Marcellus Shale, a Devonian black shale, spans the majority of the Appalachian Basin from New York through Pennsylvania, West Virginia and also extends into Ohio and Maryland (Bartuska, et al. 2012). The unconventional gas reservoir is a term commonly used to refer to ultra-low permeability formations that produces mainly dry natural gas and is not able to produce an economic flow rate without stimulation treatments. The natural gas in the Marcellus Shale is produced most efficiently through horizontal wells with multiple hydraulic fracturing stimulation treatments. Even though advances in technology have unlocked considerable reserves of hydrocarbon, the long-term production behavior of the horizontal wells with multiple hydraulic fractures is not well understood. This paper provides the results of parametric studies to investigate the impact of the hydraulic fracture properties and more specifically the impact of non-uniform fracture half-length, on the gas recovery from Marcellus Shale.
The purpose of this study is to evaluate the long-term production performance of horizontal wells with multiple hydraulic fractures completed in Marcellus Shale. A commercial reservoir simulator was used to develop the base model which incorporated the storage and production mechanisms inherent in shales. The core, log, completion, stimulation, and production data obtained from wells located at the Marcellus Shale Energy and Environment Laboratory (MSEEL) were utilized to generate the simulation model. MSEEL is a research collaboration between West Virginia University, Ohio State University, The Natural Energy Technology Laboratory, and Northeast Natural Energy. MSEEL aims to achieve a better understanding of the unconventional shale resources through application of advanced technology in drilling, completion, reservoir characterization, production and monitoring of horizontal wells. Furthermore, fracture properties (fracture half-length, fracture width and fracture conductivity) were obtained by using commercial software. Precision laboratory equipment was utilized to determine the shale properties from the core samples.
Field production data from two original horizontal Marcellus Shale gas wells at MSEEL site were utilized for history matching to establish the missing shale parameters. History matching was initially performed using production data for two years from one the horizontal wells. The matched model was then used to predict the production for the following two years to confirm the accuracy and reliability of the model. In addition, the base model was used to predict the second horizontal well production which provided reliable results. Finally, a number of parametric studies were performed with the model to investigate the impact of the hydraulic fracture properties and non-uniform fracture half-length, on the recovery.
The advances in hydraulic fracturing and horizontal well technology have unlocked considerable natural gas reserves contained in the shale formations. Reliable values of the shale key petrophysical properties including permeability and porosity are necessary to estimate the original gas-in-place, predict the production rates, and optimize the hydraulic fracturing treatments. The quantification of the key shale petrophysical properties however remain challenging due to complex nature of the shale foramtions. Unsteady state techniques are commonly used to estimate permeability of the shale samples because the shales typically have permeability values in nano-Darcy range. The measured permeability values by these techniques however suffer from a large margin of uncertainty and reproducibility problems. Furthermore, the unsteady state measurements cannot be performed under the reservoir stress and temperature conditions.
In this study, a fully automated laboratory set-up, which has been designed and constructed for the evaluation of the ultra-low permeability petrophysical properties under the reservoir conditions, was utilized to measure the porosity and permeability of the Marcellus shale core plugs. The core plugs were obtained from a vertical well drilled specifically for the laboratory research and other scientific purposes (science well) on the site of the Marcellus Shale Energy and Environment Laboratory (MSEEL). MSEEL is a field site and dedicated laboratory in the Marcellus Shale unconventional production region of north-central West Virginia. The filed site is owned and operated by Northeast Natural Energy, LLC and contains several horizontal Marcellus Shale wells. MSEEL provides a unique opportunity to undertake field and laboratory research to advance and demonstrate new subsurface technologies and to enable surface environmental studies related to unconventional energy development.
One of the core plugs obtained from the science well was used in this study for the evaluation of reliable Marcellus Shale petrophysical properties. The permeability of the core plug was measured under different gas pressures at constant net stress. The absolute permeability was then determined by applying the appropriate gas slippage correction. The porosity and the permeability of the core plug were then measured under a wide range of net stress. The measured porosity and permeability values were found to be sensitive to the stress. The permeability measurement results exhibited two distinctive behaviors with respect to the net stress that can be attributed to the natural fracture and matrix properties. The experimental results were then utilized to determine the natural fracture closure stress. The measurements also revealed that gas adsorption, when an adsorbent gas was used for the mesurements, resulted in a reduction in the absolute permeability of the sample.
The Marcellus Shale is a prolific source of natural gas and is located in much of the Appalachian Basin, The combination of horizontal drilling and multistage hydraulic fracturing have proven extremely successful in achieving commercial production from this ultra-low permeability formation. Even though advances in technology have unlocked considerable gas reserves, the emergence of unconventional shale formations as targets for exploration and development have created new challenges for resource development. These challenges arise because the impact many reservoir and fracture parameters have on the effectiveness of the stimulation treatments to increase reserves and improve production efficiency is not well established.
