Energy Resources and Petroleum Engineering Program

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 5 of 467
  • Conference Paper

    A new enhanced gas recovery scheme using carbonated water and supercritical CO2

    (International Energy Agency Greenhouse Gas, IEAGHG, 2021-01-01) Omar, Abdirizak; Addassi, Mouadh; Hoteit, Hussein; Vahrenkamp, Volker; Energy Resources and Petroleum Engineering; Energy Resources and Petroleum Engineering Program; Physical Sciences and Engineering; Physical Science and Engineering (PSE) Division; Ali I. Al-Naimi Petroleum Engineering Research Center; Ali I. Al-Naimi Petroleum Engineering Research Center (ANPERC)

    The transition for clean, efficient, reliable, and affordable sources of energy is one of the top challenges of this century. Despite significant advances in renewable energies such as solar and wind, fossil fuels will remain a primary source of energy for years to come. Natural gas, fueling power generation and other vital industrial sectors such as desalination, is expected to play a more prominent role in the energy mix for many countries around the world. The demand for natural gas, especially in the Middle East, is on the rise. With the shortfall in recurrent discoveries of conventional gas resources, improving the recovery factors from legacy gas reservoirs is crucial. On the other hand, burning gas, like other fossil fuels, is associated with significant greenhouse gas emissions that should be mitigated, which is critical to comply with the pressing call for low-emission economies. CO2 injection in gas reservoirs is appealing as it may provide a dual benefit, including enhanced gas recovery (EGR), and partial entrapment of CO2 in the subsurface reservoir. Different CO2-EGR recovery schemes have been studied in the literature and showed that key thermodynamic properties of CO2, such as its high viscosity and density at reservoir conditions, make CO2 an efficient displacing fluid for the hydrocarbon gas. On the one hand, injecting CO2 can maintain the reservoir pressure by providing voidage replacement amid gas production. On the other hand, it can displace and mobilize the gas in place towards the production wells; nonetheless, CO2-EGR may not be economically viable. The main drawback of CO2 injection is the relatively rapid breakthrough at the production wells leading to gas recovery inefficiency due to CO2 separation and recycling. This pre-mature breakthrough of CO2 is driven by CO2/gas dilution from mechanical mixing, dispersion, and molecular diffusion. Producing gas with high impurity because of CO2 mixing adds significant CAPEX and OPEX related to the surface facilities and gas treatment, especially for fields whose initial content of CO2 within the gas composition is negligible. In this work, we propose a CO2-based EGR scheme that involves two stages. The first stage corresponds to the situation when the reservoir pressure is below that CO2 bubble point pressure. At this condition, CO2 will be in the gas state, and therefore, its sweep and displacement efficiency will be poor owing to its low viscosity and density. At this stage, CO2/gas mixing will be pronounced, and CO2 mobility will be high, leading to an early breakthrough. To avoid this unfavorable situation, we propose to co-inject brine and CO2 (carbonated water) in which CO2 will be in the aqueous solution and non-buoyant. As the carbonated water is heavier than the in-situ brine, CO2 remains trapped in the aqueous phase and flows down-dip towards the bottom of the reservoir. In stage 2, CO2 is injected in its dense supercritical phase, where the presence of water reduces its mobility and delays its breakthrough time. This two-stage injection process is more efficient than the traditional continuous CO2 injection. We demonstrate the effectiveness of this recovery scheme for a synthetic model mimicking an actual gas field. We analyze the optimum transition conditions from stage 1 to stage 2 and show a significant incremental recovery and CO2 entrapment compared to the traditional CO2-EGR that relies on continuous CO2 injection. This proposed scheme serves both objectives, which are an efficient EGR and optimum CO2 sequestration resulting from delayed CO2 breakthrough. We believe that this scheme provides significant improvements to the traditional CO2-EGR and has the features to be deployed at the field scale.

  • Conference Paper

    A new enhanced gas recovery scheme using carbonated water and supercritical CO2

    (International Energy Agency Greenhouse Gas, IEAGHG, 2021-01-01) Omar, Abdirizak; Addassi, Mouadh; Hoteit, Hussein; Vahrenkamp, Volker; Energy Resources and Petroleum Engineering; Energy Resources and Petroleum Engineering Program; Physical Sciences and Engineering; Physical Science and Engineering (PSE) Division; Ali I. Al-Naimi Petroleum Engineering Research Center; Ali I. Al-Naimi Petroleum Engineering Research Center (ANPERC)

