data, final model was more difficult to interpret because it does not perform feature selection or even produce coefficient estimates. Therefore, Partial Least Squares (PLS) regression was used which gives better model interpretability as it is supervised by feature-outcome relationship. Shrinkage methods-Lasso and Ridge Regression were also used. These methods add an additional penalty term to Least Square Regression to get a bias-variance tradeoff. Cross-validation was used to select the penalty term that gave the lowest MSE. Both methods produced competitive MSE but performed better than PCA and PLS regression. In conclusion, Lasso-Regression performed the best with lowest error and good interpretability.","PeriodicalId":298489,"journal":{"name":"Day 4 Wed, October 10, 2018","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132185895","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}
Current decline models fail to capture all of the behavior in shale gas production histories. That is, upon fitting one of these models, one often sees significant and sustained deviation of the flow rate data points from the decline trend. One way to measure this "lost signal" is to look at the autocorrelation in the residuals about the fitted decline model. Indeed, with many shale gas wells we see significant amounts of autocorrelation, especially when comparing the flow rate at one time to the next (lag one). Theoretically, this serially autocorrelated error can impact decline curve analysis in two ways: 1) inefficient estimation of decline curve parameters, and 2) lost signal in the data. Borrowing from time series statistics, there are two conventional ways of dealing with these potential problems: 1) estimate the decline curve parameters with generalized least squares or generalized nonlinear least squares, and 2) fitting an ARMA model to the residuals and adding it to the fitted decline curve. This paper investigates the practical implications of these two procedures by exercising them over decline curves fit to 8,527 Marcellus shale gas wells (all wells from that play with viable data for the analysis). The study explores the effect that generalized regression methods and ARMA-modeled residuals have on six different decline curves, and performance is measured in terms of sum of squared residuals (a metric for goodness-of-fit, calculated on the training data (first 24 months of each record)) and mean absolute percent error (a standard metric for forecasting accuracy, calculated on the testing data (all production rates after 24 months)). We find that inclusion of the ARMA-modeled residuals largely improves the goodness-of-fit for any decline curve, and improves the forecasting accuracy for the Hyperbolic decline curve and Duong's model. The use of generalized least squares or generalized nonlinear least squares has little benefit in fitting the decline curves, except for the Logistic Growth model, where it improves both fit and forecasting accuracy.
{"title":"Accounting for Serial Autocorrelation in Decline Curve Analysis of Marcellus Shale Gas Wells","authors":"E. Morgan","doi":"10.2118/191788-18ERM-MS","DOIUrl":"https://doi.org/10.2118/191788-18ERM-MS","url":null,"abstract":"\u0000 Current decline models fail to capture all of the behavior in shale gas production histories. That is, upon fitting one of these models, one often sees significant and sustained deviation of the flow rate data points from the decline trend. One way to measure this \"lost signal\" is to look at the autocorrelation in the residuals about the fitted decline model. Indeed, with many shale gas wells we see significant amounts of autocorrelation, especially when comparing the flow rate at one time to the next (lag one). Theoretically, this serially autocorrelated error can impact decline curve analysis in two ways: 1) inefficient estimation of decline curve parameters, and 2) lost signal in the data. Borrowing from time series statistics, there are two conventional ways of dealing with these potential problems: 1) estimate the decline curve parameters with generalized least squares or generalized nonlinear least squares, and 2) fitting an ARMA model to the residuals and adding it to the fitted decline curve.\u0000 This paper investigates the practical implications of these two procedures by exercising them over decline curves fit to 8,527 Marcellus shale gas wells (all wells from that play with viable data for the analysis). The study explores the effect that generalized regression methods and ARMA-modeled residuals have on six different decline curves, and performance is measured in terms of sum of squared residuals (a metric for goodness-of-fit, calculated on the training data (first 24 months of each record)) and mean absolute percent error (a standard metric for forecasting accuracy, calculated on the testing data (all production rates after 24 months)).\u0000 We find that inclusion of the ARMA-modeled residuals largely improves the goodness-of-fit for any decline curve, and improves the forecasting accuracy for the Hyperbolic decline curve and Duong's model. The use of generalized least squares or generalized nonlinear least squares has little benefit in fitting the decline curves, except for the Logistic Growth model, where it improves both fit and forecasting accuracy.","PeriodicalId":298489,"journal":{"name":"Day 4 Wed, October 10, 2018","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122681169","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}
