Summary Hydraulic fractures can enhance well productivity from stress-sensitive naturally fractured reservoirs, such as coalbed methane or coal seam gas (CSG) reservoirs. Graded proppant injection (GPI) has been proposed to enhance long-term, far-field interconnectivity between the created hydraulic and short-term, enhanced natural fracture permeability, resulting from fracture fluid leakoff and lowered net effective stress. This novel study shows how applying GPI with hydraulic fracturing treatments resulting in an increased stimulated reservoir volume (SRV) can enhance well productivity and improve CSG well economics. A commercially available reservoir model and history-matched hydraulically fractured coal seam case are used to evaluate well performance differences between a hydraulic fractured reservoir and one including GPI application. A dual-porosity system and the Palmer and Mansoori model are used to simulate initial and long-term permeability accounting for reservoir depletion (i.e., increased net effective stress and matrix shrinkage). A previously validated case study is used to describe the post-embedment benefits of GPI based on the porosity model and history-matched reservoir properties. A net present value (NPV) can then be calculated for each scenario, based on the production differences and typical Australian CSG costs. Our results show that permeability enhancement is achieved beyond the hydraulically fractured region for all post-GPI stimulation cases. An optimal SRV can be found relative to permeability that maximizes the incremental NPV from GPI application. The next most significant parameters after permeability that influence the economic outcomes are fracture porosity and coal compressibility. A larger SRV yields higher cumulative gas production over 30 years with up to 7.2 times increase over gas production without GPI. This study substantially increases our understanding of how to model and understand the benefits of GPI application along with hydraulic fracturing to increase the SRV in CSG wells.
Summary In coal-seam-gas (CSG) fields, where single wells tap multiple seams, it is likely that some of the individual seams hardly contribute to gas recovery. This study aims to examine the contribution of individual seams to the total gas and water production considering that each seam can have different properties and dimensions. A sensitivity analysis using reservoir simulation investigates the effects of individual seam properties on production profiles. A radial model simulates the production of a single CSG well consisting of a stack of two seams with a range of properties for permeability, thickness, seam extent, initial reservoir pressure, coal compressibility and porosity. The stress dependency of permeability obeys the Palmer and Mansoori (1998) model. A time coefficient (α) relates seam radius, viscosity, porosity, fracture compressibility, and permeability. It is used to aid interpretation of the sensitivity study. Finally, two hypothetical simulation scenarios with five seams of different thicknesses and depths obtained from producing wells are explored. The range in properties represents conditions found in the Walloon Coal Measures (WCM) of the Surat Basin, relevant to the Australian CSG industry. Each seam in the stack achieves its peak production rate at different times, and this can be estimated using α. Seams with lower α reach the peak gas rate earlier than those with higher α-coefficient. The distinct behavior of gas-production profiles depends on the combination of individual seam properties and multiseam interaction. At a αratio > 1 (i.e., αtop/αbottom > 1), the bottom seam peaks first but achieves lower gas recovery than the top seam. An increasing αratio is associated with the inhibition of less-permeable seams and reduced overall well productivity. For αratio < 1, the top seam experiences fast depletion and total gas-production rates decrease drastically. This outcome is confirmed by a more realistic scenario with a higher number of coal layers. Poor combination of seams leads to severe production inhibition of some coal reservoirs and possible wellbore crossflow. The contrast of the seam-lateral extent in the stack and fracture compressibility play an important role in well productivity in the commingled operation of a stack of coal seams. Unfortunately, the lateral extent of individual coal seams is difficult to estimate and poorly known and, therefore, represents a major uncertainty in gas-production prognosis. The αratio analysis is a useful tool to gain understanding of modeled well productivity from commingled CSG reservoirs.
Defining pressure dependent permeability (PDP) behaviour in coalbed methane (CBM) or coal seam gas (CSG) reservoirs using reservoir simulation is non-unique based on the uncertainty in coal properties and input parameters. A diagnostic fracture injection test (DFIT) can be used to investigate bulk permeability at a reservoir level and at lowered net effective stress conditions. As coal has minimal matrix porosity and under DFIT conditions cleat porosity is fluid saturated with reasonably definable total compressibility values, the DFIT data can provide insight into PDP parameters. At pressures above the fissure opening pressure, pressure dependent leak off (PDL) behaviour increases exponentially with increasing pressure. Many authors have noted that with decreasing pressure PDP declines exponentially with increasing net effective stress. Thus, PDP behaviour can be defined by PDL. In this paper, we show how combined analyses, using typically collected field data, can be used to better define and constrain the modelling of PDP. We illustrate this process based on a well case study that includes the following data: fracture fabric and porosity reasonably defined from image log and areal core studies; DFIT data acquired under initial saturation conditions; hydraulic fracturing data; and longer term production data. These analyses will be integrated and used to constrain the parameters required to obtain a rate and pressure history-match from the post-frac well production data. This workflow has application in other coal seam gas cases by identifying key variables where hydraulic fracturing performance has been unable to overcome limitations based on pressure or stress dependent behaviours and often accompanied by low reservoir permeability values. While this is purposely targeting areas where only typically collected field data is available, this workflow can include coal testing data for matrix swelling/shrinkage properties or other production data analysis techniques.