The focus of this study is to identify what effect fracture half-length and well spacing have on the maximization of gas production and gas recovery efficiency from Marcellus Shale horizontal wells. A commercial reservoir simulator was used to develop a base model for the Marcellus Shale formation. These simulations models were constructed using information from the Marcellus Shale Energy and Environment Laboratory (MSEEL). MSEEL is a research collaboration between West Virginia University, Ohio State University, The National Energy Technology Laboratory, and Northeast Natural Energy. MSEEL is a long-term laboratory and field study that will assess the hydrocarbon production potential and the environmental and economic impacts of drilling and producing wells in the Marcellus Shale. The base model incorporated many of the complex reservoir characteristics that are associated with the Marcellus Shale. They include a horizontal wellbore, multi-stage fracture treatment, adsorbed gas, and in situ rock properties. Reservoir properties, such as porosity, permeability, natural fracture spacing, etc., were determine using history matching. The available data from two original horizontal Marcellus Shale gas wells at MSEEL were utilized for history matching the daily and cumulative gas production. The impact of completion parameters on gas recovery was investigated using history-matched models to evaluate the impact that fracture half-length and well spacing have on production. The predicted optimized production will be presented and compared to the baseline production results, and be presented in the paper.
The advances in hydraulic fracturing and horizontal well technology have unlocked considerable reserves of hydrocarbon contained in shale formations. However, quantification of the key shale petrophysical properties remain challenging. It is not practical to measure the permeability of the unconventional formations such as shales by standard steady state techniques because shales typically have permeability values in nano-Darcy range. Therefore, unsteady state methods have been extensively used to estimate permeability of the shale samples. However, the measured permeability values by these techniques suffer from large margin of uncertainty and reproducibility problems. These problems are attributed to the lack of consistent experimental protocols and the interpretations of the transient data. Another limitation of the unsteady-state measurements is that the experiments cannot be performed under the reservoir stress and temperature conditions.
This paper provides the results of the porosity and permeability measurements on Marcellus shale core plugs which were performed using a fully automated laboratory set-up for evaluation of the ultra-low permeability petrophysical properties under the confining pressure. The permeability of the core plug were first measured under different gas pressures at constant net stress. The absolute permeability was then determined by applying the gas double-slippage correction. The porosity and the permeability of the core plug were then measured under a wide range of net stress. The measured porosity and permeability values were found to be sensitive to stress. Two distinctive behaviors with net stress, for both porosity and permeability, were observed that can be related to the natural fracture and matrix properties. The experimental results were then utilized to determine the natural fracture closure pressure. The permeability measurements with carbon dioxide revealed that permeability is impacted by adsorption. The results of the measurements with were carbon dioxide also provided information for determination of the sorption characteristics that were found to be in agreement with the published values.
The reservoir fluid properties exhibit significant deviation from their bulk properties under confinement in nanopores. As a result, storage capacity and recovery can be impacted in the shale reservoir due to nanoscale pore structure. It is therefore critical to investigate the impact of pore confinement to accurately estimate the gas resources and the phase behavior in the shale reservoirs. In this study, the shift in critical properties of the typical Marcellus Shale gas confined in different size nanopores were determined. The adjusted fluid properties was then used to estimate the gas deviation factor for dry gas and to predict the phase behavior of gas/condensate. The gas/condensate fluid model was then used in a compositional reservoir simulator in order to estimate the liquid recovery from a Marcellus Shale gas/condensate reservoir developed by hydraulically fractured horizontal well.
The results indicate that the gas storage capacity of the shale is altered by the pore confinement. The degree of the fluid properties alteration by pore proximity in the shale depends on the composition of the gas mixture and the size of the nanopores. Therefore to accurately estimate the gas resources in the shale reservoirs, it is necessary to incorporate the alteration of the gas deviation factor by pore proximity. In gas/condensate shale reservoirs the fluid confinement could have either positive or negative impact on the liquid recovery depending on the pore size. The results of the investigation can provide an insight relative to production optimization and economic recovery of gas/condensate reservoirs in the unconventional resources.
This paper introduces a robust and accurate technique for the steady-state permeability and porosity measurements in ultra-low permeability shale core samples. A laboratory set-up was designed and assembled which has a resolution of one nano-darcy for the permeability and one-hundredth cubic centimeters for pore volume measurements. Extremely accurate differential-pressure transducers are used to measure the flow of gas passing through the core sample under in-situ conditions. The in-situ conditions are achieved by maintaining isothermal conditions and the application of the confining stress on the core sample. The laboratory set-up is fully automated to eliminate any human error and more importantly maintains the temperature stable within the enclosed unit.
A series of measurements were performed on a Marcellus Shale core sample under wide range of pore and confining pressures using Helium (
In this study both Klinkenberg and double slippage corrections were applied to the steady-state permeability measurements. The results indicated that application of Klinkenberg to the permeability measurements lead to negative absolute permeability for shale samples. However, the double-slippage correction resulted in physically plausible values for absolute permeability of shale samples. Finally, the measurement results with adsorbent gases indicated that the adsorbed gas layer thickness can significantly impact the gas transport and storage in organic rich shale reservoirs and needs to be considered for hydrocarbon in place calculation and production predictions.