In multi-fractured horizontal wells (MFHW), fracture properties such as permeability and fracture half-length significantly deteriorate during early production, which negatively affects gas production from shale reservoirs. Therefore, it is crucial to evaluate the temporal changes in fracture properties based on production data. This paper presents a workflow in which both flowback and long-term production data are used to quantitatively evaluate hydraulic fracture closure and changes in the fracture properties. In addition, we develop a two-phase semi-analytical model based on rate transient analysis (RTA) that assumes boundary dominated flow during the flowback period. The proposed workflow consists of three steps. First, we used the flowback data to calculate fracture properties, such as initial fracture permeability and fracture half-length, by employing the two-phase semi-analytical model. Then, we calculated initial fracture permeability by using a single-phase bilinear flow model as well as the fracture half-length and matrix permeability by using a single-phase linear flow model from the long-term gas production data. These models consider pressure dependency of permeability. Last, we compared the results that are calculated from both flowback and long-term production data to evaluate fracture closure and its effects on fracture permeability. We validated the semi-analytical flowback model and the workflow against numerical simulations. The results show that the developed model is capable of predicting fracture properties and evaluating fracture closure. Furthermore, the proposed workflow provides quantitative insights on the performance of fracture stimulation and is able to closely estimate permeability modulus using flowback and long-term production data instead of conducting laboratory experiments.
{"title":"Flowback Fracture Closure of Multifractured Horizontal Wells in Shale Gas Reservoirs","authors":"Fengyuan Zhang, Hamid Emami‐Meybodi","doi":"10.2118/191817-18ERM-MS","DOIUrl":"https://doi.org/10.2118/191817-18ERM-MS","url":null,"abstract":"\u0000 In multi-fractured horizontal wells (MFHW), fracture properties such as permeability and fracture half-length significantly deteriorate during early production, which negatively affects gas production from shale reservoirs. Therefore, it is crucial to evaluate the temporal changes in fracture properties based on production data. This paper presents a workflow in which both flowback and long-term production data are used to quantitatively evaluate hydraulic fracture closure and changes in the fracture properties. In addition, we develop a two-phase semi-analytical model based on rate transient analysis (RTA) that assumes boundary dominated flow during the flowback period. The proposed workflow consists of three steps. First, we used the flowback data to calculate fracture properties, such as initial fracture permeability and fracture half-length, by employing the two-phase semi-analytical model. Then, we calculated initial fracture permeability by using a single-phase bilinear flow model as well as the fracture half-length and matrix permeability by using a single-phase linear flow model from the long-term gas production data. These models consider pressure dependency of permeability. Last, we compared the results that are calculated from both flowback and long-term production data to evaluate fracture closure and its effects on fracture permeability. We validated the semi-analytical flowback model and the workflow against numerical simulations. The results show that the developed model is capable of predicting fracture properties and evaluating fracture closure. Furthermore, the proposed workflow provides quantitative insights on the performance of fracture stimulation and is able to closely estimate permeability modulus using flowback and long-term production data instead of conducting laboratory experiments.","PeriodicalId":298489,"journal":{"name":"Day 4 Wed, October 10, 2018","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115842805","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}
Friction reducers (FRs) are used to decrease the amount of horsepower required to move a hydraulic fracturing fluid through a formation at a fixed flow rate. Though FR viscosity is not a crucial consideration in proppant transport when used before the perforations in slick water applications, FR viscosity becomes a greater consideration in proppant transport from the perforations into the formation and an important qualifying criterion with the advent of High Viscosity Friction Reducer (HVFR) systems that require higher loadings than traditional FRs. Consistent viscosity measurement can vary greatly depending upon a number of factors, for example temperature, hydration approach, polymer concentration, brine composition, and additive interaction. A study was developed and implemented to determine the influence of HVFR by concentrated particulate and bead settling. This study investigated the viscosities of five HVFRs applying eight variables using response surface methodology. Initial study criteria were establishing consistent hydration with unique apparatus design and viscosity measurement verification. Once established, this method examined the effects of 1:1, 2:1, and 2:2 salts, singularly or in various concentrations and combinations. Experimental designs under fresh water conditions were also conducted with varied HVFR loadings (1.0 to 6.0gpt), blender RPM (600 to 12,000), and blender mixing times (0.5 to 8.7 minutes). Viscosities were measured from 200 to 6000 (1/sec). Static settlement testing in ranges of 0.87 to 3.50 pounds per gallon in 0 to 140,000 total dissolved solids (TDS) brines was conducted. Single bead settling