Carbonate reservoirs are more geochemically reactive than sandstones and can experience big changes in porosity and permeability because of mineral reactions. In this work, we analysed the calcite dissolution and precipitation in chalk reservoirs during injection of seawater and CO2 bearing fluids. We performed reactive transport simulations with injection of seawater, carbonated water, CO2 gas, CO2 SWAG and CO2 WAG. We evaluated the mineral reactions that occur in the injector and producer blocks. Moreover, the calcite dissolution rate was calculated and its relationship with flow rate was investigated. Simulation results showed that during injection of CO2 gas alone, calcite dissolution was fast but limited, and occurred everywhere. On the other hand, for the other injected fluids the dissolution around the injector was continuous and, with the exception of the seawater scenario, precipitation was observed downstream. In addition, the calcite dissolution per injected water pore volumes for both CO2 SWAG and CO2 WAG was higher because of higher dissolution of gaseous CO2 in injected and formation waters. Moreover, the dissolution rate was found to be proportional to the water flow rate which confirms the assumption that calcite kinetics are fast compared to reservoir flow. This knowledge is valuable when planning CO2 WAG projects in carbonate reservoirs. As dissolution rates increase with flow rates, high permeability zones will show faster porosity changes, which may compromise the injector wellbore integrity and may lead to a more severe and growing calcite scaling risk around the producer wellbore.
Summary Vapor/liquid-equilibria (VLE) calculations, particularly involving the phase behavior of carbon dioxide (CO2) and hydrogen sulfide (H2S), are used in scale-prediction modeling. In this work, the impact of VLE calculations for CO2- and H2S-rich gas phases and for acid- and sour-gas mixtures on scale-prediction calculations is evaluated. Three equations of state (EOSs)—Soave-Redlich-Kwong (SRK) (Soave 1972), Peng-Robinson (PR) (Peng and Robinson 1976), and Valderrama-Patel-Teja (VPT) (Valderrama 1990)—are implemented in the Heriot-Watt model and used in VLE calculations. The solubility of single-component CO2 and H2S in water and the solubility of a gas mixture in water were compared with experimental data in terms of the absolute relative deviation (ARD). The solubility data were then used in PHREEQC (USGS 2016) to calculate the impact of using different EOSs on carbonate and sulfide scales, particularly on calcium carbonate (CaCO3) and iron sulfide (FeS). Average ARDs of 6.04, 4.10, and 3.77% between experimental and calculated values for CO2 solubility in water were obtained for the SRK, PR, and VPT EOSs, respectively. Similarly, for H2S solubility in water, average ARDs of 6.49, 6.66, and 6.48% were obtained for each EOS, respectively. For the solubility of sour- and acid-gas mixtures in water, average ARDs of 13.92, 13.25, and 10.78% were obtained, respectively. It has thus been concluded that the VPT EOS performs better than the SRK and PR EOSs in VLE calculations for the analyzed data. The errors introduced in VLE calculations have been found to impact the calculation of the amount of CaCO3 precipitated, with consequences for scale-inhibitor selection. Higher deviations were found in the amount of CaCO3 precipitation for gas mixtures when compared with a single-component, CO2-rich phase. Furthermore, the large errors occurring in VLE calculations for H2S solubility have not been found to impact the calculation of the amount of FeS precipitated when H2S is in excess with respect to Fe2+. In addition to this, a case study that was performed by use of formation-water data from the Brazilian presalt revealed that the choice of EOS can cause errors of 6 kg of precipitate during each day of production. Scale-prediction calculations carried out with PHREEQC demonstrate that VLE calculations can have a high impact on mineral precipitation. Thus, it is recommended that the best VLE model available should always be used for scale-prediction modeling, particularly when mixtures of gases are present.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.