    The transition for clean, efficient, reliable, and affordable sources of energy is one of the top challenges of this century. Despite significant advances in renewable energies such as solar and wind, fossil fuels will remain a primary source of energy for years to come. Natural gas, fueling power generation and other vital industrial sectors such as desalination, is expected to play a more prominent role in the energy mix for many countries around the world. The demand for natural gas, especially in the Middle East, is on the rise. With the shortfall in recurrent discoveries of conventional gas resources, improving the recovery factors from legacy gas reservoirs is crucial. On the other hand, burning gas, like other fossil fuels, is associated with significant greenhouse gas emissions that should be mitigated, which is critical to comply with the pressing call for low-emission economies. CO2 injection in gas reservoirs is appealing as it may provide a dual benefit, including enhanced gas recovery (EGR), and partial entrapment of CO2 in the subsurface reservoir. Different CO2-EGR recovery schemes have been studied in the literature and showed that key thermodynamic properties of CO2, such as its high viscosity and density at reservoir conditions, make CO2 an efficient displacing fluid for the hydrocarbon gas. On the one hand, injecting CO2 can maintain the reservoir pressure by providing voidage replacement amid gas production. On the other hand, it can displace and mobilize the gas in place towards the production wells; nonetheless, CO2-EGR may not be economically viable. The main drawback of CO2 injection is the relatively rapid breakthrough at the production wells leading to gas recovery inefficiency due to CO2 separation and recycling. This pre-mature breakthrough of CO2 is driven by CO2/gas dilution from mechanical mixing, dispersion, and molecular diffusion. Producing gas with high impurity because of CO2 mixing adds significant CAPEX and OPEX related to the surface facilities and gas treatment, especially for fields whose initial content of CO2 within the gas composition is negligible. In this work, we propose a CO2-based EGR scheme that involves two stages. The first stage corresponds to the situation when the reservoir pressure is below that CO2 bubble point pressure. At this condition, CO2 will be in the gas state, and therefore, its sweep and displacement efficiency will be poor owing to its low viscosity and density. At this stage, CO2/gas mixing will be pronounced, and CO2 mobility will be high, leading to an early breakthrough. To avoid this unfavorable situation, we propose to co-inject brine and CO2 (carbonated water) in which CO2 will be in the aqueous solution and non-buoyant. As the carbonated water is heavier than the in-situ brine, CO2 remains trapped in the aqueous phase and flows down-dip towards the bottom of the reservoir. In stage 2, CO2 is injected in its dense supercritical phase, where the presence of water reduces its mobility and delays its breakthrough time. This two-stage injection process is more efficient than the traditional continuous CO2 injection. We demonstrate the effectiveness of this recovery scheme for a synthetic model mimicking an actual gas field. We analyze the optimum transition conditions from stage 1 to stage 2 and show a significant incremental recovery and CO2 entrapment compared to the traditional CO2-EGR that relies on continuous CO2 injection. This proposed scheme serves both objectives, which are an efficient EGR and optimum CO2 sequestration resulting from delayed CO2 breakthrough. We believe that this scheme provides significant improvements to the traditional CO2-EGR and has the features to be deployed at the field scale.

  • Conference Paper

    Physics-Informed Neural Networks for Modeling Flow in Heterogeneous Porous Media: A Decoupled Pressure-Velocity Approach

    (IPTC, 2024-02-12) Alhubail, Ali; Fahs, Marwan; Lehmann, Francois; Hoteit, Hussein; King Abdullah University of Science and Technology; Energy Resources and Petroleum Engineering; Energy Resources and Petroleum Engineering Program; Physical Sciences and Engineering; Physical Science and Engineering (PSE) Division; Ali I. Al-Naimi Petroleum Engineering Research Center; Ali I. Al-Naimi Petroleum Engineering Research Center (ANPERC); University of Strasbourg

    Physics-informed neural networks (PINNs) have shown success in solving physical problems in various fields. However, PINNs face major limitations when addressing fluid flow in heterogeneous porous media, related to discontinuities in rock properties. This is because automatic differentiation is inadequate for evaluating the spatial derivatives of hydraulic conductivity where it is discontinuous. This study aims to devise PINN implementations that overcome this limitation. This work proposes decoupling the mass conservation equation from Darcy's law and utilizing the residuals of these decoupled equations to train the loss function of the PINN, rather than using a single residual from the combined equation. As a result, we circumvent the need to find the spatial derivative of the discontinuous hydraulic conductivity, and instead, we impose the continuity of fluxes. This decoupling necessitates that each primary unknown (pressure and velocity components) be computed by the neural networks (NNs) rather than deriving the velocity (or fluxes) from the pressure. We examined three NN configurations and compared their performance by analyzing their accuracy and training time for various 2D scenarios. These scenarios explored various boundary conditions, different hydraulic conductivity fields, as well as different orientations of the heterogeneous media within the domain of interest. In these problems, the pressure and velocity field are the primary unknowns. The three configurations include: (a) one NN with the three unknowns as its outputs, (b) two NNs, one outputting pressure and the other outputting the velocity, and (c) three NNs, each having one primary unknown as an output. Utilizing these NN architectures, we were able to solve the heterogeneous problems with varying levels of accuracy when compared to results from numerical simulators. While maintaining a similar number of training parameters for a fair assessment, the configuration with three NNs yielded the most accurate results, with a comparable training time to the other configurations. Using this optimal configuration, we performed a sensitivity analysis to demonstrate the effect of modifying the NN(s) hyperparameters, such as the number of layers, the number of nodes per layer, and the learning rate, on the accuracy of the results. We introduce a novel PINN approach for modeling fluid flow in heterogeneous media. This proposed method not only preserves the inherent discontinuity of rock petrophysical properties but also leverages the benefits of automatic differentiation. By incorporating this PINN architecture, we have opened up new possibilities for extending the application of PINN to realistic reservoir simulations that capture the complexities of the subsurface.