measurements were performed in fresh water and API brine. Specific HVFR and salt matrix combinations tested resulted in highly correlated response surfaces exhibiting consistent trends. The TDS and hardness had a minor to major influence on viscosity based upon the specific HVFR examined. Brines were predominately antagonistic with respect to viscosity with few synergistic results. The influences of HVFR dosage and mixing correlated highly to the viscosity of all HVFRs, and extended mixing time durations had no influence on some HVFR combinations indicating a viscosity reduction limit. In certain regions of the design space, settling rates were related to viscosity. Selection of an HVFR system precisely tailored for a specific brine composition guaranteeing maximum friction reduction and proppant transportation performance was vital. The influence of pumping and tubular transport on the HVFR viscosity is continuous and quantifiable. Additionally, the viscosity of the HVFR in a downhole brine environment provides discernable data for assessing far end well bore proppant transport and damage potential. This study established a reliable method for gauging performance and examining measurable field variables of HVFR systems.
{"title":"Analysis of Various High Viscosity Friction Reducers and Brine Ranges Effectiveness on Proppant Transport","authors":"C. Aften","doi":"10.2118/191792-18ERM-MS","DOIUrl":"https://doi.org/10.2118/191792-18ERM-MS","url":null,"abstract":"\u0000 Friction reducers (FRs) are used to decrease the amount of horsepower required to move a hydraulic fracturing fluid through a formation at a fixed flow rate. Though FR viscosity is not a crucial consideration in proppant transport when used before the perforations in slick water applications, FR viscosity becomes a greater consideration in proppant transport from the perforations into the formation and an important qualifying criterion with the advent of High Viscosity Friction Reducer (HVFR) systems that require higher loadings than traditional FRs. Consistent viscosity measurement can vary greatly depending upon a number of factors, for example temperature, hydration approach, polymer concentration, brine composition, and additive interaction. A study was developed and implemented to determine the influence of HVFR by concentrated particulate and bead settling.\u0000 This study investigated the viscosities of five HVFRs applying eight variables using response surface methodology. Initial study criteria were establishing consistent hydration with unique apparatus design and viscosity measurement verification. Once established, this method examined the effects of 1:1, 2:1, and 2:2 salts, singularly or in various concentrations and combinations. Experimental designs under fresh water conditions were also conducted with varied HVFR loadings (1.0 to 6.0gpt), blender RPM (600 to 12,000), and blender mixing times (0.5 to 8.7 minutes). Viscosities were measured from 200 to 6000 (1/sec). Static settlement testing in ranges of 0.87 to 3.50 pounds per gallon in 0 to 140,000 total dissolved solids (TDS) brines was conducted. Single bead settling measurements were performed in fresh water and API brine.\u0000 Specific HVFR and salt matrix combinations tested resulted in highly correlated response surfaces exhibiting consistent trends. The TDS and hardness had a minor to major influence on viscosity based upon the specific HVFR examined. Brines were predominately antagonistic with respect to viscosity with few synergistic results. The influences of HVFR dosage and mixing correlated highly to the viscosity of all HVFRs, and extended mixing time durations had no influence on some HVFR combinations indicating a viscosity reduction limit. In certain regions of the design space, settling rates were related to viscosity.\u0000 Selection of an HVFR system precisely tailored for a specific brine composition guaranteeing maximum friction reduction and proppant transportation performance was vital. The influence of pumping and tubular transport on the HVFR viscosity is continuous and quantifiable. Additionally, the viscosity of the HVFR in a downhole brine environment provides discernable data for assessing far end well bore proppant transport and damage potential. This study established a reliable method for gauging performance and examining measurable field variables of HVFR systems.","PeriodicalId":298489,"journal":{"name":"Day 4 Wed, October 10, 2018","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124512118","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}
A. Goodman, S. Sanguinito, B. Kutchko, S. Natesakhawat, J. Culp
Fundamental research targeting the interactions of CO2 and fluids with unconventional shale systems is limited from the perspective of using carbon dioxide 1) as an alternative fracturing fluid, 2) as an agent to enhance hydrocarbon production, and 3) as an injection agent into the shale formation for storage purposes to avert emissions to the atmosphere. In this work, we apply in-situ infrared spectroscopy (FT-IR), scanning electron microscopy coupled with energy dispersive spectroscopy (SEM-EDS), and Brunauer-Emmett-Teller (BET) surface area and density functional theory (DFT) pore size analysis to examine the effects of CO2 and fluid on the Marcellus and Utica Shales. Results show changes to the shale at both the micron and nanometer scale after reaction with CO2 and water. These alterations could potentially alter overall permeability and fracture networks that may cause issues for future EOR activities, CO2 storage, and/or the practice of using CO2 as a hydraulic fracturing material.