The long-term production behavior of horizontal wells producing from Marcellus shale due to limited production history has not been well established. As a result a simple method for predicting the long-term production and reserves would be of interest. Several production decline curve analysis models have been proposed specifically for application to unconventional gas reservoirs. However, reliable production predictions cannot be obtained when they are applied to limited production history such as those available from Marcellus shale wells. In this study, the production and completion data from a number of horizontal wells completed in Marcellus shale were collected and used to build a generic model. The model was then in conjunction with a commercial dual porosity numerical model which included the adsorbed gas to simulate long-term production profiles for Marcellus shale horizontal wells with multiple hydraulic fracture stages. The simulated production profiles were then utilized to develop correlations for adjusting the conventional (Arps) decline curve constants obtained to from the limited production history to accurately predict the long-term production performance. The impacts of key formation and fracture properties on the decline curve constants were also investigated. Finally, the correlations were utilized to accurately predict the long-term production rates from a Marcellus shale horizontal well in West Virginia based on the early production history.
The production behavior of horizontal wells producing from Marcellus shale has not been well established due to limited production history. As a result a simple method for predicting the long-term production would be of interest to the industry. Several DCA models have been proposed specifically for unconventional gas reservoirs. However, their applicability to production data from Marcellus shale wells has not been attempted. In this study the production and completion data from a number of horizontal wells completed in Marcellus shale were collected. In addition, the properties of Marcellus shale were measured in laboratory with precision equipment designed for unconventional formations. The field data as well the laboratory measured properties were utilized in conjunction with a commercial numerical simulator to predict the long-term production behavior of horizontal wells producing from Marcellus shale. The numerical model allowed for inclusion of adsorbed gas, multiplehydraulic fracture stages, as well as dual porosity behavior. The predicted production profiles were then utilized to evaluate the applicability of the various DCA models. Subsequently, a technique was developed to estimate the parameters of the DCA model to predict the long-term production based on the early production history as well as the key reservoir parameters. Finally, the results of the DCA model predictions were compared to the history-matched simulation model predictions for confirmation.
Naturally fractured reservoirs (NFR) have been receiving more attention than ever since the beginning of the last decade due to various reasons. The current understanding is not sufficient to achieve a favorable recovery factor, due to the lack of effective characterization and the complex dynamic nature of the fractured system. The fact that NFR are found in many countries around the globe in almost every lithology is another justification for more interest. The majority of the fractured reservoirs around the world are developed. Therefore, before proceeding into secondary or possibly tertiary recovery processes a thorough understanding must be accomplished to avoid undesirable results. The fracture characterization is the first building block in any NFR study. This study utilized well test data to detect fractures between two communicating reservoirs. The study was inspired by a real field example in which two reservoirs, separated by a thick non-reservoir formation, are in hydraulic communication with each other. A refined simulation model was constructed to predict the type of response that would be observed in such communicating reservoirs during a well test. This paper presents the result of this study, which showed a unique shape on the derivative due to the communication through fractures. The investigation was taken further to study the impact of the first and the third layer permeabilities on the observed shape of the derivative. The impact of fracture length, fracture permeability, fracture distance from the well, and the existence of multiple fractures were also investigated.
A depleted gas condensate reservoir in the Appalachian Basin has been under investigation for possible conversion to a gas storage reservoir. When depleted gas condensate reservoirs are used for gas storage, the injected gas will pressurize the reservoir and will evaporate the retrograde condensate that is remaining in the reservoir upon completion of the primary production. This will result in significant compositional alteration of the withdrawn gas from the storage. The produced gas during withdrawal cycle must be processed to remove condensible hydrocarbons prior to pipeline transportation to prevent liquid drop out by retrograde condensation in the pipeline. The composition of produced gas will depend on the degree of mixing between the injected and the residue fluids.
To investigate the impact of the formation on the mixing, a compositional reservoir simulator was utilized in this study. The primary production history of the reservoir was first modeled with simulator using available data and the results of the data analysis. The phase behavior studies with Peng-Robinson equation of state (PR-EOS) provided a reliable estimate for reservoir fluid composition. Upon successfully history matching the primary production, the reservoir model was utilized to predict the composition of the withdrawn gas from the storage. Pipeline gas was injected into the reservoir and the reservoir was pressurized to its original pressure. Both vertical and horizontal wells were considered for injection to evaluate the extent of mixing among residue gas, residue condensate, and the injected gas. Subsequently, a withdrawal cycle was simulated to predict the liquid yields and gas heating contents based on the design of the surface facilities. The results of simulation study were then used to finalize the design storage field and surface facilities. The preliminary data collected after pressurizing the reservoir confirms the simulation results.