  • Conference Paper

    Physics-Informed Neural Networks for Modeling Flow in Heterogeneous Porous Media: A Decoupled Pressure-Velocity Approach

    (IPTC, 2024-02-12) Alhubail, Ali; Fahs, Marwan; Lehmann, Francois; Hoteit, Hussein; King Abdullah University of Science and Technology; Energy Resources and Petroleum Engineering; Energy Resources and Petroleum Engineering Program; Physical Sciences and Engineering; Physical Science and Engineering (PSE) Division; Ali I. Al-Naimi Petroleum Engineering Research Center; Ali I. Al-Naimi Petroleum Engineering Research Center (ANPERC); University of Strasbourg

    Physics-informed neural networks (PINNs) have shown success in solving physical problems in various fields. However, PINNs face major limitations when addressing fluid flow in heterogeneous porous media, related to discontinuities in rock properties. This is because automatic differentiation is inadequate for evaluating the spatial derivatives of hydraulic conductivity where it is discontinuous. This study aims to devise PINN implementations that overcome this limitation. This work proposes decoupling the mass conservation equation from Darcy's law and utilizing the residuals of these decoupled equations to train the loss function of the PINN, rather than using a single residual from the combined equation. As a result, we circumvent the need to find the spatial derivative of the discontinuous hydraulic conductivity, and instead, we impose the continuity of fluxes. This decoupling necessitates that each primary unknown (pressure and velocity components) be computed by the neural networks (NNs) rather than deriving the velocity (or fluxes) from the pressure. We examined three NN configurations and compared their performance by analyzing their accuracy and training time for various 2D scenarios. These scenarios explored various boundary conditions, different hydraulic conductivity fields, as well as different orientations of the heterogeneous media within the domain of interest. In these problems, the pressure and velocity field are the primary unknowns. The three configurations include: (a) one NN with the three unknowns as its outputs, (b) two NNs, one outputting pressure and the other outputting the velocity, and (c) three NNs, each having one primary unknown as an output. Utilizing these NN architectures, we were able to solve the heterogeneous problems with varying levels of accuracy when compared to results from numerical simulators. While maintaining a similar number of training parameters for a fair assessment, the configuration with three NNs yielded the most accurate results, with a comparable training time to the other configurations. Using this optimal configuration, we performed a sensitivity analysis to demonstrate the effect of modifying the NN(s) hyperparameters, such as the number of layers, the number of nodes per layer, and the learning rate, on the accuracy of the results. We introduce a novel PINN approach for modeling fluid flow in heterogeneous media. This proposed method not only preserves the inherent discontinuity of rock petrophysical properties but also leverages the benefits of automatic differentiation. By incorporating this PINN architecture, we have opened up new possibilities for extending the application of PINN to realistic reservoir simulations that capture the complexities of the subsurface.

  • Conference Paper

    Visualizing Geoscience Data Interpretation Into Photorealistic Modern Analog Using Generative AI: A Preliminary Result From Carbonate Platform Environment

    (IPTC, 2024-02-12) Ramdani, Ahmad; Perbawa, Andika; Vahrenkamp, Volker; Energy Resources and Petroleum Engineering Department, King Abdullah University of Science and Technology, Saudi Arabia; Earth Science and Engineering; Earth Science and Engineering Program; Physical Sciences and Engineering; Physical Science and Engineering (PSE) Division; Ali I. Al-Naimi Petroleum Engineering Research Center; Ali I. Al-Naimi Petroleum Engineering Research Center (ANPERC); Energy Resources and Petroleum Engineering; Energy Resources and Petroleum Engineering Program

    Geoscience datasets are fundamental for subsurface investigation. Paradoxically, they are sometimes exclusive and require subject-specific expertise to interpret and visualize. One such example is seismic interpretation. Geophysicists typically reconstruct ancient depositional settings by interpreting a myriad of seismic attributes and drawing analogs to the sedimentary process of the modern depositional environment (Posamentier et al. 2007; Vahrenkamp et al. 2019; Ramdani et al. 2021). Most of these interpretations will likely be reflection amplitude, frequency, impedance, or other geophysical attributes interpreted and "visualized" in the present-day geomorphology context (Posamentier et al. 2007; Warrlich et al. 2019; Ramdani et al. 2022b). The interpreter will then rely on verbal or written descriptions to convey their interpretation. Often, these descriptions are only well understood by fellow interpreters. Attempting to convey the same interpretation to a non-expert requires some degree of visual aid. Thus, a method to picture geophysical signals as a "depositional environment" is needed to bridge this gap. This study aims to leverage the application of generative AI as a tool for seismic interpretation. We propose a Conditional Generative Adversarial Network (CGAN)-based methodology capable of converting seismic attribute maps into photorealistic images of the modern satellite imagery analog as a visual aid for seismic interpretation.