{"title":"Characterization of the CO2-Fluid-Shale Interface Via Feature Relocation Using Field-Emission Scanning Electron Microscopy, in Situ Infrared Spectroscopy, and Pore Size Analysis","authors":"A. Goodman, S. Sanguinito, B. Kutchko, S. Natesakhawat, J. Culp","doi":"10.2118/191828-18erm-ms","DOIUrl":"https://doi.org/10.2118/191828-18erm-ms","url":null,"abstract":"\u0000 Fundamental research targeting the interactions of CO2 and fluids with unconventional shale systems is limited from the perspective of using carbon dioxide 1) as an alternative fracturing fluid, 2) as an agent to enhance hydrocarbon production, and 3) as an injection agent into the shale formation for storage purposes to avert emissions to the atmosphere. In this work, we apply in-situ infrared spectroscopy (FT-IR), scanning electron microscopy coupled with energy dispersive spectroscopy (SEM-EDS), and Brunauer-Emmett-Teller (BET) surface area and density functional theory (DFT) pore size analysis to examine the effects of CO2 and fluid on the Marcellus and Utica Shales. Results show changes to the shale at both the micron and nanometer scale after reaction with CO2 and water. These alterations could potentially alter overall permeability and fracture networks that may cause issues for future EOR activities, CO2 storage, and/or the practice of using CO2 as a hydraulic fracturing material.","PeriodicalId":298489,"journal":{"name":"Day 4 Wed, October 10, 2018","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123135192","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 natural gas from Marcellus Shale can be produced most efficiently through horizontal wells stimulated by multi-stage hydraulic fracturing. The objective of this study is to investigate the impact of the geomechanical factors and non-uniform formation properties on the gas recovery for the horizontal wells with multiple hydraulic fractures completed in Marcellus Shale. Various information including core analysis, well log interpretations, completion records, stimulation design and field information, and production data from the Marcellus Shale wells in Morgantown, WV at the Marcellus Shale Energy and Environment Laboratory (MSEEL) were collected, compiled, and analyzed. The collected shale petrophysical properties included laboratory measurements that provided the impact of stress on core plug permeability and porosity. The petrophysical data were analyzed to estimate the fissure closure stress. The hydraulic fracture properties (half-length and conductivity) were estimated by analyzing the completion data with the aid of a commercial P3D fracture model. In addition, the information from the published studies on Marcellus Shale cores plugs were utilized to determine the impact of stress on the propped fracture conductivity and fissure permeability. The results of the data collection and analysis were utilized to generate a base reservoir model. Various gas storage mechanisms inherent in shales, i.e., free gas (matrix and fissure porosity), and adsorbed gas were incorporated in the model. Furthermore, the geomechanical effects for matrix permeability, fissure permeability, and hydraulic fracture conductivity were included in the model. A commercial reservoir simulator was then employed to predict the gas production for a horizontal well with multi-stage fracture stimulation using the base model. The production data from two horizontal wells (MIP-4H and MIP-6H), that were drilled in 2011 at the site, were utilized for comparison with the model predictions. The model was then also used to perform a number of parametric studies to investigate the impact of the geomechanical factors and non-uniform formation properties on hydraulic fractures and the gas recovery. The matrix permeability geomechanical effect was determined by an innovative method using the core plug analysis results. The results of the modeling study revealed that the fracture stage contribution has a more significant impact on gas recovery than the fracture half-length. Furthermore, the predicted production by the model was significantly higher than the observed field production when the geomechanical effects were excluded from the model. The inclusion of the geomechanical factors, even though it reduced the differences between the predictions and field results to a large degree, was sufficient to obtain an agreement with field data. This lead to the conclusion that various fracture stages do not have the same contribution to the total production. Based on well trajectory, vari
{"title":"Contribution of Hydraulic Fracture Stage on the Gas Recovery from the Marcellus Shale","authors":"M. E. Sgher, K. Aminian, S. Ameri","doi":"10.2118/191778-18ERM-MS","DOIUrl":"https://doi.org/10.2118/191778-18ERM-MS","url":null,"abstract":"\u0000 The natural gas from Marcellus Shale can be produced most efficiently through horizontal wells stimulated by multi-stage hydraulic fracturing. The objective of this study is to investigate the impact of the geomechanical factors and non-uniform formation properties on the gas recovery for the horizontal wells with multiple hydraulic fractures completed in Marcellus Shale.\u0000 Various information including core analysis, well log interpretations, completion records, stimulation design and field information, and production data from the Marcellus Shale wells in Morgantown, WV at the Marcellus Shale Energy and Environment Laboratory (MSEEL) were collected, compiled, and analyzed. The collected shale petrophysical properties included laboratory measurements that provided the impact of stress on core plug permeability and porosity. The petrophysical data were analyzed to estimate the fissure closure stress. The hydraulic fracture properties (half-length and conductivity) were estimated by analyzing the completion data with the aid of a commercial P3D fracture model. In addition, the information from the published studies on Marcellus Shale cores plugs were utilized to determine the impact of stress on the propped fracture conductivity and fissure permeability. The results of the data collection and analysis were utilized to generate a base reservoir model. Various gas storage mechanisms inherent in shales, i.e., free gas (matrix and fissure porosity), and adsorbed gas were incorporated in the model. Furthermore, the geomechanical effects for matrix permeability, fissure permeability, and hydraulic fracture conductivity were included in the model. A commercial reservoir simulator was then employed to predict the gas production for a horizontal well with multi-stage fracture stimulation using the base model. The production data from two horizontal wells (MIP-4H and MIP-6H), that were drilled in 2011 at the site, were utilized for comparison with the model predictions. The model was then also used to perform a number of parametric studies to investigate the impact of the geomechanical factors and non-uniform formation properties on hydraulic fractures and the gas recovery.\u0000 The matrix permeability geomechanical effect was determined by an innovative method using the core plug analysis results. The results of the modeling study revealed that the fracture stage contribution has a more significant impact on gas recovery than the fracture half-length. Furthermore, the predicted production by the model was significantly higher than the observed field production when the geomechanical effects were excluded from the model. The inclusion of the geomechanical factors, even though it reduced the differences between the predictions and field results to a large degree, was sufficient to obtain an agreement with field data. This lead to the conclusion that various fracture stages do not have the same contribution to the total production. Based on well trajectory, vari","PeriodicalId":298489,"journal":{"name":"Day 4 Wed, October 10, 2018","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115149586","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}
Throughout fracturing treatment, millions of gallons of water are injected, but commonly less than 50% is recovered after stimulation. This study was constructed to evaluate the impact of the fracturing additives on the fluid flowback and fluid loss during hydraulic fracturing. Different pad fluids types were considered including; friction reducer fluid, friction reducer with a non-ionic surfactant fluid and 3 wt% HCl acid. Flooding experiments were conducted for core samples from the Eagle Ford outcrop to measure the brine permeability, time of breakthrough and water relative permeability. The measurements were performed for intact samples and also after flooding the samples with the fracturing fluids. A simulation sector modeling for a hydraulically fractured vertical well in the shale formation was constructed to investigate the effect of the fracturing additives on the fluid flowback and fluid loss during hydraulic fracturing. A sensitivity analysis was considered to study the effect of the formation capillary pressure and reservoir pressure on the fluid flowback and fluid loss due to counter-current capillary imbibition. The study results showed that the fluid saturation in the near fracture face shale matrix is highly reduced by the effect of the high capillary pressure. Therefore, the fluid had not flow back from the near fracture face matrix. Moreover, adding a non-ionic surfactant to the friction reducer pad fluid or using 3 wt% HCl increased the fluid loss during pumping and the fluid imbibition during shut-in, flowback, and production. Therefore, the dilute HCl acid and small well shut-in times are recommended when no flowback occurs from the near fracture face matrix due to low fluid saturation. The fluid loss from the near fracture face region due to counter-current capillary imbibition reduced the effect of the fluid saturation on the gas production. However, the high fluid saturation and the polymer adsorption may cause water blocks. Thus, reducing the gas production or leading to a complete gas block. For shales with moderate capillary pressure, a flowback from the near fracture face matrix has occurred. Hence, the friction reducer with a non-ionic surfactant fluid and 3 wt% HCl enhanced both of the fluid loss due to counter-current capillary imbibition and the fluid flowback. However, a non-ionic surfactant and long shut-in time are recommended for the hydraulic fracturing. Shales with low reservoir pressure had less fluid flowback and more fluid loss. To minimize the fluid loss during pumping and to overcome the water block problem, it is recommended to use a friction reducer fluid in the pad stage while injecting a non-ionic surfactant or dilute acid during the subsequent fracturing steps.
{"title":"Hydraulic Fracturing Design in Shale Formations Based on the Impact of Fracturing Additives on the Fluid Loss and Flowback","authors":"A. Al-Ameri, T. Gamadi, I. Ispas, M. Watson","doi":"10.2118/191782-18ERM-MS","DOIUrl":"https://doi.org/10.2118/191782-18ERM-MS","url":null,"abstract":"\u0000 Throughout fracturing treatment, millions of gallons of water are injected, but commonly less than 50% is recovered after stimulation. This study was constructed to evaluate the impact of the fracturing additives on the fluid flowback and fluid loss during hydraulic fracturing. Different pad fluids types were considered including; friction reducer fluid, friction reducer with a non-ionic surfactant fluid and 3 wt% HCl acid.\u0000 Flooding experiments were conducted for core samples from the Eagle Ford outcrop to measure the brine permeability, time of breakthrough and water relative permeability. The measurements were performed for intact samples and also after flooding the samples with the fracturing fluids. A simulation sector modeling for a hydraulically fractured vertical well in the shale formation was constructed to investigate the effect of the fracturing additives on the fluid flowback and fluid loss during hydraulic fracturing. A sensitivity analysis was considered to study the effect of the formation capillary pressure and reservoir pressure on the fluid flowback and fluid loss due to counter-current capillary imbibition.\u0000 The study results showed that the fluid saturation in the near fracture face shale matrix is highly reduced by the effect of the high capillary pressure. Therefore, the fluid had not flow back from the near fracture face matrix. Moreover, adding a non-ionic surfactant to the friction reducer pad fluid or using 3 wt% HCl increased the fluid loss during pumping and the fluid imbibition during shut-in, flowback, and production. Therefore, the dilute HCl acid and small well shut-in times are recommended when no flowback occurs from the near fracture face matrix due to low fluid saturation.\u0000 The fluid loss from the near fracture face region due to counter-current capillary imbibition reduced the effect of the fluid saturation on the gas production. However, the high fluid saturation and the polymer adsorption may cause water blocks. Thus, reducing the gas production or leading to a complete gas block.\u0000 For shales with moderate capillary pressure, a flowback from the near fracture face matrix has occurred. Hence, the friction reducer with a non-ionic surfactant fluid and 3 wt% HCl enhanced both of the fluid loss due to counter-current capillary imbibition and the fluid flowback. However, a non-ionic surfactant and long shut-in time are recommended for the hydraulic fracturing. Shales with low reservoir pressure had less fluid flowback and more fluid loss. To minimize the fluid loss during pumping and to overcome the water block problem, it is recommended to use a friction reducer fluid in the pad stage while injecting a non-ionic surfactant or dilute acid during the subsequent fracturing steps.","PeriodicalId":298489,"journal":{"name":"Day 4 Wed, October 10, 2018","volume":"48 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117117863","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}
Michael Deasy, K. Brown, Jonathan He, Wade Lipscomb, Matthew Ockree, K. Voller, Joseph H. Frantz
This paper presents a case history and lookback on the Reduced Cluster Spacing (RCS) completion design that was initiated in 2012. We review results of the initial analyses used to demonstrate proof of concept, summarize key aspects of the completion design, and discuss execution and results of the initial pilot tests and subsequent field-wide implementation. We propose a method to incorporate results of RCS models into production forecasts, and quantify the impact of RCS designs on volumes and economics over time. We begin by presenting proof of concept analyses used to justify the initial field pilot. We then discuss RCS field trials, commenting on key aspects of project design and operational execution. We compare RCS well performance to control wells using normalized production plots, discuss type curve (TC) forecasting for RCS wells, and touch briefly on more rigorous modeling of RCS completions. We present a methodology to incorporate results from rigorous models into simpler type curves suitable for quick economic analyses and volumetric comparisons. We conclude by reviewing the economic and production impact of RCS on production at the well level and field development level. Case histories are presented demonstrating the use of a production normalization process to assess the value of different completion designs. We demonstrate that RCS completion designs have been successful in terms of both volumes uplift and economic performance. We describe positive and negative aspects of the route taken to implement this strategy in the field. We conclude that the use of "uplift factors" derived from modeling can be used to efficiently incorporate detailed model findings into typical engineering workflows for volumes and economics forecasting. As a result of the work presented in this paper, RCS completions have become the standard in our Marcellus wells. This lookback will present a method to effectively demonstrate proof of concept for new completion designs and assess the field implementation of novel completion strategies. This method is demonstrated by quantifying the value of reduced cluster spacing achieved in the Marcellus. We also provide a simple way to incorporate complex model results into every day engineering and economic forecasts.
{"title":"Reduced Cluster Spacing: From Concept to Implementation – A Case History","authors":"Michael Deasy, K. Brown, Jonathan He, Wade Lipscomb, Matthew Ockree, K. Voller, Joseph H. Frantz","doi":"10.2118/191801-18ERM-MS","DOIUrl":"https://doi.org/10.2118/191801-18ERM-MS","url":null,"abstract":"\u0000 This paper presents a case history and lookback on the Reduced Cluster Spacing (RCS) completion design that was initiated in 2012. We review results of the initial analyses used to demonstrate proof of concept, summarize key aspects of the completion design, and discuss execution and results of the initial pilot tests and subsequent field-wide implementation. We propose a method to incorporate results of RCS models into production forecasts, and quantify the impact of RCS designs on volumes and economics over time.\u0000 We begin by presenting proof of concept analyses used to justify the initial field pilot. We then discuss RCS field trials, commenting on key aspects of project design and operational execution. We compare RCS well performance to control wells using normalized production plots, discuss type curve (TC) forecasting for RCS wells, and touch briefly on more rigorous modeling of RCS completions. We present a methodology to incorporate results from rigorous models into simpler type curves suitable for quick economic analyses and volumetric comparisons. We conclude by reviewing the economic and production impact of RCS on production at the well level and field development level.\u0000 Case histories are presented demonstrating the use of a production normalization process to assess the value of different completion designs. We demonstrate that RCS completion designs have been successful in terms of both volumes uplift and economic performance. We describe positive and negative aspects of the route taken to implement this strategy in the field. We conclude that the use of \"uplift factors\" derived from modeling can be used to efficiently incorporate detailed model findings into typical engineering workflows for volumes and economics forecasting. As a result of the work presented in this paper, RCS completions have become the standard in our Marcellus wells.\u0000 This lookback will present a method to effectively demonstrate proof of concept for new completion designs and assess the field implementation of novel completion strategies. This method is demonstrated by quantifying the value of reduced cluster spacing achieved in the Marcellus. We also provide a simple way to incorporate complex model results into every day engineering and economic forecasts.","PeriodicalId":298489,"journal":{"name":"Day 4 Wed, October 10, 2018","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121102704","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}
Matthew Ockree, K. Brown, Joseph H. Frantz, Michael Deasy, Ramey John
This paper reviews several Big Data analytical initiatives in the Marcellus Shale. We describe how application of Big Data technology evolved, share challenges and benefits derived from Big Data analytical processes, and discuss lessons learned. We present an overview of Big Data methods employed, show how we integrated results with economic analyses to guide field development, and summarize the significant impact on development economics. This paper will help operators, analysts, and investors "de-mystify" Big Data technology, and provide insights and guidance to those embarking on Big Data initiatives. We discuss an ongoing initiative that employs cognitive analytics to generate production type curves via machine learning and couples the results with integrated economic analyses to guide field development. Challenges associated with data management, such as automated data QA/QC, sparse datasets, interpolation/extrapolation, model training and evaluation are discussed. Benefits derived from integrating Big Data-generated type curves with economics analyses to guide well/field optimization are also presented. Our past big data experiences have taught us several important lessons. First, Big Data initiatives are journeys, not destinations, so expect to constantly feel like there is more to learn and do. Nonetheless, implementation of Big Data processes along the journey can add significant value to an asset, as demonstrated in this paper. Second, it is critical to clearly define the problem to be solved; without a crystal-clear mission statement, scope creep is inevitable, because Big Data technology is capable of so much. Finally, partnering with someone that has experience solving similar problems can significantly accelerate the process and add value. Using Machine Learning to generate forecasts allows the engineers to focus their efforts on increasing business value, rather than managing and manipulating data. In the end, we will demonstrate how a process that once took multiple man-weeks of effort was solved within a single man-day of time. Finally, we present an example of an optimization opportunity identified with the potential to yield approximately 15 Bcfe in additional cumulative production, while maximizing future drilling inventory in the Marcellus Shale. (Note – this is presented as a "theoretical" example in the body of the paper.)
{"title":"Integrating Big Data Analytics Into Development Planning Optimization","authors":"Matthew Ockree, K. Brown, Joseph H. Frantz, Michael Deasy, Ramey John","doi":"10.2118/191796-18ERM-MS","DOIUrl":"https://doi.org/10.2118/191796-18ERM-MS","url":null,"abstract":"\u0000 This paper reviews several Big Data analytical initiatives in the Marcellus Shale. We describe how application of Big Data technology evolved, share challenges and benefits derived from Big Data analytical processes, and discuss lessons learned. We present an overview of Big Data methods employed, show how we integrated results with economic analyses to guide field development, and summarize the significant impact on development economics.\u0000 This paper will help operators, analysts, and investors \"de-mystify\" Big Data technology, and provide insights and guidance to those embarking on Big Data initiatives. We discuss an ongoing initiative that employs cognitive analytics to generate production type curves via machine learning and couples the results with integrated economic analyses to guide field development. Challenges associated with data management, such as automated data QA/QC, sparse datasets, interpolation/extrapolation, model training and evaluation are discussed. Benefits derived from integrating Big Data-generated type curves with economics analyses to guide well/field optimization are also presented.\u0000 Our past big data experiences have taught us several important lessons. First, Big Data initiatives are journeys, not destinations, so expect to constantly feel like there is more to learn and do. Nonetheless, implementation of Big Data processes along the journey can add significant value to an asset, as demonstrated in this paper. Second, it is critical to clearly define the problem to be solved; without a crystal-clear mission statement, scope creep is inevitable, because Big Data technology is capable of so much. Finally, partnering with someone that has experience solving similar problems can significantly accelerate the process and add value.\u0000 Using Machine Learning to generate forecasts allows the engineers to focus their efforts on increasing business value, rather than managing and manipulating data. In the end, we will demonstrate how a process that once took multiple man-weeks of effort was solved within a single man-day of time. Finally, we present an example of an optimization opportunity identified with the potential to yield approximately 15 Bcfe in additional cumulative production, while maximizing future drilling inventory in the Marcellus Shale. (Note – this is presented as a \"theoretical\" example in the body of the paper.)","PeriodicalId":298489,"journal":{"name":"Day 4 Wed, October 10, 2018","volume":"89 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134164753","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：